Dual sequential chargers

**rmcdaniels** · May 5th, 2005, 05:01 PM

Hello all, new here, linked off of H-T. I'm loving all the great JRSC info.

I used to have a JRSC, then went turbo, now looking to do something different. Turbo is great for drag racing, but not so good for autox and road course. JRSC was great for autox, but too limited in power for drag racing and long road courses.

I'd like to use a JRSC until the turbo gets fully spooled up, then switch to the turbo to wind it up to redline. I built a setup to do this, but am waiting on my engine to get back from the machine shop to try it out, probably in a couple of weeks.

Does anyone know if this has been done before? I've seen the Mini with the twin charger setup, but it uses the SC to compress the turbo output, and I don't want to do that. I know the Lancia Delta S4 used a turbo and SC setup, but I don't know if they were configured sequentially or for blow-through.

P.S. - I also have a Whipple setup that is about 25% done, but that is requiring a lot of fabrication, so it could be some time before it is ready.

**GreddyTurboSi** · May 5th, 2005, 09:44 PM

anything is possible. Id like to see pics of your project.

**rmcdaniels** · May 5th, 2005, 10:02 PM

Here's a crappy cellphone pic of me test fitting the assembly on the engine. This was before the block went to the machine shop, but I should have the block back this weekend and get it all running next week, if it actually works.

TC3

**mgro** · May 6th, 2005, 11:52 AM

Originally posted by rmcdaniels:
<snip>
I'd like to use a JRSC until the turbo gets fully spooled up, then switch to the turbo to wind it up to redline.
<snip>
I've seen the Mini with the twin charger setup, but it uses the SC to compress the turbo output, and I don't want to do that.

Very nice projects you have going on! I especially like the Whipple idea; how did you fix the direction of rotation issue?

I think its not a bad idea to let the turbo blow into the supercharger. Its far easier to implement. The only disadvantage I can think of is a bit tricky to explain. The total boost generation is non-linear (4psi turbo+4psi Roots >8psi total boost), you can prove this with simple math. This means that the Roots starts to generate more boost when the turbo starts to spool. Considering the fact that the Roots get very inefficient at high boost levels, the turbo wastegate and the Roots drive system must be configured in such a way that the Roots generates only a part of its maximum boost before the turbo is spooled, to ensure it doesn't change into a hairdryer at max boost. This, of course reduces its added value at the time its needed most (when the turbo isn't spooled yet). If you're looking for a "mild" turbo application the added value of the supercharger is still significant IMO.

If you really want the turbocharger to take over from the Roots at a certain point you'd have build some kind of valve, an actuator to kill the Roots' drive and do some devious plumbing to get it all hooked up properly! The transition from SC to tubo (and back) would have to be very smooth- no sudden pressure transients, and no sudden changes in inlet air temperature.

**rmcdaniels** · May 6th, 2005, 04:45 PM

Originally posted by mgro:
<blockquote>quote:<hr />Originally posted by rmcdaniels:
<snip>
I'd like to use a JRSC until the turbo gets fully spooled up, then switch to the turbo to wind it up to redline.
<snip>
I've seen the Mini with the twin charger setup, but it uses the SC to compress the turbo output, and I don't want to do that.

Very nice projects you have going on! I especially like the Whipple idea; how did you fix the direction of rotation issue?

I think its not a bad idea to let the turbo blow into the supercharger. Its far easier to implement. The only disadvantage I can think of is a bit tricky to explain. The total boost generation is non-linear (4psi turbo+4psi Roots >8psi total boost), you can prove this with simple math. This means that the Roots starts to generate more boost when the turbo starts to spool. Considering the fact that the Roots get very inefficient at high boost levels, the turbo wastegate and the Roots drive system must be configured in such a way that the Roots generates only a part of its maximum boost before the turbo is spooled, to ensure it doesn't change into a hairdryer at max boost. This, of course reduces its added value at the time its needed most (when the turbo isn't spooled yet). If you're looking for a "mild" turbo application the added value of the supercharger is still significant IMO.

If you really want the turbocharger to take over from the Roots at a certain point you'd have build some kind of valve, an actuator to kill the Roots' drive and do some devious plumbing to get it all hooked up properly! The transition from SC to tubo (and back) would have to be very smooth- no sudden pressure transients, and no sudden changes in inlet air temperature.<hr /></blockquote>I figure if I blew through the Roots blower, it would only be concerned with pressure differential, vice the total pressure, so if it makes 6 PSI, then from the perspective of the blower it will only be making 6 PSI when the manifold pressure is 20 PSI with the turbo making 14 PSI. That would mean I would have to keep the SC pressure relatively low, limiting the benefit I'd receive from it. The problem I see with that setup is based on my prior experience logging outlet air temps of an M62 with a homemade bead-wire probe. It's really a blower, vice a compressor like a Lysholm SC, so the greater the pressure differential it has to maintain, the more blowby it gets between the rotors and case, and the more air has to go through the blower multiple times, which raises the temperature dramatically. The other thing I observed is that at the same PSI, the blower loses efficiency the more CFM you move with it. At or near it's operating limit (around 14K RPM if I remember correctly) the temps go crazy, rapidly escalating to well over 200F, even with relatively cool ambient air temps.

