speeder Posted June 30, 2007 Share Posted June 30, 2007 On a new Megasquirt (V3.0, MSNS, Version 29V, low impedence injectors, PWM installed on the board, no dropping resistors) install on an L28ET, the engine started right up, sounded good, although too rich. I was busy tuning to get good idle/free-rev and suddenly the engine died. Everything checks out - good tach signal, injector common volts at the injector plugs, scoped 3Msec pulses on the other side of the injectors during cranking, but no fuel being injected. Going back through my calibration I find that, although I have set the PWM current limit to 30%, I have neglected to reduce the PWM time threshold from 25.5 Msec. I disconnected the ECU and, with the injector common active, momentarily grounded the other pin of each injector, and they all opened and injected fuel. Setting the PWM threshold to 1Msec and rechecking the scoped injector driver outputs with injectors connected, I still have good looking injector driver pulses, but see no PWM action on the waveshape. I swapped in my identically configured spare ECU and saw the same symptoms. I removed my wiring harness and verified all connections solid - They are. All injectors read the correct 3 Ohms resistance across their coils. That just leaves the injectors being the problem, and I have ordered some more. But I'm still puzzled at the indications. Back to the question: By missetting the PWM, have I damaged the injectors in such a way that they work with the old voltage test but do not when pulsed by the ECU? I'd also like to know what the injector driver signals should look like when viewed with a conventional oscilloscope rather than Z-ya's excellent current representation. Thanks for any input... Quote Link to comment Share on other sites More sharing options...
z-ya Posted July 1, 2007 Share Posted July 1, 2007 Rick, Yes you can damage the injectors by setting the time threshold too high without limiting resistors. So if you had it set to 25ms, and the injector opening time is set to 12ms, the injector is allowed to draw 100% current for the entire opening time (12ms). Hear is an example of the time threshold being too high: So essentially you are allowing the injector to draw max current after the injector has already fully opened. The MS has overload protection to protect itself, but nothing to protect the injectors if the PWM settings are incorrect. It sounds like you have made all the proper checks, but still no fuel. So unfortunately, it sounds like the injectors. The voltage waveform on the - side of the injector should show 12V when the injector is closed, and zero volts when it is open. I'll try to probe one this week as an example. Quote Link to comment Share on other sites More sharing options...
speeder Posted July 1, 2007 Author Share Posted July 1, 2007 Thanks for that, Pete- I have never paid much attention to injector driver waveshapes before, and now wish I had. Before now, I had never considered the finer points of working with a pulsed solenoid valve (injector) concerning driver setup. Rather more complex requirements (Than just a square wave pulling to ground) to deal with the inductive "kick" from the coil modified by the added inductive artifacts from the saturable mass of the armature. Add to this having to look after current limiting. Looking at my - side terminals with the scope again and playing with scope triggering, I see an initial positive spike of almost 3 times the supply voltage for about 1 Msec before the signal goes negative to ground, where it is a good clean square wave. Pretty sure the big positive spike isn't normal, and must be a sign of something screwed with the mechanics of the injectors. I have just never considered an injector failure other than burning out the coil! I may try to get some scoping done on another car to compare, but your example would be appreciated- I don't have another Megasquirted vehicle together at this time. After all this, I am making plans for a test bench. -Rick Quote Link to comment Share on other sites More sharing options...
z-ya Posted July 3, 2007 Share Posted July 3, 2007 OK, hear are some screen shots from my scope. On this one, the top trace is voltage on the injector driver output, and the bottom trace is current. The current limit is set to 30%, and the time is set to 1.3ms. These are the correct settings for most low Z injectors. You can see on the voltage trace that the driver drives the output to ground for the injector peak current period, and then pulse width modulates (PWM) the output during the hold period. Hear is a zoom of the voltage trace showing the PWM. If the time threshold is too long, the injector will never go into PWM mode. The injector driver will essentially ground the driver side of the injector, allowing it to draw max current for the entire opening time. Because there isn't a current limiting resistor, all of the power is dissipated in the injector coil, which will cause it to overheat and fail. If you are using small injectors, the opening time can be long, increasing the probability of injector damage. Quote Link to comment Share on other sites More sharing options...
