HEI 8 pin hot-start problem

[hytekauto]

HEI 8 problems

Here's the setup:

1968 Dodge Super Bee, 426 Hemi, Cross Ram intake w/ 2 Holley 670 cfm throttle bodies, MegaSquirt 2/3 firmware v 2.35 with Relay Module, GM HEI 8 pin ignition module, input to HEI module from locked stock Chrysler magnetic pickup distributor. The wiring to MS is per the book: Terminal R to "tach in", terminal E from pin 36 (via Fast idle terminal) from JS10, start control via mod using Knock Enable (RPM > 300 goes high), then to proto area using a 2n2222 to provide +5 when above 300 RPM, this is wired to the HEI module B via DB37 pin 5 properly ungrounded and wired. The Tach In is the typical Opto with diodes jumped out, 560 ohm R12, C30 removed, and XG1 to XG2. This all works as advertised.

Here's the problem:

The engine fires right up, idles decently at <1000 rpm, timing is handed off to MS (as verified with a timing light). All is well for about 2-3 minutes, then little by little the timing control cuts out intermittantly, as indicated by the tachometer and the ignition LED on MS. This gets progressively worse with the cutouts getting longer and more frequent until the engine dies. Let it sit for a minute or so and the process can be duplicated. However, if you hold the throttle a bit open to keep the rpms >1500 or so, this problem doesn't occur?????? Nothing gets hot, just nicely warm. The injectors are following the ignition and cut out when the ignition does, the engine usually catches when the RPMs drop below the 300 rpm level indicating that the internal startup part of the HEI mod is working properly. I've switched HEI modules, replaced the coil and reluctor in the distributor, tried two different Mega 2 modules, tried several different combinations of parts in the Opto area, all with no change. The next step is to pull the startup control get it running without ECU ignition control and see what happens. If the reluctor/coil in the distributor is flakey, the problem should continue, if not, then there is something going on in MS. Anybody have any further ideas? Thanks!

[Bernard Fife]

Tom,

I have seen this very occassionally on my car. The car runs fine for an hour or so, but then won't run below 1000 rpm or so, until it cools down.

I believe (but don't have good evidence yet) that the signal is being compromised somewhat, possibly by heat in the HEI 8 pin module (mine only does it on while idling for several minutes with the hood closed after a long run hot, so the underhood temp climbs quite a bit). I found the decreasing the dwell helped somewhat (enough that I never saw this unless I was specifically looking for it).

The odd thing is that not everyone is having problems, and they aren't 100% consistent. The general trend seems to indicate that the signal strength at low rpms is weak if the module gets hot, though. Apparently depending on the module and the input circuit (and the tolerances of it's components), it is very close to borderline is some circumstances.

First you need to make sure you have good power to your module (and 5 V to the bypass pin) so check those connections.

On my list of things to try is using the VR input circuit with the 8-pin module. The reason for doing this is that the trigger voltage is both more sensitive than the Hall input circuit, and is adjustable. I'm thinking that a lower trigger voltage might solve this issue. (http://www.msefi.com/viewtopic.php?p=11 ... hei#115134)

I have tried this on the bench. With a square wave input, turning the R56 pot a dozen or so turns clockwise from the fully couter-clockwise position is enough to get good triggering from a 0.5+ Volt square wave under all the conditions I tested at.

So you could try a few things:

- check the wiring from the module to MegaSquirt to make sure it has good connections, solder joints, etc. to make the most of the signal that is sent.

- you could try the VR input circuit as described above.

- some people are having good success with the MS1/Extra input mods here: http://megasquirt.sourceforge.net/extra/setup-hei7.html (do only the input mods, not the output mods)

- you could 'boost'; the input signal with a transistor as Jeff described here: http://www.msefi.com/viewtopic.php?p=93 ... stor#93343 (Unfortunately the attachment is missing, I will add it below.)

Lance.

