Page 2 of 3

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 10:52 am
by kb1gtt
NormanAlphaspeed wrote:
Tue Dec 11, 2018 7:01 am
I made a PnP Speeduino for an Audi A4 1.8T, I can help with fan control (the Audi used one, idk if other VAG did), CANBUS messages (I am only missing A/C confirmation message), and well VVT is two-stage as far as I know, on and off. If I can help let me know!
If I recall correctly,the 1.8T and 2.0T are very similar. I have an A4 2.0T Quattro. It's automatic and I understand the transmission is mechanical, so does not need electrical signals to operate. I understand the electrical signals are for monitoring purposes only. However I could be wrong about that. I think a key issue is controlling the injectors. However we have made some decent progress on that front. So control over those injectors is not really a far reach any more.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 11:11 am
by kb1gtt
It sounds like we have a person who has the hardware and perhaps the time to attempt a FSI engine. It sounds like they have a breakout connector, and might be able to piece together the other bits to get some hardware. I think the software isn't really that hard. It might run with rusEFI as it stands now. We once did some speed tests and I recall we expect we can get the speed fast enough to handle multiple injection events during combustion at something like 6 kRPM. I don't think the firmware can do that now, but we did some tests and these speeds should be realistically obtainable.

This fellow was making good progress, but he went to school and has been idle in recent days. As a baby step, he was working on a test bench.
https://rusefi.com/forum/viewtopic.php?f=2&t=1393

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 6:26 pm
by 960
NormanAlphaspeed wrote:
Tue Dec 11, 2018 7:01 am
I made a PnP Speeduino for an Audi A4 1.8T, I can help with fan control (the Audi used one, idk if other VAG did), CANBUS messages (I am only missing A/C confirmation message), and well VVT is two-stage as far as I know, on and off. If I can help let me know!
This would be of great help.

What I am thinking, is to get everyting on one board that fits in the original ME7 and MED9 case.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 6:56 pm
by 960
kb1gtt wrote:
Tue Dec 11, 2018 10:52 am
NormanAlphaspeed wrote:
Tue Dec 11, 2018 7:01 am
I made a PnP Speeduino for an Audi A4 1.8T, I can help with fan control (the Audi used one, idk if other VAG did), CANBUS messages (I am only missing A/C confirmation message), and well VVT is two-stage as far as I know, on and off. If I can help let me know!
If I recall correctly,the 1.8T and 2.0T are very similar. I have an A4 2.0T Quattro. It's automatic and I understand the transmission is mechanical, so does not need electrical signals to operate. I understand the electrical signals are for monitoring purposes only. However I could be wrong about that. I think a key issue is controlling the injectors. However we have made some decent progress on that front. So control over those injectors is not really a far reach any more.
They are basically the same engines. They even share the same engine block.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 7:21 pm
by kb1gtt
I have the paper Bentley manual for the A4's. The A3 only came on a CD, and I can only open it on an XP virtual box. It's a pain to get the schematics into PDF format for the A3. Do you know of any method for getting the A3's schematics into PDF? Perhaps a windows macro.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 7:29 pm
by 960
kb1gtt wrote:
Tue Dec 11, 2018 10:52 am
NormanAlphaspeed wrote:
Tue Dec 11, 2018 7:01 am
I made a PnP Speeduino for an Audi A4 1.8T, I can help with fan control (the Audi used one, idk if other VAG did), CANBUS messages (I am only missing A/C confirmation message), and well VVT is two-stage as far as I know, on and off. If I can help let me know!
If I recall correctly,the 1.8T and 2.0T are very similar. I have an A4 2.0T Quattro. It's automatic and I understand the transmission is mechanical, so does not need electrical signals to operate. I understand the electrical signals are for monitoring purposes only. However I could be wrong about that. I think a key issue is controlling the injectors. However we have made some decent progress on that front. So control over those injectors is not really a far reach any more.
The automatic tramsmission has the TCU inside. The only communication it has with the engine, is via can-signals.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 8:00 pm
by 960
As we now has someone that already have a working PNP, I think we should go straight to making a PCB.

