Matts side projects

If you know what B.S. stands for you know what goes here.

Moderator: Matt

Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

Now for a cleaner dyno curve. This is the R34 with the GT3071R turbo, Z32 AFM and 480CC Nismo injectors now added... rising from 219rwkw with factory turbs to 278rwkw with this one and supporting mods
Attachments
P8200013t.jpg
spinning away
(85.5 KiB) Downloaded 7023 times
Pict0008t.jpg
Power vs Boost
(73.55 KiB) Downloaded 7023 times
Pict0009t.jpg
Power vs AFR
(73.81 KiB) Downloaded 7023 times
Pict0010t.jpg
Power vs Torque
(74.55 KiB) Downloaded 7023 times
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

TPS switch was way off...

I've updated software over the weekend for Nistune to display selected tables (for VLT RB30 anyway). It helps heaps with working out which tables are active on the ECU at any given time

I'll roll those changes through the rest of the addresses for other ECUs as I go
Legionnaire
 

Posts: 125
Joined: Fri Oct 03, 2008 5:45 am
Location: Moscow, CFD, Russia

Post by Legionnaire »

Hi, Matt! Nice result with your R34, although boost pressure is still falling closer to the end of rev range. What kind of boost control is installed?

I see a jump in AFR and power graphs in 5200-5600 interval. Have you adjusted stock valve timing (i mean VTC switch point)? If so, have you noticed any significant influence of switch point on power, torque, volumetric efficiency?

Regrds,
Petros.
Cheers,
Petros Katunian
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

We are using a turbotech bleed valve. Ideally we would like to use the boost solenoid on the ECU just need the time (again!) to find those things

Good spot with the VTC. We forgot to adjust that on the run and pete noticed that afterwards when going through the logs and dyno graphs. So we will need to adjust it next time and we will compare the curve against this one. Theres a few things to fix including my factory BOV not working on this one, plus see a power figure on more boost

I would like to try the autotune on dyno again also. I did it on the road in my R31 but the ranges where it alters the tune are limited
Legionnaire
 

Posts: 125
Joined: Fri Oct 03, 2008 5:45 am
Location: Moscow, CFD, Russia

Post by Legionnaire »

Maybe some kind of EBC can help with boost stability then. Even better if it would be factory boost control, but lack of time is a definite issue.

The most impressive thing with this turbo and your joint effort with Pete is that although the GT3071 makes boost later than stock one, motor actually produces similar or more power and, even more important, torque everywhere! How does it feel on the road? Does it feel laggy or is it eager to go? Does it feel fun? :)

What's wrong with factory BOV?
Increased boost at this level may be good for some torque in the mid range, but having reached 278rwkw (roughly, 278kw at wheels + ~55kw tranny loss = 333kw at flywheel = 446hp at flywheel) you are really reaching flow limit for 56 trim 71mm compressor. The good indication of this is declining boost level. You may want to upgrade to 76mm compressor later. What peak duty cycles do your injectors see curently? And what AR rear housing does turbo have?

Petros.
Cheers,
Petros Katunian
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

it feels good on the road, since its still got the same torque down low that the factory turbo had, but once its boosted the car starts to launch!

overall i still find more torque on the RB30 more 'fun' because theres more to use on a continous basis but you hit the power limit pretty quickly on that engine

BOV just isnt venting. not sure why. after turbo was fitted. checked all the vacuum lines etc. I'll pull the thing off and test it

yes the 76mm wheel would provide more flow. We did notice some compressor surge on the dyno but havent noticed it driving this week. I'll have to grab the logs for the duty cycle calcs when i get back. Pete said the duty cycle was misreported (injector time is double on R34 when reported for some reason and this affected duty cycle calcs). Not sure on the rear housing since I got the turbo with the car and just got it fitted
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Yes, it does feel good on the road. Looking at the boost curve you'd expect it to be a bit lazy but the power curve tells the story. And this is backed up on the road.

I would have liked to tune it using the factory turbo to see what we could get out of it. As it was we just ran it up, tweaked the boost a bit and got a figure. I'd love to see what some tweaking of the IGN values would produce with std turbo. As you probably know, the factory IGN maps on the R34 have a big trench in the last column. Presumably to save the engine if the boost gets turned up. Result of course being that you turn the boost up, hit the column with the very low timing figures in it and power gets killed off.

