The car runs much better now with the over run fuel cut off enabled. NO resets under engine braking conditions. It behaves wonderfully under all light to medium throttle conditions. However:

The engine breaks down under boost conditions. This is big time failure. Engine sounds just like it is in over-boost or over rev-protection. No resets happen. Lift the foot and instant return to sanity. It is currently operating with 50% filtering enabled. It runs better with more, and worse with less. I sent a datalog to Peter and he says clearly that the VR signal to the HEI module is not clean. So how to clean it up? It seems like this should not be rocket science. But I have too many questions to just jump in with a fix yet.

The most obvious thing to look at is the connection between the VR feed and the HEI module. This is a poor match. It is a not real tight/best-I-could-do-at-the-time kind of connection. Can I solder to the connecting blades on this HEI module to put a better connector in there? Or willthe heat destroy something vital in there?

Also, how do I know if the VR wires are reversed? Peter said it worked fine on his bench, but I know VR feeds WILL run if wired backwards, but not well. In this case it reads the wrong slope on the sine wave. One is sharp and straight and the other is less clearly defined.

Here is a list of questions about shielding the VR feed:


1. Is the wire inside an audio cable big enough to carry the current? Peter says yes, but second opinions are often useful. I have no idea what kind of a load is placed on this wire. For some reason Mopar chose to use pretty heavy wire. They come with RCA jacks on them, but I can cut them off.

2. And are they shielded enough to block EMI interference from the engine compartment. Apparently the best is twisted and shielded.

3. Would it work to run it outside the fender and then back in through the fender right at the HEI module? That is, Would the fender serve to shield it from noise.

4. Or if I build a special aluminum conduit of sorts to run the cable in, would that shield it?

5. If shielded audio cable will carry the current, I could also run that outside the fender and back in at the hei module, or in a special "conduit".

6. What kind of noise would be most intrusive only at boost levels? It seems perhaps a fuel induced miss can create more intense noise, based on the coast down failures we were seeing before over run was enabled.

CAn someone point me towards a twisted, shielded cable I can simply buy in bulk in consumer quantities? Are there OEM applications I could just buy from either a junk yard or a dealer? Thanks. again, as always.

Any thoughts are appreciated. I can send the data log to anyone who will PM me their e-mail address.


Sam

_________________

_________________
一期一会
Too many people who were born on third base actually believe they've hit a triple.

Sam, a couple of quick things. You are correct about the VR polarity. It wants to be one way or the other. By that I mean, the MS is going to use either the rising edge, or the falling edge. The only way to 'know' is to use an oscilloscope to trace the waveform. However, it's really easy to check: just reverse the wires, and try it. You won't hurt anything.

Also, this is a setting in TunerStudio. Under your ignition setup, you can change whether MS is using the rising edge, or falling edge of the signal. Instead of swapping wires, you can try changing this setting (don't do both - wires + setting!) to see if it helps. You will know instantly if you made a change for the better.

I have had this problem several times. I always guess, and 100% of the time I have been wrong on how I set the polarity. I do a quick swap, and all is well. I finally got smart on the Duster, and checked continuity to ground and found the correct wire the first time, for the first time. It was simple but I never knew it before - one of those wires from the VR goes straight to ground. That may be a place to check for a poor connection too - check that ground. Currently, my VR setup has one wire feeding the MS and the other just running to a frame ground.

About the current through the wire, the variable reluctor can create quite a high voltage and current depending on RPM. Higher RPM = higher voltage. I think, but I don't know, that this current is limited by resistance in the MS. My wiring is not particularly heavy, about the same as the rest of the MS sensor wiring, and I have not had a problem with it.

Other thoughts: reversed VR polarity should not, in and of itself, cause processor resets in the MS. It will happily use the incorrect slope, it just won't be timed right.

For reference's sake, my VR setup is the stock pickup inside the distributor, two wires not twisted, with a couple of butt connectors joining them. They route straight up from the distributor, to the passenger fender, where one is grounded and the other joins the MS harness. I took some care to not have them physically touching the spark plug or coil wires, but that's all I did.

I did notice that the MS was sensitive to the air gap in the VR pickup. I usually set it by 'eye' and have it work, but that was a no-go for the MS. I had to use the .008 brass feeler, and I got the feeling that it may have liked an even smaller gap.

