Kevin's Compound Sequential Twin Turbo Design

[V8KILR]

It's all apart again for final assembly, so here's a better pic of the exhaust pipes and waste-gates.

 

[V8KILR]

Got my Supra back from the fabricators today. It's only seen a few psi so far, but feels quicker on the street as you would expect as the first turbo is now a lot smaller then my old single turbo. It will still need tuning before I know how much quicker it really is.

 

Also none of the wastegates or the QSV have boost hoses to them yet, so I'll connect them all up this weekend.

 

There is one little issue I need to solve before dynoing it and that is a rattle/knock that I first noticed about 6-8 weeks ago. I thought it was the exhaust rattling at first as it louder when cold and goes away as the rpm increases and gets quieter as it warms up. On checking with a stethoscope, the front idler pulley is very noisy so I'll replace that but so is the area around the trans bell housing. Could it be the flex plate?

[mellonman]

make sure the crank pulley is tight i had a noise similar to what your describing and the actual pulley bolt was loose .

 

but to be fair it could be any of you turbo set up as it looks amazingly tight in there

 

its nice to see that its up and runing and spooling quicker i cant wait for the full results and if there are any draw backs

[V8KILR]

make sure the crank pulley is tight i had a noise similar to what your describing and the actual pulley bolt was loose .

 

but to be fair it could be any of you turbo set up as it looks amazingly tight in there

 

its nice to see that its up and running and spooling quicker i cant wait for the full results and if there are any draw backs

 

Thanks. I checked the crank bolt and that's fine. I also had that happen to me 5 years ago. That made more of a ticking sound (due to fretting on the crank key) where as this one is a rattle sound.

 

The turbo pipes are tight but don't touch each other, and I had the problem before I dropped the car off for the turbo setup so that's not the issue either.

 

I've done some more reading on flex plates with loose bolts or cracked flex plates and it seems very similar to my issue. I'm going to drop in and see my auto trans guy next week and put it up on the hoist while its running to track the sound down.

[Wez]

Timing star, they have been known to become loose, some tuners even took to spot welding them.

[V8KILR]

Timing star, they have been known to become loose, some tuners even took to spot welding them.

 

I've read that too, but if that was loose, wouldn't it be running badly due to incorrect timing?

[Chris Wilson]

Yes.

[V8KILR]

Progress is a bit slow at the moment. Got a engineering company (not my fabricator) to add some ports to the intercooler pipes and that took them 7 working days to get round to it. I put the pipes back on last night, so all the hoses to the controlling wastegates are all hooked up now.

 

I have booked a time with my trans guy on Monday to check the flex plate, as I think its causing the rattle I can hear. Once that's sorted, I can then book some dyno time.

[Jellybean]

Subscribed

[V8KILR]

I have finally got an electrical issue I was having fixed so I took the Supra to the trans guy today to check the flex plate.

 

The flex plate was the cause of the rattling noise. The bolts they used were a bit long and were bottoming out, so even though they were tight, they were not holding it properly. My trans guy took all the bolts out and shortened them by 2mm and problem solved.

 

I can now finally book the Supra to get it dyno'd. I'll update this thread when I have a booking.

[mellonman]

make sure you video it mate , cant wait to see the torque curve print out , hope it all works as intended best of luck to you

[V8KILR]

The dyno is booked for Monday 11th August. I'll take a video on the day as well.

Edited July 29, 2014 by V8KILR (see edit history)

[Chris Wilson]

Glad you edited the date, thought you'd built a time machine Lovely fabrication, hope it all works out, good luck with it!

[Noz]

Just caught up with this.

 

Awesome engineering.

 

Must of taken some fiddling around with. Great to see. A true work of art. The dyno curve and pressure curves will be great to see on this!

[Chris Wilson]

I have to say that I agree with Jagman, the complexity of this thing, from a thermodynamics and mapping point of view, is mind boggling, I still sometimes wonder if its actually technically sound. I *THINK* it should work, then I sometimes think, "Hmmm, but can it...?", I am well outside my comfort zone here though. But the mapping, the mapping...... The fabrication is awesome though, really really nice tight work.