Rear housing is 0.63 AR although looking at the boost response you could be forgiven for thinking it was bigger.

I agree about the peak power reading. Surely that 71mm wheel doesn't have a lot more in it? I'd like to find out though. :) Next time...

As usual time was an issue. We got a basic tune banged into place and managed to address a few issues. Now we've been able to sit back and contemplate a bit we're ready for another go.

PL
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Oh, forgot...

Duty cycle went to 45% at 7000rpm. Which is obviously rubbish.

I just went through the raw log file and injection time is actually being logged as 7.7ms max. Which I assume is out by a factor of 2.

Z32 AFM went to 4.6V max which sounds about right for 280rwkw.

PL
streetbike
 

Posts: 69
Joined: Sat Jan 19, 2008 11:44 pm
Location: toulouse FRANCE

Post by streetbike »

PL wrote:Oh, forgot...

Duty cycle went to 45% at 7000rpm. Which is obviously rubbish.

I just went through the raw log file and injection time is actually being logged as 7.7ms max. Which I assume is out by a factor of 2.

Z32 AFM went to 4.6V max which sounds about right for 280rwkw.

PL
I have the same problem on CA18det before 3500rpm the injection time is good, and after that it is devided by 2 on consult... do you know why ??
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

I'll leave that one for Matt but it may have to do with how CA's go from sequential to batch fired injection at a certain RPM/load.

PL
Legionnaire
 

Posts: 125
Joined: Fri Oct 03, 2008 5:45 am
Location: Moscow, CFD, Russia

Post by Legionnaire »

PL wrote:I'll leave that one for Matt but it may have to do with how CA's go from sequential to batch fired injection at a certain RPM/load.

PL
Does RB25 neo do the same? Why cant it remain sequental at high rpm?

As for standard turbo - we have done some dyno pulls on my mate's R34 GT-t, turbo has stock compressor mounted on a shaft with a metal turbine (about the same size as stock ~53mm inducer or so). Played with boost and maps. The best result we saw was at 1-1.1 bar as measured at intake manifold, so probably 1.2-1.3bar at compressor outlet. Dont have IGN maps ATM, boost dropped to 0.8 or even less bar closer to rev limiter, that was set at 7500, 98RON octane, FMIC. Managed to produce 360+ flywheel hp and we ran out of stock MAF measuring capabilities. This sounds about right for stock dimentions turbo. When I sort out some problems with my hosting here, I'll post a couple of compressor maps with plotted RB25 traces.
Looks like RB25 neos just like to swallow a lot of air at high rpm - more than compressor can supply them with. This obviously holds true with GT3071R. I wonder how it would behave with some cams, os valves and porting done to it.

Petros.
Cheers,
Petros Katunian
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Hi Petros,

I'm not sure if the RB25's go batch fired up top. The only ones I've seen do it are the CA18's. On the bench there's a distinct chnage in the sound the injectors make. I can only assume that it's so they can get good resolution at low load and then they fire the injectors twice as often up high when you need lots of fuel.

All the SR20 turbo ECU's use batch fire all the time.

I'd be interested in seeing some dyno curves for what can be achieved with std turbo. All reports I've heard say that there's not much point trying to turn boost up much over std.

The NEO's certainly seem to be a much better head design than the older RB25. A local guy has done quite a bit of work to his and he reckons the NEO head has many similarities with the RB26 head. Which would explain why they flow so well.

Apparently they respond very well to a set of mild cams (like any Nissan turbo engine it seems!). That would be the next logical step for Matt's car but he's already spent quite a bit of cash on it and it goes quite well with 280rwkw. It'd be tempting to go for 300 with cams. And I suspect it'd be quite achievable but it'd really only be spending money to get a number...

Not that there's anything wrong with that! ;)

PL
Matt
Site Admin
 

Posts: 8961
Joined: Sun Jan 29, 2006 1:45 am
Location: Adelaide, Australia
Contact:

Post by Matt »

Nah I'll do it once my problem car (R31) has been sorted out. Need to drive something...
Legionnaire
 

Posts: 125
Joined: Fri Oct 03, 2008 5:45 am
Location: Moscow, CFD, Russia

Post by Legionnaire »

PL wrote:Hi Petros,

I'm not sure if the RB25's go batch fired up top. The only ones I've seen do it are the CA18's. On the bench there's a distinct chnage in the sound the injectors make. I can only assume that it's so they can get good resolution at low load and then they fire the injectors twice as often up high when you need lots of fuel.

