130 | What is Knock - Essential Knowledge
Summary
Since knock is the biggest killer of performance engines, it’s essential to have an understanding of what it is, what causes it, and how to know when it’s occurring. We’ll cover these topics in this webinar and you’ll also hear what knock sounds like when it happens
00:00 | Hi guys, Andre from High Performance Academy, welcome to another webinar, and in this webinar, we're going to be discussing what I'd consider to be one of the most important aspects of performance, EFI tuning, and that is, understanding, highlighting, and most importantly of all, avoiding knock or detonation while we're performing our tuning. |
00:23 | And the reason I think this is so important is because, in my experience, over somewhere in the region of 15 years in the industry, and certainly a lot of the time I spent running my performance workshop was spent tuning very high specific output, high boost, turbo charged drag engines, what I repeatedly saw was that knock is undoubtedly the biggest killer of any performance engine. |
00:49 | So if we don't understand what we're dealing with when it comes to knock, obviously we can easily end up damaging and destroying a lot of expensive engines, and expensive engine components. |
01:02 | And I think we'll start just straight away with a little bit of a look at exactly how damaging detonation can be. |
01:10 | So this is a piston, JE Forged piston, out of one of our Mitsubishi 4G63 drag engines, and this was out my own car, this suffered a failure at about the 1,000 foot mark, due to one of the three injectors that we ran on each cylinder, failing. |
01:29 | So basically, it was only getting 2/3 of the required fuel flow. |
01:33 | So this sort of damage that we can see here, happened in probably less than a second, and you can see that it's lanced the whole side out of the piston, melted straight down the side of the piston, and also down through the rings. |
01:48 | So really catastrophic, and very, very quick failure. |
01:53 | So this is obviously something we need to understand and avoid if we want to make sure that our engines are going to be reliable. |
02:02 | Now obviously, that is an extreme case, and what we find with knock, is that the damage that it produces, is very, very tightly linked to the specific power level that an engine's producing. |
02:16 | So what I mean by this is, we've probably all heard a 20 year old Toyota Corolla, or a Honda Civic, naturally aspirated engine, that is being labored up a hill, probably in a gear that's too high, and the driver is sitting there at full throttle, and probably about 1,500 RPM. |
02:36 | And you can audibly hear the rattle associated with knock, or the pinging sound associated with knock, I'm going to talk about that in detail shortly, so don't worry if you don't understand what that is. |
02:47 | But this is essentially, that engine suffering from detonation. |
02:51 | And the chances are that, that engine will probably survive a long and healthy life, it might do another 100,000 or 200,000 kilometers, with that sort of detonation occurring, simply because at the point where the detonation is occurring, the amount of power, or specific power, and when I say specific power, I'm sort of really meaning power per cylinder, is so low, that there's no real chance of the detonation actually causing damage. |
03:18 | When we take that to extremes though, and we look at an engine, like my Mitsubishi 4G63 drag engine, we're making somewhere in the region of perhaps 300 horsepower per cylinder. |
03:31 | At that sort of power level, and at very, very high RPM, almost any level of detonation, can be damaging, and it can be damaging very quickly. |
03:41 | So obviously, as we're focusing primarily in the industry, and performance engines, and over the years, of course, we've seen great advances in the amount of power that we can make, and how easy and cheap it is to make more power, then this becomes a more prominent issue that we need to understand, we need to focus on, and we need to be able to deal with. |
04:05 | Okay, so we've got a bit of background there, let's move back a step though, and understand what exactly knock is, and in order to understand what knock is, we actually need to start by talking about what normal combustion looks like. |
04:21 | So in other words, what's actually happening, inside the combustion chamber, inside the cylinder, during normal combustion, the sort of combustion that we want, or expect to be occurring while our engine is operating. |
04:33 | Now a lot of people consider that, when the spark event occurs, that's obviously what initiates the combustion process, they feel that, or believe that the fuel/air charge inside the cylinder explodes, kind of little bit like a bomb, or stick of dynamite exploding inside the cylinder. |
04:52 | And all of that energy is instantly released, and in reality, that's far from the case. |
04:58 | What we actually get is that the spark will ignite the fuel and air charge right beside the spark, so right by where that spark happens, we get a kernel of fuel and air that begins to burn. |
05:12 | So combustion is initiated right by the spark plug. |
05:15 | Now what happens from there is, we get a bit of flame front that is going to propagate out through the combustion chamber, and as it moves out through the combustion chamber. |
05:24 | And this is a relatively slow process in terms of engine speed aspects, what it does is, it then ignites the unburnt fuel and air charge that's ahead of it, so we get this relatively gradual and smooth combustion process, and this results in a smooth and steady increase in the pressure inside the cylinder. |
05:44 | And it's a really important aspect of normal combustion, that's what we need to have occurring. |
05:52 | Okay, so when we get knock occurring, on the other hand, and I'll just actually, also mention that, when I'm talking about the term knock, this is also interchangeable with the term detonation, and very, very different from the other term that's often confused with, which is pre-ignition, and we'll talk a little bit about that shortly. |
06:15 | So, just if you are wondering about that term, knock, knock and detonation, spark knock, I've also heard it referred to, we're all referring to the same thing. |
06:25 | So the important point to understand about knock, the definition of knock or detonation, or spark knock, is that it is a type of abnormal combustion, that begins after the spark event has occurred. |
06:41 | That's a really key point that distinguishes knock and pre-ignition. |
06:46 | Okay, so what's happened is, we've initiated combustion, we've got the combustion process underway, the flame front is propagating out, and while it's doing that, what we're ending up with is, the pressure inside the cylinder is increasing quite rapidly. |
07:01 | And as the pressure inside the cylinder increases, so does the heat, and this is the key point, the heat inside the cylinder, at some point, if we are having our engine suffer from knock, will get so great that pockets of unburnt fuel around the outside of the combustion chamber, often referred to as in gasses, those will get so hot that essentially, the fuel and air charge spontaneously combusts. |
07:27 | And when this occurs, so this separate to the fuel and air charge being ignited by that propagating flame front, these pockets of fuel and air just spontaneously combust, due to the increasing heat inside the combustion chamber. |
07:41 | Now when this occurs, it's quite a different type of combustion event, this time the, basically the energy contained in that fuel is released almost instantly, so we get large amounts of energy being released very, very quickly, and a very, very fast moving flame front, as a result of this. |
08:02 | So, the damaging aspect of this is, we end up with large pressure pulsations, happening inside the combustion chamber, as these little pockets of fuel and air spontaneously combust, or explode, in the cylinder. |
08:15 | And you really could liken it to small sticks of dynamite exploding inside the combustion chamber, while the engine is operating. |
08:23 | And of course, the damage that knock can do is just as dangerous or damaging as that may sound. |
08:33 | So we've already seen the results of very, very high levels of detonation with the piston that I've shown you. |
08:40 | We can end up with the whole side melted out of the piston in a matter of seconds. |
08:46 | With lighter forms of detonation though, what we're going to end up seeing, one of the first signs will be a sandblasted appearance to the crown of the piston. |
08:57 | Now why this is happening is because, under normal combustion, we find that the crown of the piston is protected by a layer of unburnt fuel and air, unburnt gasses called a boundary layer, and this protects that piston from the full heat of combustion, which is kind of handy, because depending on the alloy that the piston it made out of, the piston may melt somewhere in the region of maybe 700 to 750 degrees Centigrade, and full heat of combustion, understandably, is much hotter than that. |
09:30 | So when we get detonation occurring, these pressure pulses can strip away that boundary layer of gasses that's protecting the crown of the piston, allowing that full heat of combustion to contact the piston crown. |
09:43 | And this is why we get the sandblasted, or pitted, appearance. |
09:47 | Now while there's obviously much more catastrophic damage going on, we actually can see, maybe a little bit too difficult on this GoPro image, but we actually can see signs, before we get to the catastrophic damage, of the sandblasted finish to the piston. |
10:04 | The other aspect to be wary of is on the top ring land, particularly if you have a piston that uses anti-detonation grooves, or contact reduction grooves, as they're also known, these are quite sharp little grooves machined into that top ring land, and those can also start to show signs of melting, that sandblasted appearance. |
10:27 | So that's going to be some of your first clues. |
10:30 | The other aspect of these pressure pulses is, it could be likened to someone smashing down onto the crown of the piston with a sledge hammer, and again, the damage that that creates, is just as bad as it sounds, particularly in cast pistons, we can find that, large sections of the lower ring lands can be completely broken out of the piston. |
