venderbroeck

Thanks everyone ^^ I have a micro, the sd-card module, and a RTC in there. Also a little something for the k-line interface, and you can see the lm7805. I don't have much spare time as I'm also building a logbox for someone. That's why it's still looking like that. Most of the work was reverse engineering the xdl format, and programming.

It's been a while, but I've been working on a new trinket in my spare time. Here it is: This gem is a little rough to look at still, but it makes up for that in functionality. Everytime I start my car and a sd card is inserted in the slot, it will: -Create a file with the current month day hour and minute as the filename. -determine the size of the frames that are coming in. -record the frames on the sd card -update the "last modified" timestamp of the file after every frame. And the thing that really makes this awesome is that it creates the log in the correct tunerpro .xdl format. So it can be loaded straight into TunerPro. I'm thinking of adding a usb port to make it a direct replacement for the vag com cable. I'll probably also add a piëzo speaker to signal for knock events and stuff like that for some added functionality. I'm still contemplating wether to put it all inside a obd plug housing, or to make a seperate housing with an optional obd extention attached to some screwheaders. After driving around with it for a bit I can say it really simplifies things. Just stick in the card, start the car and just forget about it. Leave it in for a week or a month, and it produces a neatly ordered row of logs for every drive you did.

? German850r: thanks, glad you like it.

Ecu only goes upto 7650. Tacho goes over 8k. I can imagine there's a way to create an internal logable parameter that goes over 7650, but as its all timer and interrupt based it might be hard. It depends on how overflow is handled. Maybe we can catch the overflow before its truncated and present it as the msb of the tach signal in the logging routine. Not sure about that though.

Tach signal is freq based afaik. You can hook it up to the ECU, but without additional hardware it will be hard to calculate rpm by polling the adc channel. You could use a freq to voltage converter to convert to 0-5 volts.

I think that kind of behaviour would definately be suppressed / eliminated. No vac leaks you're aware of? I know when shifting at 6k+ rpm with m43 it displays this behaviour aswell. It even rises in rpm sometimes when disengaging the clutch. This mod takes care of that.

That's the stock setup you're hearing. Goes to show these cars sounded pretty good from the factory to start with.

I made a little video where I explain its function a little more.

I've sold some of these in the Netherlands. I can make some more for ppl that are interested. It's m43, and m44 only though. No me7 (totally different throttle body setup), and I didn't look at the fenix ECU's yet.

It manipulates the idle control valve based on a few parameters. It also makes the ECU think it's still in full control to avoid throwing any DTC's.] Of course to get the 'clutch pedal disappear into nothingness' feel you need to shift consistently at the same speed. With a little practice this is easy to master though. I have it set to drop pretty hard as I like to shift quickly. edit: I'm going over your maf table atm btw ^^

Something a little bit different for this post. I always thought the 'throttle hang' while shifting, and the inablilty to engine brake effectively was hugely annoying. I always had to either shift slowly to shift up smoothly, or let the clutch deal with the elevated RPM's and try to live with the clunkyness of the shifts. With the swap to m44 this got alot better, but still not nearly good enough for my tastes. Naturally I got fed up by this. I started working on a solution almost 1,5 yrs ago, and I've since perfected it. from this (first quick & dirty prototype with an arduino mounted on top of it): to this (latest version with nice bosch plugs): I call it the 'RPM defloater' Its settings allow me to precisely control the RPM dropoff speed on upshifts. I have it set in such a way that the revs are matched on my normal upshifts and the clutch pedal just seems to 'disappear into nothingness' when I let go of it. I just don't feel the clutch engage, because it doesn't have to synchronise the crank with the gearbox. No clunkyness, but smooth and fast upshifts. When I let go of the throttle, my car will instantly start engine braking. I wonder why Volvo ever implemented the RPM hang feature (supposedly to aid shifting, or to keep the turbo spooled), but I'm very happy with this little mod.

Thanks The pins we used aren't populated in a stock MT car. (I believe some of them were AT related). We used some pins from an old ECU connector to populate them, and ran the usb cable connected to them through the firewall.

