Ford Mustang Forums banner

21 - 40 of 53 Posts

·
Registered
Joined
·
4,477 Posts
I had a little time so I shook the rust off and looked at the schematics again. Here's some information for you to check out to help you understand how your board is (or isn't) set up, and how this ignition control is supposed to work. If the above test fails, you'll need to dive in and check stuff out and hopefully this will help. I'll highlight actual tests to perform in bold. Unless stated otherwise hardware tests must be done with the megasquirt off.

To run TFI, megasquirt uses the high-current ignition driver circuit (Q16 the main component which is an insulated-gate bipolar transistor : IGBT : specs here) to recreate a SPOUT signal. Transistors are fancy relays which are fancy switches. This is the thing we turn on and off to create that square wave.

First describing turning it on or off. The input for Q16 is Pin1 and the access for this on the board is "igbtin". We take the signal from the CPU directly to this. There are 2 options of CPU pins that can be used for this, CPU Pin7 aka "Squirt-1" aka D14 (I will call it D14 from now on). The other option is CPU Pin 17 aka JS10 (just JS10 from now on).

Your megasquirt needs to have this connection to igbtin from either JS10 or D14. The instructions in the hardware manual say to make this connection using a jumper wire with a 330 ohm resistor inline. This is not necessarily the only way to do it, but the connectivity must be there and this will be verifiable with a DMM. 1. Verify connectivity from igbtin to either D14 or JS10 expecting ~ 330 ohms resistance, up to 1k might be ok.
If it doesn't exist create it using D14 per steps A B and C on page 106 of the manual


In order for Q16 to function it needs a ground, Pin3 should be connected directly to ground. 2. Verify Q16 Pin3 has connectivity to PAD4 with 0 ohms resistance.

Q16 Pin2 is the output of the IGBT called "IG_Out" in the diagram. When the IGBT is energized by the CPU IG_Out is effectively shorted to ground as Q16 internally connects Pin2 with Pin3. This shorting to ground gives the 0v half of the SPOUT square wave.

To get the 12v half of the SPOUT square wave a "pull-up" resistor is connected to IG_Out and a +12v source. On the 3.57 board the correct place to source this is the S12 pin. The resistor should be 1k ohms. 3. Verify Q16 Pin2 aka IG_Out has connectivity with S12 expecting ~ 1k ohms resistance.
If it doesn't exist create it per steps E F and G on page 106 of the manual.


IG_Out is connected to DB37 Pin36 directly through the board.

Because the Pin2 is energized to +12v through 1k of resistance, but shorted to ground without any resistance when Q16 is activated, Pin2 will show ground when Q16 is "on" and will show +12v when Q16 is "off". Therefore switching Q16 on and off will create the 12v to 0v square wave output, simulating the SPOUT signal that the TFI needs.

Tests you can perform with the megasquirt turned on: Be careful with any testing that requires probing the board while megasquirt is on. 4. With megasquirt on verify that you have +12v on your SPOUT wire, with the spout plug in, at Pin5 of the TFI. If not, test for it in the megasquirt starting at the source S12. Next verify that you have +12v on Q16 Pin2. This +12v should read through IG_out to DB37 Pin36 to your SPOUT wire all the way to the TFI Pin5.

Once the hardware input for the IGBT is verified, the firmware needs to be configured to use the correct CPU output, either D14 or JS10. 5. With megasquirt on connect with Tunerstudio and verify firmware matches config under Typical Settings on page 106 of the manual.
Key settings that must be right for timing control: Spark A output config matches hardware config of either D14 (possibly called LED Spark in the firmware) or JS10, Spark mode = Basic Trigger, Spark output = Going High.


Ignition capture needs to be falling edge but you'll still get timing control, with some hiccups, even if set wrong. Dwell Type should be fixed and Duty should be 50% unless you have a 94/95. For the sake of testing you should used fixed at 50% anyway even if you have a 94/95. Setting 100% duty will turn Q16 on constantly, so the SPOUT should read ground constantly. Setting 0% duty will never turn Q16 on so the SPOUT should read +12v constantly. This is the basis of the test I gave in the earlier post.