10:53 | Now sometimes this won't be apparent until a full strip down of the engine, and what we'll find is that sections of the ring land will be broken out in the piston, and will trapped, obviously, between the rings until we actually remove the piston from the engine, and then it can fall out. |
11:11 | The other aspects, we can see the same sandblasted appearance on the aluminum of the cylinder head can also end up damaging the top of the bores. |
11:20 | We can also see the sort of detonation damage resulting in leaking head gaskets, blown head gaskets, because the pressure inside the cylinder becomes so great. |
11:32 | Also, if you've got an engine that has known weaknesses in the connecting rods, we're not passing a lot more pressure down through the crown of the piston, into the connecting rod. |
11:41 | This can be enough to bend or damage the connecting rod, and finally, we can also see the pressure pulses being transferred down through the piston, into the connecting rod, and this can end up showing damage in the big end bearings, where the big end bearing is essentially hammered through the oil film and can contact the journal surface of the crank shaft. |
12:05 | So, all things considered, not a lot of good things going on here, potential for a huge amount of damage, and something we understandably need to stay away from. |
12:19 | Now I did mention that there is the term pre-ignition, which is often confused with detonation. |
12:25 | While we are talking here about knock detonation, I just want to just touch very briefly on pre-ignition, for those who are confused about those terms. |
12:36 | So pre-ignition, as it's name implies, is an abnormal type of combustion that occurs prior, or pre, the ignition event, so it occurs prior to the spark event happening. |
12:48 | Whereas knock, as we've found out, that happens after the spark event. |
12:51 | So pre-ignition essentially is an abnormal type of combustion, where the fuel and air charge is prematurely ignited, this could be due to a glowing ember inside the combustion chamber, it could be due to the spark plug electrode glowing, something of that nature. |
13:10 | And here, we're starting the combustion process way before the spark event would normally occur, and it's important to understand that the point where pre-ignition is likely to occur is actually somewhere pretty close to 180 degrees before top dead center on the compression stroke, so right at the bottom of the intake stroke, when the intake valves are closing, and the piston is just starting to make it's way up towards the top of the cylinder, beginning to compress the fuel and air charge. |
13:40 | When that pressure inside the cylinder is at its lowest, that's the point where it's actually easiest for that fuel air charge to be combusted. |
13:48 | As we move up the bore, as the piston moves towards top dead center, and compresses that fuel and air charge, and the pressure inside the cylinder rises, it actually becomes harder and harder for that charge to be ignited. |
14:02 | So pre-ignition is most likely to happen fairly early in that compression stroke, it's actually much more catastrophic, in terms of the damage it causes compared to knock, although, thankfully, it also a little less common. |
14:17 | And pre-ignition is much more likely to show up, first as a hole, essentially blown straight through the center of the crown of the piston. |
14:27 | And the reason for this is that with pre-ignition, we're igniting that fuel air charge very early in the compression stroke, so for the entire compression stroke, the engine is working against the rapidly expanding combustion charge that's occurring in that cylinder. |
14:42 | There is a lot of heat also being exposed to the crown of the piston, and because the piston get rid of that heat very easily, this is why we often see couple the heat and the pressure, that's why we often see a hole melted through the top of the piston. |
14:58 | And to give you some sort of context around this, in a high output engine, such as the drag engines I was talking about, we may end up seeing catastrophic failure, like I've shown you from a detonation, extended detonation, in perhaps a second or less. |
15:16 | Whereas, with pre-ignition, we're likely to see catastrophic failure of the engine in as much as, just a few, as few as a few engine cycles. |
15:27 | Sorry, I'll try that again. |
15:29 | We're likely to see that damage happen as quickly as just a few complete engine cycles, so very, very damaging. |
15:37 | Alright. So again, we are focusing the rest of this webinar on knock, it's just important to clarify, because I know there are a lot of people who still get confused about that situation. |
15:49 | Now, I just want to jump across now to a graph here on my laptop screen, so if we can have a look at this. |
15:55 | This is a graph I actually use when I'm training or talking about optimizing ignition timing. |
16:02 | But it does sort of show one of the effects here, we've got three points graphed here on this graph. |
16:09 | On the vertical axis we have cylinder pressure, so this is coming from in cylinder pressure sensor. |
16:15 | On the horizontal axis, we have whereabouts in the engine cycle we are, it also corresponds to our ignition advance angle, considering ignition advances, define and degrees of crank shaft rotation. |
16:28 | So on the center point here, we have our zero point, which is our top dead center point, so we've got three points of our ignition plotted, and we've got point C here, where we started the ignition timing, 13 degrees before top dead center. |
16:42 | And if we look at the pressure occurring here, we see that, that pressure really doesn't get much time to get going before the piston's moved past top dead center, and has started down the power stroke. |
16:53 | Then we advance the timing a little bit further, and we move through to point B, which in this case, is 20 degrees, now for this particular example, it happens to be our MBT timing. |
17:04 | We see our pressure rises a little bit more sharply, and we see our pressure rise and peak at about 16 to 18 degrees after TDC. |
17:12 | However, we then continue to further advance the timing to point A, in this case, 35 degrees before top dead center, and we see now the pressure is rising sharply, while the piston is still trying to come up towards TDC. |
17:26 | And remember, that pressure, as it rises in the cylinder, also creates heat, and that's what leads us to the, the auto-ignition, or spontaneous combustion of those in gasses. |
17:38 | And that's being demonstrated here by this sharp spiking that we see in the pressure trace, and that's the spiking of those pockets of fuel and air, exploding, and that's the spike in the pressure, that can be so damaging. |
17:53 | The other aspect to note here, is the peak pressure, you can see, these numbers are just arbitrary, but in our example, with the correct ignition timing, we had 80 bar of cylinder pressure, in our demonstration here, we would have gotten knock. |
18:08 | We're peaking at over 120, so, significantly higher pressures are occurring during knock, and this is another lead reason why we see so much damage from knock occurring. |
18:21 | Alright, so we'll carry on now, so I've talked about what knock is, we've given some background into the damage that can cause. |
18:29 | And now it's worth discussing what actually causes knock, and I mean, there are a variety, a wide variety, of reasons, some engines, due to their combustion chamber, designed for example, a cylinder head design, are more prone to knock, and others with seemingly similar designs of engines. |
18:48 | So there really are an unlimited number of parameters that can result in an engine that is more prone to knock. |
18:55 | However, I'll just discuss some of the key things. |
19:00 | So really, what we're getting at here is, the excessive heat in the combustion chamber is probably one of the prime drivers for knock, and anything that is going to result in higher combustion chamber heat, is obviously going to make the engine more prone to knock. |
19:17 | At the same time, this goes hand in hand with the octane of the fuel, that you're tuning on. |
19:23 | The octane of the fuel, and another way of looking at that, is just a measure of the fuel resistance to suffering from auto-ignition, or detonation. |
19:35 | So a fuel with a higher octane, all other things being equal, is going to be more stable, or less susceptible to suffering from knock. |
19:44 | Also, along the same lines of the fuel's octane rating, if we've got an engine that is consuming or using large quantities of oil, or even moderate quantities of oil, and that could be either oil directly bypassing the rings, making it's way into the combustion chamber, it could also be oil that is reentering the intake system, through a crank case breather, for example. |
20:09 | This has the effect of diluting the octane of the fuel, or reducing the octane rating of the fuel, and that will again, make the engine more prone to suffering from detonation. |
20:22 | Now I've talked about the heat in the combustion chamber, and there are a few aspects that can affect that heat. |
20:31 | So, these include a high intake air temperature, so if we've got excessively high intake air temperatures, maybe we're operating in very high ambient temperature conditions, or we've got a turbo charged or super charged engine, where our inter-cooling is not effective. |
20:48 | Higher intake temperatures will result in higher combustion chamber temperatures, which in turn will make an engine more prone to suffering from detonation. |
20:58 | Likewise, if our engine coolant temperature is excessive, this also can't remove the heat from the combustion chamber as easily, and again, makes the engine more prone to suffering from knock. |
21:12 | Then we have the obvious ones, which include high compression ratios, the higher the compression ratio, the higher the peak pressure that's going to occur in the cylinder. |
21:23 | Again, this creates more heat, makes it more likely for an engine to suffer from knock. |