I have contemplated for a while to start my own build thread, and finally decided to do so. It probably wont be a very fast moving build, but I think there might be some interesting things to come. I've been (and still am) pretty busy with the ECU software side of things, so I tend to neglect my own car a little. I'll just post up the mods as they happen. Some of the current specs of my 1994 Volvo 850-t5 (MT) -It's very onyx green -stock 15g turbo. The standard conical flange has been replaced with a straight one. -white injectors -M44 swapped with ongoing modifications/tuning. -stock intake system. This one wasn't converted during the recall back in the day. This means it still has the stock "rip kit" (although the rip kit is actually what used to be stock) -stock exhaust system -suspension was replaced by previous owner with something rather firm. Handles nicely. Some things that are in desperate need of attention: -rear crank seal leaking badly. -Clutch is worn and needs replacement. -it needs a cleaning Tuning the thing is alot easier with an Ostrich 2.0 rom emulator installed. So to kick off with something more interesting than my rather stock specs, here's how I made that happen: I first clipped all the legs of the flash chip, and removed them one by one with a soldering iron. With a little tin on the tip, you only need to touch them lightly to release them from the board. After some cleaning with desoldering strip, it looks like this: I positioned the Ostrich to where I wanted it to go, and pre folded / measured the floppy flatcable I was going to use. Then I cut one end of a floppy cable, and seperated the wires. I left them about 5 cm longer than required, so I could cut them to length and strip them while soldering. After this soldering commenced. Using liquid flux helps alot with this: I cut one of the original posts of the ECU to the proper height to support the ostrich. I then unscrewed one the the screws of the ostrich, and used a longer screw to attach the ostrich to the post. Then I then soldered some wires to a usb connector I had floating around, and soldered the other ends to some unused pins on the ECU connector. (Don't forget to cut these pins, I didn't and the ostrich didn't connect to the pc). This is what my dad did in his car, and it seemed very convenient. Then the moment of truth... It fired right up after turning the key. After cutting the aforementioned pins off the ECU pcb, usb connected right up as well. Very happy with the setup, can recommend to anyone who wants to do some tuning of his own. The ostrich looks a little scratched. It was removed from another ECU where it was glued in, but it functions perfectly.

Hmm, maybe easier at first. But the way I see it, tuning with a suitable maf, and the right table isn't all that hard. I think the maf system is generally poorly understood / implemented. I get the impression people see it as a bit of a black box while it can make your tune better if you know how to use it well. I also think the problems lots of people are having originate from having an undersized maf which then proceeds to max out. It takes a bit of work yes, but refinement, consistency, and increased drivability are big advantages of the maf system in my view.

With the load maxed as it was (much bigger maf & accompanying scaling) you don't have any control in the high load/ rpm ranges. wether it's load 18 or 12.25, to the ECU it makes no difference. That's a problem. With the new scaling each cell in each table is reached in a unique situation. This provides you with the control to substantially improve lower load drivability, and still tune for maximum power. In all gears... not just one or two. It takes a bit of effort, but you end up with a much smoother running engine in all situations, instead of a one trick WOT pony with a tiny little maf and maxed out wot sliders.

How did the clutch hold up before the tranny gave way? That kevlar disc looks an awful lot like my fibre-carbon one I just realised.

How did the car feel during this pull?

Losing info by posting alot of info? Oo

I think it's using a shift register to handle the different coils. This would mean that it can handle only one coil at a time, and no overlap would be possible. It could be made to work with software alterations, but this would not be easy. It would have to set the next coil in line after the current charging one to high at a calculated time before its turn in the shift register. Essentially, it would need to handle the 'begin charging' point of 2 coils at the same time. The firing is unaltered of course.

Hmm there seems to be some ambiguety about the dwell time specific to those coils. It seems like nobody tested them. They seem to have protected the coil from overheatingfailure caused by to much dwelltime. It just sparks when reaching 8A of coil current. Of course this is unwanted behaviour, as this will mess with your timing. I saw someone posted a stock table for those coils. I would probably use that, if I had no testing equipment. Just have to convert from ms to crank degrees.

Looks tidy, did those coils come with dwell times?

Hmmm seems like I'll be checking mine before it goes in.

clogged oil channel?

Glad it's ok Maybe something to do with the oil squirters?