I'm tempted to reprise the War-and Peace-length posts I have previously made regarding twins (compound turbos).

So here's the short version:

A b-series engine is relatively small displacement. Big power requires lots of mass flow of both fuel and air. Adding a lot of fuel is relatively easy. Adding a lot of air requires a lot of pressure to force feed the required mass of air into the engine.

Adding a lot of air gets difficult because your centrifugal compressor is limited in its ability to multiply pressure. So once you get above the existing turbo's pressure capability (PR or Pressure Ratio)-- aka "fall off the map", the compressor efficiency plummets and you get no more mass into the engine-- you actually get less.

But if you divide the work of compressing across two separate stages, to total ratio each stage must provide is less. Say you need an insane 8:1 Pressure Ratio. With two turbos, you could run each compressor at a PR of about 2.8 (the square root of 8). Or you could split it 2:1 on one compressor and 4:1 on the other. The first times the second equals the total.

Exactly how the workload is divided and how the system is setup is really the demanding part of installing a twins setup.

For peak power, a large single turbo can often match the output of a small twins setup. But the "area under the curve" isn't even comparable-- the twins setup will have MUCH more area under the curve. This represents vastly more usable power. Most larger singles will give you about 500rpm of "meat" in the torque curve (2300-2800rpm or so). A well-designed twins setup can come in hard is low as 1700 rpm.

Here's how to choose turbos, in my opinion. You first start with a realistic power goal. Let's say you want 550hp at 2700rpm. Doing the math shows that this requires about 70lb per minute of airflow into the engine. So you know the mass flow rate. How much pressure is required to get that into your engine? It varies a lot with intake manifold temps (i.e. CAC capability). If you can drop intake temps to 73F, it only takes 36psi to make 550hp. But since intake temps are more likely to fall in the 130F range, you'll need more like 41 psig of boost.

This is a PR of about 4-- within the range of some larger single turbos. So we pause here to point out that you might not need twins to make 550hp at 2700rpm.

But we don't just want 550hp at 2700rpm. We also want some response in the lower range. What if we wanted a respectably stout 1300ft-lbs at 2000rpm? This is about 500hp at significantly reduced RPM. Now we need a PR of about 4.7, and it has to be delivered at only 2000rpm. A single turbo will not do this, because a single turbo large enough to to deliver that much airflow will be tool large to deliver it at 2k rpm-- and it will likely be operating very close to the surge line.

But if you had a small turbo fully up on the boost before 2K rpm, then spooling a larger turbo isn't a problem, and you'll be splitting that 4.7 PR across two compressors.

The idea is that before 2K rpm, the small turbo is the bulk of your flow, and after 2K the larger turbo is the majority of your flow. This gives you quick response AND big power. For max response, pick a big charger that can barely flow 70lb/min. This means even a GT4294 is too big for this power level. A better choice would be the GT4094r, looking where 70lb/min falls:




Now you'll need a smallish turbo to initiate the system and get things rolling. This is where I like the smallish GT3076r:


To maximize the effectiveness of these turbos, you'll want to go fairly small on the top turbine housing and larger on the bottom one. You'll be bypassing a lot of flow at higher RPM, so a large wastegate will be required. I'd pick a Tial 44mm set at 1.5Bar (~21psi).

With the tightest housing, the GT3076 will spool very quickly and start making boost. Faster than stock (pre-VGT) spoolup, I'd wager.

At 1800 rpm or so, your GT30 will be rushing up past 2:1PR (~14psi of boost), your and your wastegate will open with you flowing bout 30lb/min.

The turbine map for the GT4094 shows that with 31lb/min, you are just now spooling the turbo, even with the largest turbine housing:


Properly wastegated, these two turbos could provide a lightning-fast setup that can make about 550hp and have a HUGE area under the curve.


If you have access to the compressor and turbine maps, you can do this kind of matching with any turbos. Garrett is the only one I know of that publishes both, so I used them here. You could just as easily call up Bullseye and order an S258 over an S366.