21:28 | We can cause or create that same effect, in a boosted engine by increasing our boost pressure. |
21:34 | And of course, we've got the situation where we've got a combination of all of the above, stacking up to make our engine suffer from knock. |
21:43 | And in that vein as well, I also generally, just for simplicity, when I'm tuning engines, I class them into two separate categories, I will refer to an engine as either being not knock limited, and what this means is, that we can quite safely tune the engine, turn in the ignition timing through to MBT, without any onset of detonation or knock occurring. |
22:08 | The other close of engine an engine that is knock limited, and in that case, what we're going to find is, while we're advancing the ignition timing on the dyno, and we are seeing the torque increase on the dyno, at some point we're going to end up having the engine begin suffering from knock, we're going to audibly hear that. |
22:28 | We'll talk, very shortly, about ways we can monitor and notice if the engine is knocking as well, so don't worry too much about that. |
22:36 | Now, in my experience with the engines I tune, primarily the engines that I tune on pump fuel, and almost certainly if I'm tuning a turbo charged engine on pump fuel, that engine will be knock limited. |
22:52 | Now I can't say that's always the case, but very, very likely to be the case. |
22:57 | The engine that we're going to use for demonstration here, the ah 350Z our Nissan Q35 naturally aspirated, 3.5 litre V6, on our local 98 RON fuel, is actually, not quite knock limited, but right on the very brink. |
23:15 | So, it's actually quite a safe engine to tune, but in a lot of cases, we're going to find that even naturally aspirated engines on pump fuel will still be knock limited. |
23:27 | Okay, so now let's move and we're going to talk about how we can detect knock. |
23:32 | Obviously if it's such a big deal, if it's so important to make sure our engine is not suffering from knock, we need to find out whether or not it's occurring, and what are the techniques we can use to do so. |
23:46 | So one of the key things about knock is the pressure pulses that we've seen on the little graph, the pressure pulses that I've talked about, what these do is create a resonance in the engine block. |
23:58 | So it's essentially like smacking the engine block with a hammer, and the entire engine block will resonate at a certain frequency. |
24:08 | And that's what occurs when knock is happening, and this is what we're trying to pick up. |
24:13 | So, particularly in the example I gave of the 20 year old Toyota Corolla being lumbered up a hill in the wrong gear at a high load, we're going to be able to audibly hear that resonance in the engine block from outside of the car, it's referred to often as pinging, and I liken it to the sound that you get if you took a metal tray, put some coins on it, and you shook it around. |
24:38 | That's the kind of sound that you're going to audibly get. |
24:42 | Now that's all well and good, in completely stock car like that, but particularly when we are tuning a very heavily modified race engine, particularly one that has a loud exhaust system, or worse still, an exhaust system that vents straight to atmosphere with no muffling, we're going to really struggle to audibly hear knock over the ambient noise that the engine is creating. |
25:06 | So using our bare ears to listen for knock is definitely not a viable option, in my opinion. |
25:13 | And certainly if we can audibly hear knock occurring, over that sound, over the exhaust noise, over the intake noise and the general engine noise, chances are that, we've already probably done some pretty significant damage if the engine is reasonably powerful. |
25:30 | So we need some other options to, to notice if knock is occurring. |
25:36 | And this is where knock detection comes in, and this really falls into two categories, we've got audio knock detection, which is where we're going to use a product to listen to a knock sensor bolted to the engine, run it through a digital signal processor, and then output that signal to headphones, we're going to have a listen to that really shortly. |
26:00 | And this is what I've used for my entire career, I've used various products, I've use the Link, G4 Plus KnockBlock, I've used the older Link KnockBlock, I've used a product called The Knock Box, which came out of Australia. |
26:13 | More recently, those who've followed us will know, that I'm also using the Plex Knock Monitor, essentially all of these products have pros and cons, advantages and disadvantages, and different sets of, sets of sort of skills, I guess you'd like to call it. |
26:29 | But, they all essentially also do the same thing, they take a signal from piezoelectric knock sensor, which you could liken to a small microphone that gets bolted to the side of the cylinder block. |
26:43 | That piezoelectric sensor detects vibration in the engine block, and it transfers that, or converts that vibration into an electrical signal, that goes into the audio knock detection equipment through a digital signal processor, that often times will allow us to focus on specific frequencies, and then output that using an amplifier to the audio headset. |
27:07 | So that's the audio knock detection system, the other option is an ECU with built in knock detection, it's essentially the same thing, although the digital signal processor is all working in harmony with the ECU, we're able to monitor, visually, the output of the knock signal, and we can tune the knock control strategy to come into play and safeguard our engine, if it is found that knock is occurring. |
27:36 | Now a good example here, or one of the things to keep in mind is that, knock happens at a relatively specific frequency, and it is possible to calculate the frequency that knock is likely to occur at. |
27:52 | There are a range of factors that will affect the specific frequency that knock occurs at, and we do need to do some testing to fine tune this. |
28:03 | We can at least get ourselves in the ballpark, by calculating the likely knock frequency based solely on the bore diameter. |
28:12 | And the calculation here is 1,800, divided by pi, multiplied by our bore diameter. |
28:20 | So what I'll do is, we'll just go through this, so if we can just jump to my laptop screen. |
28:26 | And what we've got here, is our VQ35, it has a 95 millimeter bore diameter, the equation that I just gave you as well, is in, you do need to enter your bore diameter in millimeters, so that's a metric equation, that will output the frequency in kilohertz. |
28:46 | So, let's go through that, what we want to do is, multiply pi, so let's take 3.14 by 95, and it gives an answer of 298.3. |
28:57 | And then what we're going to do is, divide 1,800 by 298.3, and it gives us a base frequency of knock for this particular engine of six kilohertz, or 6.03 kilohertz. |
29:10 | And as I said, that's not likely to be exactly correct, but it's going to get us in the ballpark, depending on the ECU or system we're tuning. |
29:18 | We can then start looking at frequencies close to that, and finding out what gives us the best signal to noise ratio. |
29:27 | The other option that a lot of systems will have is, to refer to a look at what is known as the second harmonic frequency for knock. |
29:37 | So in this case, we're looking at 6.03 kilohertz, let's call it six kilohertz for simplicity. |
29:44 | The second harmonic frequency is simply double that, which is 12 kilohertz, and often what we'll find is, that if we're looking or listening to knock at the second order harmonic, we're going to find that, that offers a better signal to noise ratio. |
29:59 | When I use the term signal to noise ratio, what I'm talking about here is, trying to highlight knock and separate that from the obvious background, mechanical noise, that's naturally occurring in our engine. |
30:13 | There's obviously a lot of noise with the pistons moving against the bores, we've got valve train noise as well, so by looking at the specific frequency, or that second harmonic frequency of knock, this is where we've seen a lot of advances in knock control recently, because the systems are able to do a better job of isolating knock, from the background engine noise. |
30:38 | So from my perspective, and I'm really quite a big advocate of this, I believe that anyone who is professionally tuning for a living, or even those home enthusiasts who want to get serious about tuning, should be looking at an audio knock detection equipment system. |
30:54 | It's really, in the big scheme of things, a relatively small expense, and I would not feel safe tuning any engine without having audio knock detection fitted to it from the outset. |
31:08 | And some of these systems are really quite cost effective. |
31:12 | Obviously we've got higher end systems, like the Plex Knock Monitor, which offer some really advanced functionality. |
31:18 | But systems such as the the Link G4 Plus KnockBlock, which are a little bit more stripped out, and basic, so very, very cost effective. |
31:27 | Particularly if you're professionally tuning, you could factor that in and it's less than the cost of just one tune, so it's an essential in my opinion. |
31:38 | Okay so, I've talked a little bit about the piezoelectric knock sensor, and again, regardless of whether we're talking about audio knock detection or we're talking about a closed loop knock control strategy inside our ECU, one of the key aspects is, where we're going to actually fit the knock sensor to the engine block. |
31:59 | This is a question we quite often get, I've had it in the forum, I've had it on our Facebook page, so I'll just cover this off again, and we need to understand here, that there is the ideal scenario, and then there's what we actually, realistically can achieve. |
32:13 | Modern engine bays are becoming increasingly compact, there's more and more equipment being jammed in there, and often it's simply not possible to fit a stand alone knock sensor, particularly if we're using an aftermarket knock detection system, just for the purposes of tuning, that will then be removed afterwards, it's simply not practical sometimes to fit the sensor in the perfect location. |
