(ECT and VSS work!) Yet another s2000 cauge project [READY]

JaredKaragen · Post by **JaredKaragen** » Fri Aug 01, 2008 11:49 pm

Yeah, in the pgmfi archives there is some info, but not enough I must say =)

Orthello · Post by **Orthello** » Wed Aug 06, 2008 7:25 am

CRX2 wrote:How do you put the VSS signal to the gauge? Direct? You could try with a Schmitt-trigger (HEF4093 I think...) and "clean" the signal from spikes and noise. And also check your supply voltage for spikes.

Supply voltage is taken from the S2000 gauge. There is a very nice and stable +5v there to supply it's own IC's. So supply isn't a problem. I did a test with a fixed signal on the output and went for a spin. Reading on the S2000 gauge kept stable too. So the noise is on the input of the converter. I went through my stocks and found a 74ls132 (NAND with Schmitt-Trigger inputs) Put one input to +5v the other to the VSS. Output then is the inverted of the input with Schmitt trigger 'filtering'. Inverted signal for VSS isn't a problem. The readings on the gauge were way more stable then before. Even at high speed and with the engine running in VTEC. But still not stable enough to my taste. The readings jumped up from time to time, indicating some input noise was is passing Vt+ level. Vt+ level is about 1.5 to 2.0 volt Vt- level is about 0.6 to 1.1 volt After searching some more through my stocks I think I found a 74HC14 (hex inverting Schmitt Trigger) I'm planning to play around with that one tonight. The Vt+ of the 74HC14 is 2.5 to 3.0 volt. Because of the higher Vt+ it should 'filter' out more noise. Maybe I'll be able to find a way to even have a higher Vt+ I want Vt+ to be as high as possible without losing the 5v VSS pulses. Getting closer and closer. And still... the hardware is F*%^& simpel. Realy! When I'm ready I'll post the schematics here. You'll all be surprised! I was like... "WTF?!? and you would have to pay $150 for that!?!?" Trust me.... Hardware is less then 25% of that!!

Orthello · Post by **Orthello** » Thu Aug 07, 2008 3:43 am

Yesterday evening I created another schmitt trigger 'filter' to eliminate all the noise below 3v. This morning I did a road test, while driving to my work, which is a 50km (30Mile) trip. The first part (30km / 20mile) went PERFECT. Only two noise spikes (with a jump in the speed reading as a result) but that was it. I was like... YEEEEHAAHAHAHAAAAAA But.. NEVER BE HAPPY TO SOON!!!!!! Last part of the trip (20km/10mile) the readings started to jump more and more. I'm almost sure this has something to do with the environment temperature. The warmer copper wires get, the better it conducts electricity (and electric noise!) Vt+ (the upper noise level threshold) is now set to 3v. Somewhere next week I want to get Vt+ to set to 4v. Getting sooooooo close now!!!!! to measure is to know

Oh.. and because of the very acurate readings during the first part of the trip, I was able to have a first indication of the mismatch with my original speedreading. With a 36 multiplier (calculated from original Honda documentation), the S2000 gauge reads 10% less then it should. So I'll set the pultiplier to 39 and do another road test when I have Vt+ at 4v

Orthello · Post by **Orthello** » Thu Aug 07, 2008 2:28 pm

DAMN.... it continues...... The way home.... Gauge was nerveus again for the first 10 minutes. The last 30 minutes the gauge was showing a very very stable speed again without any glitch, spike or whatever! BUT.... at twice the speed as it should!!! Something what caught my attention... when the environment was cold, everything did what it is expected to do. When it gets warmer, things get nervous. And when warmer then xx degrees, gauge shows double the speed. The nervous part maybe is the part where everything transitionsto 'double'. I don't think this can be any form of noise. If this is noise at the input it's a F@$# nice, constant noise. Got my point? Need to think about this more, and do measurments again in car while driving. WEIRD!!!! If the gauge/converter shows the exact same behaviour next monday morning... I'm going to get \"Freeze 75\" and cool the converter component by component. This way I can detect which one is the suspect!

RoadXY · Post by **RoadXY** » Fri Aug 08, 2008 5:57 am

Damn, never knew temperatur did so many to an relative simple board? (At least I think it's relative simple

). Are there any components running really hot then or is it the environment temp?

Orthello · Post by **Orthello** » Fri Aug 08, 2008 4:30 pm

It's just environment. And still... i'm not 100% sure. It is something that caughtmy attention. Monday morning I can say more. Hopefully the exact same behaviour repeats itself when driving to my work again.

