944 Turbo DIY TunerPro Chips

[cmurphy034]

I guess my question is more so does the DME chip have any direct reference to manifold pressure for the overboost protection you show in the YouTube video (and otherwise)? Or is that reading entirely handled through the manifold pressure reason the stock KLR sees? I believe you’re right that the MAP setup in the Focus 9 Tech DME is just an extra input to supplement the AFM reading and improve throttle response, etc.

And also another question, is the KLR chip you have in the downloads section specifically tailored to the carpokes performance chip or can that KLR chip be used for any tune I put onto the DME? And if I do use the KLR chip without an MBC or EBC is it now relying on the DME chip overboost protection alone using that “overboost” map you mentioned in the video?

Sorry for so many questions I appreciate the feedback.

The factory DME does not have any direct reference to manifold pressure for the overboost shown in my video. The video shows the DME's overboost map, which is actually triggered when engine's calculated load (as measured primarily by the AFM) exceeds pre-set thresholds. In other words, even though it is called an 'overboost' function, it is not actually monitoring boost. It's monitoring engine load as a proxy for boost. I don't know enough about the F9 MAF system to know if it incorporates its extra MAP sensor into the DME's overboost features, but I would be quite surprised if it did.

The KLR has its own, separate version of overboost protection that does use the map sensor in the KLR. It can trigger error codes and reduced boost pressure via the cycling valve when the actual boost exceeds pre-set thresholds, which is why you'd want the modified KLR chip if you are increasing boost but still using the cycling valve.

The experimental KLR chip available for download is not specific to the Carpokes performance image. You can tune your own DME chip however you want and still use that KLR chip. Keep in mind that the primary way to make more power via chip tuning is by tuning the motor for increased boost. So, if you want to keep the cycling valve and make noticeably more power, you'll need to find a way to increase boost without a boost controller. Back in the day, people would do that with a restrictor in the banjo bolt, shimming the wastegate, drilling a vent hole in the cycling valve, etc. -- none of which is particularly easier to install than a MBC -- and none of which offer any real control/adjustability over the amount of boost you get. Just something to keep in mind as you plot your course...

And, finally, yes, if you use the experimental KLR chip to eliminate the KLR's overboost protection, you will be relying solely on the DME's overboost protection -- or no protection at all if you max out the overboost tables in the DME chip. I'd hazard a guess that most modded 951's don't use either of the factory overboost protection features. Both are designed to prevent meaningful increases in boost/power, which is the whole point of tuning a chip.

Great, thank you for the detailed explanation. I’ll likely look for an MBC or EBC for this

[whalenlg]

Does anyone have KLR schematics?

I'm working on my reverse engineering of the DME and am getting stuck on the KLR delay circuit SPICE simulation that doesn't seem to do anything with simulated 0-5 or 0-12 pulses coming in from the KLR - the pulses are just getting swallowed by the 1st stage of the circuit.

Also - if any one has info on the S400 and S100 chip functions/block diagrams, that would also be helpful. But they are likely buried inside Bosch somewhere.....

[johnb]

Does anyone have KLR schematics?

I'm working on my reverse engineering of the DME and am getting stuck on the KLR delay circuit SPICE simulation that doesn't seem to do anything with simulated 0-5 or 0-12 pulses coming in from the KLR - the pulses are just getting swallowed by the 1st stage of the circuit.

Also - if any one has info on the S400 and S100 chip functions/block diagrams, that would also be helpful. But they are likely buried inside Bosch somewhere.....

I'm not aware of any proper schematics for the KLR but I have a rough hand-drawn schematic of the knock sensor signal conditioning on my site: https://jhnbyrn.github.io/951-KLR-PAGES ... dware.html

It only shows the basic configuration of each op-amp, not the actual values used, but it is an accurate description of the processing pipeline.

I did a Spice simulation for the filter stage too with component values.

I don't know what you mean by "0-5 or 0-12 pulses" - could you elaborate on what you're doing?

Which one is S400?

For S100, I presume you mean in the DME? I'm attaching the schematics of the DME that have been doing the rounds for years in case you don't have them. I have no idea who created them. They're not perfect - there's at least one mistake with either the injector or ignition coil signal from the 8051, I forget which. But they are pretty handy.

The S100 creates digital pulses for the speed sensor, one for each tooth. But for the ref sensor it latches and stays low until it gets reset by the 8051. I can point you to the code that does it if you want. I have not been able to figure out why it does this - as far as I know, the state of the digital ref sensor input to the 8051 is never read explicitly, so the latching behavior seems pointless to me. (see https://jhnbyrn.github.io/951-KLR-PAGES ... dware.html)

[whalenlg]

Thanks for the response - yes I have the DME schematics already and have Spice models for most of the circuits for the analog page working already.

