The Mega Supercharger 4AGZE thread, using twinscrew charger

[knightrous]

Put on the next size pulley and I think the boost will be more consistence

[jondee86]

Put on the next size pulley...

Working on it Gotta find the pulley first as I put it somewhere safe, and usually that means it takes ages to remember where it is hid. Then check the size and get a new belt. Hmmm... seems the memory is not what it was

The Sprintex S5-210 is rated at 940cc/rev. From memory my crank pulley is 130 dia and the s/c pulleys are 110, 90 and 75 dia. So ratios of 1.18, 1.44 and 1.73:1 respectively. According to the old Sprintex calculator, at 7600rpm this should give 50kPa, 82kPa and 117kPa of boost, give or take a few kPa.

Just dug out my order for the pulleys and it seems I wavered a bit... and actually ordered 1@105 and 1@87 dia. That changes the expected boost at 7600rpm to 57kPa (8.4psi) for the 105 pulley and 88kPa (12.9psi) for the 87 pulley. So still short of target, but creeping up on it

Should have all the bits for the new intake by Friday so I can start work on that over the weekend.

Cheers... jondee86

[jondee86]

Noticed a couple of other interesting points while studying the log for that last pull to 7400rpm. That was in 2nd gear, so not a huge amount of data, but these figures caught my eye...



1. The injector Duty Cycle hit 70% which surprised me !! I'm using a set of OEM GZE injectors, but it looks like I may have to change them out for something a bit bigger once I step up to the next pulley.
2. The Water injection looks to be doing a good job. It is currently set to start spraying at 30% IDC and track IDC up from there. I have the AuxT sensor right at the s/c discharge and the MAT in the intake manifold. As you can see AuxT increased by 17degC during the pull, but MAT only increased by 6degC.

EDIT: Got a bit carried away with that bit I have underlined Failed to notice that most of the increase in AuxT happened after I lifted and the spray turned off. While the spray was active MAT increased by 5 deg and AuxT increased by 8 deg. So need to do a few more logs to be able to tell how effective water alone is on this setup. MAT dropped as soon as I lifted (even while AuxT was shooting up) so maybe residual water or the cold ambient (around 12 deg) conditions were at work ?

I have one nozzle installed in the first bend after the s/c discharge and another smaller one in the last bend before the intake manifold. I'd say with a bit of meth in the spray I should be able to hold the MAT steady right thru a pull like that... gotta make tuning a bit easier

Cheers... jondee86

[jondee86]

So a bit of an update. Made a new intake to replace the OEM AE86 stuff that I was using, and gained 1psi of boost due to the lower restriction.



That gave me 143.3kPa @ 7490rpm which was a bit below expectations, but I was happy with the car was running. I did retard the intake cam by 2 degrees but it made no noticeable difference. Plugs are now NGK heat range 8 and that looks like where they need to be.

This weekend I installed the 87 dia pulley and went for a drive. Definitely perked up a bit, but I was still wondering what was preventing me getting closer to the theoretical boost targets, so I retarded the intake cam 2 more degrees for a total of 4 degrees, and advanced the exhaust cam 2 degrees.

I need to get a better laptop as the old one with a serial port that I bought to use with the e420c can no longer keep up with the sample rate of the e440c. But it looks like I should get 175kPa before hitting the limiter, and that is a lot closer to the theoretical target boost. Given that my intake ducting has too many 90deg bends and I am using the factory GZE intake manifold, I'm happy with that

Hit 90% IDC with the OEM GZE injectors, so I have some of 440cc/min 7MGTE injectors on the way to use when I get around to installing the smallest pulley. Still have to do some map tweaking and get a bit of meth mixed with the water.

Cheers... jondee86

[jondee86]

Update and question time I'm running the 75mm pulley now and spraying a 50/50 w/w water/meth mix. Was out for a drive today and found enough clear road for a short uphill pull in second gear. Made 195kPa @ 6000rpm so with another 1800rpm in hand, I think I'll be able to hit 200kPa. Engine MAT and s/c discharge temps are running a lot higher now that the s/c is spinning faster. I can see as high as 90 degC MAT and 115 degC s/c discharge while cruising at highway speeds at 70-80kPa. No intercooler on this car.

So I've been wondering if I need to move up from the standard plastic Bosch bypass valve ? The valve seems to have quite a strong spring, and doing a bit of reading it seems that it could take around 40kPa to get the Bosch valve open, whereas aftermarket metal valves can open with only 20kPa or even less. So at the moment I think that I am cruising with the bypass valve closed, and this is contributing to the high temps. If I changed to a GFB or similar metal valve with a lighter spring, I could cruise with the bypass valve open.