32:37 | So the locations that I'll use and sort of order of preference, the best place would be to look at where your factory knock sensors are mounted, and these are generally mounted on the engine block, relatively high up near the deck surface. |
32:52 | So this is going to give you the best opportunity to pick up and monitor knock accurately. |
32:59 | Now as I've said often, that's not possible, so my preference beyond this, would be lower down on the engine block. |
33:05 | If we've got a cast sump, I've also had really good success bolting the knock sensor directly to the sump. |
33:12 | Although it's not ideal, what we find is that, that resonance in the engine block actually transmits quite quickly through the entire system. |
33:21 | So everything that's solidly bolted to that engine block, will still show those frequencies of knock. |
33:27 | Beyond the sump, the other place that does work quite well, is the front cover, if that's a cast aluminum front cover, you can bolt it on there. |
33:37 | Then I would look at locations on the cylinder head, and this is probably my least favorite place, because you tend to pick up a lot more valve train noise. |
33:48 | So sometimes you just need to get a little bit creative, on exactly where you're going to fit that sensor, sometimes I will remove a bracket temporarily, while I am tuning, so I've got a place to bolt the sensor to, then I'll put the factory bracket back in, once I've completed my tune. |
34:05 | The other thing you'll often find as well, is that the knock sensors, as they're delivered, so I'm using Bosch donut style knock sensors, as delivered, they come with M8 hole, eight millimeter hole through the center of them. |
34:21 | And what we'll quite often find is that, the available holes on the engine block will be either 10 mil or 3/8 diameter, so there are a couple of options here, you can buy or you can make a stepped stud, which M10, or 3/8 UNC on one side, and then M8 on the other to pass through the knock sensor. |
34:43 | Of course that works incredibly well, there's no problem there. |
34:46 | The other thing that I have come to do, and have absolutely no problems with, it sounds a little bit scary, but if you're careful, you can actually drill the centre of the knock sensor out to 10 millimetres, and I have had absolutely no problem with that degrading the accuracy or the results from that knock sensor, so I have absolutely no hesitation in recommending that you do that. |
35:10 | It is really important to make sure wherever you are bolting that knock sensor on the engine, that you have a flat surface that is wide enough to take the sensor. |
35:20 | It's really important that the base of the sensor is bolted firmly to the engine, so that it can easily transmit all of that vibration into the sensor. |
35:30 | So if that's not the case, you're going to have to look for another location. |
35:36 | Okay, so what we'll do now is we'll have a quick demonstration here our 350Z. |
35:42 | And what I wanted to do was give you the opportunity to hear what knock actually sounds like, so let me just give our KnockBlock a test and I'm not sure it is powered up. |
35:59 | Okay, that is all up and running, so I've got a Link G4 Plus KnockBlock connected here to our 350Z engine, while the KnockBlock will take input from two sensors, for our demonstration, I have one single sensor fitted to the left bank of cylinders, and while I'm talking here, I'm just going to get our engine up and running, so that I can get a little bit of heat into it. |
36:27 | And what we're going to do, once I start this demonstration is, we're going to have a look at making this engine, purposely suffer from knock. |
36:38 | So in order to do this, as I've already mentioned, this engine actually is not heavily knock limited, so under normal conditions, on 98 octane fuel, I can actually quite safely tune this, and not have any problems with it suffering from detonation. |
36:55 | So what I've done here, let's just jump into our laptop screen, and I'll talk you through it. |
37:00 | I'm on our ignition table here. |
37:02 | I'm going to perform this demonstration at 1,750 RPM, and the reason I'm going to do that is, I can easily make the engine knock there, and also, because it's quite low in the RPM, the specific value is quite low, I'm unlikely to end up suffering from engine damage, obviously while we want to be as thorough as possible with these demonstrations, we're not that interested in breaking engines to bring you a webinar either. |
37:29 | So you can see what I've done here is, the two ignition timing values at wide open throttle, set to 45 degrees, whereas MBT timing around that zone, is somewhere around about 15 to 18 degrees, or something on that order, so well and truly advanced well past MBT timing. |
37:49 | Now what I'm also going to do here is, jump across to our logging. |
37:53 | And when we perform this demonstration, you can see at the moment I've got a few examples of some of the testing we did before we came live here, just zoom in a little bit on one of these, and show you what you're going to be looking at. |
38:06 | On the top graph, we have our engine speed, on the next graph down we have throttle position, then on the next graph down, we have our ignition timing, and then finally, this is the one that you're going to be interested in watching here. |
38:19 | We have our knock level, so this is the knock activity coming from the knock sensor, or the vibration, in other words, the raw data, processed data, coming from the knock sensor. |
38:31 | And at the bottom here, you'll also be able to see there is a gray line, which is our knock threshold. |
38:39 | So essentially anytime the individual cylinder levels go above that knock threshold, this is the engine suffering from knock, and it's what you can see here with the spikes. |
38:50 | So what we're going to do, shortly we'll mute my microphone, we'll start the logger running. |
38:58 | And what I'm going to do is I’m going to cycle in and out of full throttle, and you're going to hopefully be able to audibly hear the click that occurs when knock happens, and you're going to also see that occur simultaneously with these spikes as the individual cylinder levels go above the knock threshold. |
39:19 | So let's just get our data logger running, we'll mute the microphone now, and I'll get our RPM back down, and let's do our test. |
40:10 | Okay, so I can't hear what you guys were hearing, simultaneously, I'm hoping that, that demonstration did prove to be as effective as it was when we were testing it. |
40:20 | Certainly if we look now, on my laptop screen, we can see that several hits of detonation were detected, particularly it seemed actually worse than the first section where I went to full throttle, and we went in and out of the throttle a few times there, and we can see that there were several cylinders suffering form knock during that first test. |
40:43 | So we saw a little bit occurring in that last test, and what you should be hearing is the general background engine noise, so we're always going to be able to audibly hear something of the background engine noise. |
40:56 | And we should, if our knock detection equipment is doing a good job and you've got everything set up correctly, you should be able to hear a distinct tick or click, when knock is occurring. |
41:06 | Now unfortunately of course, the 350Z, being an engine that isn't really knock limited, in standard form, this isn't really the best example, or best engine to demonstrate this on. |
41:19 | And at higher RPM, audibly listening, and separating out knock from background engine noise, can be a little bit tricky, it does take a little bit of experience on the part of the tuner. |
41:32 | And even sometimes when I get an engine in that is maybe a little bit worn out, and mechanically very noisy, it can be quite challenging for me to pick up knock audibly. |
41:44 | Now I'm going to have questions and answers really shortly, so if you do have anything that you'd like me to answer, I see I've already got a few coming through. |
41:53 | Please ask those in the chat, and Colin will transfer those through to me shortly. |
42:00 | There is a fairly well, widely held theory, amongst a lot of tuners, both enthusiasts and professional, that knock detection equipment is unnecessary. |
42:11 | And the theory goes that when we're tuning on the dyno, all we need is a quality dyno that can accurately measure torque, and if we've got that, what we find is, that when the engine begins to suffer from knock, we'll start to see the torque drop away. |
42:29 | Now the reason for this is when a cylinder does begin to knock, we will see the torque on that cylinder drop. |
42:36 | Now the theory is sound, and so far as, when a cylinder knocks, yes, the torque will drop off, but my argument, or rebuttal to that is, that it's unlikely that we’ll see every cylinder on an engine begin to knock simultaneously. |
42:52 | Much more likely, is that we're going to find that one or perhaps two cylinders will begin to suffer from knock initially. |
43:00 | And while those cylinders may be suffering from knock, and yes, we may see the torque output from those cylinders drop away, we're also going to see the other cylinders on the engine that aren't knock limited at that point, quite likely still pick up torque, and in this case, we see the overall torque on the engine, will still increase. |
43:21 | Now I've had several situations where, I've had a car on the dyno, particularly my best example here is turbo charged engines running on pump gas, where we're running the engine under a moderate amount of boost, perhaps 15 or 18 PSI, and we're advancing the timing. |
43:37 | And we're seeing really good improvements in engine torque and power, but we're audibly starting to hear the engine begin to suffer from knock. |
43:46 | And we've got that situation, we need to retard the timing, and produce a safety margin between, between the timing we're running, and the timing that actually begins to create knock. |
43:58 | So that's really important for long term engine reliability. |
44:03 | Okay, so moving on, we've heard what knock is, we know a little bit more about what causes it, what the phenomenon of knock actually is. |
44:13 | We'll talk now about some precautions we can use during our tuning to help avoid knock. |
