TurboTommy

I've done a search but couldn't really find definite answers to these questions:

At what voltage output from the TPS does the DME switch over to WOT maps?

What actual throttle position does this equal to in degrees (at 0 degrees the throttle plate is closed; at 90 degrees the throttle plate is fully open)?

I'm aware that the rate of change fo throttle opening brings about this condition of WOT, but I'm more concerned about when you just gradually open the throttle and at what actual throttle position (and corresponding voltage) this condition occurs.

Is WOT and open loop exactly the same thing?

As the throttle plate moves through its arc from zero degrees to 90 degrees, does the voltage output from the TPS go up in a linear fashion? I'm asking this because in stock form the KLR uses these readings in order to control the cycling valve. As most of us know, in stock form, there is a large gap in actual throttle position between low boost conditions and when the KLR/ cycling valve combo allows full boost. Therefore, I'm wondering if this is dictated by the TPS or by the way the KLR (or the DME for that matter) is processing the voltage information?

Autothority has managed to, via the KLR chip in their kits, to bring the boost up in stepped increments in this "middle range throttle positions" (although this does not perform quite as good as the LBE). Does anybody know if they programed this "updated KLR chip" within the bounderies of the WOT maps?

The TPS output information, and how the DME and the KLR process this information, might be an important indication of this.

I'm stuck with my mods now and any of us with reliable info on this would be muchly appreciated.

Tomas L

According to the factory workshop manual the KLR sends a full load signal to the DME at 65° throttle opening.

WOT uses the WOT maps and no closed loop lambda.
You can also have open loop at part throttle using the PT maps.

It doesn't matter if the TPS voltage is linear or not. The KLR is a digital device and can convert whatever voltage value to whatever CV duty cycle. The boost vs throttle position is the way Porsche intended it to be and that has nothing to do with the voltage/degree relationship.

The KLR code contains two 3-D maps which i think is throttle position vs rpm to duty cycle maps. The APE chips has changed both maps over the entire map area, other chips just modifies one row in one map (probably the WOT TPS value) and then max out the map that probably contains max boost limit, yet others just max out the boost limit map to make it possible to run with an external boost controller.

Tomas

TonyG

Tomas L


Interesting Info...

So you are saying that the TPS is fed to the KLR and not the DME? And that the KLR is the one that triggers the WOT fuel/timing maps within the DME?

TonyG

MHT

TPS throttle pos info from the pot goes to the KLR box, the idle switch signal goes to the DME box. Open loop is simply when the ecu is not operating with reference to the O2 sensor.

Bengt Sweden

Since it is close to midnight in Sweden, (and no sunlight at all where Tomas lives this time of the year), I assist and confirm your questions. The TPS feeds the KLR and the KLR only sends a WOT-on signal to the DME.
It may be interesting to know as well that there is one pressure limiter and one flow limiter so you have to adjust the DME as well to use external Boost control.
Bengt

TurboTommy

Tomas, thanks for posting back.

"You can also have open loop at part throttle using PT maps."

What does this mean? I thought you can either have open loop (full throttle) or closed loop, but not a combination of the two.

"The KLR is a digital device that can convert whatever voltage to whatever CV duty cycle."

Yes, I understand this. But if the TPS voltage goes up in stepped increments and there is a relatively large gap of throttle arc where there is no voltage change, the KLR would not make any changes to CV duty cycle during this part of the throttle arc. I think it's important to know how the TPS voltage goes up relative to the actual throttle arc.

If full throttle is activated at about 65 degrees throttle, does the voltage still increase in the last 25 degrees? If it does, that would be an indication that the chip programer has something to work with. Also ,if there is a steady increase in voltage during the throttle arc just before the WOT
mode, that would be an indication that the KLR has something to work with in order to change the CV duty cycle.

I would assume that the KLR needs changing signals from the TPS. It all starts there. Somebody tell me if I'm out to lunch here.

Really, the big question is: at what part of the throttle arc is it possible to change CV duty cycle? I know this will probably be hard to answer unless you've tested this, or have worked in the engine management department at APE.

Glad to here from anybody on this.

Bengt Sweden

Manual

Tomas L

It's correct that you can't have a combination of open and closed loop.
Closed loop is when the DME uses the lambda sensor to correct mixture although it still uses the maps as a baseline. Open loop only uses the maps.

