V3 VR Conditioner...

[SBDJ]

Hi,

I'm having a problem with the VR conditioner on my V3 PCB. A scope shows a signal from the NE wheel at Pin 24, but the signal disappears after Q22; I get nothing at R44 or R48

I have a prototype PCB I have built as I need a second ignition input which is also doing exactly the same

Not sure where to start with this to be honest.

Q22 appears to be soundly soldered; I get continuity between all the components in the VR conditioner as far as I could tell. I couldn't see a short anywhere, but I could be very mistaken

A couple of oddities I did find where regarding the trimmers R52 and R56. First value is the MS PCB, second is on my daughtercard.

R52 (W104, Middle of PCB):

Fully wound in:
1-2: 0,100k
2-3: 0,0
1-3: 0,100k

Fully wound out:
1-2: 0,0
2-3: 20k,100k
1-3: 20k,100k

R56 (W103):

Fully wound in:
1-2: 540, 10k
2-3: 0,0
1-3: 540,10k

Fully wound out:
1-2: 0,0
2-3: 540,10k
1-3: 540,10k

I'm not sure why I'm getting different values, or which one is correct. Either way, I'm still getting exactly the same problem on both.

Can anyone come up with a suggestion as to where to look next? It's got me stumped and obviously had stuck me on my MS implementation...

[SBDJ]

Well, I hadn't realised to sink G- to ground. Upon doing so, I get RPM and cranking notification in MS, cranking RPM is about right too I reckon.

He's the output I get from the two conditioners...

http://www.omegahost.co.uk/gt4/Top-NE%20Bottom-G1G2.wav

I don't really know what it should look like

[lmr052]

Well, I hadn't realised to sink G- to ground. Upon doing so, I get RPM and cranking notification in MS, cranking RPM is about right too I reckon.

I can't verify the waveform on the PC I am on at the moment - it will have to wait. It sounds OK though - I reckon I hear a buzz from NE and clicks from G1/G2. However, if you are getting sensible cranking RPM in MS you must be getting close to a start - ie you have input .... now onto the outputs.

Are you getting spark? (from the coil or distributor)? Are you getting +5 pulses from LED17 or FIDLE modified for 5V?

How about Fuel pump relay closing (do you hear the fuel pump run at switch on)?

During cranking (with failed starting) do you smell fuel at the exhaust pipe?

I assume you have left the injector resistors in place for the time being... if not then you may have to fiddle with the PWM settings and flyback settings - I am ignorant of these as I have high impedance injectors.

I assume you have the correct Req_Fuel value for your injectors too.

Regards, Richard

[SBDJ]

Hi Richard, thanks for the response

I can't verify the waveform on the PC I am on at the moment - it will have to wait. It sounds OK though - I reckon I hear a buzz from NE and clicks from G1/G2.

Good good, that's exactly what you hear

However, if you are getting sensible cranking RPM in MS you must be getting close to a start - ie you have input .... now onto the outputs.

Weeeeeeeeeeee I'm getting somewhere between 200 and 300 rpm whilst cranking, which sounds about right to me and fits in with the MS guide as well.

Are you getting spark? (from the coil or distributor)? Are you getting +5 pulses from LED17 or FIDLE modified for 5V?

I'm not sure if I'm getting spark to be honest - whats a safe way of testing this? Using the LED settings from the MegaTune file you sent me, the centre LED on the v3, configured as IRQ Trigger is pulsing.

How about Fuel pump relay closing (do you hear the fuel pump run at switch on)?

Absolutely, it's loud enough ;) Definate smell of petrol too

During cranking (with failed starting) do you smell fuel at the exhaust pipe?

Yes

I assume you have left the injector resistors in place for the time being... if not then you may have to fiddle with the PWM settings and flyback settings - I am ignorant of these as I have high impedance injectors.

The v3 PCB is as-per original design at the moment. You're right, I have low impedance injectors, and have set them as per the guide.

I assume you have the correct Req_Fuel value for your injectors too.

I've run the little calculator to get this value.

To be honest, I ran out of time tonight - I don't want to upset my neighbours by repeatedly cranking my car over late at night I'm going to have all day and evening saturday to do this; if I can get it started by close of play Saturday I'll be over the moon.

Hopefully it's just some tinkering with values

I have pin 30 of the DB37 hooked up to the stock ECU IGT pin to the ignitor. In fact, all the ECU connections are the same as the "Standard MS Connections" in that work document you compiled.

[lmr052]

I'm not sure if I'm getting spark to be honest - whats a safe way of testing this? Using the LED settings from the MegaTune file you sent me, the centre LED on the v3, configured as IRQ Trigger is pulsing
.......

The v3 PCB is as-per original design at the moment. You're right, I have low impedance injectors, and have set them as per the guide.
...
I have pin 30 of the DB37 hooked up to the stock ECU IGT pin to the ignitor. In fact, all the ECU connections are the same as the "Standard MS Connections" in that work document you compiled.