44:18 | So the first one really goes without saying, is always use the best grade, or best octane of fuel that you have available to you. |
44:28 | This is a sort of a split aspect here, because some of you will be tuning your own cars, some of you will be professional tuners where, you're limited by what the customer is bringing you. |
44:39 | But a classic example of this was, spent a lot of time tuning late model GM LS V8s, and what we see here in New Zealand, is the Australian domestic market, Holden or HSV brand. |
44:54 | And we tuned literally hundreds of these cars over several years, so you get to get a bit of a feel for what a particular combination should produce. |
45:03 | How much ignition time it should handle, and what sort of power we are likely to see at the wheels. |
45:09 | And on occasion, what we would end up with is a car delivered to us, and you'd do the upgrades that we had discussed with the customer, put it on the dyno and it would just be horrible. |
45:21 | It'd be rattling it's head off, it wouldn't take any timing, the knock control in the factory GM ECU would be going haywire, and the end result of this, would sometimes be as much as just 30 or 40 kilowatts at the wheels, down on where we knew that car should be. |
45:37 | So a quick call to the customer, asking what grade of fuel they are running, it would usually be, the answer we would usually get was, that they were running on 91 octane, because that's what the dealership said they could run their car on. |
45:51 | So of course, draining the fuel out, putting in 98 octane fuel, and all of the sudden, everything was back to where we knew it should be, and the car was making the power that we expected and taking the sort of ignition timing that we'd expect, so that's a really important aspect. |
46:07 | Make sure that, if you're dealing with customers, that your customers are educated about that aspect. |
46:15 | Along the same lines as well, it's really important to make sure that you perform your tuning on the same fuel that you intend to run. |
46:22 | Now, I feel this should go without saying, but you'd be surprised how many times I have customers bring me cars to tune, and we talked about the fuel, and they've put an octane booster in it, or they're running 98 octane for the dyno, but had not intention of running that once they got the car back from the tuning. |
46:41 | Now that was a bit of bewildering sort of situation for me, and when I dug a little bit deeper, what I found is that these customers wanted to supply the car with a good quality fuel, so it made a big number on the dyno, and then they'd run a cheaper fuel on the street. |
46:56 | Sort of really missing the whole aspect of, I optimized the ignition timing and the boost for the fuel that was in the tank, and then they put a lower grade fuel in the tank after the car was tuned. |
47:08 | Obviously, that could entail detonation occurring, and engine damage, and of course, the finger always gets pointed back at the tuner. |
47:16 | So again, a little bit of education for your customers, if you're professionally tuning, is always advisable and can help you maintain a good reputation for a failure that would be completely out of your hands. |
47:30 | Okay, beyond those basics, of course we have the obvious is, monitor for knock or detonation while you're tuning, and the key here is, to also start with conservative ignition timing, you always want to start with safe and retarded ignition timing, and creep up on MBT timing slowly, while you're audibly listening for knock, or you're running your closed loop with knock control. |
47:56 | And that way, you can establish whether your engine is knock limited or whether it isn't, this will guide you in what you're going to be able to get away with, with your tuning. |
48:06 | Now beyond this as well, we know now that, knock is to a large extent related to the heat in the combustion chamber. |
48:14 | So there's some things we can do in our tuning, to help safeguard that engine against higher combustion temperatures. |
48:22 | In particular, I'll often use intake air temperature ignition timing modifiers, as well as engine coolant temperature ignition timing modifier tables. |
48:32 | So these are compensations that allow us to remove a little bit of timing when either the intake air temperature, engine coolant temperature or perhaps both, start getting a little bit higher than what we would like. |
48:46 | So this can also be simulated aswell, if you're tuning a turbo charged engine, in particular, it's really easy. |
48:53 | If you're tuning one of those in the dyno, what you can do is, perform your tuning, and then you can purposely blank off the inter-cooler, using a piece of cardboard or something like that. |
49:03 | And it's going to stop your dyno fan from cooling the inter-cooler, you can perform two back to back dyno runs, and that's going to allow the intake air temperature to creep up. |
49:13 | so what that's going to do is allow you to test and see if you need to remove further ignition timing at higher intake air temperatures, so always an advisable aspect to cover off. |
49:26 | It is important though, to note that if you are going to use these timing modifiers, these compensation tables, remember that they can become additive. |
49:35 | What I mean by this is, if we're suffering from both high engine coolant temperature, and high intake air temperature, and you've got compensations for both of those conditions, those will both be active so, removing, let's say, three degrees because of air intake temperature, and a further three degrees because of engine coolant temperature, you're now removing six degrees total. |
49:56 | And that can, in itself, create it's own set of headaches. |
49:59 | So it is important to understand, when you are tuning those tables, what their potential effect can be. |
50:07 | The other aspect that's a little bit easy to overlook, is that combustion chamber temperature will build up over sustained high load operation, and a really easy example of this on a knock limited engine is if we allow the engine to reach normal operating temperature, perhaps idle, with the cooling fan on the dyno running. |
50:28 | Then we got to, let's say, 3,000 RPM, wide open throttle, and we optimize the tuning very, very quickly to MBT. |
50:36 | And then we hold the engine there. |
50:38 | Quite often what we're going to find is that, as we sit there for a sustained amount of time, we're going to actually start finding that the engine begins to suffer from knock. |
50:49 | And this is because under that sustained load, we're seeing the combustion chamber temperature start to rise. |
50:54 | So that's something we need to consider, and quite often, what I'll do, particularly on an engine that I know is going to be used very, very hard, is when I've completed my tune on the dyno, and I've used a typical ramp rate on the dyno that I know is realistic for the sort of acceleration rates the car is likely to see on the road or track. |
51:16 | What you want to might do is, perform a very long run on the dyno, under full load, so maybe if I'm ramping at 500 RPM per second, I might perform one or two runs at maybe 200 or 250 RPM per second. |
51:31 | So this is going to give a much longer run, and it's going to build much more temperature in the combustion chamber. |
51:37 | And if we're still safe from knock occurring under those conditions, we can be relatively assured that when you're going to strike problems out in the real world. |
51:46 | Of course, if you are tuning for a customer that you know has an incredibly heavy foot, and you really are worried, what you can also do is, add some further safety precautions, with perhaps a road speed or a gear related trim for ignition timing or maybe boost, or maybe we want to add some more fuel to help bring that charge temperature down, that combustion temperature down as well. |
52:09 | So, there's a lot of things we can do there, hopefully at the end of this webinar you will understand the problem, and the implications of knock occurring, and I've given you some strategies that you can employ in your own tuning to help mitigate, or reduce the chances of knock occurring. |
52:27 | Let's move into some questions now, and we'll see what we've got. |
52:32 | We've got our first question that comes from Bathurst Bully, who said for many road tuners out there, like myself, that don't have access to a dyno, after finding MBT, using audible knock detection for the 98 tune, would it be okay to use the mile per hour trap speed from a 1/4 mile pass, to try and find MBT ignition advance for the E85, when trying to find MBT on flex fuel? So maybe adding a couple of degrees to see if there’s a plateau in mile per hour. |
53:01 | Yeah, look, that's, obviously if you are going to performing your tuning on the road, as we say in the courses, in our practical tuning courses, there are some compromises, and yes, that is a situation you might find yourself in, if you're on E85, in a lot of cases, you will be, you'll be knock free. |
53:22 | You're not going to be able to make the engine suffer from knock, so in those instances, well I haven't done it myself, yes, tuning at the drag strip and adding timing, and watching what will be the effect on the mile an hour, is probably as good an option as you've got. |
53:38 | There are some caveats around that though, the mile an hour is quite dramatically affected by your 60 foot time. |
53:45 | While it might seem counterintuitive, if you get a bad launch and a bad 60 foot time, this generally couples to a higher trap speed. |
53:53 | The reason for this is, the slower ET, the car actually has longer to accelerate on the drag strip. |
53:59 | So for your, your tests to work there, you're going to need to have a car that is consistent. |
54:07 | It's no use if you're seeing your mile an hour change by perhaps one or two mile an hour every pass, when you're not making any tuning changes, if you've got that sort of variation, you're going to be chasing your tail. |
54:18 | But if you've got a consistent car, then yes, that's a reasonably sensible approach. |
54:25 | Jenu has asked, I was told knock at peak torque is most deadly, and high RPM isn't that bad. Is that true? I wouldn't say that high RPM isn't that bad, I mean, my, recommendation is that we stay away from knock under all circumstances. |