Trigger angle/offset just needs to equal whatever is necessary so that what you read on a timing light is the timing you command in megasquirt with the spout plug installed. If you set a commanded 20 degrees but don't read 20 degrees with a timing light, change trigger angle until they match.
 

·
Registered
Joined
·
1,420 Posts
That last post was awesome and lets me in on the inner workings of the MS2...

Just in reading and looking over my notes and pics I am having some issue with the IGBTIN to JS10 having a 330 ohm resistor... I have this pic from when I had it apart. I see a straight wire from JS10 to IGBTIN. I will disassemble it again and verify if it does or doesn't. I also read somewhere that the factory harness has this resistor built in? Is that true?

Thank you for helping me with the issue. I am wanting to play with the car again and this little hiccup is pushing me the wrong direction, lol.
 

Attachments

·
Registered
Joined
·
4,477 Posts
Honestly I think JS10 straight to igbtin is fine. I think the resistor is mentioned to offer some protection to the CPU in the event that Q16 failed in some strange way that shorted Q16 Pin1 to Pin3; straight to ground.

Even then, I don't know if cooks a CPU for an output pin to be shorted to ground or not.

Anyway of all the resistors mentioned I think this is the least important. Also looking around the internet I'm seeing a lot of examples where this resistor is omitted so I don't think it makes a difference for functionality.

As far as the harness having that resistor, no. Any resistor in a harness would be irrelevant to this.

What you're probably reading is that the megasquirts built and sold for mustangs also come with an adapter harness to connect to the factory harness. That adapter harness will have the pull-up resistor connecting DB37 pin 36 to +12 volts. Keep in mind DB37 pin36 is hardwired to Q16 pin2 internally to the megasquirt, so you can check for 12v in either place it's effectively the same thing.

I see Q16 in place, that's good.

What I don't see is S12 to Q16 pin2. Again this is just one way to get 12v on the SPOUT wire. The way that people sell these things packaged like to do it is in the harness like mentioned above. Either way works, it doesn't matter, but it is absolutely necessary that a 1k ohm resistor is somewhere connecting the SPOUT wire to battery voltage.

I prefer S12 to Q16 pin2 or DB37 pin36 over doing it in the harness because it keeps it contained in the computer because this necessary pull-up resistor in a harness tends to be a forgotten critical component in a few years and next thing you know a harness swap later you've lost timing control and you don't know why. Full disclosure, my pull-up resistor was in the harness so I'm a sinner too!

The other jumpers I see 2 going to the top right which is the config for canbus, and the other 4 together plus the 5th that makes the 90 degree bend are the stepper motor outputs but in mustangs they're typically used for fan control.

I still think this is where I would start:
Verify your firmware settings, using JS10 etc. Then set fixed dwell to 0%, SPOUT plug removed, fire up the car on megasquirt.
Check the voltage at the spout plug ECU side with the fixed dwell set to 0%. Shut down the car, change dwell to 100%, fire it back up and check voltage again.

At 0% you should measure battery voltage, and at 100% you should measure ground. If that isn't true something is wrong in the hardware or config.
 

·
Registered
Joined
·
1,420 Posts
Honestly I think JS10 straight to igbtin is fine. I think the resistor is mentioned to offer some protection to the CPU in the event that Q16 failed in some strange way that shorted Q16 Pin1 to Pin3; straight to ground.

Even then, I don't know if cooks a CPU for an output pin to be shorted to ground or not.

Anyway of all the resistors mentioned I think this is the least important. Also looking around the internet I'm seeing a lot of examples where this resistor is omitted so I don't think it makes a difference for functionality.

As far as the harness having that resistor, no. Any resistor in a harness would be irrelevant to this.

What you're probably reading is that the megasquirts built and sold for mustangs also come with an adapter harness to connect to the factory harness. That adapter harness will have the pull-up resistor connecting DB37 pin 36 to +12 volts. Keep in mind DB37 pin36 is hardwired to Q16 pin2 internally to the megasquirt, so you can check for 12v in either place it's effectively the same thing.

I see Q16 in place, that's good.

