• Hello Guest, welcome to the initial stages of our new platform!
    You can find some additional information about where we are in the process of migrating the board and setting up our new software here

    Thank you for being a part of our community!

60-2 VR Signal woes - At my wits end here

Alex Buchka

Alex Buchka

8 cylinders
Joined
Apr 8, 2004
Location
Los Angeles
I know this is long winded but please bear with me here.

Quick run down of the specs:

b5234t 2.3l DOHC inline-5,
MSnS-Extra 029v on a V3 pcb,
dual trigger setup with 60-2 wheel + VR on the crank and 1 tooth wheel + hall effect sensor on the cam.

I set it up to drive 5 S80 T6 COP ignition coils with built in ignition modules that fire when pulled to +5v.

I've been having nothing but trouble with the rpm signal from the 60-2 wheel. The tooth logger in megatunix has been showing some very strange trigger events. The norm seems to be I crank the engine over and I see what appears to be the missing tooth, then 4-10 regular teeth, one more long tooth time almost exactly the same as the first one, 4-10 regular teeth, then one more long tooth, then anything from 30-80 short ones, then the insanity starts all over again.

I've reflashed the MS probably 5 times just to be sure I got it right, I have also gone over the settings for the wheel decoder 10 times and I'm very sure I have trigger times and such set right. Sometimes during cranking a it will read a clean missing tooth and start firing the plugs in the exact right order (verified by someone standing next to the engine bay looking at exposed spark plugs grounded on the engine), then the goofy sync behavior starts and the spark will drop out or behave erratically.

The VR and hall sensors are stock, in a completely stock configuration exactly as Volvo arranged them from the factory. I have triple and quadruple checked all the wiring to the sensors and I'm positive it's correct, I've also swapped all the sensors for other known good ones with no differing results. I've verified the function of the additional opto-isolator circuit for the second trigger and it delivers a perfect squarewave 0-5v signal to JS8 on the V3 pcb.

I wired up 6 LED's on the case of the MS; one connected to the output of the opto-isolator so I can see when the tooth is under the sensor and 5 more connected to the ignition outputs arranged in firing order so i can clearly see which outputs are working.

At first I thought it was cross talk between the wiring so I ran a new shielded wire from the VR to the MS outside the car, as far away from the other wires as possible, this made no difference.

I borrowed an old tektronix analogue scope from Adam (AHox244+t) to see what the signal looked like. I had never used a scope before but after some stumbling around I got the hang of it. The raw VR signal from the sensor looks proper and I can clearly see the missing tooth in the wave form with no apparent noise. I also scoped the signal coming out of the VR conditioner on the pcb and it looks very good, once again the missing tooth is clearly displayed with none of the strange behavior noted in the tooth logger.

I attempted to fiddle with the trim pots to see if it would make a difference. Chris Wiita (The poi) is running MS on his 960 with an identical VR/wheel arrangement and he mentioned leaving the pots turned fully CCW gave best results. After looking at the signal on the scope I agree with this. Turning the lower pot more than 3 turns made the signal vanish completely and turning the upper pot (closest to the FET's) had no noticeable effect.

I've verified all the components on the pcb are correct according to the BOM. I'm really at my wits end here. Any advice is appreciated.
 
Sounds like a pain in the butt. What does the tooth logger show in Megatunix when the cam tooth goes under the sensor? Does it superimpose on the crank trigger signal somehow?

I have come to the conclusion that the VR sensor circuit is somehow lacking in Megasquirt. I see too many posts like this. Its just really finicky for some reason. I don't really have any suggestions for you that you haven't tried yet. But, I'm currently in the process of replacing my crank sensor with a Hall type unit. I'm hoping this will give a better signal. Not that the source is the problem, but the circuitry that works with the hall sensor seems to just work. Maybe because the lm1815 conditioner is not used. I dunno. I'm kinda playing trial and error here. I should have some results in the next day or two I hope.
 
I've figured out that Megatunix doesn't display the second trigger at all. That's why I added the LED so I can verify 100%. I've entertained the idea of using a Hall sensor instead of the VR, I agree with you 100% there. It seems like a VR with MS is hit or miss. I anxiously await your results with the Hall sensor.
 
Alex, can you pull the coils and put plugs in them to test the coils firing as the engine cranks?

Did you build a circuit for the hall input?

Where is your vr shield grounded?

Mike
 
I've always had great luck with MS VR... but never with a Volvo sensor. :-P

Check Tout- Vr inverted.
Try deceasing the air-gap from the sensor to the flywheel.

Besides that, I can't help you much.
 
What a pain. As awesome as megasquirt sequential coil-on-plug is, everyone who's done it (Mike, myself and Alex) have all had like a weeks worth of problems getting it set up!

Just for an updated record, my setup (36-1 wheel with ford VR sensor, lh2.2 hall sensor, home built conditioner circuits) has been working perfectly and without a hitch for the past few hundred miles. 5 cylinders shouldn't be any different.

Alex it's messed up that a few days ago you were getting a perfect VR signal and rpm signal, and now it all just went to hell. Maybe your LM1815 bit the dust? Post your trigger wheel settings just so that we can compare what you've come up with. You got the dwell figured out right? Actually just post up your whole .msq if you can, or e-mail it to me. I'd like to take a look through it in MegaTune, maybe you're missing something. I dunno, worth a try. Maybe I could send you my whole file, you can change the outputs and wheel decoder settings, then see if it'll work for ya?