Performance wise, as I understand it, when getting back on the throttle after lifting, the Bosch valve will close sooner. This means you have boost a fraction earlier than with a lighter spring. So better response, but the downside is the s/c does more work during cruise meaning more heat and using a bit more gas.

Has anyone ever experimented with different springs or bypass valves and found any worthwhile advantage ?

Cheers... jondee86

[JustenGT8]

You won't pick the nano second difference is boost response but you will notice huge temp decrease from a bigger bypass volume...go the GFB valve for sure!

[knightrous]

Pretty hard to go past something like the Eaton BPV which is on hundreds of thousands of superchargers from OEM.

[jondee86]

You won't pick the nano second difference is boost response but you will notice huge temp decrease from a bigger bypass volume...go the GFB valve for sure!

Did you actually use the GFB valve and get this result ? I had the Bosch valve off today for checking, and noticed that the inlet and outlet tubes had been affected by the heat and were no longer round. So I am about to order the GFB, and it will be worth the coin if it does bring the s/c discharge temperature down.

If I could have figured a way to fit one of the Eaton/Mini Cooper style servo operated valves, I would have used one as I liked the design. But I didn't have the space, so the Bosch style valve will have to do. They are used on some reasonable capacity Euro turbo cars, and seem to work OK on my 1600. And if the GFB flows a bit more at the same pressure differential, that would be a bonus

Cheers... jondee86

[JustenGT8]

Nah all my sc experience has all been V8... either UZ OR LS.....no question a bigger bypass is better for intake temps though. I have used large aftermarket BOVs to good effect. If you get the flow path the right way ie boost holding closed....there shouldn't be any decrease is response you could feel or even measure.

[jondee86]

No worries... and no chance of fitting a larger valve. But looking at the GFB design it see that it appears to have a springs with a lot more coils than the Bosch. In theory that means that the valve should open sooner and higher for the same pressure differential. So gotta be worth a try.

Cheers... jondee86

[jondee86]

Since I have been putting a lot of time into figuring out what it takes to get a bypass valve working with my twinscrew supercharger, I think it is worth documenting what I have figured out so far. Maybe it will help someone else and save them a bit of time.

1. With a positive displacement s/c that has the throttle before the s/c, there needs to be a bypass to route excess air from the discharge side of the s/c back to the inlet side. You can operate without a bypass (and I did), but at cruise the inlet air temperatures will be high so that performance and fuel economy will suffer.

2. There is a huge amount of uninformed opinion on the interwebz regarding the Bosch bypass/blowoff valve. Some say it can be used and others say no. Some say air out the bottom of the valve and some say air in the bottom of the valve. No-one ever says that this valve will not work on a positive displacement s/c with the throttle body in before the s/c. I tried one with air going in both directions, and IT DOES NOT WORK !!!



3. So lets get the GFB 9301 "Bosch replacement" valve as that is a superior piece of kit. Bought one and tried it and IT DOES NOT WORK !!! Reason; (and this did not dawn on me for a long time) the MAP signal and the air under the piston are at exactly the same pressure. There is never a differential in favour of opening the valve. It will always be shut (same as the Bosch valve).

4. Both of the above valves are vacuum operated poppet valves. Manifold pressure is applied to the top of the valve, on top of a diaphragm in the case of the Bosch valve, and on top of a piston in the GFB valve. Both have a spring to hold the valve closed. Both use manifold pressure to hold the valve closed under boost, and "suck" it open when manifold pressure drops below atmospheric.

Works fine if there is a decent amount of pressure under the valve and less pressure above to provide a differential great enough to overcome the spring pressure. These valves are designed for turbo or centrifugal s/c application where the throttle is AFTER the compressor. They will not open if the pressure under the valve and above the diaphragm/piston are the same.

5. So why does almost every s/c manufacturer use a throttle plate type bypass valve ? When I actually stopped to think about it, the reason is clear. Throttle plate valves don't care about how much pressure is on the valve when it opens... because the load on the top half of the plate is balanced by the load on the bottom half, and the valve can be opened by a piss weak actuator. The only downside with them is that they don't seal tight and always leak a little bit of boost. Can't use one because they are too big and I cannot fit one into my ducting layout.