What I don't see is S12 to Q16 pin2. Again this is just one way to get 12v on the SPOUT wire. The way that people sell these things packaged like to do it is in the harness like mentioned above. Either way works, it doesn't matter, but it is absolutely necessary that a 1k ohm resistor is somewhere connecting the SPOUT wire to battery voltage.

I prefer S12 to Q16 pin2 or DB37 pin36 over doing it in the harness because it keeps it contained in the computer because this necessary pull-up resistor in a harness tends to be a forgotten critical component in a few years and next thing you know a harness swap later you've lost timing control and you don't know why. Full disclosure, my pull-up resistor was in the harness so I'm a sinner too!

The other jumpers I see 2 going to the top right which is the config for canbus, and the other 4 together plus the 5th that makes the 90 degree bend are the stepper motor outputs but in mustangs they're typically used for fan control.

I still think this is where I would start:
Verify your firmware settings, using JS10 etc. Then set fixed dwell to 0%, SPOUT plug removed, fire up the car on megasquirt.
Check the voltage at the spout plug ECU side with the fixed dwell set to 0%. Shut down the car, change dwell to 100%, fire it back up and check voltage again.

At 0% you should measure battery voltage, and at 100% you should measure ground. If that isn't true something is wrong in the hardware or config.
Finally got around to testing this. My TS will only let me go from 0%-99% for some reason, but either way here are the results. At 99% dwell we are sitting at 2.3ish volts at idle, and at 0% dwell we are all over battery voltage. I opened up the case and have not found a resistor from Q16 pin 2 to S12 nor anything connected to S12. I seen on the TFI instructions that they connect S12 to IGNOUT? I dont see IGNOUT labeled anywhere. But in talking with other MS2 users and builders this weekend, they stated that if I didnt have that pull up resistor installed I wouldnt get 12V key on engine off, is that correct? I also unhooked the DB37 from the MS2 and measured voltage at the spout wire key on engine off and I have 12v. I removed the jumper harness from the factory harness and 12v left with is, which indicates to me that the jumper harness contains my pull up resistor, correct??
 

·
Registered
Joined
·
1,420 Posts
Also verified that the harness has the pull up resistor. 330 ohms from PCM pin 57 to 36. Plugged jumper harness in DB37 and measured resistance from pin 57 to middle pin on Q16 and its 330ohm, PCM pin 36 to middle pin on Q16 is 1.0 ohms.
 

·
Registered
Joined
·
4,477 Posts
Yup you found it. It's correct if it didn't exist you wouldn't measure voltage on that wire.

It seems to me the megasquirt checks out.

I would run a new wire from pin36 to TFI pin5, for now it's sufficient to strip the ends of the wires and backprobe them into the connectors with a toothpick or something. Leave the SPOUT plug removed on the original wire and see if this fixes your timing control.
 

·
Registered
Joined
·
1,420 Posts
Yup you found it. It's correct if it didn't exist you wouldn't measure voltage on that wire.

It seems to me the megasquirt checks out.

I would run a new wire from pin36 to TFI pin5, for now it's sufficient to strip the ends of the wires and backprobe them into the connectors with a toothpick or something. Leave the SPOUT plug removed on the original wire and see if this fixes your timing control.
I will try that today on my lunch break. As a side note I verified that timing is correct with my DVOM as the spout jumper, so that had my curiosity peaked, so I put a resistor in the spout connector just to see what it would do. The only resistor I had on had was 1750 ohms. The car ran but you could tell the timing was way off as it was popping back in the intake.

Just a question but all the MS2 builds I've seen have a 1000 ohm pull up resistor, is it possible that the 330 ohm resistor in the harness isnt enough? If so I can always solder in a 1000 ohm on the board and jumper past the 330 in the jumper harness.

Once again thank you for all the help!

Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
4,477 Posts
I will try that today on my lunch break. As a side note I verified that timing is correct with my DVOM as the spout jumper, so that had my curiosity peaked, so I put a resistor in the spout connector just to see what it would do. The only resistor I had on had was 1750 ohms. The car ran but you could tell the timing was way off as it was popping back in the intake.