[hytekauto]

Thanks for the reply Lance.  It's nice to know that I'm not the only one seeing this problem.  I have a nice dual trace scope, and can easily verify that the trigger, return, and control lines are correct.  I do believe that the gain of the 8 pin HEI module varies whether or not it is in the startup mode or run mode.  This was also verified via scope, which explains why the engine will "catch" and continue to run a few times before the "problem" gets to the "kill" the engine point.  Also the Chrysler VR system is much smaller and only looks at one point of the 8 point reluctor as it passes the pickup coil, unlike the GM designed eight points in an HEI distributor.  This probably explains the lack of sensitivity at low distributor speeds.  To add more fuel this line of thought, my brother has a MSD dist in his big Chevy, which has a much larger Ford style reluctor and he has had zero problems with the same HEI 8 pin module.  My intent is to time the engine statically with the control disconnected, via the same connector used in the original GM installation, then run it to see if the gremlins return.  If they do, then I'll try the transistor amp as outlined in #93343.  That sure sounds like exactly what I'm seeing.  If not, then the means that the timing control in the MS has a problem, and I'm in deep trouble....Eventually, I'll probably just go to EDIS.  I've planned that mod for the future, as I have the machined adapter to connect the Escort wheel to the center of the pulley mount on the Hemi. 
 
If this transistor mod fixes all of the hassles I've been having, I sure believe that that mod should be documented in the HEI portion of the Ignition Control Documentation.  Thanks again for your input!
 
Posted by email.

[Bernard Fife]

I sure believe that that mod should be documented in the HEI portion of the Ignition Control Documentation.

Tom,

Yes, I agree. If you could keep notes of what you've done and how it works, and post the results here, I'll add this to the HEI docs when you are happy with the way it works.

Lance.

[jsmcortina]

There is an alternate simpler way of doing this that Joethemechanic has been trying out with improved results.

http://www.msefi.com/viewtopic.php?p=117643

James

[hytekauto]

Lance,
 
The addition of the switching transistor in the opto input did the trick!  That and discovering that the polarity of the input as documented is wrong.  The documentation for the HEI says that the orange lead for the Chrysler VR is positive, however, when used with a HEI 8 module that's reversed.  Wired so the orange lead goes to the "P" input and black to the "N" input, the timing is all over the place, and way out of phase with the cap.  By wiring the black to "P" and orange to "N" the timing is rock solid and the cap phasing is correct.  This is verified by scope.  The engine will now runs very well and I can get this thing tuned.  The Hot Rod Power Tour is only three weeks away........I would be the first to suggest that this mod for anyone using the HEI 8 ignition interface, as it works solidly.  It also has no negative effect on using the stimulator for testing either.  Thanks again for your outstanding help!
 
Tom Westcott
Hytek Automotive
Portage, WI
  Posted by email.

[hytekauto]

Thanks for your comments James.  I chose to use the transistor switch mod, because in testing I also felt that the output of the HEI 8 module was marginal.  The additional gain with the transistor totally fixed the issues I was having.  Keep up the GREAT work, as I follow your work closely.
 
 
Tom Westcott
Hytek Automotive
Portage, WI
 
'68 Hemi Dodge Super Bee
Posted by email.

[Bernard Fife]

I sure believe that that mod should be documented in the HEI portion of the Ignition Control Documentation.

Tom,

I have added the MSnS-E and the transistor mods for the Hall circuit (for those having triggering troubles) here:

http://www.megamanual.com/ms2/inputHEI.htm

I will add links in the assembly and troubleshooting sections of the manual.

Lance.

[clint78z]

I am a bit confused by this GM HEI diagram given at megamanual



It looks like the VR signal should be -ve first??

Then at PICOTECH they have the +VE voltage first on the VR Sensor I need to confirn how too hook it up to the module.


http://www.picotech.com/auto/tutorials/ ... er-1_1.png
I have a scope so it's no problem to check, but need to get them right.

[clint78z]

bump!! Nobody knows if the positive voltage of the VR sensor is supposed to go first?? Reading the Megamanual it describes it as going first, but the diagram doesn't jive !!