The DME has different fuel and ignition maps for part throttle and for WOT, this has nothing to do with open or closed loop. The connection is that at WOT the DME is programmed to operate in open loop always. For part throttle however it has the possibility to use either depending on how the code plug is coded, closed loop for counties with stricter emisions regulations. Also of interest is is that the DME has three different fuel and ignition maps for different parts of the world, this is also changed with the code plug.

The signal from the TPS is almost certainly continuos (not stepped) but probably not linear. My somewhat educated guess is that the A/D converter in the KLR uses 8 bits which means that the KLR can measure 256 throttle positions. Since the TPS signal most likely isn't linear you will not get a 90°/256 resolution, it will be higher at some part of the throttle movement and lower at some parts. For the purpose of controlling boost against TPS I think the resolution will be high enough even where the resolution is lowest.

Bengt: I live 100 km south of the arctic circle so we do have some sunlight, however today we had -27°C (-16.6°F).

Tomas

Danno

"So you are saying that the TPS is fed to the KLR and not the DME? And that the KLR is the one that triggers the WOT fuel/timing maps within the DME?"

Yes, the analog-voltage from the TPS goes into the KLR. The KLR then outputs a binary ON/OFF signal to the DME as to the WOT state. So the KLR effectively replaces the WOT switch in the 944 NA cars. Notice that the 951 doesn't have a WOT switch in the TPS? This function is simulated by the KLR.

What this means is that you can actually have the DME triggered for WOT operation before you actually max-out the TPS. The KLR actually senses velocity of the TPS as well and if you open it a certain number of degrees in less than a certain amount of time, it will trigger the WOT signal to the DME.

Also Tomas is correct in that WOT vs. partial-throttle operation as far as selecting the maps is completely different than open- vs. closed-loop feedback on the O2-sensor. The USA maps tend to be about 5-6% richer than the ROW/Australia maps. That's because the O2 will tend to lean out the mixture on US cars. Without the O2-feedback, the ROW/Australia fuel maps has to be much more precise and match actual air-flows much better. Thus their maps actually look bumper when graphed.

Tomas, are you saying you're able to read the program-code off the KLR's CPU? How did you get around the read-protect bit? Or are you just using information you got from the data-chip?

TurboTommy

So Danno, would you happen to know the voltage output from the TPS to the KLR for WOT; and in general what approximate voltage vs throttle arc?

Is it possible for the KLR chip to be programed to increase boost (via the CV) even though the actual throttle position is not within the full throttle arc? I'm only concerned with this with the stock setup (no different boost controlers or wastegates).

Is it possible that the throttle arc would be greater than 65 degrees for WOT if the throttle is opened slowly?



Thanks all for info on these difficult questions

Tomas L

Just information from the data-chip. I haven't read out the program code from the microcontroler, however I do have an idea on how it can be done. Still it will take a lot of work and for the moment it's way down on my priority list.

The factory workshop manual states that the CV valve is activated at an Air Flow Meter output of 60%. That would mean that the KLR gets a load signal from the DME. I can only find one lead between the DME and the KLR on the circuit layout that has an unknown function,it's terminal 24 "tr-signal" on the KLR. Has anyone some info on this?

Tomas

Bengt Sweden

Tomas, have a look at the page I posted. tr is Trigger signal for ignition.
Bengt

Danno

"So Danno, would you happen to know the voltage output from the TPS to the KLR for WOT; and in general what approximate voltage vs throttle arc?"

From fully-closed to fully-open, it spans a range of 0.25-3.75v. If you do a search, forget about those people who posted that the 951 uses a 0-1v TPS, thus requiring retrofitting a GM-sensor if you use an aftermarket EFI system. These people have been repeating things they've heard, rather than making their own measurements and gathering empirical data.

"Is it possible for the KLR chip to be programed to increase boost (via the CV) even though the actual throttle position is not within the full throttle arc? I'm only concerned with this with the stock setup (no different boost controlers or wastegates)."

Yes it can be done by re-programming the duty-cycle maps for the CV. However, since there's no feedback control, it's a trial-and-error process that can take quite a while. The maps would then be customized only for your particular car only based upon your turbo, exhaust, wastegate, valves, rings, etc. Using the same map on another car would yield a completely differnt boost-curve. Cyntex tried this over 10-years ago and gave up trying to reprogram boost with a chip-only method. The limitation is in the KLR hardware we're given by Bosch (more later).

