I just finally finished up a v3.0 for my car. My original plans were to isolate the tach input to another board to help isolate noise within the v3.0 board. I have always had 2 theories on this issue:

V3.0 board has a flaw or is more succeptible to noise,
Cars with tach noise have issues witht he sensors degrading and the v3.0 can pick it up.

I built the v3.0 in its factory design (sort to speak) and set it up to run just like the v2.2. Not a single spike!

Here is my setup
V3.0 running tandem with Factory ECU
Spark "pullup output" to Crane Ignition box
NO IAC
Idle 800-900
Factory crank receives power and ground from the factory ECU
Factory TPS " " " " " " " "
NO SPIKES


no special noise supression wiring only GXL harnesses i sell routed with factory harnesses.

I will drive this thing a bit more this fall to see if conditions change.

Congrats!

Thats awesome Terry, great work

Nope after driving it a day it started to acquire spikes in the grounds. All sensors saw it including the tach input.

Terry, if you're still having this issue, have you considered making a twisted pair with the tach signal wire and a ground wire to the crank angle sensor?

-----------------------------------------

-Paul

Service Manuals and Guides

I have considered it Paul but i think that will just mask the real issue. The V2.2 works fine so i think there is another problem somewhere. I still maintain that the board is the issue but i think I am alone with that concept. Odd part is its a spike in a specific area. Not all over like some people get. If I get bored i might try a new sensor.

Terry

Terry,

I'm not sure how you specifically build 420A input circuits, but generally speaking, the modifications required create the exact same circuit on both the v3.0 and v2.2 boards. How exactly did you build your input circuit?

I respect your opinion, Terry, but here's why I do not think the issue is within the board:

When we bench test our v3.0 units, there are absolutely no peculiarities in the crank signal. Up and down the entire RPM range, we measure consistent signals with no spikes.

At standard engine speeds, the crank signal is not oscillating fast enough to be generating any significant RF artifacts. Furthermore, given the internal ground and power planes, the tach input circuit is isolated from noise rather well.

In car, we still do not observe any irregularities in the signal.

Using the process of elimination, the only difference between our units and yours is the wiring harness. We use coaxial cable with a grounded sheath... For this reason, I do not think a twisted pair would mask the problem so much as solve it.

As you suggested, perhaps the sensor itself as fault... that would be odd considering it works fine with the v2.2 board, though.

-----------------------------------------

-Paul

Service Manuals and Guides

I agree with you paul. Go back and read your post....i think some of it got chopped off. I would like to see the point you started to make in the first couple sentences.


The circuit from the 2.2 board is the same on the v3.0 board. Now when my 2.2 was made, the neon circuit had the 1.5k ohm resistor added between the VCC and the 390 ohm resistor in the optoisolator circuit. I tried that......no noise at idle but some in random rpm ranges.

As you know now, the neon circuit has been changed to add 150 ohm resistor between the 390 ohm of the opto isolator circuit. This helped a little.

As for the harness, its all GXL wire and i have never seen the issue with a V2.2. The problem is with spark and FUEl. if you run fuel only then its not an issue.

the only thing i have not tried and it could be part of the problem is to run the internal ignitors. All the people running the v3.0 are runnign those. However the fairchild ignitors are obsolete and that only leads us to the bosch BIP ones.

Hong Kong has 5000 of the Fairchild ignitors but I would imagine they want to move 1000 or more to get them here.

Now I wonder if mose Neon guys are running the VR circuit, I am still pushing to use the Optoisolator.

Keep whatever ideas you have coming.....its appreciated.

Terry

Oops, I'll go back and correct that last post. You already answered the question I left out, though: I wanted to know how you built your input circuit.

We use a slightly different configuration than what's in the Megamanual. If you replace the 150ohm resistor in the Megamanual's schematic with a 1k ohm resistor, the input circuit will be electrically identical to what we use. It's much more like the first circuit you mentioned, rather than what's currently in the Megamanual.

