'Eric Wiklunds Volvo 142 Turbo'
This car shares a lot with Jonas car: MembersPage/MattiasSandgren/VolvoBtwenty
Basically the same engine configuration, only this one is turbo-charged.
- Stock B20 block
- 8-bolt crank
- forged KS pistons
- F cam (300 degree duration, 111 degree lobe separation, 11 mm valve lift)
- head ported by Jens Gustavsson
- 3K K27 turbo (51 in/65 out)
- home made turbo adapter to stock N/A manifold
- home made wastegate
- home made intercooler (from a large diesel truck)
- MSD/Holley fuel injectors 895 cc/min @ 3 bar
I repeat, the analog inputs only work as inputs for switches due to the interference present on all of the ADC channels. At least that's better than what it used to be, but analog inputs are not worth connecting. Any suggestions on what could cause the interference?
I must point out that the MCP3208 was replaced and that I assume that the new chip is working as it is supposed to. Why does the ADC then not work?
With this new chip we have only applied +5V to the inputs through a pull-up resistor and of course with the AD597 EGT chips.
Is there something besides the MCP3208 that potentially gets damaged if excessive voltage is applied to any of the ADC inputs?
The MCP3208 was replaced and it somewhat works. There is a slight interference in the ADC on all 8 channels. When I measure either EGT input, the voltage on the input looks completely normal (stable value between 40-50 mV room temp) but the calculated temperature gauge in the datalogs report a varying value. The error is clearly induced within the MCP3208 itself as the input voltages are nice and clean. Here is a datalog:
Note about the datalog: Channel 5 (misc 1) is connected with a 2.7k pull-up to +5V. I applied a soldering iron to the EGT sensor after a few seconds into the log to show that it works.
The inputs are still very useable as digital input switches for launch control, but as analog inputs they are of some concern. All soldering is double-checked. Any pointers to possible causes for the interference are welcome.
Make sure the mcp3208 never gets higher voltage than 5V. We've seen similar symptoms (inputs still useful for digital inputs, but otherwise quite broken) when someone connected shiftlight LED between 12V and an mcp3208 analog input (he also used so small resistor value that the big LED was destroyed in a few minutes when we connected it to a p259 output).
None of the MCP3208 analog inputs work, this explains the problems with the EGT inputs. The chip seems to have +5V and ground, but MegaTune reports randomly flickering values between 0 and 5.0 Volts in the datalogs.
The chip has probably been damaged about two years ago when we first installed everything, the harness was then redone because it was only used to control ignition and datalogging. I don't think there is reason to suspect anything wrong in the harness at this point.
It is very likely the problem is with the MCP3208 chip?
- the mcp3208 is very easy to damage, eg by applying higher than 5V to the input (+14V or static electricity)
- the AVR is unlikely to be damaged (though the only chip that talks over SPI is the MCP3208)
The engine now has a rebuilt bottom-end and there must have been an incredible amount of blow-by on the old bottom-end because VE increased almost 20% at higher loads, with no other changes. The turbo is much more responsive too.
The injectors have been wired in series with a 6.8 Ohm 50W resistor for each injector and a transient protection diode (Zener) has been added to the flyback wire. The total opening time has increased for a given effective pulse width but the accuracy seems to have increased due to more effective flyback control.
There is still problems with EGT readings, I will look into that very soon as the car will go on to a dyno in a few weeks and then put into storage for the winter.
The car ran 12.34 sec @ 174 km/h at the "local" dragstrip in Piteň, north of Sweden. A very nice result for such a simple setup. 110 kPa boost and very little tuning. [Movie here]
Genboard specific configuration
This is a v3.2 with production number 126.
Both triggers are configured for Hall.
MCP3208 channels 5,6 and 7 are wired to EC18 pins 6, 12 and 16.
EGT - only EGT #2 is wired in (MCP3208-4 ?), see page about EGT trouble: MembersPage/MattiasSandgren/EricsVolvoTurbo/EGT
Using power flyback for the injectors with a 30V transient suppression diode.