It just seems to me that if I already have a compressor on the engine that's relatively efficient pushing a lot of CFM at high boost levels, it would be silly to use a Roots blower to do it.

Ideally I want to make a large amount of boost (around 12 PSI)with the Roots blower for a very brief period at a relatively low RPM, then bypass it as soon as the turbo is spooled up, which should be much quicker with the Roots blower making so much boost just off idle.

I don't know if it's clear from the picture, but I've extended the throttle body (TB) and placed a 3" butterfly valve that connects the TB directly to the intake manifold plenum when it opens. The valve is operated by a wastegate actuator with a 12 PSI spring on a bracket. I also have a Hobbs switch attached to a boost control relay that I can connect to the wastegate actuator and I may use an inline restrictor to control the speed of the valve actuation. I figure I'll use the Roots blower until the turbo is making 10 -12 PSI of boost, then open the valve. It will act like a giant bypass valve, taking the SC out of the equation and the turbo will feed the engine while the SC just freewheels. Here's a logical drawing of the setup:

Twin20Charger20Drawing

The Whipple project is not nearly as far along. The Whipple is actually a compressor, in that it compresses air internally, vice externally for the Eaton blower. This makes it efficient pushing 15-20 PSI of boost, but I can't unload it with just a bypass valve. I don't want to try to deal with an electrical clutch assembly due to the speed and power involved, so I will probably use a bypass valve and live with the parasitic loss driving the compressor when off-boost. The reverse rotation issue has been dealt with for years by the people running Eaton blowers on LS1/LT1 engines, they just use a jackshaft assembly like this:

reverser

I'll probably use a Vortech shaft drive to run the compressor. Here's a pic of the shaft drive assembly mounted on my dummy engine setup:

Sdrive

I've got a Whipple 1200AX that will fit nicely in front of the transmission if I change the radiator fan to a slim pusher-type unit. Here's the Whipple:

whipple1

I just need to make a couple of brackets to bolt the Whipple to the blue shaft drive support bracket and to the transmission bell housing bolts. I'll need to extend the shaft a few inches and put a pulley on it, but that is off-the-shelf hardware from Misumi America Rotary Motion Parts. It will discharge to the front and I can run it through a FMIC. I think this will be a really great system when I get around to finishing it. It may be a while, as I just picked up a used B18A block that's in pretty good shape for $50 and now that Eagle is selling inexpensive forged rotating assemblies I'm thinking about building a LS/VTEC 95mm stroker engine using extended 84mm sleeves and a deck plate to get the R/S ratio to 1.6. I figure I can run it with 12:1 CR pistons and make 250WHP with a ported B16 head that I have. I think that will be pretty potent for autocross and road course racing, and a 100-shot of nitrous will make it pretty quick at the drag strip.

What I really need is a wealthy sponsor so I don't have to work full-time and can build car stuff all day.

[ May 06, 2005 01:47 PM: Message edited by: rmcdaniels ]

[ May 07, 2005 01:05 PM: Message edited by: rmcdaniels ]

**rmcdaniels** · May 7th, 2005, 04:03 PM

The butterfly valve isn't the simplest thing, but because the SC will be adding to the turbo boost, the IM plenum pressure will always be higher until the valve opens, so the flapper would never open. As it is, I'll need to open the valve in a controlled manner, or the boost will suddenly drop as the SC is bypassed before the turbo fully spools up. That's part of the reason for using the Hobbs switch and the boost control relay. I figure it will start to open the valve, then if the pressure drops it will stop opening the valve until the pressure builds again. That way I will get a stepped opening of the valve that will keep the pressure from fluctuating too much, at least in theory. I may also hook it up to one of my AEM EMS outputs and control it with the ECU.

I am concerned with the fitment of the Whipple, as it will be tight. I'll have to make a sharp 180 into the compressor to clear the header and a sharp 90 out of it to clear the radiator, plus it will be tucked right up against the clutch slave cylinder (I'm going to have to modify the blue bracket a bit). I may have to just move the radiator over to the AC heat exchanger spot and use one of the long AC heat exchangers from a newer Civic mounted up behind my FMIC. And of course it will have to line up perfectly so the force from turning a shaft at 14K RPM driving a blower that will take about 30HP to spin won't make the whole assembly violently fly apart. But all that stuff is why I back-burnered it for the twin charger idea, as it used mostly off-the-shelf parts with the exception of the TB extender flanges, which were really cheap and easy to make.

**mgro** · May 8th, 2005, 03:32 AM

Originally posted by rmcdaniels:
<snip>
It just seems to me that if I already have a compressor on the engine that's relatively efficient pushing a lot of CFM at high boost levels, it would be silly to use a Roots blower to do it.

