UrS4 2.2T Avant project

[Thuppu]

Well des, you have to try it. I was first thinkin of a setup like you are suggesting. But I think this setup is brilliant in all its simplicity. I'll say it again, when a roots doesn't work against pressure, it will spin freely and won't produce heat.

Lets say the supercharger flows 600cfm and the turbo flows 900cfm. The turbo gets 600cfm from the supercharger and then the pressure between supercharger and turbo has drop to zero. When the pressure drops below zero(to vacuum) the one-way valve starts to open up and the turbo gets the last 300cfm thru the valve.

[quattro-nick]

For the coolant patch, select NTC 256/1900.this is when you use the origional white sensor on the back of the head
if you use an other sensor for iat or coolant than you can fill in a 17-point curve.
this can be done in "inputs"-CLT or MAT sensor curve dialoge
you can choose "flash table" (this uses the path that you have chosen by upload a firmware) or "17 point curve"(fill in the correct resistance)

maybe this is also interesting
https://www.youtube.com/watch?v=20qqavckWdw
they show a flow diagram

[diesel des]

[QUOTE=Thuppu;850273]Well des, you have to try it. I was first thinkin of a setup like you are suggesting. But I think this setup is brilliant in all its simplicity. I'll say it again, when a roots doesn't work against pressure, it will spin freely and won't produce

Thomas it's a great system and you have done a great job making it fit! I have good response with my stroker and small efr turbo but this setup of yours would be even more responsive and with a big turbo will make more power too. What's not to like!

I still think that even when a supercharger is spinning with equal pressure on both sides it will heat the air slightly and cause pumping losses. The supercharger is still doing work by moving air even though it's not required and the supercharger is quite inefficient compared to a turbo.

It's hard to say by how much this loads the engine and heats the air. I wonder if you put a throttle type valve on the intake of the supercharger and had it operated with a solinoide, and a pressure switch on the supercharger outlet so when it's outlet went to zero pressurr eg turbo spooled and sucking then the throttle shut and stopped the compressor pumping air. That should unload the compressor and might be easier than a clutch.

Just ideas..

[Zoig]

Basically same layout that is used on 1.4 TFSI engine.

Lancia Delta S4 used the alternative layout: turbo-cooler-supercharger-cooler-intake

[heromanni]

I quess if the heat is really a problem. Just switch to a bigger ic

Not seeing too big a problem there.

Anyway great outcome and most of all great that this got done!

[Error404]

This is awesome, well done Thomas. I wondered what crazy **** you were thinking when you mentioned the charger a few weeks ago

[Thuppu]

Alex, I'll guess you was thinking; "now he has lost it for good".

Did a small 50km trip today to a friend. Easy to drive steady 80-100km/h with a -0,5bar vacuum in the intake, but in the second you open little more throttle you get like 0,5-1bar of boost. I did flash the ecu with a new firmware with the AAN coolant- and MAT patch. Now I get movement in the MAT temps. With a warm engine that was turned off, the MAT temp heatsoked up to 50C degrees in about 15min. In steady traffic at 90km/h MAT was something like 13-15C. Outside temp was 6-7C. There was much traffic today so I couldn't test to the limit. But at WOT up to 6000rpm in 3rd gear I got max 22C charge temp. In fact at first the temps started to go down to 10-12C and then slowly rise up to 22C. I think there will not be any problems with the charge temps. And I will keep this pulley on the supercharger.

[quattro-nick]

did you make a fast acceleration with the car from 1 gear untill...........?
if you've been on the dyno, can you show us both results with and without compressor?

[Thuppu]

No, just steady WOT in 3rd from ~2000-6000rpm. But of course need to do better logging when there is less traffic. But there will propably not be any big problems. Someone was telling me I will get something like 200C charge temps with this boost level and pulley setup. Not exactly.

Will see, I may do it if I can change the belt without removing the IC and radiator.

[diesel des]

Good results Thomas. ! The low load cruseing issue with boost building at slight throttle could be sorted with a throttle at the intake of the supercharger. Or a bypass whatever is easiest.

[diesel des]

Thomas some info I found on the net about supercharger bypass systems using a simple wastgate. This is to make light load driving better.

The ideal thing to use for a positive displacement supercharger bypass is an external turbocharger wastegate. This should be mounted so it bypasses the flow directly around the supercharger when open. It does not need to be very large.**Try to imagine how much boost pressure you would lose if you drilled a one inch hole in your supercharger pipework!**

It absolutely must be mounted so it flows in the correct direction.**That is, increasing boost pressure would try to force the poppet valve off its seat.

Now most if not all external wastegates have two pressure inlet ports, one above and one below the control diaphragm. And the internal wastegate spring always holds the wastegate shut.

