I, Mikael Pihlblad (vmsfarbr in #avrboard), am currently converting my Audi 100 Turbo Quattro -90 into running on a GenBoard/VerThree. (3.0 that is)
Ignition advance doesn't move with spark table. Looking in megatune it looks like it is using different advance when changing the table, but with a strob-lamp the ignition doesn't move a single degree. If changing ign_tdc_delay 10 deg the ignition moves. Any ideas?
Yes, FF doesn't mean 'disabled' in the knock section. I don't have much more information but maximum knock advance and knock retard should probably be 00 instead of FF.
-- Problem solved, Cylinder count was set to 1 instead of 5, changing this got rid of the problem //Emil
config is here: http://www.garmic.nu/100tq/vems/config.txt
my_make is here: http://www.garmic.nu/100tq/vems/my_make
Running on firmware STABLE_1_0_12 and using a 5 slot hall distributor as trigger.
Cars starts and runs smooth but with a static advance of around ~20 deg
When changing in the spark table we changed a lot of values at the same time not only current map/rpm (to avoid faulty rpm/map values)
Old WBO2 question
Now WBO2-heating seems to work. Added missing R98 10k resistor to nernst OPA. Now Ri readings are good.
We adjusted wbo2_nernstdc_target to C0 to get nernst voltage to be even close to 4.45 volts, with this value we get nernst voltage at ~4.35, so we should need to increase wbo2_nernstdc_target even more.
Or is anything else wrong? Everybody else seem to run with 8D here... It can be OK. Maybe your nernst amplification is lower than (270k+75k)/75k for some reason ... Just check if the voltage mean value after the nernst amplifier matches configured wbo2_nernstdc_target.
Except for this the sensor maintains Ri in bound and seems alright...
Not yet latest question: (WBO2)
When connecting the sensor and running the mde02 command the heater pulses starts. The sensor got warm and then the heater pulses stopped. Is this correct?
And it starts again after a while? This is just fine since it maitains the temperatur of the WB02
No it does not start again... The reading of Ri via wbo2log.pl says a value between 2-20 not anything near 1200 (as the documentation says) How do I debug this further?
Measure the nernst signal with scope. Before the OPA and after the OPA (at ADC input). Is the nersnt OPA doing it's job? (amplifying * appr.4 referenced from 5V). Check that the squarewave signal mean value and amplitude conforms to your nernst DC and Ri_target settings. If the signal rails, the measured Ri will be small, and the heating will stop. Stopping for the heater is correct behaviour, you need to see why the signal rails (if so).
The car needs to pass the 4 gas exhaust emissions test. To do this we need to get EGO-correction to switch fast around lambda 1.0, since we don't want it to run lean either. Now the EGO-correction doesn't seem to be fast enough. (Jonas, feel free to fill in)
The threshold values are:
- CO: maximum 0.5
- CO@2500 RPM: maximum 0.2
- HC: maximum 100 ppm
- Lambda: 0.97 - 1.03
I think it's impossible or very hard to pass recent NOx and CO and HC treshold at the same without a catalytic converter.
The plan is to make it with the catalytic converter, it is currently mounted.
The incredible-EGO controller can be made fast with proper configuration (look for speed_limit = config.ego_pid_kp; in ego.c).
Seems like the incredible-EGO is just for WBO2, we are currently running with GenBoard 2.2...
Took the car out for a little drive today, runs really smooth and was incredibly easy to start. So this fall the car will be taken apart and get some rebuilds done... GenBoard/VerThree with ignition and WBO2 is a must for next season...
Went to bilprovningen to "borrow" their 4 gas meter to do some emissions testing. This test will be hard to pass... When running the car lean we get CO values that are in bound, but lambda is much to high (~1.20). When running richer CO starts to go high (>2.0). Any suggestions? Will a more precise EGO-correction that switch fast around stoichiometric do the job?
Changed dead old NBO2-sensor for another old, but working one (from same Volkswagen Jetta that donated its hydralic valve lifter earlier).
Mounted the catalytic converter and did some fine tuning. Enabled EGO-correction and adjusted the values for that, works very nice now, when having a quicker sensor. The car runs almost as smooth as with the original K-jetronic :)
After some minor problems with running on Emils GenBoard/VerTwo compared to running on Jonas, the car is up and running again. The current plan is to get the car to pass the smog test...
Worked on building a new config with lots of comments, based on Jonas, Miks and Daves configs. Made a first ignition map based on dumps from MAC-11 ECU.
GenBoard/VerThree work isn't advancing to fast =)
Tonight we uploaded new firmware and config, got the InputTrigger working with the built in InputTrigger Stimulator function, and found out there must be a way to say which injectors should be driven... Then we went out and did some ethanol tuning on Jonas Peugeot 309 gti.
WBO2 Question: Are the heater pulses supposed to be stable? When connecting the sensor and running the mde02 command the heater pulses started to come quite irregular. The sensor got warm and then the heater pulses stopped. Is this correct?
