Rpm signal loss

[madkaw]

When I see- ground wires for coils, being connected to the sensor ground for the MS ,and that is all connected to a chassis ground, I see everything that MS tells you NOT to do. Sensor ground is for sensors-ONLY. I didn't use a chassis ground for anything, everything goes to the engine block. Maybe it's just your wording, but thats the way I read it.

[seattlejester]

Ah, I see what you mean, I meant grounded to the chassis, not a pre-existing chassis ground, my wording was indeed misleading.

 

I believe I did read that the sensor grounds should be dedicated. I also read that the LC-1 unit will not work reliably without grounding to the engine, but that is attached to the chassis and I have not had any problems with the unit in the past.

 

I can try moving the coil ground and seeing if that clears up anything. Thanks guys, I have a whole new variety of things to take a look at. 

[madkaw]

grounding is everything from what i read. You might want to rethink the basics -as in grounding- before you troubleshoot anymore. I might have gone overboard(see my ms3x thread), but I didn't want this to be an issue and MS stressed this a lot.

Good luck and hope I helped

[seattlejester]

Oh just realized it was you in the other post. Sorry for referring to you in the removed pronoun form.

 

I have read that grounding to the head may be preferable for the coils, so I may play around with it. The grounding cable for the battery directly attaches to the head, so that would be a stronger ground then flowing through the chassis into the little 8/10 gauge wire that the chassis is grounded with. Granted, the engine is also grounded to the chassis, so the grounding path is pretty much all the same. I can appreciate some overkill though.

[madkaw]

It's not the same. I think you are underestimating the importance of proper grounding. Not trying to be a dick, just emphazing this. You bypass all the noise created in a chassis ground by going directly to the block. 

There is a thing called a ground loop you need to research. 

[winstonusmc]

Ground everything but the coolant temp, air temp, and throttle possition in the same place on the block. Runa a wire from everywhere on the DB37 connectors that is listed as a ground and tie them together and ground them on the block. Then ground all other sensors and coils and such to th block. The reason you ground to the block is the main power source of the car, the alternator, is grounded to the block. I have built quite a few OEM harnesses and they ground this way. The temp sensors and TPS ground through the ECU to get a better reference. Might need to ground the MAF/MAP through the ECU as well. Ground is important, but no need to go crazy like those stupid ground systems from 15 years ago. If you are worried about it, run a 4 or 8 gauge wire from the ECU grounding point on the harness to the battery, not like there isnt a 2 gauge wire for the starter on the block anyway.

 

the only grounding points that go to the chassis are lights, radio, etc. Even some gauges should be grounded to the ECU, like tac and speedo. Seattlejester is correct about a 8/10 gauge wire from the battery to the chassis, or from the block to the chassis, same same.

Edited April 10, 2014 by winstonusmc

[seattlejester]

^That's my thought, the "grounding kits" fad made me turn off of over grounding the car. I can understand having a noisy ground, it is a thing, and a noisy ground would play havoc with sensitive electrical equipment and the coil and injectors can indeed be noisy, but I doubt it will exhibit the same problems I am encountering (rock steady sensor input, variable coil firing, rough running). To be fair, I did take your advice and ground the coils to the head. Still the same problem. The capacitors did not seem to work, so I may hold off on them for the moment.

 

Once again I can appreciate it, input is very important to me, but if possible I would like to look past the possibility of the ground for the moment as it turns out grounding isn't the problem.

 

Akrev, I owe you a meal.

 

The coils are all firing at the same time. I had the tach wire grounded to the chassis with a loop so that I could read off of it with my timing gun. Changed the output to test spark C, and started reading spark pulses. Only to find the wire was on spark A. Switched them around and the same result, switched to B same result. All 6 coils/3 inputs are firing when any of the spark outputs are triggered. Spark output A, B, or C, fires all the coils if the tach output is to be trusted.

 

So to be clear.

The problem I am encountering is all the coils are firing at the same time, making the engine run really really badly.

1-6 are paired, 2-5, are paired, 3-4 are paired as A, B, and C respectively.

When testing singular outputs all the coils seem to fire.

Testing the other outputs (D, E, etc) do not trigger the coils. 

This is limited to A, B, C

The tach wire was wired the same, so 1-6 is paired, 2-5 is paired, and 3-4 is paired.

 

Logic says it can only be 1 of 3 things. 

1. Wiring is messed up, all the triggers are tied together

2. Board is messed up, somehow all the triggers are tied together

3. Settings are messed up, somehow the ECU is triggering the three outputs together

 

Wiring:

I took the wiring apart to confirm.

Following this link

http://www.miataturbo.net/useful-saved-posts-8/upgrading-coil-plugs-all-years-cop-writeup-12704/

My wires are for each coil

1. Common ground

2. Split trigger signal (1-6, 2-5, 3-4)

3. Split tach signal (1-6, 2-5, 3-4)

4. Common power

 

I have checked and they are all going where they should be. Short of unplugging everything and testing continuity, I am willing to say that they follow the wiring diagram correctly.