All the SR20 turbo ECU's use batch fire all the time.
That's some interesting info, thanks. Why cannot they stay sequentially injected at higher revs?
I'd be interested in seeing some dyno curves for what can be achieved with std turbo. All reports I've heard say that there's not much point trying to turn boost up much over std.
Well, I'd like to post dyno graphs here - but unfortunately i can't! Dyno operator promised to send dyno printouts by email, but never did.

As for stock 25 neo turbo and its potential - there is a whole saga here. I too was interested and have done some research. It began when my mate blew his stock turbo pushing it to 15psi, so he needed a replacement. I didn't know anything about stock turbo at that time other than the fact that R34s has bigger turbine housing than previous R33 turbos. He asked my advice about what he should get and i sat down trying to do some maths on RB25 flow. This resulted in a set of Excel tables and calculations, and curves representing RB25 flow data plotted against various compressor maps taken from garrett website (garrett was chosen because 1)nissan turbos are garrett licenced and 2)turbo rebuild was performed by regional garrett dealer and there was a HUGE amount of spare parts from all kinds of turbos).

Among other compressors that I looked at there was a compressor from diesel turbo - GT3267 - which is now removed from garrett site for some reason. This turbo was equipped with 67mm compressor in 48 and 52 trim (46.4mm and 48.3mm inducer respectively). Map for the latter one amazed me in that compressor had excellent peak efficiency - 77%, had very good flow potential for its size and reached maximum flow with adequate efficiency at low pressure ratios, i.e. you dont need to run high boost to get power.

I plotted RB25 flow curves on it - see attach and sorry for large pic :shock: . Diagonal red curves are flow curves produced by linear interpolation of 7 pressure ratio points, each line represents theoretical 2.5 litre engine operating at fixed rpm at different pressure ratios (1, 1.33, 1.66, 2, 2.33, 2.66, 3) and constant volumetric efficiency of 90%. Narrow lines represent 2500, 3500, .., 7500 rpm, bold lines - 2000 - 7000 rpm.
As you can see, this map covers RB25 flow requirements very well, is not prone to surging at low revs (allowing smaller turbine section for quicker spool) and shows that compressor can deliver more than 40lb/min (10.36lb/min = ~100hp) of air above ~2.25 pressure ratio. We decided to ask for this compressor in our custom-(re-)built turbo. :) I've found out that this very compressor was common on nissan 6.0 litre diesel truck engine, so we referred to nissan diesels when specified our requirements.

Now is the most interesting part of this story. We dismantled stock turbo to measure its wheels, turbine wheel was badly damaged so no measurements were possible. Measured stock compressor and guess what - 48.3mm inducer, 67mm major diameter :D. See another attached pic. This pic is not a photo of actual compressor we measured, but it was found later on skylinesaustralia and confirms the validity of our own measurements.

Now we had the necessary info for compressor side - in that we did not actually need to change things. Turbine wise, if i remember this right, we chose GT28 53.8mm turbine wheel in bored out stock OP6 turbine housing so turbo now can be actually rightfully called GT2867R. It produced numbers I mentioned before.

Conclusion to all this, as i see it - stock turbo is GOOD for stock spoolup and power at 9psi (my own turbo makes this exact boost figure). It is BAD for increased boost and power owing to restrictively small and brittle ceramic turbine wheel (that was not intended to work at overheat/overrev anyway) - the fact that rebuilders had to machine turbine housing for it to accept even rather small GT28 wheel, proves this. But considering that compressor is already VERY GOOD and up to the task, add a bit turbine wise to reduce exhaust backpressure and reversion, add efficient free-flowing intercooler and you have factory-like response and good street power and torque at 15-17psi of intake manifold pressure. I honestly don't believe that it can make 400bhp at whatever boost level, but 360-370 bhp from stock compressor is entirely achievable, just work on turbine side.
The NEO's certainly seem to be a much better head design than the older RB25. A local guy has done quite a bit of work to his and he reckons the NEO head has many similarities with the RB26 head. Which would explain why they flow so well.
I have seen some claims made by Australian and NZ head porters that RB25 head has better low lift intake flow and better exhaust flow than RB26 head. Like this:
TK wrote:My stock valve RB25 v's RB26 with 1mm os valves:

