Never ending VR issues with 4AGZE

[mhiscott]

Well it's getting to the point where I've seemingly tried everything to make this work. My latest attempt at fixing my tach spikes has been to try a v3.0 with it's built in conditioning circuit, but I'll get to that later in the post.
My engine is an ae101 DLI 4agze and EVERYTHING aside from the VR signal (Using stock cas 24 tooth wheel with 2 opposite teeth ground off) is working fine. When using the v2.2 pcb and MT 2.25 with msns-E code 29q2, I had a hicup at ~2200 RPM. At this point the tach would shoot to around 4000 rpm. Aside from that, things were working fine. I was using an LM1815 to condition the NE signal and two vb921s to drive my coils. At this point, I was using the stock Toyota Cas wiring (sheilded) with the un-used pickups in the CAS gutted to cut down on interference. To make matters more strange, my after regrounding everything as diligently as possible, I decided to try messing with my dwell settings a bit. I ended up changing the dwell from 5 to 4.2 (MSD GM DIS replacement coils). Upon trying this, the engine worked flawlessly through all RPMs. I figured this had fixed the problem, but the next day, when I went to build a tach output circuit, after finishing up, the spikes were back. I undid everything I had done that day, and the spikes remained. I'm guessing what happened was something was moved slightly to the point that interference arose. I also made sure to ground my NE negative signal to the same point the LM1815 was grounded on the MS PCB. Still no luck.
I was frustrated at this point (trying to work out this issue for close on a month, checking everything from wire routing to alternator noise) and decided I'd try a v3.0 MS. Before doing this, I upgraded all my CAS wiring to an ultra shielded cable with the NE and NE - wires seperately shielded and the shielding grounded as before. I retained the same settings as I was using on the v2.2 and used the built in v3.0 conditioning circuit to condition the VR signal. I used the built in vb921 as well as a second vb921 to drive the coils. At first the engine would not start with the two VR pots turned all the way counterclockwise as suggested in the instructions. I saw no RPM signal at all. The engine did start, however, with both pots turned fully clockwise. It hardly stayed running however. That was last night and I decided to research the forum to see if anyone had similar experience. I found this thread:
http://www.msefi.com/viewtopic.php?t=11253
And used the mods described by lmr052 on the second page which involved removing the pots. After doing this, the engine started and idled fairly smoothly. However, after ~2.5k rpm, the tach was spiking terribly and was spiking downwards instead of up to a higher rpm as was the case with the v2.2. So right now I've tried both ECUs, upgraded the signal wiring, checked all grounds, pretty much all I can think of, but still no dice. If anyone could help with this at all I'd be very grateful. Sorry for the long message but I'm nearing the end of my wits with this problem.

Thanks for the input!
Mike

[lmr052]

Mike,
Can you get a look at the wave form that is being output by the conditioner? I suggest it may be a matter of a bit more tweaking of those resistor values or the pots. A PC oscilloscope can be a cheap choice to capture the waveform.

I suspect it is a matter of the conditioner missing triggers on some of the teeth... or not sensing the ground off teeth. Looking at the waveform at various rpm levels will tell you what is happenning.

I have found that spinning up the CAS (or distributor in my case) with a drill allows you to see what the conditioner outputs and how MS reacts, especially at the magic rpm where the miss occurs.

Good luck