 

Board:

I opened up the board to take a look. 

 

To my understanding the factory case has 1 native coil driver, so another two would have to be setup (a third was added to drive the boost controller). Based on the silica insulation, it looks like the boost control driver a BIP373 unit was added and two additional inputs were attached to the input, ground goes to ground, and output goes to PAD11 outputting DB5 for the boost control.

 

PAD3 is outputting through DB10, and PAD2 is outputting DB11, PAD1 is outputting through the DB37. I haven't traced where PAD2&3 are from so no progress there. I would imagine it would be going to the other two coil drivers.

 

I think it would be a good idea to test continuity between PAD1-3 as that would make sense on how the coils were triggering at the same time. The only point the come together however look like good clean solder points.

 

I did notice that the jumpers are not in the correct location for the 36-1 trigger wheel

 

• Build the VR conditioner circuit, as described in Step 51 of the MegaManual. All our preassembled Megasquirts with the V3.0 or V3.57 board come with this circuit installed.
• Jumper TachSelect to VRIN. (V3.57: Set JP1 to the 1-2 position.)
• Jumper TSEL to VROUT. (V3.57: Set J1 to the 3-4 position.)
• You may need to adjust the VR trim pots, R52 and R56, when this is installed on a running engine.  A usual base setting is to turn them all the way counterclockwise.
• Jumper IGBTOUT to IGN to send to IGBT ignition coil driver signal out of pin 36 on the DB37. (not needed on a V3.57)
• Cut out R57 if fitted (This won't be there on my units, though.).

JP1 looks to be in 2-3 position

J1 looks to be in the 1-2 position

 

The settings are what I followed below

• Set Spark Mode to "Toothed Wheel."
• Set the trigger angle / offset to 0.
• Set Ignition Input Capture to "Falling Edge."
• Set Spark Output to "Going High (Inverted)." Setting this wrong can overheat the BIP373s or damage the coils.
• Set Number of Coils to "Wasted Spark."
• Set Spark A output pin to D14.

Under Trigger Wheel Settings:

Some settings are the same for all engines.

• Trigger Wheel Arrangement: Single wheel with missing teeth
• Trigger Wheel Teeth: 36
• Missing Teeth: 1
• Wheel Speed: Crank Wheel

The only difference is my angle is set to 80 (8 teeth ahead) and my sensor is set to rising edge, falling edge would give me no signal, rising edge would give me an excellent signal.

 

I have noticed that 80* is specified for 1-2 or 4 cylinders. That isn't how the ECU identifies the number of cylinders does it? That is done in the basic settings tab no?

 

Settings:

My MSQ is up a page or two back, nothing has changed. I believe the only change that could be a problem is having spark set to D14 instead of J11 or something of that nature

 

 

Any thoughts? I'm planning on tracing the board to check where all the inputs are and checking continuity. Other then that I am a bit stumped. Unless the output wires are continuous to each other.

Edited April 11, 2014 by seattlejester

[AkRev]

Good to know, I will have some time to take a look at this in the morning and see what I can find out.

[seattlejester]

Thanks would appreciate it. If someone has their megasquirt unit handy, could they see if there's continuity between PAD1-3? My multimeter is saying there's continuity between the pads. I am getting continuity from each individual output to 3 drivers, the one on U5, Q11, and Q3. Or the 1st, 3rd, and 7th driver mounted on the top going from left to right. This is with testing the output (middle leg) on each of the pad.

 

The lower the value the lower the Ohm's I believe for my multimeter.

 

So...

PAD1 is continuous to U5, Q11, and Q3 (1260). 

PAD2 is continuous to U5, Q11, and Q3 (1260).

PAD3 is continuous to U5, Q11, and Q3 (1260).

 

My brain says, that they should be separate. Like I am getting continuity on the output pins, which would make complete sense as to why and how they are firing the way they were.

 

Additional testing...

 

I am also getting continuity from the heat sink to the drivers.

U5 middle leg (003) & right leg (600)

Q11 middle leg (003)

Q5 right leg (003)

R38 left leg (003)

Q3 middle leg (003)

Q1 right leg (003)

 

These same continuous pins have continuity to the spark output as well. Although at a higher rating. And all of the following have continuity to ground, and power....I can't tell if I am retarded right now or if something is really wrong. How would the board even power up with everything connected like that? 

 

I've moved JP1 to 1-2 and J1 to 3-4. I do not have a resistor for R57. DIY says to remove it, msextra to add the resistor for a pullup hall sensors. Not sure what to do there.

Edited April 11, 2014 by seattlejester

[AkRev]

Alright, Sorry if I am missing something, but let me make sure I see this right. Do you only have one BIP373? (My wife and I just had a new addition to the family this week so my eyes might be missing something.)

You will need one BIP373 for each 2 cylinders to run wasted spark. (http://www.diyautotune.com/tech_articles/using_bosch_bip373s_with_megasquirt.htm)

"

• Now, you will be constructing duplicates of this BIP373 circuit for each coil output you need. For a 1 cylinder, you'll use 1 output; for more cylinders, you will use 1 output for every 2 cylinders. You can mount the additional BIP373s on a second heat sink stacked on top of the first, attached on top with long screws. Or you can mount the BIP373s to the case."