54:44 | Now the aspect of peak torque though, is at peak torque, that is where we have maximum cylinder pressure, so in terms of aspects such as bending connecting rods, certainly if you've got knock occurring, you remember back to the graph that I showed, when we have knock occurring, we're seeing much higher levels of cylinder pressure, than what we see when we've got our timing at MBT, so at peak torque, this really sort of, magnifies the situation, if you like, and if we've got a mechanically fragile engine, we're much more likely to have the engine suffer from damage at that point. |
55:23 | However, because we end up seeing peak power being produced higher in the rev range, that's not to say that we can ignore high RPM knock, we definitely don't want that occurring, you are definitely going to damage something quickly there. |
55:39 | Jenu’s also asked, when an engine’s still stock, but making twice as much power as stock, for an example, an Evo engine, do we need to adjust the stock knock sensor sensitivity, or only when the engine hardware internals are upgraded? Okay, that's a pretty good question, and the problem I have with answering this is, my own experience on those sorts of engines, with stock ECUs is actually quite limited. |
56:09 | We sort of had a line in the sand, where once we were wanting to go significantly further than stock we would probably almost certainly be running on an aftermarket ECU. |
56:20 | But from my own experience with the knock control strategies, in aftermarket ECUs, what I can say is, that the thresholds for knock will depend on the specific power output. |
56:36 | Essentially if you can think about it, if the engine is making more power, generally, it is going to be, mechanically, a little bit noisier. |
56:44 | And the way we can see this as well, is a lot of people, when they are setting up, configuring and tuning closed loop knock control, think that the easiest way to set the threshold is just to simply pull all of the timing out of the engine, make sure that there's no timing in it, hence, there's no chance at all of the engine suffering from knock. |
57:04 | Perform a dyno run like that, and set the threshold safely above that, and start adding timing. |
57:11 | Even audibly, if you listen to an engine being run on the dyno, that has no timing in it, it sounds flat and horrible. |
57:18 | As soon as you start adding timing, the engine starts sounding sharper, and that noise is also being shown up in the background noise trace on the knock input. |
57:30 | So, often that will be the case. |
57:32 | The real answer here though is, even with an ECU that has closed loop knock control, or knock logging, I always prefer to back up what I'm seeing in the logger, with audio knock detection equipment. |
57:46 | I've had several cases where an ECU has falsely detected knock that definitely wasn't occurring, in this case, we'd need to effect the filtering of the knock sensors or the threshold, depending on the ECU and what control we've got. |
58:00 | I've also had several factory ECUs which were not picking up knock, obviously a much more dangerous scenario, not picking up knock that I could audibly hear. |
58:10 | So that's why I always recommend using audio knock detection equipment, even on a stock ECU. |
58:17 | Skyliner has asked, electronic knock detection versus det cans, what do you prefer? And will you ever hear knock, but not see it? Or see it, but not hear it. |
58:27 | So I think I probably just answered that question there, I really, really like running an aftermarket ECU, or a factory ECU with a quality knock detection strategy, this technology has come so far in the five or six years, and it really, when it is set up well, it really is amazing. |
58:48 | Now this also, you need to understand that, it is not band-aid for not doing our job properly in the first place. |
58:55 | We still need to correctly tune the ignition timing table, the knock control strategy is there for those odd occasions that we couldn't encounter on the dyno. |
59:04 | Maybe we get a bad batch of fuel from the boondocks somewhere, maybe the conditions the engine is running in, ambient conditions, just are so far outside of what we could possible see on the dyno, that this creates knock, we've got that knock control strategy there as a safety backstop. |
59:22 | However, that strategy is only as good as the configuration, that needs to be set up correctly in the first place, and this is where I use audio knock detection, Or as you're calling it det cans, to help me with that configuration. |
59:37 | Jenu’s asked whether E85 at 100, 105 octane, if we're turning a low, 300 wheel horsepower car, can we turn the whole knock sensor off, instead of adjusting, if the car will never use any other gas? Okay, let me just process this, and create an answer that I think is going to be sensible and safe. |
59:59 | The answer I'm going to give you is, it depends. |
01:00:03 | I know that might not seem very helpful, but I'm going to back this up with an example. |
01:00:08 | Under most circumstances, I would say, yes, you probably certainly could, but it does depend on the particular engine. |
01:00:16 | Now why I say this is because we have a Toyota 86, which is pretty much in that exactly scenario, it's heavily knock limited on pump gas, it runs 12.5 to one compression, and around nine or 10 PSI of boost pressure on E85. |
01:00:33 | So we're making somewhere in the region of 300 wheel horsepower, however, on E85, due to that very high compression ratio, it is still knock limited. |
01:00:44 | So under those circumstances, you can't say across the board, that we just turn the knock control strategy off. |
01:00:50 | What it's really going to come down to is you're assessing for yourself, if the engine is completely, completely knock free, there's no way we can create knock, then obviously the knock control strategy becomes a little bit irrelevant. |
01:01:05 | But, again, just can't say that in every situation. |
01:01:10 | RLP01, what's a safe method to verify true or false knock? Again, audio knock detection, you're going to be able to audibly hear, if there is actually knock. |
01:01:21 | And once you're familiar with using audio knock detection, and you've used it on a few engines, in most instances, there is very, very clear distinction, between knock and background engine noise, so this will help you verify whether the knock control strategy is being triggered by some background engine noise that just happens to coincide with the similar frequency tune knock. |
01:01:46 | Jenu has asked, is it true 115 octane will never have any knock under normal circumstances? Again, it's really impossible for me to make a blanket statement that says, yes, that's true, because there's so many aspects that can affect the reality there, I spent a lot of time before E85 was a common fuel, I spent a lot of time tuning high boost drag engines, where we were running on proper grades of race fuel, such as VPC16, and VPQ16, VP Import. |
01:02:22 | Those ranged to the 118 to 120 plus octane, and certainly with a lot of the engines that I was tuning, from Mitsubishi 4G63 in particular, 38 to 45 PSI of boost, the engines were not knock limited. |
01:02:39 | But that was a case where we're running 8.5, to 9:1 compression, certainly you couldn't say that, that would hold true if we were at 11, 11 1/2, 12 to one compression. |
01:02:49 | So again, you really need to verify for yourself, and often, with those sorts of engines, you're going to get to a situation where it's prudent to be a little bit conservative with the ignition timing any way. |
01:03:02 | What we know is, that if we look at the relationship between torque and ignition timing, as we start from a very retarded ignition timing, we see a very sharp increase in our engine torque, initially as we advance the timing. |
01:03:17 | As we start moving toward MBT, we're going to see that relationship start to taper off and plateau. |
01:03:24 | And if we want to get the absolute most out of the engine, obviously we can continue until we reach that peak. |
01:03:31 | But what we find is that, at the start of that plateau, where we first start to see the relationship between torque and timing start to flatten off. |
01:03:40 | Now we're probably at that point within maybe 5%, or maybe 6% of peak torque anyway, and it gives us a nice safety buffer, so if we're mechanically worried about the strength of the engine, then often we're going to have that safety margin, and we're not going to be really fighting for those last few horsepower, last few pound-feet of torque. |
01:04:02 | Anyway, it really comes down to the situation, what you are tuning. |
01:04:06 | Next question is, what about piston slap under 2,000 RPM, that triggers the knock sensor when you're moving off, this happens. |
01:04:16 | Yeah look, this is definitely a problem, and really the situation there is, if you've got audible piston slap, I'm not going to say this doesn't happen, because clearly it does, it's a problem with the build of the engine. |
01:04:31 | You shouldn't be having such excessive piston to bore clearance that, once the engine is up to operating temperature that the piston audibly rocking in the bore is something that you can hear. |
01:04:43 | So that's the first problem, it probably indicates that the engine maybe is a little bit worn, or the tolerances are a little bit looser than ideal. |
01:04:49 | And sometimes there's a limit to what we can work around, when we're audibly listening for knock though, we will be able to hear, again, if we are genuinely getting knock, or if we are finding and engine mechanical noise. |
01:05:04 | Now a good way to test that situation is, if it's under a low RPM situation like that, and we're hearing it consistently, what I'll try doing, initially, is pulling out quite a large amount of timing, maybe five to 10 degrees in that area. |
01:05:18 | And test it again, and see if I'm still getting that same noise, and if I am, I can be relatively confident that, that is actually a mechanical engine noise. |