Keep at it buddy!

John
 
where are you grounding your shield wire, and the rest of your grounds for that matter? are your grounds related to analogue devices separated from those associated with the digital devices? and are your grounds to coils seeing a ground somewhere on the motor or are the wires going back to the ms box - i have no idea how you set those coils up?
i used the vb921 drivers and put those grounds to the chassis, rather than directly to the motor, with the hope of isolating them somewhat.



when you "see" the missing hole, is it a pull down towards ground, or up? i have not gotten this far with mine yet, just hooked everything up in the most sensible way and now i'm hoping for the best :).


my purpose in asking is that it's possible that this is a ground-loop issue.


http://www.epanorama.net/documents/groundloop/in_circuit.html


check to see where you have everything grounded, and look for items grounded to the same spot from different paths, you have a multimeter right? the coil grounds are paricularly worth isolating first though, you mentioned that you saw an acceptable signal at the conditioning circuit, did you see this when the coils were firing?
 
Last edited:
FYI on my car everything MS related is all grounded to a grounding block on my ms cutting board, which has a 12ga wire directly to the battery. That means, MS, coils, injectors, sensors, everything. No problems yet! Although I have heard of this issue before, so maybe different setups are more susceptible to ground-loop issues.
 
so as you describe it the analogue device grounds don't go through the digital device grounds before reaching the ground block or plate or whatever ... so your setup would not apply to my concerns, since it sounds like you are grounding everything in parallel ... not in series mixing digi and analog.
know what i mean :)

as for the coil negatives i just wanted those as far away as i could get them, since i think they ought to see some real current spikes, no?
 
My setup is like John's, everything goes to one big grounding block by the MS.

Here's a pic:
<a href="http://picasaweb.google.com/kbuchka/1979Volvo242/photo#5159308878775024338"><img src="http://lh4.google.com/kbuchka/R5mMFw3netI/AAAAAAAAHKY/X1GI3RNuQDw/s800/DSC01094.JPG" /></a>

The leftmost bolt has a B21 battery cable running from it straight to the engine block. From that point on the block there's also a grounding strap going to the chassis. On the other side of the engine bay we have another battery cable connecting the chassis, engine block, and negative battery pole.

Justin, I really don't see how mixing 'digital' and 'analog' grounds make any difference. What does that make the megasquirt? It's got plenty of ICs mixed in with discrete components.
 
mixing digital and analogue grounds causes problems sometimes heard as noise in a line because the digital signal occurs at an edge, while the analogue occurs progressively. those edges cause the noise.

read the article i linked to, its a rather common problem in audio systems.

it may not be a problem here of course, but since you have that shielding grounded somewhere on a system that sees spikes to ground, in particular those coils firing, then i don't see how one could ignore it, the vr signal is being processed so i don't see how hard it is to imagine that a shielding cable that sees some activity at ground couldn't effect the rather quiet vr signal enough that the conditioned signal shows incorrect "ons" or "offs", know what i mean?

the buzz that you hear in the stereo is often caused by shielding cable seeing some oscillating voltage noise rather than a perfect ground (remember there is resistance in these cables).

think of how feedback systems work, in taking continuous measurements on the quantum level ... some analogy to that is what i had in mind here (we talked a little about it a couple of weeks ago)
 
Also, which he hasn't posted here yet: he's scoped the input pin on the chip, and its a perfect 58 tooth/gap that it should be. I have a feeling theres some weird code issue. On the hardware front: he's running exactly what I had on MS1/e and my **** runned. Actually...

edit:// nevermind, that jackass isn't on AIM. Alex: what's your input connected to, VRin to inverting or noninverting? Because my **** worked on MS/1 with VROutInv, and whatever settings in the code that worked. Yet MS2 simply didn't work unless I switched it to VROut, and toggled the software setting. I vaguely remember the explanation of this is that the conditioner pots simply don't work on VROutInv properly. However, you've been having this **** for so long I don't remember if we even worked on this or not. Christ.
 
Don't feel bad, most VR sensor input conditioners are lacking. I work as a controls engineer these days developing controllers for ABS systems and IVTs, and even we have plenty of issues with them.


First off:
Determining if it is a noise issue or not is painfully easy. Unplug all of the injectors and coils and crank it.

If it is a noise issue, the signal should be interpreted absolutely perfectly at this point because there is no current flowing to even create noise. This will determine once and for all if it is an external issue or an internal issue.

Try that and report back. You can fix this.

Honestly the easiest way to fix it is to probably adapt a hall sensor for the crnak sensor as well. :grrr:
 
Ok, here's the latest:

The MS doesn't detect the second trigger (hall sensor) when it's flashed with 029v, but all the coils fire. If we flash it with 029y4 it recognizes the hall input, but the spark outputs act really wonky.

Alex is inside re-soldering a wire that hopped off the opto-isolator. As soon as he's done with that we're testing the alpha firmwares that came in between the two we've tested.

EDIT: Also, the tooth logger in Megatunix sucks beyond belief. Don't trust it when the RPM is anywhere below 1000.
 
We're in the process of trying every firmware we can download. So far the pattern seems to be that anything past 029v will run the hall sensor perfectly, but won't fire the coils properly.
 
You must log in or register to reply here.
Back
Top