jpurban

I just sent an e-mail to TRW, but I'm not expecting much... I guess I'm cynical -- Corporate customer service usually disappoints, right? We really need to talk to a TRW engineer with access to their MC specs/design sheets.

I came to the "special bypass" conclusion by accident... I was trying to bleed my PSM system manually using a 12V source to power the PSM boost pump (Who needs a PIWIS? Forget about the non-functional Durametric.). With the GT3 MC, it only worked when the pedal was partially depressed (blocking the bypass). When the pedal is not engaged, the PSM pump tries to build pressure, but the fluid just circulates via the "standard" bypass circuit in the GT3 MC. That is the source of the error -- no pressure builds. So, we actually need a MC without a standard bypass circuit.

We may be out of luck because I suspect a "no bypass" design would be a safety hazard in a non-PSM system. I suspect PSM has a built in bypass function outside of the MC. So, I doubt many options would exist, except for those other MC that are also used in a PSM-like application.

I could be wrong, but this is my current working hypothesis, for whatever that is worth.

246GTSA

Thank you GVA-SFO and jpurban, the bypass makes perfect sense, especially with regards to pressure build up (or lack thereof)
I will be taking the Carrera back again tomorrow for further investigation, I might have to refit the original MC, which would be painful as the GT3 MC has such great feel!
Will let you know the outcome.

DGI

https://rennlist.com/forums/parts-ma...stock-trw.html

Price gouging?

Nickshu

Not really considering they are over $500 from Porsche.

DGI

The link is for the TRW part. If it was $350 for the "Porsche" part i'd grab one up

Nickshu

They are identical parts, Porsche's are made by TRW. You just pay the premium to have a Porsche part number stamped on it and to have it packaged in a Porsche box. I do get that some people prefer the Porsche branded stuff.

cds72911

Also, Aaron (the guy at Rennpart) is a good guy. I find it hard to imagine him "gouging" anyone. He's always treated me very fairly, gives good prices, is very communicative, and packs stuff well/securely. He's an enthusiast, and active member of Rennlist too.

DGI

I understand this. I'm just saying that not long ago, they were around $200-250 for the TRW part and ~$450-500 for the Porsche Part. This is the TRW part thats ~$100 more than it was not long ago

Im sure he is, and im all about capitalism... but im also a cheapskate LOL Too many projects to be handing out extra hundreds if I can avoid it.

I sure wish I got my ROW 030 suspension kit for $900 back in the day as opposed to the $1200 I paid for it about a month ago. If this is the new price for the M/C, I'll pay it eventually but i'm still hoping for a deal

Nickshu

Agreed. Basic supply and demand. There is low supply and high demand. Aaron would be a bad businessman not to up the price a little bit.

moburki

Does anyone know if error 4460 is limited to PSM equipped cars? Nickshu, did you get this error on your 996.2 (iirc) car?

Nickshu

I had no error messages on my 2003 Carrera 2 with PSM when using this master cylinder. None at all. Searching here the problem seems to be with Turbo cars. The PSM pump location is different on Turbos...it's right under the master cylinder. On my C2 the pump was over on the passenger's side of the frunk. Not sure if the system is different or just the location of the pump but that could indicate some sort of difference in the two systems.

RennPart

Anyone that needs a TRW master give me a call directly, I have two left. Yes the price is a bit higher, but I haven't been able to get them in the past several months and actually needed one for a GT3 in the shop recently so I don't want to give them away. Not sure when more will come in as they're still on BO with TRW. Often times when this happens the price ends up going up, recently they've been blaming it on "tarrifs."

GVA-SFO

@jpurban, about error 4460 and bypass : I'm still thinking and to tell you the truth, I'm not yet to the full understanding point about the "bypass" issue !

The ABS ECU is triggering the Error 4460.
What are the possible sources (sensors) involved to have the ABS ECU triggering this error ?
In my current understanding, there is the brake switch sensor, and there is the pressure sensor, ..that is located inside the PSM pump assembly.

What I’m missing in this decision chain ???
Most probably the reading of the 4 wheels sensors (rotation speed of each wheels), but I doubt these are involved or are at the source of our specific problem.
(Btw, I wonder what would be the error code if one of the wheel sensors would fail.)
Also, I do not think that the pressure sensor is sending “analog info”.
I think this pressure sensor gives an “off/on” info, and this, according to a threshold level.
(But, ..I may be wrong on that !)
Anyway, even if it would be an analog info, I do not think this would have effect in “our” case. I.e.: if the GT3 Master Cylinder would give more pressure, than the OEM MC,
Btw, this would easily trigger the ABS on every braking !
The larger piston should give more quantity of fluid on the move, but, the pressure should remain about the same, ..this for sure, until you apply a brake pressure to a level that would then trigger the ABS.
(Which, in my case never did happen when I have the 4460 error (seen in this case with dash board: ABS and PSM lights on)!

So, at my current understanding level, the amount of pressure is not the problem, but the buildup of the pressure (our perception that the braking point is firm and efficient) that happen to be much more earlier in the travel (position) of the brake pedal.
I think in the ABS/PSM ECU, the function diagram for a normal operation should be : first, to see brake pedal action, i.e.: brake light on, and then, “a bit” later, pressure building up, that do trigger pressure sensor above the threshold level.
To my understanding, in the 4460 error seen in our cases, I still think that the ECU get the “brake on” too late, i.e.: the ECA sees that there is already pressure reading from the sensor, ..or: ..thinking that something is wrong with the pressure sensor (i.e.: error if pressure is up, ..when brake pedal is “off”, or the ABS (or PSM) pump is generating a pressure when it should no.