As for #6 above, I don't see how boost would create or increase noise. I can imagine this scenario: if you do have the VR polarity reversed, then essentially your timing is being affected. I assume that under boost, your timing tables are pulling timing out. I could see this bringing the timing to a point where it doesn't want to run anymore. When boost disappears, timing comes back in, and masks the problem. Just a guess. You could test by changing the timing tables around temporarily to give you the same retard you have under boost, at some other manifold pressure. Might be quicker just to swap those VR wires for a test though


Hope some of this helps,

-GH

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

You've also probably already seen this, but this is a pretty good (great) treatise on signal noise by B&G:

http://www.bgsoflex.com/megameet2008/me ... gnding.pdf

Page 42 shows bringing the VR ground directly back to the MS circiut board to C31 to cope with resets caused by the VR... that's a v3 board though I think.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

Thanks for the thoughts.I will ponder them all. One thing I learned from Lou,: change just one thing at a time.

I found a better connector and put it in the circuit. It seems to run better, but rain is threatening once again, and I will put off testing.

I will read the link you provided also.

George, do you have any filtering enabled? If so, which one, and how much.

Sam

_________________

Sam you must have the MSExtra firmware installed? I'm using the straight MS II firmware, and it does not have software noise filtering. MSExtra does, I think.

Or are you referring to the ignition prediction settings? I'm using Alpha-Beta-Gamma filter with respective values of 90%, 80%, 10% which were the defaults.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

I am suspicious that this only is a problem under boost. No reason the noise on the VR wires should be higher in that range. It could be that higher RPM (not boost) is the issue?

My guess would be that fuel is still the problem under boost, OR it is just that the VR wires are swapped. I would swap wires first before doing anything else.

Always look for the simple problems/solutions first.

Lou

_________________
Home of Slant6-powered fun machines since 1988

Lou you might be right. Your thinking parallels mine. Here is the latest. I test drove the car with the new connector after the rain, and things are improved. It starts better, idles smoother, and behaves nicely in most normal driving. But I can still induce fatal failure, but now in different cells. Here is the skinny:

I can now get up to 5 lbs of boost in 4th gear before it breaks down, somewhere around 2 K.

It exhibits the same fatal miss at 70KPA and 3 K rpm in 3rd gear. This is the first time I have had the engine that high in the rpm range. It is also the first time I have gotten that high in the fuel map at low rpms. So I am now into new parts of the VE table. I did not take my laptop with me, and did not even pay attention to the AF ratio gauge. It was a quick run down the street and back.

I have run out of time for a few days, and will not be able to do anymore tuning for awhile. The autotune is fabulous, but when the fuel is way, way off, you must kind of sneak up on the cells that are way off so they do not kill things. So far I have driven about an hour with it in autotune, and the standard cruise and driving range of the map is getting pretty good. AT those times I was able to approach the boost ranges of the map carefully in high gear and get those cells to tune without going into failure misfire. I may try this in third gear as well when I get time. You can see the numbers on the AF gauge come down as you cruise in a cell that is too rich. Pretty cool.

GEorge, my ECU is MS II extra, 3.57. And it does have an entire screen devoted to filtering. I do not understand the different types. I am using what Peter selected and have it set to 50%. He wants it to be able to run with no filtering on, which would require a real strong VR cable shield. I should look at the filter screen and see if I can see any correlation between your filtering and mine.

Does anyone think reversing the VR wires would damage anything in the HEI? Just a dumb, paranoid question. Here is another maybe silly question. For those who have had the VR wires reversed, and then corrected them, what behavior change to you experience at idle and cruise? Was it obviously better or worse, and how? Did the ignition module overheat when wrong?

Sam

_________________

I don't believe you will damage anything in the HEI module. I ran just such a setup in my circle track car. (Chevy Nova) I used the VR pickup in the distributor to trigger a Mallory ignition. I had it wrong the first time (of several), reversed the wires, and improvement was instant.

I have had one vehicle (318 Ramcharger) run pretty well with them reversed, with the odd miss in the midrange. Swapped them and it was fixed.