This is sounding VERY VERY promising..... fuel but no spark. If I was gambling man.... I would guess you need to add pullup resistor to 5V, especially with your comment the V3 PCB is as per original design.

Your IGT (pin 30) should be swinging between 0 and 5V during cranking. Can you check that. On the V2.2 board, I removed the LED and connected a 680 ohm resistor in its place, with the leg closest to the 2N2222 connected to IGT. You can also leave the LED in place and not overload the transistor too.

This arrangement should give you spark... assuming your igniter and coil wiring is correct and your igniter and coil are earthed.

The safest way to test for spark is to mount a spark plug somewhere that is thread is earthed (use a hoseclip?, plastic zip tie, jumper cables??) and connect it to either the coil or to a plug lead from the distrbutor so you can see or hear it spark when cranking. You may want to disconnect the injectors or fuel relay while doing the spark test so you don't load the engine up with fuel.

Also, how are the temps and MAP values showing in MT? Are they close? Don't forget to calibrate the TPS too.

I am expecting the next post to be a "SHE RUNS"... I think you are that close!!!

Don't forget to check the flyback components for temprature once it runs - I would hate to hear of a failure.


Regards, Richard

[lmr052]

He's the output I get from the two conditioners...

http://www.omegahost.co.uk/gt4/Top-NE%20Bottom-G1G2.wav

I don't really know what it should look like

This link shows images of my condioner outputs (second set) http://megasquirt.sourceforge.net/extra/wheel2.html

I don't like what I see on your wave forms.

I am not convinced your signals are clean enough for reliable running. I am seeing signs of cross talk (ie "glitches" - a technical term) on the G1/G2 signal and this should not be there at cranking speeds. They will become worse at higher speeds.

I would be tempted to add 2 x 220 ohm resistors across the inputs - G1 to G- and G2 to G-. You may also need to increase R45 on the NE input to 15k. This would mimic my input set up for input resistors as the V3 circuit is an evolution of the one I used. My triggering seems to be very solid and stable.

I would also recommend you make sure the 100k pot is set to have 100k across pins 1 and 3 of the op amp, and the 10k pot is set to have the wiper (pin 2?) at 0V. This is the setup Bruce recommends and has modelled.

Bottom line is you may get it started but I don't think it will rev very well until the conditioners are sorted.

Regards, Richard

[SBDJ]

I am not convinced your signals are clean enough for reliable running. I am seeing signs of cross talk (ie "glitches" - a technical term) on the G1/G2 signal and this should not be there at cranking speeds. They will become worse at higher speeds.

Crosstalk is also present between G1/G2 and NE as well - if I run with one input disconnected, I get noise on the the output of the disconnected VRs.

I would be tempted to add 2 x 220 ohm resistors across the inputs - G1 to G- and G2 to G-. You may also need to increase R45 on the NE input to 15k. This would mimic my input set up for input resistors as the V3 circuit is an evolution of the one I used. My triggering seems to be very solid and stable.

Cool, I'll give that a try. Have you done anything with NE before the input stage, and I've connected G- directly to ground; is this correct?

I would also recommend you make sure the 100k pot is set to have 100k across pins 1 and 3 of the op amp, and the 10k pot is set to have the wiper (pin 2?) at 0V. This is the setup Bruce recommends and has modelled.

Done I'll tap out the opamp with a meter later - have you seen my resistance observations above?

Bottom line is you may get it started but I don't think it will rev very well until the conditioners are sorted.

Thanks mate

[lmr052]

Crosstalk is also present between G1/G2 and NE as well - if I run with one input disconnected, I get noise on the the output of the disconnected VRs.

Have you done anything with NE before the input stage, and I've connected G- directly to ground; is this correct?

Done I'll tap out the opamp with a meter later - have you seen my resistance observations above?

Crosstalk and noise is possible if the an input is left floating. Unconnected inputs should be grounded

Yes connect G- to ground. I increased the input resistance to 15K for NE.

I saw your resistance observations but I don't know which is Pin 1, 2 or 3 - hence my response around what the outcomes of the settings should be. You can replace the pots with fixed resistors... well only the 100K. The 10k can be omitted and the base of the transistor tied to 0V - after all - you only want the transistor for its fixed voltage between the emitter and base.

I had a spare distributor I used for my bench testing to fine tune my resistor values. You may want to mark your distributor and remove it and use it for bench testing with an electric drill.

Keep at it.

Regards, Richard

[SBDJ]

Hello

I'll let you know after more testing. I should have some time tomorrow to get a fair bit done. I'm hoping to get it started and idling, that would be a major achievement in my book

This diagram shows which pins on the pots are which...