01:05:27 | Another way to do this is go the opposite way, which actually requires a little bit more care, and physically advance the timing, and try and make the engine knock at that point, so you can then get that audible distinction for yourself, of what is knock and what is just background engine noise. |
01:05:44 | Now, how to deal with that, really depends on the situation. |
01:05:48 | The ECU you're tuning, in some instances, you may be able to change the filtering, or knock threshold. |
01:05:54 | In some instances, you may be able to affect the RPM range that the knock strategy or knock sensor is operating on, and simply exclude that area. |
01:06:03 | While we'd like to have it operating everywhere, chances are, you're probably not going to be doing damage down at 2,000 RPM anyway. |
01:06:13 | Next question comes from DCL, who's asked, I have a formula knock amplifier, and a Bosch knock sensor, any suggestions on how to best use in conjunction with the AEM Infinity knock detection within the ECU? Can I do a deliberate knock event at low power, verified by the formula, and correlate that with what the AEM hears? Yes, absolutely, that is exactly how I go about it, and we haven't actually done a webinar at this stage on the knock control in the Infinity, I don't believe. |
01:06:44 | However, if you want a general, or basic guide on the principles, or how I go about doing it, feel free to check out those webinars in the archive on both the Link G4 Plus, as well as the Haltech Elite, and the, MoTeC M1 ECU, about how to get a bit of an idea on my strategy for doing that, and how I use the audio knock detection to integrate that. |
01:07:09 | Dave has asked, in your years of tuning, have you found any fuel additives to help suppress knock? I'm not referring to stuff down at local gas stations, I'm referring to the likes of Torco, or race gas fuel additives? Look, I have tested this stuff, and I was never particularly happy with the results. |
01:07:29 | When I say this stuff, what I mean is, I have tested some octane boosters or additives for fuel. |
01:07:35 | Obviously I haven't tested across the board, and my testing, certainly I couldn't say it was to a laboratory sort of grades. |
01:07:42 | There are a few octane boosters or additives I have used, that have definitely given a small improvement in the engine's ability to take timing, and hence power. |
01:07:54 | My concern with any of these additives though, is if we perform a tune based on that additive being in the tank, in a certain ratio, or a certain mix, and then for whatever reason, the ratio or the additive, amount of additive in the tank changes, then this puts a question mark as to whether or not our tune is now safe and relevant. |
01:08:17 | So I'm always very cautious with that, I can't say that they don't work, but certainly you want to be very careful around their usage. |
01:08:26 | Years and years ago, I had a customer who ran a time attack car here in New Zealand, he wanted me to perform a tune on 50% C16 and 50% pump fuel. |
01:08:38 | I advised him strongly against this, I explained the risks involved, in terms of making sure that it always had exactly the correct mix of fuel. |
01:08:47 | Despite my warnings, he wanted me to do this, so I did that for him, and he, I think managed to get two weeks out of the car, and then blew the engine up, and melted all the pistons from detonation. |
01:09:01 | So it's a very real problem, and you do need to be very careful about it, particularly if you are tuning for customers as well, then you need to consider how that potential problem could affect your reputation. |
01:09:13 | Obviously, you can explain all the risks to the customer, they can say that they understand that, of course, when the engine blows up, because they haven't got that mix correct, you're still the one that's going to have your name dragged through the mud. |
01:09:26 | So you need to really consider that. |
01:09:29 | RLP has asked, what kind IAT targets are there, roughly, in a turbo application, as the air charge temperature increases the likelihood of knock? Okay, well there's not an intake air temperature target, as such, because you're post inter-cooler intake air temperature is understandably going to be very highly related to the ambient air temperature, what's going to happen is, your turbo charger is going to compress the air and heat it up, so it's going to heat it above abient, and then hopefully our inter-cooler is going to pull it pull it back down. |
01:10:02 | But the inter-cooler is never going to be able to pull it down below ambient. |
01:10:05 | And in fact, it's never going to get it quite down to ambient. |
01:10:09 | So for me, I'm generally going to rate an inter-cooler as being a pretty effective, and I'll be happy with everything, if we can get down to within maybe 10 or 15 degrees, Centigrade I'm talking here, of ambient, I'd call that a pretty good place to be. |
01:10:25 | And the situation here is that, obviously if you live in an area where we've got sub-zero ambient air temperatures, then you're going to be doing your tuning under those conditions. |
01:10:35 | If you're living in area where you're seeing 35 to 40 degrees Centigrade ambient air temperatures, then that's the area you're tuning under, so you're going to know that you're knock free in those circumstances. |
01:10:47 | A slightly tougher situation is where you get a large spread in temperature, perhaps sub-zero in winter, and perhaps into the 30s in summer, in those situations, the ideal is the check under both conditions, and see what you're actually getting, and ensure that your tune is safe. |
01:11:03 | Of course at colder temperatures, the engine is less prone to suffering from knock, but of course the air density is higher, so naturally what this may mean is that the engine would want less ignition timing anyway, so probably getting a little beyond the scope of today's webinar with that. |
01:11:27 | RLP has also asked what physical parameter to your knowledge, determines the resonant frequency of knock? So really, the key aspect here, as I mentioned, with that formula is the bore diameter. |
01:11:44 | That is the main driver behind what that base frequency of knock is, there are almost undoubtedly a million other aspects that would have a much more minor effect on the knock frequency, but I certainly, I couldn't give you any firm data behind what those would be. |
01:12:03 | Previl has asked, is the anti-knock additive MTBE, or Torco race concentration bad for the engine components? What's the orange residue that it leaves? So the Torco race concentration that you're talking about there, it's not something that I've personally had a huge amount of experience with, as I sort of mentioned before. |
01:12:25 | I try, for the most part, to stay away from additives, and work solely on a particular fuel, so if my customer really wants to go nuts, and make some serious power, then we'll be talking about doing a tune for specific fuel, either E85, an ethanol blend, or perhaps a proper race fuel, I have seen the orange residue with some of those additives, that you are talking about, predominantly you see that on the spark plug. |
01:12:53 | I don't honestly know what that is, I couldn't say if it is bad for the engine components, what I would say is, it's probably, in the short term, not a problem for engine components. |
01:13:06 | What I have seen with a lot of heavily leaded fuels, is they can actually be really damaging to engine components, if the engine is left for extended periods of time. |
01:13:17 | Now they don't tend to actually damage the components per se, but what I've found with the few engines that have sat, after a race, maybe running something like C16 race fuel, that haven't been started or run for an extended period of time, particularly in an area which is high humidity, what it tends to do is create rust on the bores, on the engine components, and the exhaust system. |
01:13:39 | So that's something to watch out for there. |
01:13:45 | Jenu has asked, when knock occurs during a ramp run, must we stop the run, or can we finish the run, and then adjust the map. |
01:13:53 | Okay, this is really going to come down to your experience and making a call at the time. |
01:14:00 | The safest option, understandably, is simply to abort the run, back out, and make the necessary changes. |
01:14:06 | So when I say it comes down to your experience in making your judgment call, often if I'm tuning a relatively low powered car, particularly one that I'm very familiar with the engine, and I know the engine is mechanically pretty robust. |
01:14:20 | If I've got a very small amount of light detonation, and it's momentary, so not sustained, if I just move through an RPM range, and I hear a couple of hits of detonation occur, often I will actually stay in that. |
01:14:34 | The difference is, if I'm tuning a very high powered car, I'll be aborting that run straightaway, or if that knock is sustained, so if it keeps on going, no, I'm going to abort the run. |
01:14:44 | Get out of it, make the changes. |
01:14:45 | Always, safest to do that, and then, repeat the run, rather than staying in it, and hoping that the knock will stop. |
01:14:54 | Magic Mike has asked, is there a relationship between fuel temp and knock? Look, that is something that I haven't actually experimented with myself, I've probably only been in a situation over the last couple of years where we've had a couple of cars which have fuel temperature sensors fitted onto them, and this has become more prevalent, as fuel composition sensors, which include fuel temperature, have become more common. |
01:15:23 | So I've never done any real heavy experimentation, and also, in order to do that, would also require a fuel system where you're getting quite large variations in the fuel temperature. |
01:15:34 | Not uncommon of course, in a race car, where it's passing a huge amount of fuel through the system and heating it up. |
01:15:41 | The up shot of all of this, is I would envisage that yes it is going to be a driving factor. |
01:15:47 | Anything that can help increase the combustion temperature, the combustion charge temperature, will have an effect on knock. |
01:15:56 | And certainly increasing the fuel temperature, could have that effect, so can't give you a firm answer. |