If there is a kind of bypass in the master cylinder (or in the PSM assembly), I can only think that this should just delay the pressure build up. I do not think that there his another sensor that would read specific information from a bypass circuit.

Or : at this point, I would still think that the ECU when checking these two info, is not happy if the pressure if up too early versus the pedal position, taken from the brake switch !
The fact that the PNM166 has a larger piston, it is very clear that the pressure build up earlier in the travel position of the piston.
And, to me, it seems logical that If we have “pressure” on, before “brake on”, this situation should trigger an error.

I'm still not 100% sure of the above theory.
In the mean time, I will try to trim one of my two brake switches, in order to try to verify the above.

jpurban

@GVA-SFO,

The 4460 error is related to the analog voltage of the brake sensor on the back of the ABS unit (not the PSM boost pump). The sensor has a value of about 0.6V at 0 psi. The non-zero voltage allows the system to verify the sensor is present during initialization. At max pressure, the voltage is about 2.2V, about 3,000 psi, if I recall correctly.

The brake light switch is a binary switch, open or closed. If it fails to close the circuit by the time the ABS pressure sensor is reading a materially higher voltage, an error is set. This sets a 4460 error. It is corrected by making sure the adjustable plunger is set properly (usually) or replacing the switch once the internal contacts wear out.

If you think this is the reason the PMN166 sets the 4460 error, then test it yourself. Start your car, but don't drive it. Repeatedly press the pedal slowly and see if the error sets without exceeding 30 mph. I've done this while monitoring the pressure voltage and the stop light activation -- no error sets. So, I don't think the issue is related to the brake/stop light switch.

This is confounded by the 4460 error set during the PSM boost pump check. The PSM boost pump check is done after start, vehicle above about 30 mph and no brakes applied (brake pedal/stop light switch is open). If the analog pressure does not rise when the PSM pump is engaged, then error 4460 is set. I believe this is the issue we encounter when changing to the 997 MC. The 997 does NOT have a separate PSM boost pump, which dramatically alters the braking circuit, particularly the need for a non-standard bypass circuit.

The 996 cars have the Bosch 5.7 ABS unit, which is also found in BMW (ESC nomenclature), Audi and VW cars of the same period. There are minor differences, but all are generically Bosch 5.7. Check for yourself... You'll find BMW has two types of master cylinders, ESC and non-ESC. The master cylinder design is different. My hypothesis is this... 996 bypass is in the PSM boost pump -- not the master cylinder. Note that both are connected to the reservoir.

The 997 cars have a different ABS system, which does not have a separate PSM boost pump. Therefore, they use standard master cylinder designs, with typical bypass circuits that allow fluid to flow back to the reservoir when brakes are not applied.

Unfortunately, I couldn't find any PSM/ESC master cylinders with pistons larger than 25.4mm. So, we can't source a master cylinder from another vehicle that uses the Bosch 5.7 system. I think we're out of luck unless we build a custom MC. You might be able to do this by using the 997 piston/plunger assembly in a 996 housing that has been bored from 25.4mm to 27mm, but I'm not sure. We really need someone to disassemble the 2 units side by side to determine what makes them different and then go from there. Unfortunately, I tossed my old 997 MC when I reinstalled the 996 MC -- Kind of a "F*%$ this" moment.

If someone has an old/extra 996 PSM MC and a 997 MC, I'm happy to disassemble them and compare the internals. The plunger/piston is easy to remove -- It is held in place by a snap ring. The only difficult part might be disassembling the two reservoir-to-piston channels -- There is a screen in there that might be press fit.

John

Look at the path from the charge pump to the master cylinder below... If the MC has a normal bypass, then no pressure can build when the brake is not applied. Also look at the front circuit bypass in the MC.

Bosch 5.7 System



Sensor schematic

Max brake pressure reads 2.2V. Full scale deflection is 5.0V.

Zero brake pressure in idle is 0.6V.

GVA-SFO

@@jpurban,
..you did a very important study on this subject, I'm very impressed John !
Thanks a lot to share this very precious information with all of us.
OK, from your reading, I understand and fully acknowledge that the pressure sensor is analog. So, the threshold level to trigger decision is set in the firmware code.
This does help me to understand.
If I catch it right, the 4460 error (in these cases) is not triggereing by a "braking" situation, but by a testing phase.
I did effectively got a error when driving, kind of "cruising" mode, at about 25 to 30 mph (not using the brakes), where suddenly, the PSM and ABS lights turned on together.
According to your study, it seems that the error is triggered by an "internal testing process", rather than a "normal" braking situation !
With a Durametric, we can manually start (and stop, as, it should NOT be running too long !) the PSM pump (this as well as the ABS pump).
Would you think (I can try that one of these days), that ..doing so, would also trigger 4460 ?
(This would be a clear proof that you are completely right with your findings !)

If this is the case, one tricky way to turn around this problem would be to suppress this testing in the firmware, ..but I agree that suppressing such test, ..is may be not a very clever idea !
Still, this way could be less expensive ..than boring a standard OEM master cylinder and equip it with the piston of a PMN166 !

Need to think again !
Anyway, thanks a lot again,

PS : We just lost your photos/diagrams ! (Images server down ?)