I hadn't thought about the ignitors, but you're absolutely correct: they're another major difference. We exclusively use the Bosch BIP373's and so far they've been great.

As for the Neon market, we supply them with the majority of their Megasquirt ECU's these days and of course, we only use the Hall/opto input circuit. Just like the 420A DSM crowd, we haven't been receiving any complaints from the Neon guys about spikes in their tach signal.

So, I suppose the list of things to try is:

* Change 150ohm resistor to 1k ohm resistor.
* Try new crank angle sensor.
* Replace Fairchild IGBT's/4G63 ignitor with Bosch IGBT's.

-----------------------------------------

-Paul

Service Manuals and Guides

The ignitors could be the overall issue. I dont think its the Fairchild ones. See I still use the old style setup Paul. My MS doesnt have any Ignitors, instead it drives a crane ignition box directly. The V2.2 works flawlessly. This was from when their wasnt much to work with way back when.

I will toy with these Ideas and get back to you.

Terry

Paul, the source of the tach spikes in the past were people like me trying to run the ignition box directly off the v3.0. Install the ignitors and all is well.

Now if you get a chance to investigate tell me why it works with the v2.2 but not the v3.0?

Good job and thanks for the help.

I found the output information on the MS material that directly states and diagrams you to use the ignitor when using the v3.0 board only. When using the v2.2 you dont need to, instead you use the LED transistor output Like we have been doing since the beginning.

Since you have studied the v3.0, why is that Paul? It must be the design of the board because the basic circuitry is the same...no?

-----------------------------------------

-Paul

Service Manuals and Guides

No you missed what I was explaining. There is a page in the MS documentation that shows all possible ways of outputing spark from both versions of boards. I was reading it when i was looking into the LS2 coils i have.

Mine is a direct input into an ignition box. The documentation states the way you do it with a version 2.2 and 3.0 must be different. The version 2.2 you use the NEG of diode 17/19 and run it directly to the ignition box.

On the v3.0 when you do that you get tach spikes. on the version 3.0 you must run the internal ignitors then output the ignitor to the ignition box.



I am asking, since you studied the design of the multilayered 3.0 board, why is that? The components used in the v3.0 vs the components used in the v2.2 are the same values and schematics if I recall. the only difference is the board design.

Terry,

I found the page to which I believe you're referring:

http://www.msextra.com/manuals/MS_Extra_Ignition_Hardware_Manual.htm#msdout

You're over-thinking why they suggest using the internal ignitor with the v3.0... the simple reason is that the v3.0 board was equipped with one whereas the v2.2 board was not.

In other areas of the manual, specifically where multiple outputs are required to trigger a multi-channel ignition box, internal ignitors are forgone in favor of the same configuration the v2.2 board would use for a single channel.

The v3.0 board is really designed rather well, so I still suspect the problem may lay within the circuit configuration and not the physical layout. The only differences between the v2.2 and v3.0 boards that I see are the LED driver transistors. The v2.2 board uses 2N2222A transistors, which have a higher collector current rating than the v3.0 board's 2N3904 transistors. Either way, this difference shouldn't cause a problem.

If you've run a jumper directly from the cathode of D17 to the DB37 connector, I recommend adding a 1k ohm pullup resistor if you haven't already. If that doesn't work, try adding a 330 ohm resistor inline between the LED and DB37 connector to limit current and thwart any possible loading affects.

Let me know if either of these solutions work...

-----------------------------------------

-Paul

Service Manuals and Guides

Well, I already have the internal ignitors installed and its fine. So right now now we are just discussing possible causes.

The previous setup was the cathode of LED17 with a 1k pullup resistor. Results were not good and all the previous posts on 2gnt regarding tach spikes were with this setup.

I did not try a 330Ohm resistor.

One last thing you might try is swapping the v3.0's 2N3904's for 2N2222A's. I really don't suspect the transistor type as the cause of this issue, but if you do this step we can absolutely rule out the hardware differences.

-----------------------------------------

-Paul

Service Manuals and Guides