PNP ECU for Vw/Audi ME7 #51

Hardware inside and outside of the ECU
User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

Do you have an oscilloscope? Can you capture the raw analog signal?

I believe the TLE8888 is 5V. I would suggest you keep below 5V. Also when you touch the leads you can include a 1k or similar current limiting resistor. When you exceed the 5V rail, your rail diodes will attempt to clamp the signal by shorting the pin to the 5V rail. AKA 12V on a input pin likely means 12V shorted to 5V rail.

DG- appears to have 6.8k to GND. Do you have this installed? If you install this, then you should toggle DG+ with less than 5V and get an output signal. I guess DG- is VRIN2. I do not think it matter very much, as long as it is consistent. If you change it the output signal will invert. See what happens if you have VRIN2 to with resistor to GND.
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

I have 5K at R13 R15 R16 R19 and R20
ME7_PNP_Trigger_Circuit.jpg
ME7_PNP_Trigger_Circuit.jpg (323.3 KiB) Viewed 3259 times
So I did not apply 12 directly at the inputs.

It has been working fine with a jumper-wire from a MCU output directly to one of the inputs

If you look closer at the 6.8K resistor, it's also a 4.64K resistor to 12v.

To me it looks like a voltage divider, so it will actually be a pull-up DG- ??

It seems to be pulled down with ignition off, and up with ignition on.

I also had a 1nF cap at C12. Do I need any capacitor there, and what value?

User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

Is VCC 5V or 12V. If it's 5V, then I expect that DG- is close to 2.5V. As we'll see below, it appears DG- is closer to 3V.

I drew up the attached in QUCS V0.0.19. I started with the below simulation.
https://sourceforge.net/p/daecu/code/HEAD/tree/Hardware/trunk/simulations/Inductor_measure/

For some reason my VR simulation did not play nice, so I disabled that part of the simulation, and put in the voltage source titled "Vsim" instead. See attached simulation. Sorry had to 7zip it to get it allowed in the forum. It complained *.SCH was not an allowed file extension.

I think this is a reasonably close simulation to the reference design. I see that if the Vsim is 1V it does not work, but if it's 10V then it appears to work. I see it also works with Vsim at 5V.

Any how, this might offer some more insight into how the reference was working.
Attachments
VR_TLE8888_sim.PNG
VR_TLE8888_sim.PNG (35.31 KiB) Viewed 3252 times
VR_TLE8888.7z
(1.31 KiB) Downloaded 70 times
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

I never tried simulation software, but it looks impressing!

How will it work with the actual circuit?

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

russian wrote:
Thu Apr 11, 2019 1:21 pm
Just a reminder that I have Lada M86 with TLE8888 if tracing would help. If you continue tracing I can validate with DMM.

See https://github.com/rusefi/rusefi_documentation/blob/master/oem_docs/Lada/M86/M86%20TLE8888%20traced.png
See https://github.com/rusefi/rusefi_documentation/blob/master/oem_docs/Lada/M86/M86%20TLE8888.jpg
I have been trying to find out where the TLE8888 is used OEM for a long time:-)

Some better pictures of the trigger circuit would be wery helpful.

I think this is it?
Lada_Trigger_Circuit.jpg
Lada_Trigger_Circuit.jpg (67.41 KiB) Viewed 3231 times
Last edited by 960 on Thu Apr 11, 2019 5:02 pm, edited 1 time in total.

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

Also, where can I get a ECU?

I see the part number just exist at some russian sites.

What brand is it?

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

Done some more testing today with different trigger setups.

Still nothing at trigger counter.
Test_Audi_A4.jpg
Test_Audi_A4.jpg (1.46 MiB) Viewed 3229 times

User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

take note that the scope GND strap is connected to earth ground, which is likely connected to your USB ground. This can cause problems. You might want to consider using the scope in differential mode. AKA 2 scope probes, then you take CH1 - CH2 and you get the difference.

Is your VR capturing correctly now? Are you still having problems with the VR capture?
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

The lada circuit looks like(For what I can see) 10K in series, 2 x 40K in paralell + one capacitor.

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

kb1gtt wrote:
Thu Apr 11, 2019 5:14 pm
take note that the scope GND strap is connected to earth ground, which is likely connected to your USB ground. This can cause problems. You might want to consider using the scope in differential mode. AKA 2 scope probes, then you take CH1 - CH2 and you get the difference.

Is your VR capturing correctly now? Are you still having problems with the VR capture?
This is a ScopeMeter, not connected to computer.

I had a nice slope measuring like at the picture (directly from sensor), just a few Mv but still.

Later I had it after the resistors at the TLE8888 inputs, but there I had almost nothing.

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

I see from the TLE8888 datasheet that if I connect it as OEM, the TLE will recognize it as a hall sensor.

It's figures of different setups at page 205 in datasheet

User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

The inputs have a mild pull to 2.5V. If you connect a hall, to one lead and let the other input float, that input will float to 2.5V. Then your hall will pull from 0V to 5V. When ever your input crosses the floating 2.5V the output will change state. There's not any real figuring out, its more of how you wire it. The internals of the chip are the same for VR and hall.
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

kb1gtt wrote:
Thu Apr 11, 2019 10:36 pm
The inputs have a mild pull to 2.5V. If you connect a hall, to one lead and let the other input float, that input will float to 2.5V. Then your hall will pull from 0V to 5V. When ever your input crosses the floating 2.5V the output will change state. There's not any real figuring out, its more of how you wire it. The internals of the chip are the same for VR and hall.
Ok, I was thinking about this part, but after reading some of it I dont think anything needs to be changed?