"Is it possible that the throttle arc would be greater than 65 degrees for WOT if the throttle is opened slowly?"

Yes, it is... You can tell this based on O2-sensor feedback. On some aftermarket chips that re-program the WOT maps only, there's a wide disparity between the last highest-load row of the partial-throttle map and the WOT map (effectively a higher-resolution single-stripe of the fuel-maps). On the dyno or using the O2-gauge, you can see it jump from a lean mixture all of a sudden to a too-rich mixture. This lack of matching the maps has caused a lot of tuning problems when people try to use signal-interceptors/massagers. Because a 4.0v signal on the AFM can reference a partial-throttle data-cell under certain conditions (like 90%), yet that same voltage will reference another spot on the WOT map, at say 91% and result in a wildly different fuel-mixture (TonyG's had 1st-hand experience with this).

" I haven't read out the program code from the microcontroler, however I do have an idea on how it can be done. Still it will take a lot of work and for the moment it's way down on my priority list."

yeah, I've got a hacker group that may be able to accomplish this. But it'll be very time-consuming and expensive. The idea is to mill off the top of the chip's casing to expose the wafer. Reverse-engineer the layout. Then find the exact location on the wafer that's the CPU's microcode and zap the read-protect bit with a laser.

" I can only find one lead between the DME and the KLR on the circuit layout that has an unknown function,it's terminal 24 "tr-signal" on the KLR. Has anyone some info on this?"

So far, I cannot find any communications between the DME & KLR as far as operating parameters or data goes. There is no load/airflow signal at all that makes it to the KLR from the DME. The only thing that goes between them is a piggyback massage function on the ignition timing. The DME sends out a transparent trigger signal from the flywheel sensors. The KLR intercepts this and passes it back to the DME. The DME then uses this as the actual trigger signal and fires off the coils based upon the values in the 3D maps.

When the KLR senses knock/overboost, it retards the timing signal from the DME. It basically holds onto it for an amount of time it wants to retard the timing, then passes it back to the DME. The DME then fires off the coil when it receives the trigger signal from the KLR with no idea that it's been delayed/retarded. The KLR is basically a piggyback electronic-boost controller and timing computer that sits ontop of the DME. It adds the extra functions needed to manage boost, while the DME handles fuel & ignition just like before on the NA car.

I suspect it's a cost-saving measure on the part of Porsche and Bosch. Rather than coming up with a completely new computer that does fuel, ignition and boost (with feedback control), like on the Mustang SVO of the time, they just used off-the-shelf DME units (with minor mods) and piggybacked a boost/timing controller on top of it.

TurboTommy

Thanks Danno for your input.

One of the reasons I'm asking all these questions, is that on my air/fuel gauge, I'm seeing moderate enrichment and no dithering at half throttle. This can't be an indication of already being in the WOT mode, can it?

Then, at full throttle it goes lean. Now, I can rectify this by increasing the fuel pressure. But the point is that that the injector duty cycle at full throttle is less than it should be. I'm starting to assume that the part throttle maps have a different code deciphering procedure than the WOT maps when we're trying to develop custom chips (increase in the injector size). All I did was give the instructions to decrease the injector duty cycle right across the board (for example: double the injector size; therefore, half the duty cycle). Yet, the result was richer at part throttle, leaner at full throttle.

Also, there doesn't seem to be a clear distinction between part throttle and WOT. I can be at significant boost and still seem to be at part throttle ( however, O2 not dithering), via the KLR chip, which is good, I guess.

I have a plug and play MAF.

Like I said before; the only conclusion I can come to is that chip guy has a hard time making equal linear adjustments to the chip. Is this possible?

Merry Xmas to all.

Danno

Just a quick note since I'm hanging out with the family....

There's a response-curve that can be drawn with injectors at various duty-cycles. Injectors are not linear devices. That is... sending them a signal of 50/50 on/off will not result in a 50% of full-rated capacity flow. Neither is 25% or 2%. Depending upon the injector brand-model, sometimes it lower than calculated, sometimes its higher. And on a single injector, it can be higher AND lower than calculated, depending upon where you are on the duty-cycle curve.