Dyno Lessons and Setup Progress
The dyno taught me a few things about the setup and highlighted areas that need work. Since the first real-world tests will all be straight-line racing events, I decided to focus on the torque converter first — especially while the car was still running and not in pieces.
Torque converters are interesting, particularly because I’m currently not using a transbrake. That means I have to load up the converter on the footbrake at launch. Peak torque on this setup doesn’t hit until around 5000 rpm, so I need a fairly inefficient converter that slips long enough to let the engine build rpm, so I can get some boost and generate some torque — all without overpowering the rear brakes.
8HP75 Diesel Converter vs 8HP45 B48 Petrol Converter
The OEM diesel converter that came with my 8HP75 was designed for a low-RPM, high-torque engine, and typically stalls around 2700–2800 rpm on a footbrake launch. Unfortunately, paired with a 2JZ that makes almost no usable torque under 4000 rpm, it only stalls at 2000–2100 rpm — far too low.
The next option was a petrol converter from a 2.0L B48 engine. On the footbrake, this got me to around 3500 rpm and 0.5 bar of boost — much better, but still not quite enough for this setup. On a smaller turbo setup, it might’ve been perfect. I’d heard rumors that this converter only had a single lock-up clutch, and later confirmed that after cutting it open — not ideal for anyone running big power.
Pure 4500 Stall Converter
After that, I tried a Pure Transmission 4500 stall converter from the USA. It uses a billet stator and some other internal mods to bump stall speed. But after installation, results were maybe a few hundred more RPM stall compared the B48 converter, so still some work to do there to get desired results. Think further mods to the stator/pump design and maybe a sniff of nitrous to get some more low end torque going.
I didn’t have time to test any more converter options before the next dyno session and events, so I stuck with the Pure converter for now.
Small Upgrades and Tweaks
-
Added a quick-release radiator drain — saves a ton of mess when draining coolant, as any Supra owner can appreciate.
-
Installed a Honda radiator temp sensor into the original drain hole, so I can monitor coolant temp right at the radiator and better understand what the thermostat is doing.
Exhaust Changes + Boost Creeping
The boost creeping above 7000rpm made me scratch my chin a bit as the obvious answer is that the wastegate priority on this manifold design just can't keep up with such a free flowing exhaust turbine and lack of back pressure. Short of starting from scratch with a whole new manifold design, I wanted to try some small changes to see if the situation could be improved or not.
The billet collector has 1.5'' diameter outlets for the wastegate outlets, but previously my wastegate collector used 1.25'' inlets, so I was running 1.25'' pipe from the outlet of the turbo collector down to the wastegate collector. I noticed there was some small overlap of metal where 1.25 pipe was almost acting as a sleeve inside a 1.5'' outlet port on the collector. So the obvious thing here was to use a new billet wastegate collector that would allow the use of 1.5'' tubing between the turbo collector and wastegate, allowing for slightly more flow and seamless transitions from collector to wastegate tubes, just to encourage some flow.
At the same time as this, the wastegate outlet tube was recirculated back into the main exhaust downpipe. This is mainly, so the engine is as quiet as it can be on boost for future track endeavors where there are noise limits and to protect my hearing a bit. I also love tuning engines that don't have screamer pipes, as the knock traces are usually less noisy and a bit more reliable, comparable to what I'm hearing through the det cans. There might also be a slight venturi effect going if I'm lucky and the angle of the recirc pipe allows for exhaust gases to pull a slight vacuum onto the wastegate tube.
Picked up another turbine housing, this one slightly smaller. 1.01 A.R compared to 1.15 A.R. Will try and induce some back pressure with the smaller housing and do some back to back tests on the dyno.
Also in non boost creep related stuff, I will be running the per cylinder EGT & Lambda's properly this time round, so they have all been treated to some heat sleeve protection and booting. Moreover, changed the location of the exhaust back pressure sensor port, just so I can get a cleaner run to the sensor with some coiled copper piping.
Had to opt out of making a billet junction box for the lambda sensors. I still want to do it, but it's a lot of money just to make my life marginally more practical when playing with the lambda sensors, so might revisit it if feeling flush in the future. For now, just ran all the wiring as above. Looks a bit messy, but what can you do.
Finally moved the coolant ports for the turbo smart straightgate and turbocharger slightly. Now feeding the system via a port on the radiator outlet pre thermostat, and returning coolant into the system on the GE water pump bypass tube. I did notice that I was able to get the wastegate reported temperature to the temperature limit fairly easily previously, so hopefully this updated coolant routing will provide cooler water straight to the wastegate first, rather than being fed by the heater system pipework. Each end of the pipework is now flexible as well, as direct hardlines to each point were awesome until it came time to service it and there was no flex. So now I've got hardlines where It's really hot and flexible rubber where it's a bit cooler.
ECU and Wiring Updates
As I wasn't getting as much value from the Motec as I'd hoped, and trying to organize a dev ECU with the right sort of base code package to start from fell on deaf ears at Motec HQ and no developer I contacted had a solution they could provide. I decided not to waste any more time with it.
So I rewired the modular section of my harness to support a Syvecs S7+ and X20 expander, replacing the Motec M150. Luckily, they use the same connectors and I’d left most of the wiring long enough that I just had to re-pin a most things and extend a couple wires to reach the expander. Minimal hassle — glad I didn’t twist and loom this section permanently!
Originally, my Motec wiring ran three separate CAN buses to distribute load, but the S7+ works best with two buses for my setup, even though the third bus could maybe be used. So I restructured my wiring and settings to the following:
-
CAN1 Receiving some misc can data from other devices, as well as translating some CAN data into slave items, that I can then reference on the ECU calibration to bring in some extra sensor data, specific to my setup.
-
CAN2 handles all critical data and is the main hub for can communication between all major devices.
Even with everything running, I rarely see more than 65% bus load — so there’s still headroom there and CPU load is barely even trying. Crazy.
The best part? All devices now report directly to the S7, so I can pull a single log from the ECU and see everything. No messing with different M1 Packages, logging packages, unlocking pay walls etc or dealing with third-party software — Syvecs just works. I honestly wish I'd never sold my original S7+ and gone M150, as I've just ended back where I started. Anyway, lesson learned, will have to keep M1 relegated to simpler setups until someone brings out some better packages that are a bit more with the times.
Wheels, Tyres, and Some Surprises
Originally, I wanted to run a new set of 18" drag radials from an Australian company. But after ordering, the tyres never arrived. With only a week left before the event, I canceled the order, got a refund, and started hunting for 15" wheels to fit a classic 275/60/15 Mickey Thompson setup.
I found a cheap set of wheels with similar offset to my existing ones, bought some new MT Street R tyres, and installed refurbed NA rear brakes and discs. But when I tried to mount the wheels, I discovered the inner barrel was a few millimeters smaller than typical forged wheels — they hit the calipers. I ended up shaving a lot of material off the NA brakes and brackets to make them fit. Not ideal, but it worked in the end.
Alignment
So to kick off real world testing, I don't have anything fancy going on with suspension. All stock arms, mix and match of 30-year-old stuff. £500 yellowspeed meh coilovers from a few years ago when my old teins shit the bed. NA rear roll bar, almost as stock as you can get really.
Had to jack up the rear coilovers length quite a bit to give me a working clearance around the new 15's and with that out the way had an alignment done to see what we could do.
Nothing groundbreaking, but decent enough to go fast in a straight line.
Michael Yazgic @ 0x33