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

Wow nice install. If you can provide a video that would be great, I'd like to see the corruption, as it might give me an insight.

I can add a "software reboot" option so you don't have to do a physical power reboot, but fixing the auto-reconnect would be a better choice. The auto-reconnect feature works fine on the 99 USA ECU when I tested it, however it may be that the other ECUs require a longer pause of silence before attempting to reconnect.

Edit: I just checked my software code, and currently it waits 0.3 seconds before it attempts to reconnect. By doing a hardware reboot, there is at least a 2 second pause (and that is reported working). I will make v3.11 that includes a longer pause.

Longer autoreconnect pause version:
http://lcdbc.xp3.biz/updates/UPDATE311.zip

Quote:

---

Originally Posted by Unlogic

I'm seeing similar issues as fastbikes76 on my 1996 eurospec TT. The unit works fine for about 5 minutes then gets disconnected from the ECU and no matter how long I leave it on it never manages to reconnect. However if I cycle the power to the LCDBC with the engine still running it reconnects just fine again until it disconnects again after a few minutes.

I'm also seeing the same false knock readings of up to 27-28 counts during light load and cruising but it's not setting of the alarm which I have set at 8 so I presume it's false knock.

I've noticed that the unit randomly slows down before disconnecting.

If it's any help I could film the display while I take a short drive, showing you the whole cycle of power up, successfully connecting, working fine, slowdowns, graphics corruption, disconnect, failed reconnect attempts, power cycled, successfully connected.

Could it be that our Eurospec ECU communicate slightly different in some way causing software issues?

Apart from the issues above I must say that I'm really impressed with this product both the quality of the components and the functionality is really great when it works. I replaces a whole bunch of gauges and make it so much easier to keep an eye on the ECU parameters while driving.

---

Yes it does sound like we are having very similar issues ..... The only real difference is my car and ECU are both JDM .

I am going to redo a couple of earth points on the centre consol and move the power wire to see if it helps the false knock .

Does you boost fluctuate much in higher gears ? I can't see any reason for it at all ( apart from poss the false knock) and really need to sort this out.

The MAIN reason I got this was for boost reduction on knock as I'm running 16psi boost on a high comp NA engine with 30% Methanol, so I really need the brilliant safety net this gizmo offers :)

I will update the unit when I get 5 min and sort out power and earths too, then I can report back with my results.

Thanks for all the help Blackstealth and I'm sure we can sort these little issues :) :)

Quote:

---

Originally Posted by BlackStealth

Wow nice install. If you can provide a video that would be great, I'd like to see the corruption, as it might give me an insight.

Longer autoreconnect pause version:
http://lcdbc.xp3.biz/updates/UPDATE311.zip

---

Thanks for such a quick response!

I'll install the update tonight and give it a try. I'll see if I can rig up a camera and make a video at the same time.

If the disconnect issues are caused by noise or interference would it help if shortened the cables going to the OBD2 connector as much as possible and added a ferrite bead?

Quote:

---

Originally Posted by fastbikes76

Yes it does sound like we are having very similar issues ..... The only real difference is my car and ECU are both JDM .

Does you boost fluctuate much in higher gears ? I can't see any reason for it at all ( apart from poss the false knock) and really need to sort this out.

---

Unfortunately I'm not using the boost control feature right now so I can't test this for you. I'm still relying on the stock ECU/solenoid setup to control my boost.

Here's some info for those that are interested in making a similar install as the one I did.

I used the stock panel for the heated seat switches (available on euro spec cars) it has the Mitsubishi part number MR175496,

http://pics.unlogic.se/3000gt/projekt/IMAG2903.jpg

http://pics.unlogic.se/3000gt/projekt/IMAG2908.jpg

http://pics.unlogic.se/3000gt/projekt/IMAG2965.jpg

It's was just a matter of chopping up the panel, gluing it back together again and then putting some paint on it.

That looks very nice. Way to go

So while I'm sitting here thinking of all the cool stuff this does, I had two thoughts I wanted too put across and see what you think ??

Firstly is there a way to increase the cold engine temp from 160f to at least 185f and even higher if possible ? I would rather not make any boost until the stat opens at around 160ish and if we can set it even higher (ie hotter than the car will get) it can be used as a another form of Valet Mode for those who don't have an aux fitted ??

Secondly, how difficult would it be to Create a rev limiter at cold coolant temps ? We all know garages like to rev the crap out a nice car when it's in for repairs etc. so if we could have a parameter that we could set a rev limit and coolant temp range that would be awesome. I would set mine to 3000rpm below 170f for example , that way I know the engine won't be revved like hell when not in my control by garages or the wife lol.

Just a couple of thoughts but don't go out your way if its not something dead easy and straight forward to add.


Cheers
Elton

Turn off the boost modifier features when diagnosing boost:
RPM offset -- all zero
Temperature offset -- all zero
Cold engine feature -- off
-DC%/Knock -- zero
PSI Start = 1psi
PSI End = 2 psi

With those essentially turned off you, boost is regulated by Duty Cycle % only.

