964 AWD system understeer vs 993 AWD system.
#31
Rennlist Member
this is getting awfully technical and is getting hard for me to follow... but what i would say is that i love the 4wd systems in modern performance cars like mitsu evo/nissan gtr/subaru sti as well as of course the better (non haldex) audis... with my newly acquired 964 c4 i can feel the lovely solidity and grip it affords compared to the classic light nose feel of a rear drive 911...but i am trying to figure out how to extract/coax out the very best the car can offer when driven near/at the limits of adhesion in a sporty manner
it is well known and legendary that the car's 4wd system comes from the 959 - i am just trying to understand how it compares to the modern, 'smart diff' systems that can almost instantaneously direct drive force to any of the four wheels as needed, then also apply inside wheel brake steer the correct corner to promote the car to pivot the right way towards an apex
clearly the 964/959 does not have the brake-steer system engineered into it... but it seems to me it also just reverts an all or nothing center diff lock to primarily prevent the car from spinning (as that was the primary criticism of older porsches by the consuming public through the ages) when it senses any loss of traction... and that in itself is awfully crude and not at all conducive to 'fun' at-the-limit behavior
it is well known and legendary that the car's 4wd system comes from the 959 - i am just trying to understand how it compares to the modern, 'smart diff' systems that can almost instantaneously direct drive force to any of the four wheels as needed, then also apply inside wheel brake steer the correct corner to promote the car to pivot the right way towards an apex
clearly the 964/959 does not have the brake-steer system engineered into it... but it seems to me it also just reverts an all or nothing center diff lock to primarily prevent the car from spinning (as that was the primary criticism of older porsches by the consuming public through the ages) when it senses any loss of traction... and that in itself is awfully crude and not at all conducive to 'fun' at-the-limit behavior
#32
Three Wheelin'
Hi ,
not sure if I understand the way it transfer all to the front diff, and how it could be controled, mechanically via a **** like the lightweight or Ruf cars or with an ECU, actually I have seen on a Ruf turbo 964 it only has one **** for front and back setting ,
one question ; how does the Bosh hamer activates the slaves for bleeding then ?
on the thread concerning converting C4 gearbox to C2 there is some info about a lateral slave that could be set manually or by cable .
I try and dig the picture .
not sure if I understand the way it transfer all to the front diff, and how it could be controled, mechanically via a **** like the lightweight or Ruf cars or with an ECU, actually I have seen on a Ruf turbo 964 it only has one **** for front and back setting ,
one question ; how does the Bosh hamer activates the slaves for bleeding then ?
on the thread concerning converting C4 gearbox to C2 there is some info about a lateral slave that could be set manually or by cable .
I try and dig the picture .
#37
Myles, Yes I agree. When the rear diff locks the car's dynamic changes abruptly and although it may be saving your bacon it may also be disrupting the driver's expectation of the car's trajectory unnecessarily.
I am thinking of 2 possible solutions. One is to try a solid state accel offered by JDSPorsche. I have read that this device may be less prone to what I'll call 'false triggers'.
The other is to disable the lateral accel altogether. This would leave the wheel spin function intact while disabling the oversteer function. I am guessing removing the accel and putting a resistor across the appropriate pins would do the trick to fool the ECU. One could also open the accel and immobilize the pendulum (the factory accels are mechanical).
I have an old accel that was 'fuzzed' up with iron metal filings, a common failure mode, that might work as a guinea pig if you (or anyone else) is interested.
Sadly, I have sold my C4 and cant try this out myself. I truly miss these discussions even though I am now more of a voyeur than a participant on the forum.
Dan
I am thinking of 2 possible solutions. One is to try a solid state accel offered by JDSPorsche. I have read that this device may be less prone to what I'll call 'false triggers'.
The other is to disable the lateral accel altogether. This would leave the wheel spin function intact while disabling the oversteer function. I am guessing removing the accel and putting a resistor across the appropriate pins would do the trick to fool the ECU. One could also open the accel and immobilize the pendulum (the factory accels are mechanical).
I have an old accel that was 'fuzzed' up with iron metal filings, a common failure mode, that might work as a guinea pig if you (or anyone else) is interested.
Sadly, I have sold my C4 and cant try this out myself. I truly miss these discussions even though I am now more of a voyeur than a participant on the forum.
Dan
#38
Three Wheelin'
You mean , disable the yaw sensor ?
sily question
oes anyone knows how is the front diff LSD working ?
pressing the PDAS button for 10 seconds to disable the system may not be a great idea then, does it diseable the whole system ?
can the longitude lock be unplugged , so the lateral lock still works, and the system stay at 31/69 all the time .
( refering to the C4 gearbox converted to C2 thread)
sily question
oes anyone knows how is the front diff LSD working ?
pressing the PDAS button for 10 seconds to disable the system may not be a great idea then, does it diseable the whole system ?
can the longitude lock be unplugged , so the lateral lock still works, and the system stay at 31/69 all the time .
