Paul_H

This link was recently posted for a fuel cut defender:

http://www.mr2.com/ARTICLE/FCD.html

Incidentally the circuit diagram omits to mention that logic gates HEF4011BP requires voltage supply at pin 14, and ground at pin 7

This is what it looks like when built, you may be able to make it prettier or smaller



I measured the output of the MAP sensor (MY99), here is a table of values which may be useful to you if you are calibrating a FCD.
These values are for Denso sensor part no 220-12AA090 079800-4110.
N.B. The usual disclaimers apply - ther is no warranty on anything written here - if you blow your motor up, it's your fault not mine!!

Bar Map sensor output (volts)
0.0 2.29
0.1 2.44
0.2 2.58
0.3 2.73
0.4 2.90
0.5 3.03
0.6 3.18
0.7 3.33
0.8 3.47
0.9 3.63
1.0 3.78
1.1 3.92
1.2 4.07 - Set to "defend" here on MY99?
1.25 4.13 - Fuel cut on MY99?
1.3 4.21
1.4 4.36 - Set new fuel cut just below here on MY99?
1.5 4.52
1.6 4.67
1.7 4.81
1.8 4.93
1.9 5.01
2.0 5.01

I measured the output of the sensor by applying pressure via a large plastic syringe with a boost gauge connected via a T-piece, and reading the output on a voltmeter. Worked a treat!





The output of the FCD follows the input i.e. as the voltage from the MAP sensor rises, so does the output voltage from the FCD. There is an error here - the voltages are not exactly the same, and this is present right across the range.

Here is what happens at 1.0 Bar. The yellow meter on the left is reading MAP sensor output (input to FCD) and the black meter on the right is reading the output from the FCD (sent to the ECU)



Here are the readings approaching 1.2 Bar (approaching fuel cut on MY99)



And here is the FCD "defending". The yellow meter (MAP sensor output) is at fuel cut level, notice that the FCD has dropped the signal that the ECU receives so that fuel is not cut. I expected this signal to be "clamped" i.e. to remain at a constant level below fuel cut, but this FCD has reduced the signal significantly. Anyone got any comments on this?



And here is the new fuel cut - notice that the output from the FCD has risen to a level where the ECU will trigger a fuel cut.



You can calibrate these settings as you desire - obviously if you take the settings too high then you run the risk of damaging your engine. As I understand it, from postings here, the fuel cut level on a MY99 with PPP is about 19 - 20 PSI (just under 1.4 Bar?). So maybe set a new fuel cut here - and run up to 18 PSI max via a Dawes Device or similar....?

I've not run this FCD on my car....just at the testing and calibrating stage. Does anyone have any comments on
(1) the slight error, across the range, between FCD input and output
(2) the level of the FCD output while "defending"
(3) sensible limits for calibration of the FCD to keep the engine in one piece

Cosie Convert

Wow Paul, you've been busy ! This looks very usefull, just that glitch at fuel cut point to fix (don't ask me how) such a big drop in output may trigger the ECU to advance the ignition timing which, as you know, is not a good thing at this level of boost

cc

WREXY

Sorry, I couldn't hold back, but Paul, you may find some of the fins on your intercooler have been bent after resting those gadjets on it.

Cheers,

Wrexy.

Paul_H

Wrexy

No they haven't, and yes you should have held back

john banks

I posted the link Paul - good job for trying it.

I think your FCD targets are very sensible.

I am puzzled also why the voltage drops in the defender zone and why there is an offset across the range.

The important one is at 2.29V for there to be no error as the ECU measures atmospheric pressure here and this *COULD* affect things?

If there is error elsewhere I don't think it will have a huge effect to be honest as most of the ignition timing is based on the MAF not MAP sensor, but it would be worth seeing if your ignition timing changes from standard up to say 15 PSI?

Incidentally, this could also be done with a microcontroller with ADC and DAC, and I may integrate something to do this into my EBC design.

Will post some more after a visit to the in-laws if I think of anything else.

Anyway, dead chuffed there are other electronics bodgers about although I'm sure I bodge it up and you do it properly

Paul_H

Cheers John, but you're too modest; and I can bodge with the best

The offset across the range is present even at atmospheric pressure i.e. with the MAP sensor outputting 2.29V, the FCD outputs a little less. The FCD consistently outputs less, but this does seem to remain constant.

I would be interested to know if anyone has actually measured what
a commercially available FCD outputs?

I've not yet summoned up the courage to connect the FCD to my ECU. With +12V and -12V on the FCD board, I would hate it to send these voltages to the ECU if the circuit were to go faulty - can you think of a way to limit the output of the FCD to +5V maximum (i.e. the maximum output of the MAP sensor)...or am I just being paranoid?

john banks

I think you are right to be paranoid! Best way to be with this sort of thing.