[V8KILR]

OK, I've finished on the dyno for today, so here's an update. This is a diagram of the system I had built by Ronnie Lim at RSL Automotive in Auckland.

 

 

The two turbos I'm using are a MasterPower R545 (5459) 1.00 AR rated at 600bhp for the #1 turbo and a MasterPower R7677 (7677) 1.05 AR rated at 1000bhp for the #2 turbo. As this system works on the exhaust side just like a compound setup does, I decided that compound sized turbos should work best for my sequential system.

 

I had several issues on the dyno today, one of which was boost leaking from one of the #2 turbo pipes where it joins the #4 wastegate, but I managed to fix that.

 

One issue I didn't manage to fix today is that the #3 wastegate does not close when the #2 turbo boost pressure matches the #1 turbo, like I thought it would. Instead of closing nicely, it fluctuates between wanting to close and wanting to open. The only thing I can put this down to is that the #4 wastegate is fighting against it closing. One of my earlier designs had a swing check valve instead of the #4 wastegate, and that may be the solution to this issue? The #3 wastegate not closing results in a massive boost leak from the #2 turbo, hence power and boost does not go up much once the #2 turbo kicks in.

 

Unfortunately the spool of the 54mm turbo doesn't seem to be any better than my old 69mm turbo at low boost and at around 8psi boost, the old turbo leaps ahead. This is the most important issue I need to resolve, as getting more power down low was the main reason for me doing this setup. Getting less back pressure down low is probably the key to this.

 

Here is a dyno comparision between one of the runs today (280 rwkw run) and one of my old runs (480rwkw run). You can see the #2 turbo adding boost at 14.5 psi and 137 kmh.

 

 

The back pressure was much as expected. On that 280rwkw run it was 37 psi back pressure with 18 psi boost, so was around 2:1. As the boost should have been much higher if the #3 wastegate had closed, it will probably end up around 1.5:1 once that issue is sorted. When the boost is turned up even further, it will go higher of course. Most of the back pressure is a result of the #1 turbo and turning the boost up on the #1 turbo does push the back pressure even higher.

 

Overall, not as good a result as I had hoped as I have two major issues to try and resolve.

[V8KILR]

Here are some pics of the current setup.

 

 

 

[Shane]

It's not easy being a pioneer!

 

Good luck with it and keep the posts coming, watching with interest. Great workmanship and greatest of respect to you for what you have done so far.

[mellonman]

Thats a shame dude , but there will always be hurdles if your the first doing something great R&D though,

Just an idea could you not change wastegate 3 and 4 to electric solenoids ? would be alot easier to control

[V8KILR]

Thats a shame dude , but there will always be hurdles if your the first doing something great R&D though,

Just an idea could you not change wastegate 3 and 4 to electric solenoids ? would be a lot easier to control

 

Thanks. Unfortunately I don't have any free outputs on my ECU which limits my options a bit.

 

Thanks to some suggestions by dazzadude (from Aussie) on the USA forum, here's what I think will make my setup work correctly.

 

 

Hobbs_2 opens at 1 psi boost to stop the #4 wastegate opening under vacuum. Once it's at 1 psi boost it then will function based solely on the pressures to the top and bottom ports. The pressure to the Hobbs_2 comes from the manifold.

 

Hobbs_1 shuts off the flow to the bottom of the #3 wastegate at 15 psi boost. This will force the wastegate closed and merge the flow from the #2 turbo. The pressure to the Hobbs_1 comes from the #2 turbo so that it running fast enough to not surge.

[V8KILR]

I've come up with an even simpler and better method to control the two compressor side wastegates. It only requires one Hobbs switch set to close the flow at 15 psi and when I increase the spring rate on the #4 wastegate, it is very likely that the check valve in the diagram is not needed at all.

 

Using this method it forces the #3 wastegate to close and the #4 wastegate to open at exactly the same time, when #2 turbo boost hits 15 psi. It also eliminates any issues from pressure on the valves faces by forcing the wastegates open and closed.

 

[V8KILR]

Here's how the Hobbs has been wired up. Dyno booked again for next Monday.