( refering to the C4 gearbox converted to C2 thread)
#39
Advanced
Thread Starter
Kos11-12 The issue with disabling the system is it creates open diffs. You tube how open diffs Work if it is unclear there is some very good videos. But it’s not what you want for any kind of racing application.
I zoomed in in on the c4 lightweight and it appears to me behind the ***** there are master cylinders. I also zoomed in on the ruf, it also looks like it has hydraulic connections behind the *****. In the pic it has an arrow to the **** that says regulates ratio of pull to the front wheels. That’s obviously for the Center diff. The **** to the right looks like it says latitudinal. It’s kinda hard to make out tho. But that would make sense to control the rear diff.
I realized my early suggestion for installing pressure reducing valves and routing excess pressure to the resivoir would not work as it would be very difficult to bleed the system and I’m sure you would get air in the system.
Goughary can you attach a link link to the setup with the bolt. I’m interested to read it.
Also so another correction to one of my statements. When I said the centre diff is 31 to 100 percent diff bias to the front. 100% would mean the front wheels would spin at the same as the rear wheels because the diff would be locked. I didn’t mean it would be a fwd vehicle.
John McM that’s a super interesting video.
Dlpalumbo my initial thought was was to disable the lateral yaw sensor. I will try something to trick the control module and see what happens.
I zoomed in in on the c4 lightweight and it appears to me behind the ***** there are master cylinders. I also zoomed in on the ruf, it also looks like it has hydraulic connections behind the *****. In the pic it has an arrow to the **** that says regulates ratio of pull to the front wheels. That’s obviously for the Center diff. The **** to the right looks like it says latitudinal. It’s kinda hard to make out tho. But that would make sense to control the rear diff.
I realized my early suggestion for installing pressure reducing valves and routing excess pressure to the resivoir would not work as it would be very difficult to bleed the system and I’m sure you would get air in the system.
Goughary can you attach a link link to the setup with the bolt. I’m interested to read it.
Also so another correction to one of my statements. When I said the centre diff is 31 to 100 percent diff bias to the front. 100% would mean the front wheels would spin at the same as the rear wheels because the diff would be locked. I didn’t mean it would be a fwd vehicle.
John McM that’s a super interesting video.
Dlpalumbo my initial thought was was to disable the lateral yaw sensor. I will try something to trick the control module and see what happens.
#40
Three Wheelin'
Exactly ; so open diff is not good as wheels spins at the same speed .
My Other question is; what is the lsd on the front diff , if it’s not great I guess when the car runs at 100 % ( 50/50 front and back)
then the car is not turning well ,
so idilly living it to 31/69 is best with a rear diff locked .
Then can we stop the longitude lock to work by disconnecting it like Jack did on the C4 to C2 gearbox convertion ?
My Other question is; what is the lsd on the front diff , if it’s not great I guess when the car runs at 100 % ( 50/50 front and back)
then the car is not turning well ,
so idilly living it to 31/69 is best with a rear diff locked .
Then can we stop the longitude lock to work by disconnecting it like Jack did on the C4 to C2 gearbox convertion ?
#41
OK, so I’m right and I’m wrong The system isn’t fully variable, it is on/off but it pulses to get fine control.
However mine doesn’t pulse, because something is broken…
The second paragraph says 'a finely metered set of pressure impulses'.
So that confirms that the system is full on/off, however it is using Pulse Width Modulation to feather the effect.
As says dlpalumbo the system works just like ABS.
So from an engineers point of view the statement 'torque ratio can be changed infinitely from 0% to 100%' from Adrian’s book is wildly inaccurate. The statement 'on average, a continuous range of torque distribution can be achieved' from the Allrad book is correct.
Under acceleration in the wet I can feel the center lock pulsing as it locks up to stop the front wheels spinning, so that is working.
The first paragraph says 'pressure is increased until revolution variations return to within the permitted limits'.
So the solenoid will lock continuously until the differences of wheel speed drop back to almost the same.
Whenever I get massive PDAS induced understeer it is the second case.
For instance a couple of weeks ago I hit a patch of mud in the middle of a wet corner and the inside rear wheel started spinning. The rear diff locked and stayed locked until I lifted.
Since it locked and stayed locked that would indicate that there was still a big speed difference between the wheels even though the diff was locked.
Logicaly that would indicate that my diff plates are worn, or that there is air in the system, or some other problem and so the diff is not able to lock 100%.
If the diff had locked properly then the wheel speeds would have matched almost instantaneously and then the system would have released the lock and started to use PWM to give fine control.
When cornering the system uses the two accelerometers and the speed to calculate the trajectory so that it knows what speed difference to expect between the rear wheels. So another cause could be a sticky/tarnished accelerometer confusing the calculation.
