In the early to late 1990s, it was common for some halo cars
to run twin turbocharger systems. We saw this with the Nissan
GTR from the R32 to the R34, the Mazda RX7 FD and the Supra
JZA80. Broadly speaking, there are two types of twin turbo
systems; the parallel systems as on the RB26 Nissan engines,
and the sequential twin turbos as on on the RX7 and the Supra.
This article will only discuss the sequential twin turbo
system on the RX7 FD.
Parallel twin turbo systems
Sequential twin turbo systems
The two turbochargers are both the same size on the standard
RX7, however at low RPM, only one turbo is active. At high
RPM, or the full-power condition, both turbochargers are
active. This allows rapid spooling of the small turbo at low
RPM, while not being limited by its restriction at high RPM,
because at high RPM you have two of them working and therefore
double the flow capacity.
When only the primary turbo is running, all the engine exhaust
goes through either the primary turbine, or bypasses the
turbine through the wastegate. The compressed air from the
primary turbo is fed into the intercooler after which it goes
up to the throttle and intake manifold. Note that in this
case, the compressor exit from the secondary turbo has to be
blocked off, otherwise the compressed air from the primary
turbo would feed back through it, which would mean it wouldn’t
produce any boost.
So in this condition, the inlet charge control flap is closed,
and the turbo control valve is also closed, to prevent exhaust
going through the secondary turbine.
At high RPM, we must open both of these flaps, to allow the
second turbo to share in the exhaust from the engine, and to
contribute compressed air into the inlet. So in this
condition, both flaps are open.
In the RX7 system, there are two chambers which check valves
going to the inlet manifold. One is pressurised with boost
air, and the other stores vacuum. The solenoids that control
these flaps are configured so that each flap actually has two
solenoids, that each connect either boost or vacuum to one
side of a servo diaphragm. Therefore in one position, one side
of the diaphragm has boost and the other has vacuum, and in
the other position the two are swapped around. Therefore it is
necessary for both solenoids to be working, on all the flaps
which need to be controlled (both the intake and the exhaust,
and there is a second one in the exhaust I will describe
later), and the hoses all need to go to the correct locations.
The tricky part comes with the transition. You can imagine
that if you’ve just been boosting hard on the first turbo, and
you engage the second turbo straight away by moving both the
exhaust and inlet flaps at once, then you’re going to get a
dose of turbo lag until the second turbo comes up to speed so
it’s producing the same boost as the primary turbo. This gets
worse the more boost you have on the primary turbo, for two
reasons. The first reason is that the first turbo loses boost,
because you’re diverting exhaust to the secondary turbo to
spool it up. The second reason is that the secondary turbo
actually needs to spin up to a higher speed to match the boost
of the primary turbo.
Mazda weren’t silly so they added in what they call a
“precontrol valve”. This allows a small amount of air to be
diverted through the secondary turbine. It does cause a drop
in boost of the primary turbo but it’s not too extreme.
Secondly, there is a blow-off valve (charge relief valve) on
the outlet of the secondary turbo, which is under ECU control.
As you’d know it’s possible to overspeed a turbo by not
“loading” it, ie letting it just breathe into the atmosphere
and not deliver any boost. So what is done is for a very short
amount of time, we can open this precontrol valve, and the
charge relief valve at the same time, to get the secondary
turbo to spin up. Once it’s spun up, we can switch over the
intake flap and main exhaust flap to allow both turbos to do
the job.
Dyno graph with PWM precontrol valve
Sample dynograph
How long we need to spin up the second turbo for depends
really on the turbo inertia, and that determines the time. So
doing this over an RPM window won’t be very useful because
that RPM window will correspond to a different time in
different gears, boosts and so on. So instead we actually tell
the ECU the time for which we want to spin up the second
turbo, and the RPM we want to finish by. The ECU looks at the
current rate of change of RPM and the current RPM to work out
when it needs to start the precontrol period. Previously we
have used a short period and just turned on the precontrol
output hard, but it turns out that starting earlier and pulse
width modulating the precontrol output period tends to keep
the boost up and smooths out the power curve.
In terms of the settings, the ECU needs a minimum MAP and a
minimum TPS to activate the twin turbo function. Then there is
the RPM to enable the secondary turbo, and the pre-control
duration as described just before.
The values we have used are 4400 RPM and 1200 ms (or 1.2
seconds).
This feature is a bit redundant now since turbo technology has
come such a long way, and there are many single turbochargers
that behave as well as the factory twins, but many series 8s
are still running the factory twins so this function is
needed.
Lastly, we need to talk about the outputs. The easiest way to
do this is with a plug and play ECU. But if you’re wiring in,
then please observe the following:
1) The charge control valve and the turbo control valve need
to be opposite polarity. Therefore enable one in the software
as Twin Turbo, ann the other as Twin Turbo, inverted
2) The precontrol output needs to be inverted, and as I said
above, making it PWM even at say 25 Hz does help with the
power curve.
3) The charge relief valve can be connected in parallel with
the twin turbo output control