Point taken. At the boost levels you name in your post I think you have to go the "switch" route. Perhaps you can implement a simpler solution then the butterfly with sensors and all, how about a spring loaded flapper valve that automatically opens when the turbo boost overtakes the supercharger boost?

The Whipple project is not nearly as far along. The Whipple is actually a compressor, in that it compresses air internally, vice externally for the Eaton blower. This makes it efficient pushing 15-20 PSI of boost, but I can't unload it with just a bypass valve. I don't want to try to deal with an electrical clutch assembly due to the speed and power involved, so I will probably use a bypass valve and live with the parasitic loss driving the compressor when off-boost.

Hmmm, looking at the performance charts of the very nice Whipple unit you have there I'd definitely ditch the entire twin charge concept!! It can easily make 20psi and will be as efficient as the turbo will. And it will spool better then the Roots.

The reverse rotation issue has been dealt with for years by the people running Eaton blowers on LS1/LT1 engines, they just use a jackshaft assembly

Yeah, I've seen the Vortech solution but for the life of me, I couldn't figure where to put the compressor! I made a cardboard dummy of the compressor based on the mechanical drawings and couldn't squeeze it in between the radiator, header and tranny??? And I've already got a slimline fan!

What I really need is a wealthy sponsor so I don't have to work full-time and can build car stuff all day.

I hear you... [img]smile.gif[/img]

Mark.

**mgro** · May 8th, 2005, 05:24 PM

Originally posted by rmcdaniels:
The butterfly valve isn't the simplest thing, but because the SC will be adding to the turbo boost, the IM plenum pressure will always be higher until the valve opens, so the flapper would never open. As it is, I'll need to open the valve in a controlled manner, or the boost will suddenly drop as the SC is bypassed before the turbo fully spools up. That's part of the reason for using the Hobbs switch and the boost control relay. I figure it will start to open the valve, then if the pressure drops it will stop opening the valve until the pressure builds again. That way I will get a stepped opening of the valve that will keep the pressure from fluctuating too much, at least in theory. I may also hook it up to one of my AEM EMS outputs and control it with the ECU.

Yeah, I see your point on the flapper valve, that wont work in this case.
I think that a Hobbs switch may not be the best sensor to determine the changeover point. The boost made by the blower rises with RPM, so its not constant. Also, in some gears the turbo will spool faster then in others. Why not use an additional MAP sensor instead? That way, you could already start bypassing the blower as soon as the turbo makes any boost at all by opening the butterfly valve just a little bit!
Perhaps the EMS has an additional A/D input that can be used for this, or you could add a small microprocessor to control the boost (many PIC processors with A/D and D/A converters out there for a couple of bucks).

I am concerned with the fitment of the Whipple, as it will be tight. I'll have to make a sharp 180 into the compressor to clear the header and a sharp 90 out of it to clear the radiator, plus it will be tucked right up against the clutch slave cylinder (I'm going to have to modify the blue bracket a bit). I may have to just move the radiator over to the AC heat exchanger spot and use one of the long AC heat exchangers from a newer Civic mounted up behind my FMIC. And of course it will have to line up perfectly so the force from turning a shaft at 14K RPM driving a blower that will take about 30HP to spin won't make the whole assembly violently fly apart. But all that stuff is why I back-burnered it for the twin charger idea, as it used mostly off-the-shelf parts with the exception of the TB extender flanges, which were really cheap and easy to make.

Well I hope you get around to the Whipple project someday soon! I'd be really interested in the fabrication and modifications needed to squeeze one in there. Keep us posted?

**rmcdaniels** · May 8th, 2005, 06:02 PM

Originally posted by mgro:
Yeah, I see your point on the flapper valve, that wont work in this case.
I think that a Hobbs switch may not be the best sensor to determine the changeover point. The boost made by the blower rises with RPM, so its not constant. Also, in some gears the turbo will spool faster then in others. Why not use an additional MAP sensor instead? That way, you could already start bypassing the blower as soon as the turbo makes any boost at all by opening the butterfly valve just a little bit!
Perhaps the EMS has an additional A/D input that can be used for this, or you could add a small microprocessor to control the boost (many PIC processors with A/D and D/A converters out there for a couple of bucks). [/QB]

Actually the more I think about it, the more I think you may be right about the Hobbs switch/solenoid arrangement. My objective is to make 20 PSI without turbo lag and the resultant non-linear torque ramp, so why not just use a checkball-type MBC from the plenum to the butterfly valve actuator and set it at 20 PSI? That way as the cumulative boost hits 20 PSI, it will open the valve until the pressure drops, then as boost from the turbo builds, it will open it more until the valve is wide open and just the turbo is making boost. An analog feedback loop like that is probably going to be smoother than any digital loop I could set up on short notice, and I could use a bleeder valve to dampen any oscillations.

Later I could set up a couple of boost control relays and program the EMS to do it, but I think the MBC idea would get it functional quickly.