#42
Burning Brakes
if i wound change something at the AWD system. I would disable the connection from the solonoid-valves to the activation cylinders and would replace this linkage with a manual "sputnik" solution like was done in the lightwheight car. I expect this as a quite easy solution, when a proper sputnik adjustment wheel **** item is available anywhere?!
This should just work as a brake-bias valve works with some measurement gauges and then it is just the piping, what has to be done...
This should just work as a brake-bias valve works with some measurement gauges and then it is just the piping, what has to be done...
#43
The fineness of the pdas response is in direct proportion to the sensor inputs. The abs sensors monitor tire rotation whose rate of roation can change quickly. Thus the pdas can respond in fine time increments as described above. The Dynamics are different when the car rotates or drifts. In this case the mass of the car dictates a much slower response. The pdas will lock the rear end until the sensor detects that the car has stoped rotating. This process has a much slower response rate (closer to a second in my experience) and you will feel the car change direction, i.e., go straight, or understeer. It takes a couple of cycles to settle down.
If you disable the lateral Accel you will only defeat the rotation response, I.e., locking the rear diff. I expect both diffs will still lock in response to wheel spin.
The car's Dynamics should then be more classic 911, with easily induced oversteer.
The lateral pdas response has kept my car from transitioning to oversteer which can be frustrating if your trying to induce rotation to compensate for the inherent understeer. It has also kept the rear end where it belongs when it unepectedly stepped out on me. If anyone figures out how to disable the lateral Accel, maybe incorporating this with a pushbutton on the steering wheel would provide interesting flexibility. When pressed, the button would inhibit the lateral Accel to allow oversteer until the driver decides he's executed the maneuver successfully or he needs to pdas help.
If you disable the lateral Accel you will only defeat the rotation response, I.e., locking the rear diff. I expect both diffs will still lock in response to wheel spin.
The car's Dynamics should then be more classic 911, with easily induced oversteer.
The lateral pdas response has kept my car from transitioning to oversteer which can be frustrating if your trying to induce rotation to compensate for the inherent understeer. It has also kept the rear end where it belongs when it unepectedly stepped out on me. If anyone figures out how to disable the lateral Accel, maybe incorporating this with a pushbutton on the steering wheel would provide interesting flexibility. When pressed, the button would inhibit the lateral Accel to allow oversteer until the driver decides he's executed the maneuver successfully or he needs to pdas help.
#44
if i wound change something at the AWD system. I would disable the connection from the solonoid-valves to the activation cylinders and would replace this linkage with a manual "sputnik" solution like was done in the lightwheight car. I expect this as a quite easy solution, when a proper sputnik adjustment wheel **** item is available anywhere?!
This should just work as a brake-bias valve works with some measurement gauges and then it is just the piping, what has to be done...
This should just work as a brake-bias valve works with some measurement gauges and then it is just the piping, what has to be done...
I don't think that simple replacing the solenoids with a brake bias valve will work :
When you brake you apply pressure to the inlet of the bias valve, which only lets a proportion of that pressure to the outlet. When you lift off the brake pedal the inlet pressure disappears and the outlet pressure can escape back through the valve to the inlet side.
The PDAS system has a constant pressure on the inlet side (pump and accumulator). So using the bias valve you could increase the pressure going to the diff lock but you wouldn't be able to reduce it because the pressure on the inlet side will always be higher than on the outlet. You would have to stop the pump and wait for the system to depressurize first.
The solution in normal hydraulic systems is to have a calibrated return pipe on the outlet side going back to the reservoir. This allows the fluid on the outlet side to escape, but obviously requires the pump to run constantly to keep the pressure up. I would imagine that the 964 pump would overheat or wear out rapidly because it wasn't designed for that sort of use...
What you could do is to keep the solenoid valves and add a brake bias valve after them to be able to limit the amount of locking effort.
When you brake you apply pressure to the inlet of the bias valve, which only lets a proportion of that pressure to the outlet. When you lift off the brake pedal the inlet pressure disappears and the outlet pressure can escape back through the valve to the inlet side.
The PDAS system has a constant pressure on the inlet side (pump and accumulator). So using the bias valve you could increase the pressure going to the diff lock but you wouldn't be able to reduce it because the pressure on the inlet side will always be higher than on the outlet. You would have to stop the pump and wait for the system to depressurize first.
The solution in normal hydraulic systems is to have a calibrated return pipe on the outlet side going back to the reservoir. This allows the fluid on the outlet side to escape, but obviously requires the pump to run constantly to keep the pressure up. I would imagine that the 964 pump would overheat or wear out rapidly because it wasn't designed for that sort of use...
What you could do is to keep the solenoid valves and add a brake bias valve after them to be able to limit the amount of locking effort.
#45
Race Car
I was thinking the same- to add a bias valve after the solenoid to reduce locking - this might be interesting. Would slow down the immediacy of the locking and maybe make the car more predictable