And it starts again after a while? This is just fine since it maitains the temperatur of the WB02
No it doesn't... The reading of Ri via wbo2log.pl says a value between 2-20 not anything near 1200 (as the documentation says)
Setting config values with Menu System: mttgXXcYY where XX is address and YY is value
We started measuring between heater(-) and GND which read 3.993 volts. After that we measured between heater(-) and heater(+) which read 0.039 volts.
ISP and Firmware upload:
ISP programmed the CPU and uploaded new firmware. Finally we are back on track. We had alot of problem with ISP, because we tried to use the miso/mosi pins instead... but with some hints from marcell we were back on the track...
This time after ISP:ing the bootloader, we uploaded the new firmware with 19200 instead of 57600, worked much better!
> perl bin/prog.pl vems.hex Ew 19200 :COM1
Programming error when uploading new firmware, time to make an ISP-cable...
I strongly recommend using only 19200 baud, especially if you have problem. Ms windows is especially broken in port operations (it often crashes for me when using parallel port at the same time). If it verifies OK, than it should be OK, even if it verified bad (probably due to comm error) earlier.
Problem with hall sensor solved, realigned the engine once more (fifth time I think... Never had a problem with this earlier?). Now I aligned the distributer based on the TDC sensor on the flywheel with the gap in the hall sensor instead of based on the markings on the distributer. No more fault codes!
The engine started almost immediately. Just a little adjustments on the afterstart enrichments was needed.
Of we went tuning! We drove over 100 km tonight, and I'm really satisfied with the result, for being done only with "seat of the pants" tuning...
Logs, configs and pictures can be found here: http://www.sdf.nu/~kakmnstr/msavr/audi100tsq/
Problem: The original ECU says fault code 2113 (Hall sensor error or misalign). We tested the hall sensor and it seems to be Ok. We also realigned the distributer (with hall sensor inside) so that it gives us sparks. But if we crank again it want give us sparks, if we try once again it does. Strange! Any suggestions? Low voltage problem?
MembersPage/JonasAndersson helped me with the soldering of the tiny components today! Almost done except filter for the WBO2.
The fuelrail will arrive tomorrow! Also the v3 genboard hopefully will arrive this week. So it's getting closer now...
One step forward and two steps back :(
The fuelrail was gas-welded a little to hard, the injector seats wasn't round anymore so the O-ring seals wouldn't stop it from leaking, some weldings were leaky too, big disapointment, since taking so long time... Should have done it myself in the first place... Today I've ordered a CNC Machined fuelrail made for the 10v 5 cylinder engine. Feels like a fire hazard isn't far away if one tries to "save" money on stuff like that...
The only thing missing for a test run is the fuel rail. I am working hard on getting hold of it today... I want to start the car :)
Soldered the D-sub connector for the GenBoard v2.2 card. Talked to Johan who is welding the fuelrail for me, and he said I could have it finished tomorrow! So maybe maybe we can do the first fire up on late sunday evening :) To night we will do some dry testing with the genboard v2.2 card to see that we get rpm and stuff.
Cranked the car on the starter engine and scoped some signals. The TPS seems to have one pin brooken. Fixed the throttle wire and the rest of the pressure hoses.
Ordered the GenBoard/VerThree with accessories from the webshop today.
Got a throttle body from a Ford Sierra 2.0i at the local junk yard. Not to big, but round at least! Which helps alot! It also hade a small and very nice TPS mounted.
Made an adapter in 4mm alu to get throttle body to fit with the original intake manifold, since the original throttle body has a 3-bolt fitting and the ford one has a 4-bolt version, this was necessery to get it on at all. This helped with clearance between throttle body and fuelrail with about 20-30 mm. Also made up some fuel injector cables and cables for the TPS.
Also talked to the friend how is gas-welding my fuelrail, he said he'll get it finished this week... I really hope he does :)
Got the injector cables through the torpedo wall today, wasn't easy and I got alot of scratches on my hands... But now it's there! Also moved the original ECU up into the "glove compartment"? with the original wiring moved there to, which gives us a little more space and longer cables to work with...
Routed the fuel hoses and some of the pressure hoses. Mounted the intake manifold and did some measuring.
- 5 cylinder MC mk2 with dual knock sensors.
- KKK K24 turbo
- Exhaust without silencers.
- Alu profile fuelrail from Javad @ 034efi.com
- Injectors from Volvo T5R (firstname.lastname@example.org bar)
- Fuel pressure regulator from SAAB 9000 (standard Bosch 3bar)
- Conical airfilter.
- Throttle body from Audi Urquattro.
- Construct new boost control, better than last years.
- Get the car to run on the GenBoard/VerThree instead of the GenBoard/VerTwo
- InputTrigger/AudiTrigger and start playing with Ignition
See also: http://www.garmic.nu
- Hi nice work swap engine for the 20vTurbo 3b or a V8