Just a question but all the MS2 builds I've seen have a 1000 ohm pull up resistor, is it possible that the 330 ohm resistor in the harness isnt enough? If so I can always solder in a 1000 ohm on the board and jumper past the 330 in the jumper harness.

Once again thank you for all the help!

Sent from my SM-G930P using Tapatalk
No actually if anything 1k resistor would cause problems.

When you put a resistor in place of the SPOUT plug you are putting that resistor in series with the circuit. The pull-up resistor is only tapping the circuit on one leg, and connecting 12v to the other leg.

For this purpose you want the pull-up resistor to have a resistance value as low as possible without being too low that we aren't limiting the current to reasonable levels. Think of it like filling a fragile pitcher with water, if you give it the full force of a river it'll pop. If you just give it a trickle through high resistance, like a small straw, yes it will fill but it will be slow and laggy to fill. If you give it the right amount of resistance it will fill rapidly but will still be protected and won't break. at 4,000 RPM * 8 ignition events per RPM you need it filling rapidly ;) As long as 330 ohms isn't causing Q16 to fry, it's doing the job.

By the way, when Q16 is ON it shunts all the voltage on the SPOUT wire to ground, by using a resistor in place of the SPOUT plug you add lag to both how quickly the battery voltage from the pull-up is "seen" by the TFI module, and you add lag to how quickly the voltage can empty to ground.
Ideally we want 0 resistance on the SPOUT wire because we want the transition of voltage in the square wave that the TFI sees to happen as rapidly as possible.
Another way to think of it is adding resistance to this line takes the corners off the square wave and makes it look more like a sine wave.
 

·
Registered
Joined
·
1,420 Posts
No actually if anything 1k resistor would cause problems.

When you put a resistor in place of the SPOUT plug you are putting that resistor in series with the circuit. The pull-up resistor is only tapping the circuit on one leg, and connecting 12v to the other leg.

For this purpose you want the pull-up resistor to have a resistance value as low as possible without being too low that we aren't limiting the current to reasonable levels. Think of it like filling a fragile pitcher with water, if you give it the full force of a river it'll pop. If you just give it a trickle through high resistance, like a small straw, yes it will fill but it will be slow and laggy to fill. If you give it the right amount of resistance it will fill rapidly but will still be protected and won't break. at 4,000 RPM * 8 ignition events per RPM you need it filling rapidly ;) As long as 330 ohms isn't causing Q16 to fry, it's doing the job.

By the way, when Q16 is ON it shunts all the voltage on the SPOUT wire to ground, by using a resistor in place of the SPOUT plug you add lag to both how quickly the battery voltage from the pull-up is "seen" by the TFI module, and you add lag to how quickly the voltage can empty to ground.
Ideally we want 0 resistance on the SPOUT wire because we want the transition of voltage in the square wave that the TFI sees to happen as rapidly as possible.
Another way to think of it is adding resistance to this line takes the corners off the square wave and makes it look more like a sine wave.
Would a bad ground on the one leg of Q16 cause this issue?

My only reasoning for that thought is, correct me if I'm wrong but the TFI has to see 0v then 12v switching. If we never fall to 0v does the TFI "reset"? So I dont get spark with the spout in. Then I add a load on that circuit which makes the TFI side of the circuit fall to 0v because I have something drawing power. Again I may be way off base but it works in my head lol.


Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
4,477 Posts
Yeah if you go back up to my post with the numbered diagnostic tests in bold, one of them is to validate that Q16 has a ground with basically 0 resistance.

Adding a resistor in series wouldn't change anything if the ground on Q16 was no good, because your load is sharing the same ground.
 

·
Registered
Joined
·
1,798 Posts
what are the odds one of your grounds in the factory harness is bad? Sounds like seijirou's test would validate.

edit: derp, he beat me on the response
 

·
Registered
Joined
·
4,477 Posts
what are the odds one of your grounds in the factory harness is bad? Sounds like seijirou's test would validate.

edit: derp, he beat me on the response
No derp at all, that's an excellent suggestion. My test only verifies the ground path from Q16 to the internal ground in megasquirt at PAD4. I didn't say to inspect the ground from the MS to the car and I should have. Good catch.
 