speeder Posted July 3, 2007 Author Share Posted July 3, 2007 New injectors are in, the car runs, and my injector driver waveshapes look just like your voltage representation. With the old injectors the positive voltage spike was about 2X higher, and no PWM action was evident. Thanks Pete! This whole sorry episode has shown that injectors do indeed have a failure mode that does not involve burning out the coil. Ya learn something every day. Quote Link to comment Share on other sites More sharing options...
Administrators BRAAP Posted July 4, 2007 Administrators Share Posted July 4, 2007 OK, hear are some screen shots from my scope. On this one, the top trace is voltage on the injector driver output, and the bottom trace is current. The current limit is set to 30%, and the time is set to 1.3ms. These are the correct settings for most low Z injectors. You can see on the voltage trace that the driver drives the output to ground for the injector peak current period, and then pulse width modulates (PWM) the output during the hold period. ........ If the time threshold is too long, the injector will never go into PWM mode. The injector driver will essentially ground the driver side of the injector, allowing it to draw max current for the entire opening time. Because there isn't a current limiting resistor, all of the power is dissipated in the injector coil, which will cause it to overheat and fail. If you are using small injectors, the opening time can be long, increasing the probability of injector damage. Great stuff Pete. The scope of the injector and the PWM explains a LOT about low impedance injectors. This shot would be a nice addition to your sticky… Rick, Glad to hear you got your car back up and running and the PWM set correctly. Quote Link to comment Share on other sites More sharing options...
speeder Posted July 4, 2007 Author Share Posted July 4, 2007 Pete, I'm envious of your access to that kickass scope! Do you own it? Quote Link to comment Share on other sites More sharing options...
z-ya Posted July 4, 2007 Share Posted July 4, 2007 I'm a Sales Engineer for LeCroy, so I've got quite the selection at my disposal. Using a $20,000 scope debugging Megasquirt problems certainly is overkill Quote Link to comment Share on other sites More sharing options...
Zmanco Posted July 5, 2007 Share Posted July 5, 2007 Pete, I know there are scopes that utilize a PC and its USB port for the inputs. What do you think of these as a low-end solution for automotive work? Without expensive probes I know they wouldn't show the kind of detail you have, but what about for general trouble shooting? There are so many times I wish I could see what kind of noise was on the input and/or output of a device such as MS. If there was a reasonably priced one (say $200), it might be a good addition to our tool box. After all, if you're running with MS, you probably already have a laptop. Quote Link to comment Share on other sites More sharing options...
hoohaa Posted July 6, 2007 Share Posted July 6, 2007 I'm eyeing this thread with interest to see what's mentioned about o'scopes. I would love to buy one and learn to use it, but if it's going to be too expensive to find one that actually does what we need it to then I'll have to pass. Quote Link to comment Share on other sites More sharing options...
speeder Posted July 6, 2007 Author Share Posted July 6, 2007 I use an ebay special Navy surplus Tektronix 465. I think I paid 100.00 for it. More than adequate for injection work. Probes are cheap on the surplus market, too. Quote Link to comment Share on other sites More sharing options...
z-ya Posted July 6, 2007 Share Posted July 6, 2007 I use an ebay special Navy surplus Tektronix 465. I think I paid 100.00 for it. More than adequate for injection work. Probes are cheap on the surplus market, too. That is an old workhorse of a analog scope. Great scope, especially for $100. You should be able to find a current probe that will work with it. A 0 - 10A range, with a 1:1 ratio (1A = 1V) probe will work fine. You can also use it for setting the proper coil charge time (dwell). All of the scopes made today are digital. Once you get your waveform in the digital domain, the sky is the limit as to what you can do with it. A 100MHz scope is fine for most all automotive work, unless you are debugging an ECU. Real time scopes range in bandwidth from 100MHz to 18 GHz today. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.