The pots themselves are fine, I was just wondering if the strange values were the result of some dodgy soldering, faulty component elsewhere or a short elsewhere. It's confusing to say the least.

I may pull R52 and replace with a fixed, and drop R56 with Q23 base grounded anyway if they are not needed for VR sensor operation. Couple of less things to go wrong ;)

[SBDJ]

Just to update you;



I've managed to clean the signal up to this; does it look any better?

[SBDJ]

I'm a little confused by your pull-up resistor for the IGT connection.

Are you saying that instead of using the standard MS Pin 30, you've set the LED to IRQ Trigger, taken the LED out, replaced with a resistor and triggered from that?

I've seen several comments of people soldering a pull-up resistor across D9 on the v2.2. Obviously the FIDLE circuit is slighty different on the v3.

I can't tell if it swings between 0v and 5v; it does something but it's far too quick for my multimeter. Perhaps scoping the output might be better, but won't really give me a tangible voltage.

As for the coil wiring; afaik that's not directly connected to the ECU so shouldn't be affected. The ignitor has two connections - IGT and IGF. I'm don't think the feedback is needed, should I ground this or leave it floating?

[SBDJ]

Sorry mate;

I've just been looking through your MSQ file and am even more confused

You have FIDLE Out from the MS connected to the IGT connection, but your MSQ file says you are using FIDLE for idle control...?

Also, my spark output LED is always lit?

I must be really misunderstanding how this works

[SBDJ]

Hello again

You're gonna love me with all these posts when you come back online Richard

It starts...

But it doesn't idle and stalls immediately.

I've left the LED in place, and taken the IGT connection from the LED. Should I have a resistor in place - I've been told these igniters are rather sensitive

[SBDJ]

It start's and IDLES

[lmr052]

It is interesting watching the evolution of this thread... the emotions, frustrations and joys.... and the effects of 9 hour time difference and some 10,000 miles of distance!

Congratulations on battling through to success !!

The VR signals look very good now. I have addressed many of the questions in the PM. Also check that PM for tuning comments. Waht did you end up doing to get the signals in the post above?

Re the error in my document - it must be out of date as I expected to use FIDLE as an output earlier on. I use FIDLE for idle control, and and I use the LED 17 (V2.2 PCB) for ignition out. For idle, I use the 3 wire IACV adapter circuit.

Now - onto the tuning bit and the other "extras" (idle control, boost control, etc).

Regards, Richard

[SBDJ]

I used your resistor values, and they gave me what you see, so big thanks to you

I'm glad I've finally got it idling, as you say now onto the part you usually pay lots of money for - the tuning.

Given that my signals now look OK, perhaps I can finally put it all in a case and do the loom properly as well - resistors soldered everywhere at the moment

I'm definately going to be documenting this

[lmr052]

Any updates?

My experiments with V3 Conditioner on the bench (once I twisted the alligator jumper wires to the distributor) show the following mods seem to work OK

G1/G2 Conditioner
1. Duplicate R45, R46 and C31 for the second input
2. Add 220 ohm resistor across each of the G1/G2 inputs to ground
3. Replace R52 pot (Hysteresis) with fixed 27k
4. Add 1k resistor on inverted output (to protect U1 & U7B).
5. Connect output from added 1K to U1
6. Omit R56 (10k pot). Connect the base and collector of Q23 to 0V

NE Conditioner
1. Replace R45 with 15k
2. Replace R52 pot with fixed 100k
3. Connect inerted output U1. If using V2.2 PCB, remove Opto. It seems to work OK with the pullup resistor and noise cap still in place but these could be removed as well.
4. Omit R56 (10k pot). Connect the base and collector of Q23 to 0V

I have yet to test in a car but the bench test looks very nice. Has anyone got any other tips for Toyota 3VR trigger setups?

Regards, Richard

[muythaibxr]

I'll probably use some of your tips to clean up some nasty interferance from the G1/G2 sensors on the Ne sensor... I just cut 2 teeth out of the Ne wheel, and I'm still getting pretty nasty noise from the G sensors in the form of extra teeth showing up on the conditioned output to the processor....

[suberimakuri]

I'm going to try your suggestions tomorrow night Richard. I think I can build the G1/G2 conditioner in the proto area.
Thanks for spending the time on this!!

I never got a chance to trial different values before engine blew at the track.

[lmr052]

I'll probably use some of your tips to clean up some nasty interferance from the G1/G2 sensors on the Ne sensor... I just cut 2 teeth out of the Ne wheel, and I'm still getting pretty nasty noise from the G sensors in the form of extra teeth showing up on the conditioned output to the processor....

Try the resistors on the G1/G2 only so that they have some load and the induced voltage is kept under control.

At long last, I spent some time over the weekend with internal interconnect wiring - just need to wire in the 25pin expansion and comms cable so I can do some real life testing on this set-up.

Regards, Richard