No reason to waste time and money on a frankenso, if all has been figured out with Speeduino.

I have no experiense with KiCad, but if someone could help would be great.

The PCB is 165 mm wide all the way, the length is 119 mm at the middle.

So you can use the scan to get the rest right.

If someone could do this, that would be great.

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 8:00 pm
by NormanAlphaspeed
Hey guys, all the mods I made are in my Github repository, https://github.com/ElDominio

All code that I added to the Speeduino firmware is in alphaMods.h (fan control mostly)
CAN code is in the CAN ME7.5 directory, hope it all helps!

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 8:43 pm
by 960
NormanAlphaspeed wrote:
Tue Dec 11, 2018 8:00 pm
Hey guys, all the mods I made are in my Github repository, https://github.com/ElDominio

All code that I added to the Speeduino firmware is in alphaMods.h (fan control mostly)
CAN code is in the CAN ME7.5 directory, hope it all helps!
Great!

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 9:02 pm
by AndreyB
960 wrote:
Tue Dec 11, 2018 8:00 pm
As we now has someone that already have a working PNP, I think we should go straight to making a PCB.
But this was NOT a rusEfi PNP - this was a speeduino PNP, which is a different open source project https://speeduino.com/wiki/index.php/Speeduino

Making a PNP board to fit the original case is a great idea but I wonder who are the people who have
1) time
2) KiCad knowledge

I am afraid that this is where this would fall apart :)

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 10:16 pm
by 960
russian wrote:
Tue Dec 11, 2018 9:02 pm
960 wrote:
Tue Dec 11, 2018 8:00 pm
As we now has someone that already have a working PNP, I think we should go straight to making a PCB.
But this was NOT a rusEfi PNP - this was a speeduino PNP, which is a different open source project https://speeduino.com/wiki/index.php/Speeduino

Making a PNP board to fit the original case is a great idea but I wonder who are the people who have
1) time
2) KiCad knowledge

I am afraid that this is where this would fall apart :)
Yes, I know that is a Speeduino project.
But from that we now know what hardware is needed to get everything working.

That makes the prototyping-step unnessesary.

I have no idea how much work it is to design the PCB in KiCad.

I thought we could just edit the PCB to fit, and add the 121 PIN pattern, but I saw now that Frankenso was a little longer.

If not, the job would have been to add the 121 pin pattern instead of the existing, and mapped all the pins right.


What I think is, a drop in replacement PNP, would be both easy to sell, and also worth a lot more(as a fully assembled).

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 10:33 pm
by AndreyB
For a second lets assume that the board with 121 connector is designed. What would be the next step?

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 10:33 pm
by AndreyB
For a second lets assume that the board with 121 connector is designed. What would be the next step?

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 10:52 pm
by 960
russian wrote:
Tue Dec 11, 2018 10:33 pm
For a second lets assume that the board with 121 connector is designed. What would be the next step?
The next step would be for me to program/set up everything to be fully working.

All can messages etc.

I have to test everything, and then make some maps/base maps for it on the dyno.

So after that, you will have a PNP drop in replacement that would be easy to sell as a finished product.

Most people dont have the skills to solder, or wire a aftermarket ECU.

This is the cheapest alternative PNP:

http://www.maxxtuning.eu/products/engine-control-and-electronics/plugin-ecu/audi/audi-s3-or-a4-18t-me-75-maxxecu-race-plugin-standard

I also looked at the specs for the cpu they use: ARM Cortex-M3 120MHz

Re: 2.0T VW FSI potential test mules.

Posted: Tue Dec 11, 2018 11:14 pm
by AndreyB
960 wrote:
Tue Dec 11, 2018 10:52 pm
The next step would be for me to program/set up everything to be fully working.

All can messages etc.

I have to test everything, and then make some maps/base maps for it on the dyno.
You've missed the part that someone would have to fund the first batch of these boards - that's about $200 just for boards pack of 10 or 20 pieces, with high probability of missing something since there was never a PNP prototype and we are jumping from never running rusEfi in any Audi with ETB and CAN into fabricating a board with connector. Also soldering and etc.