[seattlejester]

Congratulations! I really appreciate you taking the time to look, I know time is at a premium during the first few months.

 

I think there are more than 1, tracing everything on the board is making it a bit difficult.

 

I requested and paid for DIY to install 3 additional drivers, 2 additional for spark, 1 additional for boost control. Boost control is easily identified to the side. 

 

According to the parts list for building a general board

 

It says q3 and q11 are Transistor PNP 6A 100V HI PWR TO220AB

 

U5 is Regulator LDO TO-220

 

Which makes me think they should not be the ignition* drivers. 

 

The parts list says Q16 should be an actual ignition driver, I don't have anything in that slot at all.

 

The Q16 slot does have a wire on the backside from the output (where the middle leg would be if it was there) to PAD1

 

I guess the other two drivers would be wherever PAD2, and PAD3 are connected to internally. The schematic makes it look like PAD1 is for injection (squirt-1), PAD2 is for warmup enrichment (warmup-1), and PAD3 is for acceleration enrichment (accel-1). Am I reading that incorrectly? 

http://www.megamanual.com/357/hardware.htm

Second to the bottom schematic.

Edited April 11, 2014 by seattlejester

[NewZed]

Just curious on the whole ordering from DIY process, but did they send any paperwork, like a schematic or a list, of what they produced?  Seems like you shouldn't have to guess on these things.  They wouldn't send a mystery board.

[seattlejester]

^I realized that I never did upload the form, I attached the order/mod form they sent above. 

 

I feel like I am kind of freaking out at the moment, if I seemed panicked in my response, my apologies. I think I am going to walk away for a little bit grab lunch and try to calm down.

Edited April 11, 2014 by seattlejester

[NewZed]

Thanks for that.

 

I'm no expert on MS or electronics in general but I do know that transistors and their assemblies of diodes, capacitors and resistors are weird beasts.  Some of them are not "on" or active unless powered up.  So trying to diagnose your board by measuring continuity across unknown circuits without the power is probably not the way to do it.  There's probably a method out there somewhere for troubleshooting or verifying a wasted spark ignition setup.

[seattlejester]

Yea, I'm beginning to get that feeling, I think the continuity test was just a way to make me panic prematurely. 

 

I will go back to things I can test. Wiring I will take a look at, the msq I will go over again, to my understanding though, the PAD1-3 should all be outputting through the LED's on the board so they should be on/off in series during cranking, so something to look for there.

 

If I can't find anything after that, I may be ready to throw my towel in and take it to someone more familiar with the process.

[AkRev]

Well, I feel dumb now. 

 

 

Think i got it figured out. You are using logic level drivers. You are not "Going high."

http://www.msextra.com/doc/general/sparkout-v357.html#3coils

Please update your settings as show for the Logic level ignition output. 

 

I am sorry this is short (time is a premium) your build sheet specifies Logic Level, and your coils are that also (assuming from the Miata build COP) MS3 has logic level built into it. You have your settings at "Going high" This refers to using the BIP do drive the ground supply to run the coil. Going low is for Logic Level.

 

Please check my reasoning and Madkaw please chime in if I am wrong. 

[AkRev]

My Link was messed up, Go to the top of the page of that link.

[seattlejester]

That looks exactly like how it is wired, Q16 removed, middle pin wired to PAD1. 

 

The lack of any other obvious looking BIP units also points to logic level, and the coils (based off of a quick search) indeed have built in igniters. 

 

I really hope it wasn't that simple.......could that really cause the coils to be running simultaneously/poorly?

 

Really good catch though!

Edited April 11, 2014 by seattlejester

[AkRev]

I hope so.

As for your car running poorly, yes that would do it. I really hope things have not been damaged from doing so. Please test them out and see if this solves the problem. I apologize for for not catching this earlier, as soon as i opened the link on the Miatatubo forum I knew what was wrong. BTW, the 10k capacitor is optional on the ground, my in-law just had one heck of a scare when he did the polarity backward on his and it exploded. (http://www.miataturbo.net/build-threads-57/93-miata-stolen-flipped-build-thread-75474/page103/)

[seattlejester]

I believe the coils have built in protection so if anything they should be safe. I'm hoping that the board is doing ok, nothing looks out of place, and apparently the continuity readings are indeed moot with the presence of the pull up resistors to pull the voltage up to 5v. 

 

Oh wow, yea, I did take extra care to make sure the polarity was correct, I haven't really had a chance to use them yet, the guy at the parts store sold me a 50v 10,000uf unit saying the voltage is just a measure of the max voltage it can hit. It's fairly sizable though.

 

Small world, I'm actually familiar with the build, voted the most thorough rebuild ever I believe. Actually read through the entire build a few weeks ago.

 

Hopefully some news tomorrow. Boy is my face going to be red if that was the source of all my problems.