AFR Sample Chambers - Take 3 - R
Turbo Mini Forums

Wasn't really sure where to post this since the board got re-organised for EFI, so I'll start again :)

Based on Paul's design (which was based on Denis's concept, and maybe others....) here's how it can fit into a standard T3 setup.

Proper subframe temporarily fitted and clearances look OK



The entry to the downpipe stub took a couple of attempts to get right (I had to resort to mild stell MIG for the tacks as my cylinder of argon is now totally empty so no TIG.....) so you can probably see the openning is a little large - basically I got one of the angles wrong the first attempt.



However, the end result (tacked only) shows plenty of clearance for the slave cylinder and hose and may even accept a bulkhead brake bias valve with a small re-alignment (I wont be using one, mine will be on the rear subframe)

And plenty of room to run the sample tubes under the stabiliser bar, and the thermocouples as well.



So I just have to temporarily fit this and tack it all up to the downpipe stub to prove I can lift the engine in and out wilh all this stuff connected......



EDIT - just to show it from underneath - I went to align the rest of the downpipe but it won't slide in the adaptor as the stubs are in too far - but you can see how much I got the alignment wrong the first time and the crappy MIG tacks....

Edited by Rod S on 1st Jul, 2009.

Schrödinger's cat - so which one am I ???

By the looks of it, you should be able to weld it all together and still be able to get it on and off in one piece.

I'm fitting mine to the 998 in the next couple of days, so fingers crossed that the principle works.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Schrödinger's cat - so which one am I ???

Got the bottom bits to align eventually





It's a real compromise between tunnel and gear selector with the 2" exhaust...

So tack welded, removed and re-welded....

I'm having to use MIG temporarily (although with a coil of stainless wire now) as I can't get to refill my argon until the weekend, so the welds will need tidying up a bit later.





Then after one more trial fit in the shell (no photos, but I really did....) back out to fit the sample tubes.



I'll do the drilling/tapping later tonight or tommorow but it's going to be a tight fit to get both AFR sample tube and a thermocouple in the middle one....

Schrödinger's cat - so which one am I ???

Manifold off and seperated from turbo....

I spot faced my chosen areas first, the drilled


Seemed like a good idea at the time but the drill still wandered and by the time it was drilled/tapped, things weren't quite where I had intended...


Close enough for a first attempt though...

Everything fitted




Technically not finished - I still have to make the connections from the sample tubes to the sample chambers..... tomorrow.

If I can lash up a suitable turbo oil feed and drain, I'll test run it on a carb in the next couple of days just to prove the sample system works.

Schrödinger's cat - so which one am I ???

Looks good but do you think there might be some condensation collecting in the bottom of the sample tubes that is going to be sprayed on the sensor upon start up? And if the sensor is already hot it might not be a good thing.

I don't know if that a real issue or not but it might be good to think of some way for the sensor heater not to be turned on much before the engine starts up.

Jean

http://www.jbperf.com/

Jean, good point - Paul's design puts the sample tubes over the top so no condensation loop......

Hmmm.

My relay pack at the moment is configured to do exactly the opposite, ie, the MegaSquirt cannot receive power until the TechEdge stuff has powered up.... I may need to reverse my logic and add a time delay.....

Original logic was based on "never run the engine unless the sensors have their heaters powered"......

Edited by Rod S on 3rd Jul, 2009.

Schrödinger's cat - so which one am I ???

Right,

Been out, looked again, thought about it....

Because (luckily) I haven't actually connected the sample tubes to the chambers yet, what I will do is

1 - put the sample tube in slightly below the sensor so any condensation gets shot at the chamber wall and drains away harmlessly, and

2 - put the tube in off-centre (radially) to create swirl so the sensor still reads the actual sample even though the tube is a bit low.

Let's see how that works.....

Schrödinger's cat - so which one am I ???

Looking good.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

OK, slightly below the LSU (sensor) and at a slight angle to get the swirl in the chamber....



Unfortunately the flash on the camera wouldn't "look" inside or you would just see the bottom of the LSU....

Schrödinger's cat - so which one am I ???

OK, that's it for tonight,



Now a quick shower and down the pub.....

Schrödinger's cat - so which one am I ???

Hi
I like it :) It'll be interesting to see the results/answer to the problem.
Cheers
Gavin :)

VEMs Authorised Installer / Re-seller. K head kits now available!

WB/EGT gauges. Click here for customers write-up

Visit www.doyouneedabrain.co.uk

My Mini build diary

Well, I finally rigged up a suitable turbo oil return yesterday, and put an HIF44 onto the inlet manifold I got from Phil on Sunday (I bought it to chop the flanges off to use on my next injection manifold but it was ideal for a test on the T3 exhaust manifold, no packers needed on the studs/nuts) but, as soon as I started the engine.....



A nice flood of oil....

I've been messing around with various Avonbar fittings for the hose (see here)
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=302473

But I still can't get it leak free.