I don't know what you mean by "0-5 or 0-12 pulses" - could you elaborate on what you're doing?
-modeling inputs to the circuit - a square wave going from 0v to 5v or 0v to 12v and repeating to make sure I can detect some output from each circuit, measure the voltage, delay, inversions, etc.
The knock detector schematic interesting, but I really just need to know if pin 32 of the DME is a 5v or 12v signal.

As far as the DME schematics - I suspect there's an schematic error on the input stage of the timing delay circuit. Might need to visually inspect a DME. My spare is a NA, so that's kind of out (probably won't have KLR circuitry).

Thanks for the S100 info - I can probably fake that functionality with your description.

S400 is the injector driver IC - I think I can fake that one by just bypassing the input to the output.

[johnb]

Thanks for the response - yes I have the DME schematics already and have Spice models for most of the circuits for the analog page working already.

I don't know what you mean by "0-5 or 0-12 pulses" - could you elaborate on what you're doing?
-modeling inputs to the circuit - a square wave going from 0v to 5v or 0v to 12v and repeating to make sure I can detect some output from each circuit, measure the voltage, delay, inversions, etc.
The knock detector schematic interesting, but I really just need to know if pin 32 of the DME is a 5v or 12v signal.

As far as the DME schematics - I suspect there's an schematic error on the input stage of the timing delay circuit. Might need to visually inspect a DME. My spare is a NA, so that's kind of out (probably won't have KLR circuitry).

Thanks for the S100 info - I can probably fake that functionality with your description.

S400 is the injector driver IC - I think I can fake that one by just bypassing the input to the output.

The ignition signal from the KLR to DME pin 32 is 5v (see scope traces here: https://jhnbyrn.github.io/951-KLR-PAGES ... iming.html)

I just had a quick look and I think the mistake I was thinking of was that the schematic seems to show p1.0 of the 8051 (which is pin 1 on the IC) as the O2 sensor input, when in fact it's the injector signal output.

There isn't really a "timing delay" circuit - the delay is in the software. I would expect the signal that's returned from the KLR to be handled exactly the same in NA and Turbo versions. All it needs to do is drive the coil

[whalenlg]

Thanks for the response - yes I have the DME schematics already and have Spice models for most of the circuits for the analog page working already.

I don't know what you mean by "0-5 or 0-12 pulses" - could you elaborate on what you're doing?
-modeling inputs to the circuit - a square wave going from 0v to 5v or 0v to 12v and repeating to make sure I can detect some output from each circuit, measure the voltage, delay, inversions, etc.
The knock detector schematic interesting, but I really just need to know if pin 32 of the DME is a 5v or 12v signal.

As far as the DME schematics - I suspect there's an schematic error on the input stage of the timing delay circuit. Might need to visually inspect a DME. My spare is a NA, so that's kind of out (probably won't have KLR circuitry).

Thanks for the S100 info - I can probably fake that functionality with your description.

S400 is the injector driver IC - I think I can fake that one by just bypassing the input to the output.

The ignition signal from the KLR to DME pin 32 is 5v (see scope traces here: https://jhnbyrn.github.io/951-KLR-PAGES ... iming.html)

I just had a quick look and I think the mistake I was thinking of was that the schematic seems to show p1.0 of the 8051 (which is pin 1 on the IC) as the O2 sensor input, when in fact it's the injector signal output.

There isn't really a "timing delay" circuit - the delay is in the software. I would expect the signal that's returned from the KLR to be handled exactly the same in NA and Turbo versions. All it needs to do is drive the coil

Thanks - 5V. Yeah - I know the timing delays come from SW, but they have to get from software on the KLR to the coil through the circuit that just isn't working in my simulation.
BTW - NA's don't have KLRs to pull timing, so those DME boards likely have a similar circuit that goes from DME digital board to the analog board to the coil. I'll double check the schematics you sent to see if there are revised from my 1988 version.

[333pg333]

None of this is pertinent to my car however I’d like to thank and congratulate Tom and others in this post for investing in these wonderful cars. Really impressive!

[Tom]

None of this is pertinent to my car however I’d like to thank and congratulate Tom and others in this post for investing in these wonderful cars. Really impressive!

Thank you Sir! Most of the thanks goes to @johnb, @whalenlg, and @Dave W. for cracking the code, and helping to figure out the hard parts. I'm kind of the marketing department -- they're Engineering.

[333pg333]

Well congrats to all involved!

[Tom]

This reminds me, I'd like to find a volunteer or three to test out my baseline performance chip for the 951. Only catch is we'd need a stock car with wideband (and preferably a boost controller) so we can log the AFR and fine tune as needed. If you have or know of anyone with a stock 951 who might be interested, please let me know! I'd supply the chips for free, and dial them in based on the wideband logs. I'll even throw in a pre-wired knock counter to add more data to the mix and make it worth you while!