Pietro

The other day I have tested the lambda voltage while driving. I just wanted to chjeck if my lambda sensor was still ok.

The usual voltage was about 450mV. While decelarating with fully released pedal it went to about zero.

I was also impressed by the fact that at full pedal in full accelearion the signal was a mere 600mV. Only in few times it reached the 900mV which is usually obtained with rich AF ratio.

In my mind this means that even in open loop operation, that is when the pedal is full blast, the DME is pretty good at selecting just a slightly enriched mix. just slightly then.

As I remember peak power is obtained with lambda =.92 that is a +8% enriched mix.

I have seen that it is possible to change the open loop preprogrammed mix up to 6% by turning a small rotary switch on the side of the DME. 6% more fuel at near optimal lambda should bring ~3% more power or ~7 horses.

Did anybody test this at all?

cheers

Pietro

Pietro

incidentally , I have also looked at the fuel and advance maps shown in

<a href="http://www.stig.uklinux.net/index.php3" target="_blank">http://www.stig.uklinux.net/index.php3</a> page=tech&article=chips

and found that the performance chips, at full throttle give about 6 degrees more advance, on a wide rev range, and ~3% more fuel at high revs &gt;4000.

Adrian

Dear Pietro,
I cannot wait until you have to have an emissions test. However thank you for advising that you have been inside the DME. The operating range of the O2 sensor is 150 mv to 900 mv.
Ciao,
Adrian
911C4

PS: I highly recommend that others that read this post that you do not follow this example unless you know exactly what you are doing. There are specific warnings from PAG and Bosch about attempting such adjustments. If 7 HP as claimed in the post above, was available so easily please believe me when I say, you would have known a long time ago because you would have been told. This forum and its original participants have been here for years.

Pietro

Dear Adrian,

thanks for your nice words.

1-The adjustement of the enrichment and advance on the DME unit does not affect the standard german AU smog check as this is performed at constant throttle and in this mode the DME operates in closed loop with lambda control. The operating range of a lambda sensor depends on its age. A new one is as you say. As a lambda ages its ranges drops down.

2-It is not a secret that the stock DME runs leaner that "optimal" as far as performance is concerned, that is when in full-throttle open loop operation. The reason why this is the case are
a)pollution control laws: while the normal smog check does not check gases during driving cicles, the initial registration of a model requires this test. In California-USA, where I used to leave before, the EPA was quite strict also in the driving-cicle tests.
b)conservative gasoline assumptions. Also Anderson's book mentions this as a reason for the rather conservative maps on the 964's DME.
c)fuel consumption. I guess this does/did not concern Porsche.

3-Also Anderson's book, as well as others (i.e. Probst's book on Bosch injection), mention that the main reason why aftermarket DME chips give more power is a richer A/F mix during full throttle together with a more advanced ignition (on top of the higher rev limit which I do not particularly like for obvious reasons). Plenty of discussions on this in the efi-diy mailing list.

4-I had checked this in the past on 2 of mine previous cars. a 911SC and a Ferrari 308QV. Both US models. Both had K-Jetronic with lambda. In this injection the way to run slightly reach during full throttle is quite easy:
a)with the lambda sensor connected to the K-jet control unit, you set up a tested while driving which shows you the voltage of the lambda. The best would be to have a wide range new racing lambda which can give you a decent S/N ratio also at low lambda's.
b)Then you drive the car at full throttle and adjusting the screw in the fuel distributor enriching gradually the mix until the voltage on the lambda sensor is as wanted (that depends on the sensor used also). Of course the angine is OFF when the screw is adjusted or else the sensor plate will get stuck. Anyway there are details elsewhere on how to do this.
This procedure does not affect closed loop lambda operation which is going on during "normal" driving (other that in change of the duty cycle of the control valve). In this way there's no impact on the smog check and the car runs better.

5-I know the motronic (or late L-Jetronic) much less that the K-Jet CIS. However with the motronic DME I am not aware of any mechanical way to do this enrichment of the A/F mix (not 100% sure though). Infact The airflow sensor is completely open at high rev and, with full throttle, the DME works in feed-forward based on rpm's, throttle, and other measurements (head temperature for example). So we need a new eprom which costs 500$ and just remaps the fuel and advance.

6-When i found out, through the rennlist, that there's a "stock"-simple way to adjustment of the A/F mix on the DME I thought that this was an interesting point to pursue. Again it is obvious (and written) that this adjustement cannot affect the operation in closed loop as the stock lambda sensor changes voltage too rapidly close to lambda=1 and to control actively other A/F mixes cannot be achieved by this DME unit. So this adjustement can only be designed to affect the feed-forward open loop operation in full-throttle. So no risk for fail the AU(smog) check.

For this reason, rather than simply playing around with the rotary switch on the DME, I did drive the car to check the lambda voltage first. I have done this by running a wire connection from the DME plug to a portable oscilloscope, but it can also be done if you have a hammer that will allow you to read the lambda sensor while running the car. Infact I found very interesting points out of this test.

I thought that some of these points may be of interest to others in the list as I could not find them in previously discussed threads. However :sorry if they are old stories.

On top of what reported in my previous messages, I also have noted that when MY engine is hot, the actual voltage in the lambda sensor during full-throttle drops well below 450mv. That would be running a lean A/F mix during acceleration. When the engine is cold this does not take place and the lambda voltage at full-throttle jumps from the normal closed loop swing of 400-500mV to 800mV or more. This is explained by the cold engine enrichment given by the temp sensor connected to the DME. I admit that I was surprised that this took place as I did not expect that my car, with such a DME, would run lean on acceleration! I have tockeck this more as it may be a sign of a leak somewhere or another injection small problem.

So after this long message, I hope that somebody found this interesting. If not....sorry!

cheers

Pietro

Riccardo

Pietro, altough this is almost going into rocket science for me, its very interesting to hear about the 'hidden' aspects of how our cars operate. Thanks for sharing, this is what Rennlist is all about, be it from new or long term members!

Ade - C4 91

It is very interesting Pietro. I don't think I'll be opening up my DME to have a go, but it's interesting to know your findings.

cheers.

Ade.

johnfm

Pietro

please continue with your observations and contributions. There are always more things that everybody on this board can learn - there are always more than one way to skin a cat!

I, for one, am very interested in people like you taking a look at 'known' sytems, like Motronic, from a fresh perspective. Call me cynical, but often the 'official corporate line' from Bosch or Porsche may not be the whole truth, for a number of commercial, safety and legal reasons.

Keep up the fine work!!