My LCD screen project, boost, WBO2, Knock, etc

[JasonY]

Any reason it would have connection issues with a 98 SL ecu(for flash tuning) when it worked perfect on the stock and overclocked 97 ecu's?

We have one of the OG units w/o any updates.

Jason

[BlackStealth]

Any reason it would have connection issues with a 98 SL ecu(for flash tuning) when it worked perfect on the stock and overclocked 97 ecu's?

We have one of the OG units w/o any updates.

Jason

Make sure config menu ECUclock=100% for non-overclocked ecu.

Unlogic: I've ran my LCDBC for more than 10 minutes without it losing connection on the bench. So there's no software issue. Strange how yours disconnects after 5 minutes.

[Unlogic]

Things is going to quite a long post. I won't receive the new MAX232 chips until tomorrow or Wednesday however I went out and got myself a new multimeter so that I can see what was going on with the old chip. I also mounted the unit back in the car again and did some driving to provide you with more info regarding the disconnects.

Let's start with the serial problems. The new multimeter i bought (UNI-I UT61E) has a RS232 interface for logging so I used that to test if my FTDI-based USB-to-serial adapter worked and it connected just fine.

Next I rigged up the multimeter to check how much current the LCDBC was consuming. Without the MAX232 chip installed it consumes 49 mA:



With the MAX232 chip installed but nothing connected to the serial port it consumes 63 mA:



But when I connect the serial port to a computer it immediately goes up to 400 mA!



Even if I disconnect the serial cable it stays pegged at 400 mA. To get it back to 63 mA I have to cycle the power to the unit multiple times and sometimes even remove the chip and reinstall it again.

To be absolutely sure my USB-to-serial adapater wasn't faulty I went to my server room and dug out this old Dell PowerEdge server with built in serial ports but I got the same results. As soon as I connected the serial port and powered on the server the current went right up 400 mA (the server is powered off on the photo below).



With that tested I removed the MAX232 chip and installed the LCDBC in the car again to test the disconnect issues both me and fastbikes76 have encountered.

If I turn on the ignition but don't actually start the car the LCDBC remains connected just fine. I rigged up a camera and filmed it for 18 minutes without it getting disconnected a single time.

However if I start the car it's a different story. I did some test driving and from a cold start it took 8 minutes before it disconnected. Anticipating this I had removed the radio and pulled out the cables so that I could disconnect the LCDBC from the OBD2 adapter cable while the car was running just as a test.



When I disconnected and reconnected the OBD2 adapter while the LCDBC was in "disconnected mode" it started flashing the overboost alarm for some reason but it still didn't reconnect even after multiple tries with various delays. My overboost alarm is set at 1.15 bar and the car is just idling at a gas station on the video.



So I restarted the LCDBC and drove on, 6 minutes later it started slowing down and finally disconnected again. Restarted the LCDBC again and this time it took less than a minute before I noticed corruption on the max knock/boost screen however the unit stayed connected for the rest of the 4 minute trip home.

Here is a video showing a random pixel on the max knock/boost screen together with a false knock reading:



I'll report back with more information when I receive the new MAX232 chips but I hope the information above is useful for you when it comes to solving the disconnect issue which is really crippling the usability of the LCDBC right now.

[JasonY]

Todds is doing the same thing with the false knock/overboost readings. Wont connect at all w/o a restart of the car. Then holds fine for 5-10 then spazzes out. Worked fine with the non flashable ecu for a few years.

Jason

[BlackStealth]

In the Peak PSI, and Peak Knock menu the adjacent 8 character empty space will draw Boost vs RPM graph -- so that's where your mysterious graphics are from.
When you push down the button it will reset the values to zero and clear the graph.

Unlogic there's no point in putting the defective MAX232 chip back in -- when it locks up, it is power hungry and is unable to go back to a normal state. Also when you physically disconnect/reconnect the OBD2/MUT cable you're introducing electrical noise briefly; the analog to digital conversion is picking this noise and creates a false overboost message. I could make the LCDBC software reject the first overboost reading if this a problem for you.

I need to know if v3.11 software with it's longer pause can successfully reconnect.

Without having access to a car with these connection problems, it's very difficult to solve this problem. The only option I see remaining is to copy what Shiver91-vr4 did, and make the low signal even lower.

[fastbikes76]

Ok so too add down more info to try work what happening I noted the following today...

I bought the correct ftd serial to USB cable and still can't connect to evoscan through the LCDBC , I can go straight into evoscant via OBD port directly.

After evoscan has tried it's 5 reconnect attempts it then disconnects the LCDBC which itself won't reconnect until the ignition is cycled.

I have experience slowing down to quite a degree once, but also noted I have to have my screen refresh rate right down at 2 for it too show me any real time info... Otherwise it's just like a badly dubbed movie lol.

I also don't get a proper boost vs rpm graph, I sometimes get quite a few dots and can make out what my boost was, other times there may only be 2 random dots ?

The absolute main issue for now is the random disconnects, I know the new software may make it reconnect better .. But it shouldn't be randomly disconnecting in the first place as the log file would be corrupted every time I guess ???

Cheers
Elton

[fastbikes76]

Just been trying some more and get this from evoscan

ImageUploadedByTapatalk1378221996.284574.jpg

ImageUploadedByTapatalk1378222007.507314.jpg

And these are my settings (which had always worked )
ImageUploadedByTapatalk1378222041.449370.jpg

I set the screen refresh back to 20 as if I blip the throttle to say 1400rpm or so, the rpm will barely move off idle on the screen as if I haven't touched the throttle at all. The only way to get reasonable screen info is to set the refresh at its lowest setting.

It's is starting too seem like we are all suffering the same issues of..

Screen slowing right down
Random peak knock counts of 27-28 ( NO knock on evscan)
Constant disconnects
Unable to auto reconnect unless ignition cycled
Unable to connect to evoscan via LCDBC

Hopefully we an all work together and add as much info as possible to work towards a resolve

[Unlogic]

My MAX232 chips were miss sent so they'll be a day or two late. However I went and purchased a DB9 extension cable and some transistors today. I'll see if I can build a filter using the image you posted earlier.

Can I replace the 2N3904 with a BC337 and the 2N3906 with a BC327?

[Unlogic]

Screen slowing right down
Random peak knock counts of 27-28 ( NO knock on evscan)
Constant disconnects
Unable to auto reconnect unless ignition cycled
Unable to connect to evoscan via LCDBC

That list really sums it up.

Seems like your expansion serial port doesn't want to play ball either just like mine.

[fastbikes76]

Ok so here is something totally random !!! After trying 20 times earlier I came out to try something else and no evoscan connects every time !!???


There is a difference in what evoscan is displaying and what the LCDBC shows.

Coolant on evoscan is 89c and 83c on LCDBC
Air temp is 38 vs 41
Timing is upto 3 degrees out on idle but seems to match at higher rpm