UrS4 2.2T Avant project

[prj]

The Eaton M62's electric clutch is the most common source of failure on those and it is difficult to find them new.

The clutch has been abandoned in favour of recirculating bypass valves on the newer TVS chargers.
This is also the case on the 3.0 TFSi Audi engine, where the bypass valve is electronically controlled by the ECU.
In this freely-spinning bypass mode the TVS supercharger has a parasitic loss of less than 1hp.

You don't have to use a fancy electric valve, it is also possible to use the engine vacuum to open it, but it is better with a solenoid that is PWM mapped against torque request (or throttle position), as that will improve response and on-off transition. PD superchargers are otherwise very surgy at the moment that the valve closes.

[Thuppu]

Yes, I have heard that those are prone to fail, thats why I don't wan't an electric clutch on my. Like you told Dmitri, the loss is really low on these when they don't have to work, exactly what happens when the turbo sucks all the cfm the supercharger is producing. Thats why I don't see the idea of an electric clutch on a roots twincharge setup like this.

[prj]

On full boost yes, but when cruising around you still need an openable recirc valve, otherwise the supercharger will be working all the time when driving on part throttle.

Basically 1 valve will solve both the turbo issue and the supercharger parasitic loss issue.
As a quick hack you can use two one-way valves and T them into the bypass valve. But electronic control is better.

[Thuppu]

In fact the dump is dumping air out on idle and light part throttle, so there isn't much pressure the SC has to work against. Yes I know, this isn't the finest system... But, I would love to have you setting up and mapping this thing.

But I still have to resolve the fuel pumps cut out that I get every here and then. Have to try change the relay. It's funny because they always start up after switching off and on the ign. again. I never happens on light driving, only on hard launch/high boost levels.

[prj]

The 3.0 TFSi uses a stepper type valve with a feedback.
So there is a position sensor and then input to move the flap one way or the other way.

A little like a modern electric VNT actuator, although those come in both PWM and Stepper types.

A simple solution with two valves like you are using now is keep a standard dump valve to recirculate between holset inlet and holset outlet.
And a second electronic valve, which can be either some clever PWM thing or for the simplest solution a 2.0 TFSi bypass valve with the USP motorsport adapter flange (same connection like standard dump valve) to recirc between the supercharger inlet and outlet.

For control there are two conditions to satisfy:
1. For cruise area, it would be the best to have a TPS x RPM map, and then values open/closed. Or if you are using a more clever valve, then PWM (I am not sure what would happen if you were to PWM the 2.0 tfsi dump valve, I don't think it would like it).
2. For high boost area it would be nice to have simply a single value MAP limit from which the valve opens again. The supercharger is providing a fairly fixed amount of boost, so this limiter can be set a little over this value (this means the turbo has spooled up), and it will set the supercharger to recirc mode.

So for example you floor the car from low RPM then first the valve closes so the supercharger can provide boost.
Then after the big turbo spools up, the valve opens fully to allow the turbo to feed with minimum restriction.

The only thing that will maybe be a little annoying is partial throttle. When the recirc valve gets switched you will have at some TPS position a surge of power.

Funny enough I think I could manage this all with the stock ECU, but I am not sure about VEMS, as I can not edit the firmware and I don't know if there is enough configuration options to do this. Maybe use nitrous control for this or something.

[quattro-nick]

dimitri,do you have experience with this type of compressor?
it's like a belt driven turbo
this is nicer if you have no room for something bigger

[prj]

Using a centrifugal supercharger is a waste of time in a twincharged project, because of the way the centrifugal supercharger delivers boost.
The only type of compressor really worth using in a twincharged application is a positive displacement compressor.

[Thuppu]

If you think about it, I don't see any benefit in re-circ the air in the top end. All air that goes thru the SC will be sucked in the turbo + then the gravity operated one-way valve also opens to give the extra air the turbo needs(this works without any outside control and just when needed). This seems to provide a really smooth powerband and transition between the SC and turbo. If the re-circ valve is open in the top end, it will propably not re-circ any air as the turbo sucks it all and will propably also suck air thru the re-circ valve.

[Thuppu]

Fuel pumps cut out seems to be gone. First checked the OE "372" relay under the dash, it gives the control signal to my 70A "power" relay that feeds the fuel pumps. It looked ok, cleaned the connectors, but no luck. Then right after drive I touched the 70 amp relay and it was burning hot. When I lookt at it closer I saw that one connector cover was molten. I'll guess there was a bad connection in the connector because it havn't cut out afterwards when I took it off and put back in. The relay propably heated up and loose connection, when the pumps started to draw big amps at WOT. Well, I did buy a new 80A relay, the old one may have been damaged...

Also my fuse box seems to got some heat issues, the cover have molten on the side that drives the 2x044 pumps with 40A fuse. I have a second 20A fuse for the Walbro in tank pump.

[Thuppu]

Ok, seem that the pump cut out is still here with the new relay. Yesturday couldn't get it to show up anymore, but today it was present multible times.

Can't see any errors in VEMS, no strange behaviour. Cutted out at quite low rpm, ~3000, and not even at full boost/throttle. Anyone, look at the first log I posted before, if you can see anything I can't.

[diesel des]

Could you disconnected the vems feed to the relays and put a permanent 12v feed on the coil side of the relay to hold it on. Then go a careful drive as yiu have no crash protection and pumps will continue if engine cuts.

This would rule out if it's vems or the fuses/relay.

I had vems do weird things with boost control and once I reloaded the firmware it was fine. Worth a try.

[Thuppu]

Yeah, have to test that. It cuts the engine in the second the fuel pumps stops, BANG. Maybe it's also cutting the injectors..?

I use the stock AAN wiring for the control feed, so I'll guess there are some things that can cause the failure in there, not only VEMS. I re-flashed the ecu when I needed the MAT- and coolant patches. It made no difference in the cut out. But with a cold car, it seems to cut even easier.

[A1QSHIP]

The valve in this video is what you want. It is PWM controlled.

https://www.youtube.com/watch?v=vyhX5MpRz18

HTH, Chris

[Thuppu]

Drove home from work and tested to drive to get the cut out again, no problems to get it "activated". But now I realised I don't have to cut off and on the ignition. While the car is rolling I just need to press the clutch and let the engine die, and when I release the clutch the pumps and engine fires up. What does VEMS need to run the fuel pumps? Does it need some trigger signals like Motronic?

[Thuppu]

Problems seems to be resolved... I hope.

I did override the fuel pump control with a wire connected directly from the battery... and it still cutted the pumps, lol. So there wasn't anything else than the fuses left, so I unplugged them... tadaa!



The 40 Amps fuse looked funny. Also the fusebox had some damage, so I changed everything to new. Didn't get cut outs anymore. Everything was so strange because you could babydrive all day, but when got more torque the cut out appeared.