Cold engine temperature feature was copied from the stock ECU, and they figured 160F was safe enough. If you still want it raised higher, I can do that. Also I cannot tell the ECU not to go above XXXX RPMS, so that rev limiter idea is not possible.

I have made OBD2 noise filters in the past and it did not help for those few that suffered from noisy OBD2 communication port. The enhanced OBD2 cables are shielded, so if you want to shorten or add a ferrite bead it is certainly possible because all the electronics are inside the LCDBC.

If any of you guys are electronic wizards, I can give you a schematic on building a serial OBD2/MUT cable using a couple resistors and transistors. Shiver-91vr4 also combated the same problem, here's a snip it of his reply (I hope he does not mind me sharing this):
"No I didn't end up using a low pass filter. I ended up changing my transistor circuitry for pulling the line low and put a very small resistor with huge wattage on it so I could pull it as close to 0v as possible. The problem with my previous design was trying to pull it to ground with a higher valued resistor and leaving some voltage still on the bus as to not burn out the resistor. It worked 90% of the time, but there were issues with ground noise.

I still had issues with it though, and the only way to cure it was to have the customer make sure their car was wired correctly. Most people would do a fuel pump hotwire the incorrect way and it would cause feedback on the signal wire."

I believe Shiver-91vr4 changed Resistor 1 from 560 ohm to a smaller value.
http://i782.photobucket.com/albums/y...D2MUTcable.jpg

Antero, I've been using the MC33290 for years in my Spyder TopLink product (interfaces to the roof computer). It is an RS-232 to ISO 9141 interface chip with built in short circuit protection. It is the proper chip for the ISO-9141 OBD interface used by our cars. It was discontinued and there is a new chip, MC33660, that is pin to pin compatible with what appears to be the same specifications. MC33660EF-ND is available at Digi-key for $0.89 each.

Thanks for making me look at this. I did a last time buy of MC33290 a year ago and thought I would have to re-design the board when they ran out.

Quote:

---

Originally Posted by fastbikes76

HI,

Just fitted this lovely bit of kit and im very impressed with it !! I do however have a few teething issues I need your advice on please

1.) boost control

The car makes the required boost fine, BUT it cycles very badly when building boost in higher gears like 3rd onwards, its almost as if i'm lifting on and off the gas constantly. I am using the Mac valve as recommended and have the connections the correct way around going from Google.

2.) Screen

The screen refresh rate seems very slow on mine and I have had to drop it too its lowest setting of 2 to actually get 'live time' information. It also sometimes slows down so bad that if I turn the dial it can take upto 2 seconds to change to next menu.

3.) connection

The car connects perfectly something and then will drop out and attempt to reconnect 4 times randomly for no reason. Once its dropped out it wont actually connect again until car is restarted. Having checked all the settings I am happy they are correct for my car...saying that Evoscan used to do something similar before too so perhaps this is more of a car issue ?

4.) Knock

The screen with the WB, BAR and Knock registers no knock and the audible alarm is set to 5, however if I go to the peak memory showing BAR and Knock it records 28 counts of knock. I know 100% fact its not knocking as I'm running 30% methanol and usually 1.1bar, today I was testing at only 0.75 and recalling a peak of 28counts. No audible warning and no live count of knock shown ?


The car in question is a 97 JDM GTO SR which uses the exact same ECU as your federal spec 3 plug ecu's.

Any help would be greatly appreciated.

:)

---

SO , some progress today...Damn I LOVE this thing !!! :coo3l:


1.) boost control is now fixed !! It turns out the barbs I got for my Mac valve were the correct 1/8th npt thread BUT the side the vacuum hose fits on had a pretty large inlet (6-8mm). I ordered a replacement set which only had around a 5mm hole and now have solid stable boost control. :)

2.) Screen. Not done this today although I did move the earth to the ECU bolt and also added an extra earth from chassis to engine (I had done a wire tuck and only had one small earth on engine)

3.) Connection..still dropping out although not tried the update yet. Will hopefully do it tonight , or wait until unlogic reports back on his lol

4.) Knock..still recording random peaks of 24-28 although no knock warning or reduction in boost as set.

so just number 3 and 4 to try figure out and all is well.

Thanks very much for all your time and effort you put into not only the product but your exceptional backup you provide us too. :coo3l:

The boxed version LCDBC uses SI9241AEY chip that does a proper TTL serial to ISO 9141 conversion, along with protection mechanisms built-in. It is equivalent to the MC33290, just different pin outs. Digikey for $2.76. :)

I provided a diagram of the transistors/resistors version in case someone wanted to tweak their own.

OBD2 (ISO9141) and MUT specifications are:
Low signal range: (0.00V to 2.40V)
High signal range: (9.60V to 13.5V)

I believe Shiver-91vr4 changed R1 to a low value to get that low signal as close to 0V as possible to make it more immune to noise. A while back I did some electronic simulations, I could bring that signal down very close to 0, using R1 as 200 ohms (if memory serves me correctly).