Intake
Lift Flow(rb26) flow(rb25)
.1 -- 81.1 -- 89.2
.2 -- 164.9 -- 172.4
.3 -- 232.2 -- 223.4
.4 -- 265.9 -- 257.2
.45 -- 272.7 -- 267.5

Exhaust
.1 -- 75.6 -- 76.7
.2 -- 134.9 -- 140.1
.3 -- 160.9 -- 185.4
.4 -- 173.8 -- 197.8
.45 -- 178.9 -- 199.8
Taken here: http://www.performanceforums.com/forums ... t=67236980. I don't know who is this guy, but figures look good stock and can be easily improved upon even further (unlike toyota JZs with their assymetrical ports and poorer flow figures in both stock and modified form)
Apparently they respond very well to a set of mild cams (like any Nissan turbo engine it seems!). That would be the next logical step for Matt's car but he's already spent quite a bit of cash on it and it goes quite well with 280rwkw. It'd be tempting to go for 300 with cams. And I suspect it'd be quite achievable but it'd really only be spending money to get a number...

Not that there's anything wrong with that! ;)

PL
Which raises a question as to why did not nissan put warmer cams on their turbo engines in the first place. Maybe some ecological reasons?

Selecting a set of suitable cams for turbo engine is a problem I'm trying to study now. I've read about some interesting effects seen with unusual combination of duration/lift in/ex cams on audi engines. I wish I had some spare money to experiment with this myself! Otherwise I have to stick to tried and proven set of poncam type-B's
Matt wrote:Nah I'll do it once my problem car (R31) has been sorted out. Need to drive something...
That is the main concern with "problem cars" or any cars that are tempting to modify. I'm thinking of buying something like corolla or micra - boring, absolutely unmodifiable, dedicated winter car before doing anything serious to skyline.

Petros.

P.S. Sorry for thread jacking, Matt.
Attachments
stock_comps_1.JPG
(78.89 KiB) Downloaded 6803 times
ve90_gt3267_52T_048AR.jpg
(112.38 KiB) Downloaded 6803 times
PL
 

Posts: 957
Joined: Fri Feb 03, 2006 3:29 pm
Location: Adelaide, South Oz
Contact:

Post by PL »

Hi Petros,

Excellent work! It's great to see people actually doing some research and working out the real reasons behind things. Yes, the stock turbo certainly works well at around stock boost levels.

So if you guys rebuilt the turbo does this mean they're plain bearing or do you have the facilities to rebuild ball bearing turbos? I thought they were pretty much throw-away and you had to buy a new core from Mr. Garrett?

Interesting that the main restriction is the turbine side on the std turbo. This makes sense with what happened when we fitted the 3071 to Matt's car. It made more power at less boost - particularly noticeable before full boost arrived. I assumed this extra power before full boost was due to the less restrictive exhaust side of the 3071 and this would appear to be correct.

Cams. Back in the old days we were always told that turbo engines work best with std cams and that increasing duration was a waste of time. This has been proven incorrect on modern turbo engines. I don't know why. I can only assume that it's because older engines used bigger cams to get the required power in the first place.

Modern engines make power other ways - with free flowing 4 valve heads and high compression, allowing the use of much milder cams. Which give the economy and emissions figures we now have. So fitting bigger cams gives good power gains but your emissions figures would go down the toilet. Fuel economy also suffers but with the wonders of programmable EFI we can still maintain pretty decent fuel economy.

I mainly do SR20's and fitting a set of Poncams (256 degree/11.5mm lift) will give 20rwkw gain every time. With no power loss anywhere in the rev range.

Sequential vs batch fired injection. Sequential fires the injectors only on the intake stroke. When you go batch fired they are also fired when the intake valve is closed. This sounds wrong at first but the fuel it happy to sit behind the intake valve for a moment before it opens again next stroke. I was surprised too. But this then allows much greater fuel delivery at high load.

What gets me is how it only appears to be CA18's that do it. S13,S14,S15 SR20DET's are all batch fired from idle. Yet, all DE's (and the oddball GTiR) are fully sequential. Beats me..

PL
Post Reply