Orthello · Post by **Orthello** » Mon Aug 11, 2008 4:17 am

Did a road test again this morning. No changes to the hardware or software. Drove about 25 kilometers when readings became nervous again. At that point I turned on the fan at full power and at its coolest, blowing over the converter hardware. Guess what....... readings became more and more stable again. This is a themperature vs. noise issue for sure. There is a noise at the input which is very very very close to the upper schmitt trigger treshold. I received tips and tricks from some people to put capacitors here and there. (I knew I had to play with caps) Especially at the input to debounce the reed type VSS and elliminate other types of noise. Although I'm almost sure I don't have bounce issues. If the nervous behavour is a bounce issue it would be a constant problem and wouldn't be temp dependant. Right? I'll add some capacitors tonight, or tomorrow night at least.

Orthello · Post by **Orthello** » Mon Aug 11, 2008 5:37 pm

Added a 100n cap, 22k and a 100ohm resistor for debouncing. Tomorrow morning I'll put the fan to high and warm to make the converter go mad. Let's see what it does. I tested with a normal swich. With and without the debounce stuff. Man... that was a day and night difference. We'll see tomorrow. For those interested in the old school reed type VSS signal: 50Hz with 2ms bounce marked in white.

Orthello · Post by **Orthello** » Tue Aug 12, 2008 3:56 am

WATER WATER WATER!!!!! Man... it was @$%#$@#$ hot in the car. Only to see if the converter was able to stand temperature changes after the alterations of last night. . . . TATADATAAAAAAAAAA . . I drove 50km without any problem. No dips, no spikes, no nervous behaviour, no nothing!!!! Not in a cold breeze and not when creating a mobile sauna. High revs, low revs, lights on, lights off. I tried to create noise in every way I could think of. No result. Which is perfect in this case. So now it is ready for extensive road testing. Next and final step..... get the ECT working.

JaredKaragen · Post by **JaredKaragen** » Tue Aug 12, 2008 8:10 pm

Sweet deal! Awesome news for sure! The ECT Shouldn't be tooo bad; maybe if you can get your hands on a S2K ECU and a set of harness plugs to test-bench it you could read the signal directly fromt he ECU to the cluster.... might be the easiest way...

Orthello · Post by **Orthello** » Wed Aug 13, 2008 3:14 am

I went to the dealer yesterday. They have three shops. One of them has an S2000 for sale. The shop owner in my city adviced me to call the shop that has the S2000 for sale, tell them my story and mention his name. There will be a very small chance that I can measure directly in an S2000. Keep my fingers crossed. EDIT: Ok... called the guys... They are willing to help me. \"Cool project. Come round. Show your thing! And you're very welcome to measure anything you want!\" So... this weekend I try to do a first fit of the S2000 gauge and next week I'll go to the city of Zwolle to do some S2000 measurements. Probably on Friday, August 22nd YEHAAAA!!!!

White_Teg · Post by **White_Teg** » Wed Aug 13, 2008 8:31 am

modifry on s2ki(dot)com forums wrote:Looks like 5v pulses, 30ms wide, with a variable spacing relative to the temperature. 3 bars is about a 1 second spacing between pulses, longer when colder, but the difference in spacing is minimal. A 2 bar reading was 1097ms, 3 bars was 1043ms.

They were looking for a solution to make the stock ECT gauge work when using an AEM EMS, which does not provide an ECT output.

White_Teg · Post by **White_Teg** » Wed Aug 13, 2008 8:32 am

Here's the link to the thread: http://www.s2ki.com/forums/index.php?showtopic=232488 (I could not put a URL in my last post)

Orthello · Post by **Orthello** » Wed Aug 13, 2008 8:36 am

DAMN... I have search the entire net. Inclusing s2ki. Didn't found anything. I created a generater to create a PWM signal. I also have an LCD attached to it to read the set freq and duty at that time. Tonight I'm going to play around with this. Very very thanks TEG!

Orthello · Post by **Orthello** » Thu Aug 14, 2008 4:58 pm

YES!!!!!! Based on the information from s2ki (Thanks again TEG!!!!) I was able to have a first quick test. Guess What? ECT WORKS!!!!! I was able to have the 1st, 2nd an 3rd segment to be lit. Because I'm busy the next few days I'm not able to do some more playing around. DAMN!!!! My hands are itching!!! Can you believe that? Hahaha. Now I need to find the EXACT specs of the transitionpoints from segment to segment. That is what I'm going to do monday or tuesday. EDIT: Got itching hands and did some testing: ECT spy movie