Printable View
When I looked at the spec sheet of the MC33660, it's 99% identical to the OBD2 chip I already used, which isn't good enough for the cars that are having problems. The transistor/resistor circuit has proven to work better than the OBD2 chip, although not 100%. The USB OBD2/MUT is said to work 100%, I just need fastbikes76 to confirm his cable uses the same LM339 chip.
Here are some pics I took
Attachment 5947
Attachment 5948
Attachment 5949
The chip is LM393 and then it looks like either 38M or 3BM below it as well as STK4.
If you need any more pics let me know :)
Can you take a sharper picture of the bottom pic? Your USB OBD2/MUT cable is different from mine, so I just ordered another cable from ebay in hopes to have the exact cable you have.
Your cable uses transistors/resistors design, a comparator chip, and FTDI chip (?). Also interesting is that they had to use two beefy resistors, it seems they are pulling lots of current.
HI, I tried too take some clearer pics but everythin is so small i couldnt get the camera too focus !! I will email some hi res pics and see if you can see what you need.Quote:
Originally Posted by BlackStealth
On a side note i managed to update the later software and the delay version too. I am using the car tomorrow night again so will see what happens. for now the connection to Evoscan is still strange and will only connect as per my 10 steps i keep banging on about..however thats not really an issue for me as i know the routine now and can get it too connect and log just fine.
Regarding the delay feature...is that for the ecu/logging corruption and what should i set it too ?? I dont know what it does or what the units are lol.
We will get there in the end :)
I just found some high res pictures of that cable, so you don't have to. I'll start drawing a schematic of it.
The delay version are in milliseconds. First try 5ms, and see if it makes any impact or improvements.
Hallelujah brother !!!!!!! We have some success :D
I updated with the later version for better reconnects as well as adding the delay software for the logging. Today I had ZERO disconnects...yup, not a single one ! and when it did disconnect from stalling, it did reconnect itself. SO cross that off the list ! :coo3l:
Now regarding the logging....I tried various ms delays and here's what I found
5ms...fractionally better but still very corrupted
10ms..less corruption and moved higher up to early 4000's rpm
15ms even better still and corruption now only above 5k
20ms is about the best I can go before going back down the slope again. very very minor corruption and although not ideal I can do a clean enough log to check boost and afr's
So that being said I can start using this box properly now !! I have a 45min each way drive tonight so I will see if I have any freezing or drop outs but for now I would say that's done and dusted.
All we need to do is sort the logging side a bit better and we sorted :coo3l::coo3l:
That sounds great I'll have to try this update right away!
With firmware 3.11 my car starts to drop out even before I reach 3000 rpm:
http://www.youtube.com/watch?v=iwYUDOG5wog
Screen refresh set to 5
http://www.youtube.com/watch?v=0ZBB_2gTf_A
Screen refresh set to 60
The delay setting definitely helps, flashed it tonight and with the delay set to 10 the unit seems to work up until about 4600 rpm.
http://www.youtube.com/watch?v=oWsHcb5NgRQ
I tested setting the delay to 20 but that would cause the unit to drop out before 2000 rpm.
Had to take care of some family matters after that so I didn't have time to test any other delay values but it seems we're on the right track here and it's nice to see some fotware development to tackle the issues we've been experiencing.
I suspect that difference in delay values experienced by me and Elton could be due to the fact that he's using the built in OBD chip and I'm not.
Before my garage time was up for today I did try to log via CoolTerm using TxtMode1 at 57600 baud and it worked flawlessly. My laptop connected just fine without any magic. I'll see if I can buy an Evoscan license and test that tomorrow.
We are definitely heading the right way now.
Tonight I used the car all night and didn't have a single disconnect, and it seemed to work pretty well.
I think it's mainly the OBD to serial cable that's giving a the grief. Once black stealth decodes the other cable he ordered in sure we will see even less glitches.
I am far happier now than I have ever been with it :) :)
I looked at some 94 ECU MUT datalogs and it revealed something interesting. When the ECU hasn't had a request for data for a while, it will reply back with the answer almost immediately (0.1 mS - 1mS); however if the ECU just serviced a request not too long ago, it will wait ~11.4 mS before responding. By adding those extra delays, the ECU will reply back with an answer almost immediately making it much harder for noise to make an impact.
fastbike76, it's going to take 2-4 weeks for China mail to arrive here with the USB OBD2/MUT cable; you can eliminate this wait if you sacrifice your cable. The components need to come off the PCB in order to see all the traces, a hair blow dryer will melt the solder, and the parts can be scraped off. Or we can wait until I get my cable.