Friday, January 04, 2008 - 08:04 pm, by: Phil Stuart(Ericdog)
Hi Guys,
Having read much about the use of zener diodes as "overboost" fuel cut defenders (FCD's), I decided to do some testing. This applies to my 91TT, but should apply to most soarers near that era.
For those that don't know much about MAP sensors and FCD's, its probably better if you just do a search rather than me just repeating what has been said before.
For those that just want a quick run-down on the MAP sensor, the ECU (computer) in your car monitors the amount of boost pressure/vacuum using a device called a MAP sensor. Typically, this little sensor outputs a voltage between 0 and 5 volts depending on boost or vacuum in the inlet manifold. It will put out 2.5 volts when there is neither boost or vacuum, hence its output goes above 2.5v on boost and below 2.5 on vacuum.
If the ECU "sees" this voltage coming from the MAP sensor reaching about 4.82 volts, which I understand corresponds with about 14(ish) PSI boost, the ECU cuts the fuel to the engine, generally by shutting off the fuel injectors.
So, if you want to run boost around or above 14PSI, you will run into problems.
To stop this fuel cut from happening, many people suggest putting a 4.7 volt zener diode across the MAP sensor output, in the hope that the 4.7 volt zener will clamp the voltage at 4.7 volts.
This brings me to my tests:
What I have found is (from tests) that the output impedance (internal output resistance) of the MAP is about 2.2k ohms. This means that the MAP sensor does not have a lot of "power delivery" capability, and is easily "loaded down".
One test example was done with the MAP output at 2.5v (key in the on position, motor NOT running). By simply "loading" down the output of the MAP sensor with a 2.2k (2200) ohm resistor made the voltage drop to about 1.3v. The ECU does not load down the MAP voltage as it has a very high load resistance (>100000 ohms).
By applying pressure to the MAP sensor to bring the output up to 4.0 volts, putting a 4.7v zener across the output of the MAP sensor made it drop to 3.45 volts !!!
This drop in voltage is due to the zener diode having a small amount of leakage current.
Now, this will certainly make it work as a fuel cut defender, but the ECU will now be seeing a voltage that is not correctly proportional to the true pressure in the manifold, and could cause a change in performance.
So, in summary, although a zener diode will work as a fuel cut defender, I wonder what effect the incorrect MAP-to-pressure voltage being sent to the ECU will do.
There are many different types of MAP sensors, and some may have a lower internal resistance and may not suffer from this problem.
As another experiment, I figured I could make a "pseudo" zener diode, that has no (negligible in this application) leakage by connecting a number of standard 1N4001 diodes in series. I found that if you connected 8 or 9 (I found 9 the best) of these 1N4001's in series, it produced a very good 4.7(ish) volt zener diode and the best thing is that there was no leakage at all that loaded down the output of the MAP sensor, yet the diodes functioned fine to clamp the MAP output so that the fuel cut was disabled and I could run more boost successfully.
I know this sounds strange, as a 1N4001 has a 0.65v forward voltage, and that 8 of them would make 5.2volts, and it shouldn't work, but if you take into account the high internal output resistance of MAP sensors, it worked both on the bench in my workshop, and in my 91TT soarer.
If you want a reliable fuel cut defender, I would steer clear of zener diodes and go for the adjustable design such as the kit that jaycar sells, the circuit listed elsewhere here (do a search) or a commercial unit.
I hope this little piece of info, and testing has been informative.
Friday, January 04, 2008 - 08:36 pm, by: Daniel Clarke(Dieseltrain)
All it does is tell the Ecu its running Less boost to avoid boost cut. BUt the ECU runs enough fuel to make about 225rwkw ON stock injectors with the voltage clamped @ 4.18 volts.
4.32volts is actually the boost cut threshold. ANd most adjust to 4.25v to completely avoid it/
I tried so many different voltages on the dyno.. Tha main issue is not fuel related, But by reducing the voltage to the ecu. It actually runs more ignition timing. So its the advanced timing which makes it ping, not the lack of fuel ( as we tested ).
Ive since fitted a Blitz Ecu and @ 17psi made almost 240rwkw on stock turbo's and running out of fuel with the stock injectors.
Phil Stuart Tinkerer NSW Soarer 1x TT JZZ30, 1 x RAV4 and a pork chop
Friday, January 04, 2008 - 09:05 pm, by: Phil Stuart(Ericdog)
Thanks, Daniel,
Yes, I understand how it works, and what its doing.
It was more-so a document of my tests, especially the zener leakage, and to help those guys that might want a bit more info..
I actually read somewhere else on here that the cut was at 4.82v, yet you say 4.32v ? There seems to be a bit of conflicting info about. Just so I know, have you actually measured it on a car at 4.32 ?
Saturday, January 05, 2008 - 03:57 pm, by: Cihan Aday(Cihan)
Phil, 4.82v was and early figure, incorrectly stated by myself on a few occasions - related posts have since been altered Turns out my multimeter was nearly out of battery which tends to increase the output.
The cut occurs at 4.33-4.35v, and corresponds to nearly 13.8psi.
Sunday, January 06, 2008 - 07:25 am, by: Blake Gloyn(Blakenz)
Thanks Cihan, this would explain why my 4.7 v zener diode did not defeat boost cut when i tried it a couple of years ago. i also tried a lower one(3.3? or similar) and the car idled but wouldn't accelerate.(rev)
Phil Stuart Tinkerer NSW Soarer 1x TT JZZ30, 1 x RAV4 and a pork chop
Monday, January 07, 2008 - 10:19 am, by: Phil Stuart(Ericdog)
Thanks for the correction, Cihan. It explains why some of my tests didn't correspond with 4.8v.
I agree that an effective clamping circuit is the way to go, but would still be concerned about a zener alone due to leakage, even a 4.3v one, it will work as an FCD, but, as it will alter the pressure/voltage curve of the MAP output, and in extreme cases, such as Blakes', effect vehicle performance.
Tuesday, January 08, 2008 - 05:21 am, by: Scott Ferguson(Scott_ferguson)
Harry lemmans had another circuit that i found a while back using and adjustable voltage shunt regulator. More robust circuit and a bit more reliable. Might be worth searching for.
Phil Stuart Tinkerer NSW Soarer 1x TT JZZ30, 1 x RAV4 and a pork chop
Thursday, January 17, 2008 - 01:35 pm, by: Phil Stuart(Ericdog)
I actually ended up using 8 x 1N4001 standard diodes in series, and it works a charm. No leakage (as with zeners). I did try a few zeners (4.3v, 4.7v and 5.1v - even the 5.1v one had bad enough leakage to alter the pressure-to-voltage curve of the MAP sensor), where the array of 1N4001's didn't.
I would imagine that Harrys circuit would not have leakage (I must get the adjustable "zener" and try it)
The 1N4001 array clamps ok at 4.1volts, but will still allow the voltage to go a little higher if you do something silly like try to run way too much boost. I have it running at 18PSI successfully, and did wind it up to 20, just for a test, and all was fine. Since they are only the stock turbos (ceramic wheels) I don't want to stuff them. The odd use at 18PSI seems fine, but I wouldn't like to drive at these high boost levels for a lengthy time!
Just one thing when increasing boost, make sure that all the intake hoses (turbo to engine) have good hose clamps and that they are done up tight. One of mine popped off the first time I ran it at 18PSI, but I found that the clamp around the hose wasn't fully tight.