And on top of that who exactly is interested in selling and _supporting_ these PnP boards? My limited experience with Frankenso is that selling and supporting is a lot of hassle. I am kind of on the fence every time if I want to order new batch of Frankenso or not - my primary interest is developing all this as a hobby, I do not see this becoming a job simply because the margins are not that huge - selling ECUs is a hard business, much simpler to sell sandwiches (easy for me say since I never sold a sandwich in my life)

It is my opinion that in order to increase the chances of Jared adopting the PCB for 121 connector and specific case you would have to show Jared a video or a term mule running rusEfi on any hardware - you do not have to buy anything from me, you can make your own proto board with prototyping boards and smaller boards from eBay.

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 12:29 am
by 960
russian wrote:
Tue Dec 11, 2018 11:14 pm
960 wrote:
Tue Dec 11, 2018 10:52 pm
The next step would be for me to program/set up everything to be fully working.

All can messages etc.

I have to test everything, and then make some maps/base maps for it on the dyno.
You've missed the part that someone would have to fund the first batch of these boards - that's about $200 just for boards pack of 10 or 20 pieces, with high probability of missing something since there was never a PNP prototype and we are jumping from never running rusEfi in any Audi with ETB and CAN into fabricating a board with connector. Also soldering and etc.

And on top of that who exactly is interested in selling and _supporting_ these PnP boards? My limited experience with Frankenso is that selling and supporting is a lot of hassle. I am kind of on the fence every time if I want to order new batch of Frankenso or not - my primary interest is developing all this as a hobby, I do not see this becoming a job simply because the margins are not that huge - selling ECUs is a hard business, much simpler to sell sandwiches (easy for me say since I never sold a sandwich in my life)

It is my opinion that in order to increase the chances of Jared adopting the PCB for 121 connector and specific case you would have to show Jared a video or a term mule running rusEfi on any hardware - you do not have to buy anything from me, you can make your own proto board with prototyping boards and smaller boards from eBay.
I agree with you with lot of it.

But the hazzle and profit cant be compared between what you have and a finished PNP.

A fully working PNP would also easily sell for around 1000 $ i belive.

99.9% of the people that wants a aftermarket needs it to be PNP, cause they dont have any electronic skills.

The market is there, as even big companies like Link and MaxxEcu has made PNP units for these.

Also the hazzle will be much less with PNP ECU's.
Because then you only have to deal with those cars/engines you have ECU's for, and not something new every time, that you dont even know if people has wired right.

The important piece is to make sure all the needed hardware is in place. So if a customer has something not working correct it will just be a firmware-upgrade.

I think the right way to go is starting with the ME7, because there is so much knowledge and information about it.
That is also the ecu that is the same in so many cars for so many years.

I really belive in this project, but it's not mine and the KiCad/PCB part I dont have any knowledge at.


Also we have the original ME schematics/components i posted earlier, that can be used to make sure nothing is missing.

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 4:05 am
by Abricos
960 wrote:
Wed Dec 12, 2018 12:29 am
I think we poorly understand What means plug and play ...
in my imagination PNP looks like this
Image ImageImageImage


if you want that your ECU looks how you want and fit in a standard case from ME7.5 you need to do your own design !!!!

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 5:29 am
by 960
When first making a PNP, why not make it a seller?

This is a ordinary ME7.1.

With some other stickers, 6000,- :-)

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 5:40 am
by AndreyB
c'mon, rusefiis light years behind bosch interms of models and features. Not even funny to compare.

Time to research is more important than money probably. Dp you have time? Get a frankenso and start wiring and evaluating whats missing.

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 7:28 am
by NormanAlphaspeed
russian wrote:
Wed Dec 12, 2018 5:40 am
c'mon, rusefiis light years behind bosch interms of models and features. Not even funny to compare.