 

[V8KILR]

The good news is that the reverse flow out the #2 turbo intake was resolved by using #1 turbo pressure to the controlling #3 and #4 wastegates instead of manifold vacuum. Also the Hobbs switch on the two controlling wastegates works correctly and closes one compressor wastegate and opens the other one exactly as planned, so the transition to the #2 turbo is controllable by #2 boost pressure and works as planned.

 

Unfortunately, there is now another issue and that is surging from the #2 turbo after it is online. The reason it didn't surge last time is that it was leaking boost through the #3 wastegate which doesn't happen now. I tried transitioning the turbo between 10 and 20 psi and from 4800rpm to 5800rpm, but it didn't seem to make much difference. There are two possible reasons for the surge that I can think of:

 

1. The #2 turbo is spinning too slowly and doesn't have enough drive pressure to maintain flow at the set boost pressure. In other words, it's trying to run to the left of the surge line on the compressor map.

 

2. The engine cannot consume all the airflow from both the #1 and #2 turbo at the set boost pressure and the #2 turbo surges first as it's the bigger turbo.

 

I'm not sure which one (or both) of these reasons is the cause of the issue?

 

 

There are a few possible fixes I can think of. I'm not very keen on options 1 or 2 as the #1 turbo doesn't flow enough air (so doesn't make enough power) on it's own at just 20 psi boost.

 

1. Have a controlled boost leak from the #2 turbo up to the point where it can be fully bought online, perhaps around 7000 rpm?

 

2. Drop the spring pressure in the #2 turbo wastegate to 10 psi so it can be controlled by the ECU to leak exhaust gas up to the point where it can be fully bought online, perhaps around 7000 rpm? This may not work anyway as there may not be enough drive pressure. Probably would need to work in combo with option 1.

 

3. Reduce the size of the #2 turbo to be closer to the #1 turbo. Currently they are 5459 and 7677 respectively. A 61-67mm rear turbo may have a surge line sufficiently far to the left of the map to not surge when brought online. That means the turbine side would be much smaller as well which may not bode well for back pressure. MasterPower do make a 6168 with a 1.22 divided housing, a 6568 with a 1.22 divided housing and a 6777 with a 1.05 (turbine size I'm using now) or 1.22 divided housing. One of these might turbos work better to eliminate the surge issue if they spin up a lot quicker?

 

4. Change the setup from sequential to compound (which will be pretty easy to do) and run 45 psi boost which is 1 bar boost from each turbo. This should work okay as the turbo sizes are already setup to what should be a reasonable combination for compound turbos.

[V8KILR]

I've done a few changes to the setup and here's what it looks like now.

 

 

As you can see I've changed it to a compound setup as the racing season is starting here soon and I think that's the quickest way for me to get the spool and power results I'm after. I haven't given up on the sequential and I still have all the unaltered parts for it, so I'll look at it again next winter.

 

I took it to the dyno today and it was looking promising until I started blowing off the boost hose to the QSV and the vacuum hose to the power steering. They both blew off twice and with the QSV not opening, the back pressure sky rockets, so I gave up on tuning for the day until I can stop these hoses blowing off. I got up to 35 psi during testing and 370rwkw at part throttle on one run. Here is one of the early dyno runs (with both wastegates running off spring pressure) showing how much quicker the boost comes in compared to the single turbo setup I had. Increasing back pressure at higher rpm causes the boost to drop when running off the WG springs only. Later runs allow steady boost when under ECU control.

 

 

As you can see, it reaches the same boost level almost 40 km/h earlier than the single 69mm turbo did. That run had the big turbo at 15 psi (PR = 2.0) and the small one adding 14 psi (PR = 1.5) for a total of 29 psi boost. This is quicker spool than the sequential setup as well and shows the advantage of compounding the boost and not having to wait for the second turbo to come on line.

 

I have noted that the torque flat lines from 80-120 km/h though and I'm thinking that is probably my torque converter not holding at that rpm? I have a stronger one in the garage that I may need to swap to.

Edited September 20, 2014 by V8KILR (see edit history)

[Guest Scotty 88]

Sounds awesome