I opened up the tune for the kid's Neon which is running MSExtra. I see the filter page. That one is set up for what I think are the defaults, which is Filtering ON and tach period rejection OFF which grays out the percentage fields. So I have no experience with the filtering.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

Sam's datalog:



Looking at your datalog, you are without a doubt getting spikes on your crank signal. The place I have the log stopped at is where your RPM signal went from about 1900 RPM to over 6200 RPM in an instant. Obviously this isn't possible. The 107.1 KPA indicates a few PSI of boost. There is one significant spike from 1700 to over 3000 a few seconds earlier when you are at about 98 KPA or just on the edge of making boost. The RPM signal jumps up and down like crazy there (I doubt actual RPM was changing much).

Your throttle opening never gets over 41% so you were less than half throttle (assuming a calibrated TPS)

Your Air-fuel ratio went very rich (10.6:1) when this was happening, probably due to misfires.

Everything else looks pretty normal, considering what was going on. The AFR looks like the tables could stand a little smoothing but I would not even begin to try to tune it until the crank signal problem is sorted. CLT and MAT look jaggy but that's just a variation of 2 degrees for MAT and 20 for coolant, from start to finish.

Below is a sample datalog from my engine during a 1/4 mile pass for comparison. You can see the sawtooth RPM as the auto trans shifts. The teeny weeny jagged edges are variations of 1-2 RPM in the signal.



Grounds - I'm thinking grounds. Do try the VR swap, but looking at the RPM signal, I begin now to doubt that's the problem. Eliminate it as a possibility, and we'll move on.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

Couple of screen captures from the tuning file:



Looks like the noise filter is set to off, although there are settings for tach period rejection. You might try setting the filter to on.



Ignition trigger setup. This is totally different than mine. I'm assuming Peter set this up right for what you have. You are using the 7-pin HEI and the MSD ignition still, right?



Rev limiter setup. Interesting thing here is when the MS sees that RPM spike over 5000RPM, as it did in the datalog above, it will cut fuel and spark. This is probably why it's misfiring when the RPM signal is noisy. You may want to set this to 'none' for now until the signal noise issue is resolved. Also, you may want to change the type to "Fuel cut and spark retard" instead.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

Thanks George. Peter says it goes rich BECAUSE the fuel formula thinks the engine is running at 6200 RPM and delivers this fuel. You can watch the marker bounce up and down in the VE table during these events. That looks like a very poor signal to me when compared to your log. I am suspecting now something not set right in the dizzy. Maybe I should set the air gap in the reluctor.

I will do another data log in a few days and send it to you. WE can see if the new connector made any general improvements. I will also swap the VE wires and do a data log of that as well. The next few days are booked with other stuff.



Sam

_________________

Thinking about your setup, that VR input is grounded through the body of that HEI module. I would double check that ground for sure, from the connector where it plugs into the dist.

_________________
It's a Slant thing. Even I don't understand.

1974 Duster, EFI /6 soon to be turbo...

Get that Monkay! Get that nasty thing!!

Why are you using an HEI? It was my understanding that version of MSII could be driven directly from the lean burn distributor.

Maybe Matt can verify.

MS2 can take the distributor input and drive coil directly with the right base board but noise becomes even more of an issue. The HEI module (or msd box, etc with tach output) acts as a buffer.

Sam, RCA wire would be very tricky to use. Current is miniscule so gauge isn't a problem but RCA wire is thin enough to be difficult to work with and strip. Don't mickymouse a conduit. Shielded speaker wire will work well here - thisis what it looks like. Try a hardware store in the house wiring section or ratshack. 18 or 20awg if you can find it will be suitable. 22 or 24 will be just fine as well but again that gets difficult to work with.

Peter is correct, if megasquirt gets a 6200rpm input because of a noise spike, it will apply 6200rpm's worth of fuel according to the map. MS doesn't know the difference between noise and true rpm on a digital level, it relies on the hardware to filter it out.

Scrutinize your sparkplug wires. Rev it up in the driveway while its dark to see if you notice spark jumping. Mix a few drops of dishsoap with water in a spray bottle and mist it on the plug wires while revving it in the dark - this will allow you to see any sparks jumping out of the plug wires. A plug wire may only leak under higher rpm.