Alright.

I've just been doing some preliminary test drives and setting up the tuning on my megasquirted, turbo charged, gsxr1100.

I have a good line of sight to the turbo from the driving seat and it seems to be glowing a fair amount. The turbo and headers are not wrapped and it is a td04-16T (from a volvo T5R). This is all fitted into a Sylva Phoenix (see photo) and so doesn't get great cooling from the air-flow. At present the water cooling to the turbo isn't installed, as the bike turbo guys say it isn't necessary.

I suspect that having changed the timing wheel that the ignition is out by a few degrees which is putting extra heat into the exhaust.

But it would be helpful good to know what sort of turbo temperatures (i.e. level of glow) I am looking for. I know EGT is important, but don't have a feel for it.

Any top tips?

Thanks,

Matt


p.s. I think my next job is to install heat shielding to the bonnet!

what temps are you monitoring ?

you shouldnt really be able to see it glowing much during the day .

EGT should be around 800.c

water cooling is for reliability , maybe insulate the turbine aswell .

Absolutely crackers to not install the water cooling IMO, the bearings and thermal design were made to have the water cooling in place, it's only a matter of time before it goes south.

Incidentally I've had my turbo and headers glowing so brightly with standard ECU/tune that you could see the INNER gas pathways of the manifold through the glow, so it's not the worst situation, but I'd certainly not be happy with that if I didn't have water cooling to drag some of that heat away.

Garrett tech paper

EGT's are hard/impossible to tell from the manifold colour, you need to get a gauge in there if you're questioning it. The manifold will glow at about 500C, whether it gets there depends on the rate of heat transfer as well as the port temperature, you could be pushing 1000C out at low revs and never see the manifold glowing, likewise you can see the manifold glowing when the port temps are well within control.

[Edited on 24/5/11 by coyoteboy]

Just as Mr. Coyote has said ^^^^^^

Mine glows cherry with little abuse and egt's are in the 800's post turbo. 900's pre I should think and the bay lights up after a good thrash! Lol. It's been doing that for the 6 years and 40,000miles I've had it! Just a thought,retarded ignition will see egt's go through the roof and will kill pistons and burn exhaust valves so I'd get an egt fitted if you haven't got one already!

Thanks for those tips, very useful.


Any good sources for CHEAP EGT sensors that my megasquirt will be able to understand?


I need to take a careful look at the ignition timing. Its a pain on a bike engine because you need to take an engine cover off to see the trigger wheel (where the timing marks are). Generally I get sprayed with oil when I do this.



Matt

The MS will need the EGT amp circuits creating and adding if they are not already in your system?

I know the pain of the timing on a bike engine, I'm going through that right now. However I've resorted to measuring the static sensor-tooth timing and relying on the MS to be correct rather than checking and adjusting as I too can't get the small timing cover off due to damage.

No worries on the circuit, it is described in the manual:



I have previously built a window into a spare engine cover for dynamic timing, but it started leaking.

As I have a 12-1 wheel and just built a new signal conditioner (LM1815, much better than the last one as I sorted out the filtering rather than just using the circuit described on a web-page). I think I may have switched the polarity, falling versus rising edge detection in the circuit, which would put me out by half a tooth (i.e. 15degrees). Bit of a foolish error if that turns out to be the case. I can take a look at the trigger logs tonight, which should be clear.

That is one problem of static timing, nothing like a proper timing light to eliminate the assumptions (like the ones above!)


Any cheap sources for k-type EGT? McGill* do them at £15, from other sources they seem to run to £60-70. Looks like they can be found on loads of production cars, so maybe a scrappy hunt, any clues?



Matt







*I have had good experiences with McGill for rose joints, but some of their stuff is too cheap to be true and I had a fuel regulator that I wasn't confident of.

[Edited on 25/5/11 by matt_gsxr]

I was thinking more that the EGT amps are about 20 quid each IIRC!

The probes - not aware of any cars with them as standard (that use the standard K type) but that's possibly just my lack of looking. You can buy a probe capable of 1200 from someone like farnell or CPC for about £10 and then get the gland for it which is about another £4 from the same place, the main issue is location - ideally you want it as close to the ports as possible so you'll have to have the manifold and turbo off so you don't get swarf in them while drilling for the required bung/tapping.

Looked at Farnell. More like £20 + £8 I see what you mean.

LABFACILITY - MA-ISK-S30-100-P1-2.0-C2-T - SENSOR, T/COUPLE, MI, K, S/S, 3X100MM
LABFACILITY - FC-133 - GLAND, 1/8" BSPT, 3MM PROBE, SS

I don't mind taking things apart. Very easy in my case.

Another project.

Thanks,

Matt

p.s. regarding water cooling. Mine isn't a ball bearing turbo, so I agree that water cooling can't do any harm, but owners of these same turbos on the same engines claim it isn't necessary in practice. Turbo was a £80 e-bay special so not big budget.

quote:

---

p.s. regarding water cooling. Mine isn't a ball bearing turbo, so I agree that water cooling can't do any harm, but owners of these same turbos on the same engines claim it isn't necessary in practice. Turbo was a £80 e-bay special so not big budget.

---

Dont forget to get enough length on the thermocouple leads to get to wherever your MS is.

Turns out the triggering was perfect, but I had set the megasquirt trigger alignment very wrong.

Was running 18deg retarded initially. Tweaked at roadside to 8deg retarded (compared to normal map), and only tonight realised that I am retarded! Had set timing up on the missing tooth rather than the first tooth after the missing tooth.

Not very clever. At least the engine has survived, and I have discovered a way of checking turbo temperature. FWIW direct visibility of the turbo is a good design feature for anyone turboing a car, diagnostically useful and cool to boot!

Matt

if you have a multi meter with a k type thermocouple , that can do the job for a little money.

youll need a 1/4 bsp socket welded into collector if youre running just one probe .

i was going to mention that running retarded can be a problem!!!

that 1/4 BSP sockt could also be used for back pressure too ...... so youll knoe if manifold design is good or not .

Thanks for that,

I suspected timing was out and at the side of the road advanced it by 5 and then another 5, but hadn't appreciated that I was still so far out. I destroyed a previous set of barrels from detonation, so I didn't want to repeat that error. In the end it was a stupid error and a good argument for a careful approach when test driving.

I think I will take it easy on the boost (12psi), so will not go with EGT quite yet.

I know there is a risk to valve seats and such, but strangely enough gsxr1100 cylinder heads are about £50 and I have a couple of spares! The bigger cost than the cylinder head is the replacement oil, filter and coolant and the head gasket (although to be honest I have found you can reuse them). I asked on the oldskoolsuzuki forum and the view was that either one per cylinder (drag bike thinking) or don't bother.

Thanks for your help with this, its great to be learning.

Matt