The TLE8888-1QK integrates four detection modes:
• auto detection mode for VR sensor signals
• manual detection mode for VR sensor signals
• semi auto detection mode for VR sensor signals
• detection mode for Hall sensor signals

User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

Ah, yes I can see that now. I can see how the peak detection and gain adjustments could add jitter to a hall signal. So it would be best to hold those at fixed values. I think those parameters are sent via SPI bus.
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

Finally got a trigger signal today, and engine started.

Solved by a 1K pull-up at DG+.

Still some other things left to sort, but at least its running :-)

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

video or it never happened :)
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

russian wrote:
Fri Apr 12, 2019 2:30 pm
video or it never happened :)
Agree :-)

I can't claim it ran fine or more than a few seconds.

I had a screwdriver in the throttle and the fuel pump jumped to run permanent.

I'll film it next time, but need to sort out some more things first.

Now I know it works, and all sensor readings look fine.

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

960 wrote:
Sun Apr 07, 2019 2:54 pm
The analog inputs only have voltage dividers.
I did not have the right ones then, but the part numbers are YC324-JK-0710KL and 744C083153JP.
4 of each.
Hold on, if that's only dividers - would it work?

For example we have a 5v TPS which is a potentiometer. If you try to divide the output without op-amp isolation, do not you get a complex system of four resistors which are all affecting each other? Connecting your divider to the TPS would affect TPS reading would not it? Do I miss some magic mathematics which would account for that?
Attachments
960_2.3_analog.png
960_2.3_analog.png (11.86 KiB) Viewed 3147 times
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

russian wrote:
Sat Apr 13, 2019 5:18 am


Hold on, if that's only dividers - would it work?

For example we have a 5v TPS which is a potentiometer. If you try to divide the output without op-amp isolation, do not you get a complex system of four resistors which are all affecting each other? Connecting your divider to the TPS would affect TPS reading would not it? Do I miss some magic mathematics which would account for that?
I dont understand what you mean.

Those arrays you see there replace 16 resistors, why would they affect each other?

I have perfect readings for all the analogs here.

Arrays are even used for the OEM analogs.

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

960 wrote:
Sat Apr 13, 2019 5:37 am
I dont understand what you mean.
Yes you have reading but I believe that the readings are (a bit?) screwed. Maybe not so bad for TPS.

Say you have 5v TPS which is a potentiometer - really a resistor divider between 5v and GND which has some current in the middle. You have red board divider also between 5v and divider - it also has some (smaller?) current in the middle. You connect these two points and current starts to flow both ways?

Maybe measure TPS output voltage with and without red board dividers connected. I believe red board would affect the voltage on TPS output wire.

Maybe measure CLT output voltage with and without red board dividers connected. I believe red board would affect the voltage on CLT output wire.
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

Ah, now I understand.

I think kb1gtt would be the right one to answer this?

With 10K and 15K dividers, all my readings are spot-on to specs.

Even CLT and IAT are right.

For the potmeters, I guess they have low resistance,
but I dont know why the temperature sensors shows right.

The ECT sensor have 4.4K at 10 degrees and 200 Ohm at 100 degrees
IAT the same

TPS has 650 to 1320 Ohm

User avatar
kb1gtt
contributor
contributor
Posts: 3684
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: PNP ECU for Vw/Audi ME7

Post by kb1gtt »

I expect this can work. However the input will likely be non-linear. The inputs will likely need some software which translates this this curve to the real value.
Welcome to the friendlier side of internet crazy :)

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

Does the Lada ECU have a TC1762?

That's a 3.3V MCU I belive.

How does the analog circuits look?

960
contributor
contributor
Posts: 333
Joined: Mon Dec 10, 2018 1:22 am

Re: PNP ECU for Vw/Audi ME7

Post by 960 »

kb1gtt wrote:
Sat Apr 13, 2019 9:14 am
I expect this can work. However the input will likely be non-linear. The inputs will likely need some software which translates this this curve to the real value.
Would it be better with 100K and 150K for example?

It's strange that the temperature sensors shows right when cold, when they have 4.4K.

That's when they should show most wrong.

How will it look simulated?

User avatar
Dron_Gus
donator
donator
Posts: 202
Joined: Wed Nov 13, 2013 1:11 pm
Location: S-Pb

Re: PNP ECU for Vw/Audi ME7

Post by Dron_Gus »

200 Ohm VS (200 Ohm || 30 KOhm) ~= 395Ohm. 2% make no much sense.
4.4K VS (4.4K || 30KOhm) ~= 3.8 KOhm. 14%.
Knowing additional parallel resistance introduced by input divider you can compensate it in SW.

In case of TPS it will be more complicated to compensate this.

User avatar
Dron_Gus
donator
donator
Posts: 202
Joined: Wed Nov 13, 2013 1:11 pm
Location: S-Pb

Re: PNP ECU for Vw/Audi ME7

Post by Dron_Gus »

960 wrote:
Sat Apr 13, 2019 12:06 pm
Would it be better with 100K and 150K for example?
Check DS, there is some limitation from ADC side. Source of ADC signal should not have to big output resistance.

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

Thank you for the 2x5 header on the next design :)
Attachments
ME7_PNP.jpg
ME7_PNP.jpg (251.77 KiB) Viewed 3109 times
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

User avatar
AndreyB
Site Admin
Posts: 10851
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City

Re: PNP ECU for Vw/Audi ME7

Post by AndreyB »

Are there still any unused pins on the main CPU or are they all used?

If there are unused pins can you route them to a header to be exposed as spares?
https://rusefi.com/s/howtocontribute
very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
my skype is arro239

Post Reply