·
Registered
Joined
·
1,420 Posts
Any idea which pin the ground exits from the MS2 via the DB37? Or should I open it and measure resistance between Q16 and all the pins to find it?

Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
4,477 Posts
Any idea which pin the ground exits from the MS2 via the DB37? Or should I open it and measure resistance between Q16 and all the pins to find it?

Sent from my SM-G930P using Tapatalk
Q16 connects to the ground plane on the board, shared by everything that connects to ground inside your megasquirt. That ground plane is connected to the DB37 on pins 1,2 and all pins 7 through 19.

Typically about 6 of these pins are connected and bundled together to create a single ground, you'll see that bundling in your harness. One of them is often reserved for "sensor ground" so if you see one that is by itself that's why.
 

·
Registered
Joined
·
1,420 Posts
Verified DB37 pin 1 to pad4, .1 ohms.
Pad4 to Q16 pin 3, .1 ohm.
Q16 Pin 2 to DB37 pin 36, .1 ohm.
Q16 Pin 1 to JS10, .1 ohm.

All chassis grounds show .2-.4 ohms.

Jumped straight from Q16 pin 2 to the back of the TFI with spout out and the car stalls.

What I have noticed is the D14 LED flashing very quickly, like engine RPM while it's running with the spout out. I see some people trigger off D14. Would it be an option to reroute jumper and run the ignition off D14 instead of JS10?



Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
1,420 Posts
Q16 connects to the ground plane on the board, shared by everything that connects to ground inside your megasquirt. That ground plane is connected to the DB37 on pins 1,2 and all pins 7 through 19.

Typically about 6 of these pins are connected and bundled together to create a single ground, you'll see that bundling in your harness. One of them is often reserved for "sensor ground" so if you see one that is by itself that's why.
I seen that and was worried when some of the pins in my chassis side DB37 were omitted. Along with jumping from TFI to Q16, I will add that the car runs/controls timing with a stock PCM in place so it seems to me that it has to be a MS2 issue.

I am about to the point of spending way too much time and effort on this to be worth it and will either order a new MAF and tune the stock PCM with my TwEECer or order a MS2PNP. I'm not in the mood to spend the money since I already have this system but I cant see my car sit any longer either after it just got a new engine, heads, cam, intake, converter, trans, exhaust, etc etc etc.

Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
4,477 Posts
What is your timing set to and what is your trigger angle?

Right now the D14 LED is doing it's "default" job of flashing on every fuel injection event, you could re-route the jumper and configure D14 for spark but I very highly doubt the problem is specifically on JS10. I know for a fact that output is functional because your duty cycle test showed the correct difference of voltage in the correct manner which required JS10 to be working.

I'm pretty sure your megasquirt works, what I'm wondering now is if the spark settings are resulting in spark that is so advanced or so retarded related to TDC that you get a stall.
 

·
Registered
Joined
·
1,420 Posts
What is your timing set to and what is your trigger angle?

Right now the D14 LED is doing it's "default" job of flashing on every fuel injection event, you could re-route the jumper and configure D14 for spark but I very highly doubt the problem is specifically on JS10. I know for a fact that output is functional because your duty cycle test showed the correct difference of voltage in the correct manner which required JS10 to be working.

I'm pretty sure your megasquirt works, what I'm wondering now is if the spark settings are resulting in spark that is so advanced or so retarded related to TDC that you get a stall.
Can I email you my tune for review? My base timing is set to 10° BTDC, trigger angle is 0. Total timing in the timing table varies around 28 degrees at idle. When I install the DVOM as the spout the timing advances to around 28 degrees total timing.

Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
1,420 Posts
I've even held the throttle so the engine is running 3500+ RPM and had someone install the spout and it is like you turned the key off. No backfires or pops like it's trying to spark out of time.

Sent from my SM-G930P using Tapatalk
 

·
Registered
Joined
·
4,477 Posts
Yeah PM me your email address and I'll send you an email to reply to.
 
21 - 40 of 53 Posts
Top