01:16:03 | Next question, if there's no option to log knock via the ECU, ie in a Haltech Sport, what should be done then? Simply rely on the sound you may hear, keep in mind, we're not experienced like you boys? Okay, I started my career on ECUs, that probably for the better of eight years, had no built in knock control strategy. |
01:16:26 | So it's certainly not an essential aspect to be able to log knock, it's nice, it certainly helps back up what you were hearing during a run, but it's certainly not an essential. |
01:16:39 | So predominantly, I am trusting my ears, I'm trusting what I hear. |
01:16:44 | Quite often what you'll have, depending on the ECU you're running, is the ability to add a mark into a log file, by perhaps pressing the Space Key or a key on the keyboard. |
01:16:55 | And what I'll often do, when I'm performing a run on the dyno, I'm not looking at my laptop screen, I'm generally watching the dyno screen, I'm looking at the air flow ratio and the boost, I'm audibly listening for knock, and I'll just hover my finger, my hand over the Space Bar on the keyboard, and if I hear something, like knock occur, or for that matter, if the engine just starts to do something I'm not happy with, I can just tap the Space Bar, and add a marker to the log file, so it will help me be very accurate on zoning in on exactly where that occurred, later on, before I'm making, when I'm making changes. |
01:17:30 | But I think you really sort of nailed it there with, you know, a lot of it does come down to experience, and you're going to need to simply do a lot of this. |
01:17:39 | You're going to need to test and listen to what engines sound like, you're going to have to find out what knock sounds like on your particular engine, and physically just get comfortable with the equipment, comfortable with what it all sounds like, and comfortable with how to get the most out of it. |
01:17:55 | And this isn't going to happen instantaneously, but you're going to have to just get started, so you can start building up that experience. |
01:18:06 | Bill has asked, does not sound the same using det cans, or head phones versus the bare ear. |
01:18:16 | No, I'd say it probably is safe to say, it doesn't quite sound the same, but, the situation is, what I talked about earlier, where on a loud performance engine, it's unlikely that you're going to be able to audibly hear knock occur with your bare ears, when the engine is running under load. |
01:18:40 | And certainly, if you can audibly hear it with your bare ears. |
01:18:43 | You're already probably doing damage, and the situation I've been in a few times, where we've got an engine that's got quite a quiet exhaust system, and you can audibly hear it knocking from the cabin, just with your bare ears, what you find is that if you then test with the audio knock detection equipment, what you're going to find is that, it was probably starting to suffer from light detonation, maybe two or three degrees of ignition timing earlier than when we start audibly hearing it occur. |
01:19:10 | So to more clearly answer your question though, when you're using audio knock detection equipment there is two aspects, you're going to hear the background engine noise, which sounds a little different to what you're hearing from the cabin. |
01:19:23 | Hopefully you heard that in the demonstration we performed, so there's that aspect to consider. |
01:19:29 | But you're going to hear the detonation or knock occurring, it will sound generally like a click, or a tick that's quite audibly different from that background engine noise. |
01:19:41 | The other aspect why this may differ is, if using a product such as the Plex Knock Monitors, has a lot of build in audio filters that will focus on a specific frequency. |
01:19:52 | And these audio filters will change both the background noise as well as the audio that you'll hear when knock occurs. |
01:20:01 | MCR has asked, what do like more for knock detection, Plex or Link? Look, I like them both, which might sound like a bit of a lame answer, but I really genuinely believe that they both serve two completely different markets. |
01:20:16 | Were talking about products that are of vastly different price points. |
01:20:19 | I can't remember off the top of my head here, what the actual specific prices are, and they’ll probably vary dependent on your market, but here in New Zealand, I believe that the Plex Knock Monitor version two, is somewhere in the region of double, if not more, the price of the Link KnockBlock. |
01:20:38 | Okay, so, on the one hand, the Plex Knock Monitor, definitely a much more sophisticated product, it's got a lot of advanced functionality. |
01:20:45 | However, that also takes a lot longer to set up and get the most out of, so if you want to use the Plex Knock Monitor and get the most out of it, you probably need to dedicate somewhere in the region of about 30 minutes or more, to actually getting that set up on your engine, and in particular, to get the most out of it, you really need to provide an ignition input to the Plex Knock Monitor, so that it can do RPM based knock thresholding and individual cylinder knock detection. |
01:21:13 | The Link G4 Plus KnockBlock, on the other hand, is much more straightforward, all you're getting is audio knock detection, one of the really nice features I like about the Link G4 Plus KnockBlock is how quickly you can install it on an engine and get it up and running. |
01:21:28 | The other nice feature with it aswell, is it has a built in battery, power supply, and that lasts for literally days. |
01:21:36 | So with Plex Knock Monitor, you do need to hook up power to it as well so, different products for very, very different purposes, both very good at their designed purpose. |
01:21:50 | Next question is, is there a correlation between knock frequency and knock sensor voltage output on the OEM knock sensor as read through the Haltech? Okay, so the important point to note here, is it's the frequency that's the important point, not the voltage, okay, now, I'll try and sort of explain this. |
01:22:11 | This probably would have been better if I actually had some graphs to look at, so what we're going to find is, if we look at a scope trace from the knock sensor, what we're going to find is that, as the RPM increases, the amplitude, or voltage, of the signal from the sensor is increasing. |
01:22:28 | So the voltage is relative to the size, or magnitude, of the vibration, but really, that's not the important point, what we need to do is, this is what the digital signal processor part does, is it analyzes the frequency that those amplitude, those waveforms are coming through at. |
01:22:48 | And that's what we're trying to look at, so they're very, two very, very different aspects, and I'm hoping that, that helps you understand that. |
01:22:56 | Looking purely at the voltage, really isn't that useful, because it's not doing anything to help with that background engine noise, and we're simply looking at a signal that will increase with knock, but it will also increase with just engine, RPM and engine speed as well. |
01:23:20 | Next question is, from my understanding, MBT occurs prior to knock, meaning, if I'm getting my six kilohertz on my knock sensor, I've already passed MBT, is there an amount of timing that I should dial back to be closer to MBT? Alright, James, so, let me try and get this straight. |
01:23:42 | First of all, MBT will occur prior to knock, on an engine that is not knock limited. |
01:23:51 | So this is a classic example, using our 350Z here, for our demonstration. |
01:23:54 | I mentioned that the point that I was testing 1,750 RPM, MBT timing at that point was somewhere in the region of about 16 or maybe 18 degrees. |
01:24:05 | I had to go all the way to 45 degrees of ignition advance, in order to get enough knock occurring that we could use it as a reasonable demonstration. |
01:24:15 | So under those conditions, an engine that is not knock limited, by it's very definition, yes, we will reach MBT before we reach knock, and in some instances, we may not be able to make the engine knock regardless how much timing we put into it. |
01:24:31 | Okay, we have got audio back again. |
01:24:33 | Alright, I'm not quite sure where, where you lost me there, so, I'll try and just back up a little bit. |
01:24:44 | So the point I was just getting across, is that, unfortunately, knock is not the only thing that is occurring at 6.03 kilohertz, and it is going to be some background noise from our engine also occurring at that frequency, and this is one of the reasons why it's important to have a threshold for our knock that increases with engine RPM, because that background noise is also increasing with engine RPM. |
01:25:13 | So yeah, it's not as simple as just, saying if we've got anything occurring at 6.03 kilohertz, that's knock, that's a problem, we need to back off our timing. |
01:25:23 | So I'll just again, reiterate as well, on an engine that is not knock limited, yes, we're getting to MBT timing before knock occurs, that's by it's very definition. |
01:25:34 | However, if the engine is knock limited, then MBT timing may occur beyond that knock threshold, so really, two very separate types of engine, whether the engine is knock limited or isn't, and very different strategies that we need to apply, when we're tuning them. |
01:25:52 | Alright, hopefully that was clear enough for you James, I apologize for the audio glitch there. |
01:25:59 | Alright guys, that's brought us to the end of our questions, and we got a huge number there, so hopefully, that's been of some value to everyone, hopefully it's given you some more insight into a topic that I know is still a little bit confusing to many, I know there a lot of misinformation out there, in the industry, and as I said right at the start, this for me, is the biggest killer of performance engines, so it's an essential element to make sure that you're monitoring, make sure that you properly understand, and definitely make sure that isn't occurring. |
01:26:31 | As always, if you've got further questions, please ask those in the forum, and I'd be happy to answer them there. |
01:26:37 | Thanks for joining us everyone, I really hope you enjoyed today's webinar, and look forward to seeing you all next week. |
![](/resources/themes/base/production/images/contact-popup-taz.jpg)