Maybe it needs a bit more bench testing first to be sure.

I haven't looked at the circuit in detail, but could there be a reason for the offset there? You could email the creator of the circuit if his address is on the page and ask him - certainly when it is your own baby it is 100 times easier to work this out than for other people to do, and I'm sure he would be interested in us using it!

If not the only alternative is to go through the circuit a unit at a time and work out where the offset is coming from?

Did you see the DIY FCD page I also posted - that had the some commercial plots IIRC? Or was it just stylised examples - do you have the Scoobynet link to the page I posted the link on?

john banks

Thinking about it and looking at the circuit, the logic will not be the cause. The only source is the analog circuits before including and after the sample and hold amplifier. My initial thought was an offset from the sample and hold amp. Then I figured that you could also get offsets from every op amp - (offset null rings a bell in my head but only the 741 had this - perhaps a higher spec op amp is required?). Also if the gain on the op amps is set from resistor ratios this could easily explain differences from the tolerance of the resisitors which could be remedied by using pots and tuning them with the voltmeter?

Just a few initial thoughts to go on! Might all be crap, but I wager one of the above!

T-uk

john,

send your ECU to StephenDone,I reckon he could work out how to raise the fuel cut from it .

john banks

Yeah, for starters measure R4 to R7 and please post the values on you DVM. Either that or offsets - can't be anything else. I would have a hunt through the sample and hold amplifier datasheet - I had a look before at it when I posted the circuit and I found it through google but don't have the link now.

You could also test to see if there is an offset from the input to the outputs of the op amps to home in on the culprit(s).

The author has put contact details on his page helpfully.

Don't need to raise MINE - you need to raise YOURS John so perhaps you should send your ECU

[Edited by john banks - 2/3/2002 11:21:13 PM]

T-uk

but mine will not tell him how to raise the fuel cut unlike yours ,compared to a standard ECU.

did you try your EBC taday?

Paul_H

John, I have emailed the author as you suggested, here is his reply:

Thanks for contacting me.

I will have a look today on a sample here and inform y on the results on the following of input/output. I think there is a small difference but not as much as measured in the postings... anyway I will have a look again and let you know.

In the meantime I have some printed circuit boards without parts on left over and I can send them to anyone who would be interested to distribute them to persons interested to build the unit free of any charge... let me know is there is such an interest... the board is very small and nicely set. The resistor feeding the zener diode that sets the voltage reference has to
be much lower than the valus eroneously stated, and a small by-pass capacitor has to be inserted on the + 5 VDC. to filter is out.

You will hear from me soon ...

Best regards....

Kostas Chryssos

and later......

I had some measurements taken and seems the nonlinearity derives from the AD781. All the other amps follow input/output correctly. There is a slight missmatch on the + 12V and -12V supply to this chip but I dont think the missmatch derives from this. There is no feedback on the chip to adjust so it could either be so and left so in my opinion this is not really a problem as it is within the tolerance of most ECU systems on cars on which the
tolerances are far wider, or a slight adjustment could be done by changing the 10K feedback resistor of the amp following the AD781 to increase slightly the gain to correct the missmatch. Again I dont think this is really required.

Kostas

What a great guy! I've emailed him back to thank him and also to express interest in the circuit boards - I'll post details as soon as I have them - and I'll keep working on it

Paul_H

Update as of 04 Feb:

Another email:

The resistor to feed the zener should be selected in such a way as to allow about 25 milliamps to flow through the zener. A good guess would be 220 to 330 ohms maybe less... current should not be too high for the zener could get hot. There is actually not much current required on the circuitry there so just a small value to get the zener do a good stabilization at the 5 Volts level. If the resistor is too high the voltage will be under 5 V something not wanted.

The capacitor is not really required.

I will pack for y and send y a number of pcboards. There is no cost involved and make good use of them. I may also include a stuffed sample for testing and seeing how all components are installed.

Let me know if I can be of any further help...

Regards ...Kosta

I think a resounding THANK YOU! from all at Scoobynet is required here, don't you? Stunning!

john banks

Top bloke!

Paul_H

Thread now closed - please see update:

http://www.scoobynet.co.uk/bbs/threa...threadid=70491

Reno1986

Hello there,

I'm from Germany, and I was searching on the internet for a FCD DIY.

I found this thread and Kostas website http://www.mr2.com/ARTICLE/FCD.html

My problem is that I can't enlarge the pictures on Kostas website to see what the built.

Are there any circuit boards left? On the update thread I was reading that he was sending some of them to BugEyed.

I have an engine swap done - Legacy Turbo engine to Legacy MK 2 (Hubertus edition with OEM bull bar - looks really nice )

Thank you
best regards

Reno