6. What is the ideal valve for me ? Answer: a double chamber poppet valve, and I just happen to have one in the shed.



This is the OEM 4AGZE bypass valve. The diaphragm has atmospheric pressure above and manifold pressure below. The valve is "sucked" open when manifold pressure goes below atmospheric, and the "suction" force under the diaphrogm is added to the atmospheric pressure above... once the pressure differential generates enough force to overcome the spring. Awkward piece to incorporate in my system, so relegated to last resort.

7. I wanted to use the GFB valve as I had 200 bux into it and it could fit the inlet and outlet I already had welded to my ducting (for the Bosch valve). What I needed was a way to get the valve to be open when manifold pressure was low, and closed when the engine hit boost.
BRAINFART !!!! Get a solenoid valve to control the GFB valve... and I just happen to have a TVIS solenoid and vacuum tank under the bench. So now I have this...



Changed the plumbing to the solenoid valve from this...



... to this, and fed the vacuum tank with manifold pressure as well.



Used a spare AUX OUT on the ECU to control the solenoid. Programmed it to switch the solenoid so that the vacuum tank is connected to the GFB valve below atmospheric, and connected to the MAP line above atmospheric. Didn't know if this would actually work or not, so decided to take the spring out of the GFB valve to make it easier to hold the valve open.

Went for a drive and WADDAYANO... it actually worked I could feel a definite bump in acceleration when the solenoid closed as manifold pressure went into boost. Checking my logs when I got home, I could see the cruise s/c discharge temperature had dropped by 25-30degC !!! Gonna put the spring back in today and go for a drive. I want to see if it still works with the spring in, as I'd rather have a smooth transition into boost, rather than a sudden bump.

8. So how does that work... I hear you ask. Well this is my theory. At startup, the vacuum tank is holding a pressure of between 20kPa (overrun vac) and 30kPa (idle vac) which it applies to the GFB valve.



The GFB without the spring may be open, with the spring in, it will most likely stay closed. When the car starts driving the manifold pressure will rise to (say) 50-60kPa and this pressure is applied to the underside of the piston. Now there is a differential force (50Kpa>30kPa) in favour of opening the valve. Even with the spring in, I think it will be open ? I just have to check what pressure is needed to overcome the spring, from memory around 20kPa.

When the manifold pressure hits 95kPa (or whatever I have the ECU set at) the solenoid opens and manifold pressure is applied to the GFB valve... which immediately closes the bypass to prevent any loss of boost. Thinking ahead, I should be able to get a smoother transition into boost by lowering the solenoid switching pressure. I'll need to experiment with that later.

EDIT: Yussss... with the spring back in the GFB valve, it seems to work much the same as with the spring out. Not a scientific comparison, but temperatures are lowered by somewhat similar amounts... maybe a bit less. I would expect that, as when the spring is out the piston is either fully open or fully closed. With the spring in the valve, there will be times when the valve is only partially open/closed so the bypass airflow rates will be different.

Cheers... jondee86

[knightrous]

Nice work.
Will you be adding some logic into this to include some bias from TPS?

[jondee86]

Will you be adding some logic into this to include some bias from TPS?

Haven't really thought about anything fancy in the way of control. The GFB valve is always going to switch ON/OFF because that is all the control that the solenoid has.

What I have noticed is that getting into boost with a sudden throttle movement in 4th or 5th gear produces a nudge in the back. But the same thing in 2nd gear hits much more firmly. Probably due to drivetrain and vehicle inertia having more effect in the higher gears. This is not a problem when just cruising around, but I can see it being a bit unsettling in a hillclimb or street sprint.

No bypass gives the most aggressive ramp up as the engine gets on boost sooner. So I might look at keeping the bypass closed full time above (say) 3000rpm during competition. But before that I will experiment with different MAP switching points to see if that helps tone down the "on boost" hit.

Happy to listen to suggestions tho, as this is all new to me

Cheers... jondee86

[jondee86]

But before that I will experiment with different MAP switching points to see if that helps tone down the "on boost" hit.

So I changed the switching points from 100kPa going up and 95kPa coming down, to 90 and 85kPa., and that improved the transition. Now you really don't notice the valve closing in 4th or 5tth, and 2nd is just a nudge. Can probably lower the switching points a bit more to make it even better.

Before I do that I need to get out on the highway and log MAP pressures at a few different steady state cruising speeds, say 90, 100 and 110 kph. The s/c discharge temp fluctuates quite a bit with throttle use as can be expected. but the MAT is substantially lower than before, and does not move much at all in round town driving.

Overall I am very pleased with results to date

Cheers... jondee86

[trdee]

thats some nice work dude