Time to research is more important than money probably. Dp you have time? Get a frankenso and start wiring and evaluating whats missing.
I have a Frankenso and I never got around to build it hehe, only now I'm starting to dabble in the STM line of micros (today I got interrupts to work on an STM8, that's how behind I am lol), and I hope to be having the Frankenso built sometime soon so we can put it on our test car so I can finally start getting my feet wet in STM32/ChibiOS/whatamieventalkingabout!

And sure, it is behind, but with the core it uses and enough community brainpower behind it, it can sure be decades instead of light years! lol

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 9:00 am
by 960
Well, I dont think it is much more advanced.

I have all the files for it.

This is a export of the map list:

Also hidden in the binary says it's simple :-)

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 10:37 am
by kb1gtt
Lots to catch up with here. Sorry for my late replies. Normal life constraints take most of my time. For me this is a hobby, the day job and family obligations come first. I designed Frankenso 10 minutes at a time, each night after the family went to bed.

My concern with making a PNP layout is the high risk that some key circuit will have some key layout problem. When this happens it basically means you start the layout over again from scratch, which is very time consuming, and can be reasonably expensive, as you have to spin the entire board. My suggestion to do some kind of prototype board(s) is to ensure the layout is well understood, and to prevent expensive and time consuming re-designs. That said, if there is a Speeduino PNP reference, that would be helpful. Is that layout posted some where? Are there pictures of the finished PNP some where? Can I reference it? Right now we do not have a circuit for the high pressure injectors. We have seen the ETB circuit work, so I'm confident with that, but we have not seen the high pressure injectors working.

If I do a layout, it would take like a month or so, and then it typically takes several weeks for the boards to be MFG'ed and assembled. Would a February-ish time frame work for testing hardware? Also it would be a good idea to write down a list of things to check. Something like the below, but with more details.
1. Ohm, the inputs, to check for dead shorts and opens
2. 12V power with bench supply current limited and validate 5V on the output
3. blah blah
12. Cycle ETB open and close, tune PID's accordingly
13. Pulse injector 1, then 2,3,4, etc.
14. etc etc

Also what features are needed or desired? AKA is AC a circuit that should be considered? What about extras like NOX, or similar additions? We may need more than just the high voltage injectors tested out.

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 9:21 pm
by 960
what is the actual size of frankenso.

Some places says 100x100, but the drawing is bigger?

Re: 2.0T VW FSI potential test mules.

Posted: Wed Dec 12, 2018 9:26 pm
by AndreyB
960 wrote:
Wed Dec 12, 2018 9:21 pm
what is the actual size of frankenso. Some places says 100x100, but the drawing is bigger?
Info added to https://rusefi.com/wiki/index.php?title=Hardware:For_Sale
Frankenstein is 100x100, not Frankenso.

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 2:52 pm
by 960
Ordered now :-)

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 3:09 pm
by 960
All Signals Description:

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 3:38 pm
by kb1gtt
There are lots of signals in there. I think they are all discrete type signals, with the central processor as the place that keeps the security. Is it practical to simply replace the central processor?

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 4:54 pm
by 960
kb1gtt wrote:
Sat Dec 15, 2018 3:38 pm
There are lots of signals in there. I think they are all discrete type signals, with the central processor as the place that keeps the security. Is it practical to simply replace the central processor?
That was my first idea actually.

Just make a adapter to replace the Infineon C167.

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 4:59 pm
by 960
In the A2l files there also is a lot of information.

And decompiling the hex files in IDA gives the rest

Re: 2.0T VW FSI potential test mules.

Posted: Sat Dec 15, 2018 5:02 pm
by 960
960 wrote:
Sat Dec 15, 2018 4:54 pm
kb1gtt wrote:
Sat Dec 15, 2018 3:38 pm
There are lots of signals in there. I think they are all discrete type signals, with the central processor as the place that keeps the security. Is it practical to simply replace the central processor?
That was my first idea actually.

Just make a adapter to replace the Infineon C167.
But there are also spesifications for each and every signal, so sholdn't be a problem