The trick here is to connect two control air lines, between the wastegate to directly across the throttle body. The idea here, is that the wastegate diaphragm senses the pressure drop across the throttle body, and increasing plenum vacuum opens the wastegate against the internal spring.

It should be obvious that at wide open throttle, the wastegate diaphragm sees zero differential pressure, no matter how high the boost pressure rises. That is the secret....**The air bypass is load sensitive and depends more on the combination of airflow, load, and throttle position, and**not so much on just raw manifold air pressure.

At engine idle there will be sufficient vacuum to snap the wastegate wide open. The supercharger bypass will completely unload the supercharger at engine idle.

At light throttle constant vehicle speed there should still be enough vacuum to hold the bypass open far enough so there is zero measurable boost upstream of the throttle.

A little bit more throttle, and the wastegate closes slightly further and boost may just rise slightly.**More throttle still and the wastegate fully closes, giving full available boost pressure.**it is all wonderfully smooth and progressive in action.

Same when you lift off, the wastegate opens. For a more violent type of throttle movement, the wasteagate slams shut instantly at full throttle, and springs instantly wide open on throttle opening, as in gear changing.

Any boost spikes up stream of the throttle force the wastegate off its seat, so it acts as an over pressure relief valve.

Getting it working properly depends entirely on fitting a suitable spring to the wastegate, and that requires a great deal of thought and attention to get right.

The first requirement is that the spring be stiff enough to hold the wastegate closed against full boost pressure without leakage. But it also has to be weak enough to open the wastegate against plenum vacuum at part throttle.

The problem is the ratio between poppet valve area and control diaphragm area. A ratio of at least 2 x diameter (4 x area) would be a minimum requirement, and many cheaper wastegates do not have sufficient difference in areas to work properly in this application.

But you can work all this out for yourselves. You may wish to be able to drive around with the supercharger unloaded with plenum vacuum higher than perhaps 5 inches Hg.**That is 2.5 psi (roughly) But it must also seal against say? 10 psi of boost. You can see that if the two area have only a four to one ratio, it might be very difficult trying to juggle a spring that can achieve both these requirements.

So try very hard to find yourself a wastegate with a huge diaphragm diameter, and a reasonaby small poppet valve flow diameter. The Chinese gates with the crap diecast bodies and**nasty low temperature rubber diaphragms work fine in this application, provided the area ratios are sufficient. And that can take some searching.**The more expensive "brand name" Japanese gates are mostly excellent, but far from cheap, not all meet the area ratio requirement so be careful.

It may be possible to build a "Frankenstein" hybrid wastegate by mixing parts, to get a higher than standard area ratio. The larger the ratio, the less critical spring selection will be.

While setting all this up, drive around with the wastegate on the passenger seat or floor, and watch what it does. It will respond instantly and progressively to throttle opening, just like the needle on a vacuum gauge does. Get the spring sorted, then install it onto the engine.**I think you will be extremely pleased with the results.

This system works equally well on supercharged or twincharged engines. I have been fitting this to street vehicles for well over eleven years now, with great success, and others that have copied it are also very pleased with the results.

[Dima ABY]

Gentlemans!
You are playing with science and physics.That is amaizing.This build is amaizing.Thats educational,crazy and fun me a lot.
Thuppu the gods of Olympus was told me that you will make it 100% efficent.

[quattro-nick]

did you made some progress today thomas?

[Wizard-of-OD]

Basically same layout that is used on 1.4 TFSI engine.

Lancia Delta S4 used the alternative layout: turbo-cooler-supercharger-cooler-intake

Not really
there is a bypass on the supercharger but yes you get the general idea.

1. Turbo ---> Supercharger ---> intake.
2. Turbo + Clutch driven Supercharger ---> Intake (clutch is controlled by a GPO)
3. Turbo + Clutch driven supercharger ---> Y-valve ---> intake

there are any number of ways of doing this. I did a kit for a gentlemen whom you probably know if you are in that part of the world. Was a 1.8T bored and stroked to 2.0 and we did a clutch set up to use a Rotrex C30-94 + a GT3071R.

[Thuppu]

If some of you guys are doing a similar setup, the Merc Eaton M62 has a electric clutch, but then again it's hard to change the bulley setup. The pulley is also quite big so I guess the boost level won't be much. But they are quite cheap on ebay.

I like to keep this setup as simple as possible for good reliability, and the drivability isn't bad at all. It's like using a sinter clutch, first it feels hard to use, but soon you don't even remember it. While you are on boost everything is smooth, it's just the area in low speed, when the throttle is opening and the dump is closing, that feels funny because it jumps from vacuum to boost quite quick. I'll guess there can be a inprovement when everything is mapped properly.