I am obviously now going to have to machine up my own block to hose adapter.

On the positive side, there was plenty of gas flow up the sample tubes (sensors out at the moment) despite my concerns the "outer" sample may be less than the "inner".

On the negative, there is another issue with the oil feed hose and sample tubes on a standard T3 manifold,



The problem is on the right hand sample tube (didn't come out well in the photo), not the smaller thermocouple.

Off out now to machine up my own block to oil feed adapter........

Schrödinger's cat - so which one am I ???

Ouch......

The sample tubes certainly get gas flow, I have a burnt hand to prove it....

AFRs look about 1.5 to 2.0 apart as expected.

I'll try to get the laptop connected and log what happens under changes tonight.

Schrödinger's cat - so which one am I ???

Just a very quick screen print of AFRs only, I'll connect RPM and the thermocouples tomorrow.....


The AFRs are about 1 1/2 apart as expected from a carb but both samples respond instantly to a blip on the throttle - I was expecting a slight lag on the outer reading (which I thought I saw in Paul's graph) due to the sample being 1/2 the size, but there is nothing obvious.

The only minor point of interest is the AFRs are quite a bit higher (weaker) than when I did the first test (ages ago) on the same engine but with the N/A MG metro exhaust manifold with twin widebands. The mixture settings etc. on the carb are unchanged, its just bolted to a turbo inlet manifold instead of a standard one and obviously the exhaust manifold is completely different. There are no air leaks I can see or feel - strange ????

However, the main point is that this demonstrates the sample chamber principle works - two different engines and two different ways of plumbing it in - I'm nearly confident enough to put the turbo engine straight in and map Jean's code on it rather than spend extra days on the N/A engine.

Schrödinger's cat - so which one am I ???

If you keep it off-boost to start with (as you certainly intend), I don't see a problem with plumbing in the turbo right away. You have the advantage of having seen Paul's trial and you know that your sample chambers work.

I'm a big fan of doing things step by step but putting the turbo in right now seems like a logical step and not a big risk if you make sure it stays off-boost at first. And you'll have the advantage of seeing right away if the added exhaust restriction seems to have an effect on AFR readings (which it shouldn't with the increased delta P).

Jean

http://www.jbperf.com/

I would agree with Jean.

Just leave the wastegate open until you have mapped WOT at 100kPa. Then build boost slowly.

I'm really pleased that this has worked as without it we would have been flying blind. EGTs don't seem to be if any use because of the siamesed centre exhaust port.

Apart from Jean's code, this is the next most significant development to getting port injection of the 5 port to work correctly.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Yes,

I think the risk is low enough.

I'm wiring up the thermocouples at the moment (just came in for a break) - I may as well as I've bought them and have the fittings in the "test" manifold (the proper one is away with Benross at the moment) - at least they will show if I get similar temperatures to you.

And, yes, it is significant in that we have both made it work in different configurations, the concept is so simple so all credit to whoever it was who originally suggested it.....

Although I followed your tube sizes, I originally thought the sample tubes may be too small (hence I made mine with the option of fitting slightly larger ones), from my burnt hand yesterday 5-8 seconds after the engine started, they may be slightly too big.

I have a couple of spare thermocouples so I may strap them on the sample chambers and see how hot they are getting. The TechEdge stuff gives me up to 6 thermocouple inputs (3 per controller).

Schrödinger's cat - so which one am I ???

Well, I take back everything good I said about TechEdge.....

Couldn't get the EGT thermocouples to work properly so I had a look on their website and there is yet ANOTHER design fault on their PCB..... I did the "fix" but no luck so I assume I have fried two MUX chips as well as the comms ones I fried a few months ago due to their other PCB design fault ......

What a bunch of Ozzie ......s

Schrödinger's cat - so which one am I ???

Sorry to hear your having problems with the techedge, the quality of instructions and variations on the board wound me up a treat and i'm not even sure if it actually works as we havent tested it yet

Other than that it all appears to be going well alot of trial and error but it'll all be worth it in the long run.

Edited by sturgeo on 16th Jul, 2009.

You only have to look here to see the problems they are causing

http://wbo2.com/2y/kit.htm

Both my PCB and thermocouple problem aknowledged....

Sad really, it would otherwise be a nice bit of kit.

But why should people like me be their R&D programme....

Schrödinger's cat - so which one am I ???

While I agree with you that this is sad, it shouldn't have hurt the mux. The worse it should have done is mess up the temperature reading. If it did fry the muxes then there is something else.

Jean

http://www.jbperf.com/

Well I got the T/Cs working on the second controller - the one I hadn't "fixed" - and all is well, except I have lost track of which is inner and outer.....

Anyway, everything reponds as I would expect.....

The EGTs are really high, no wonder I burnt my hand yesterday, but at least the rush of air out of the turbo when I blip the throttle, makes it all seem worthwhile.



You can now see the small lag on the outer sample with a quick pulse, but nothing to worry about.

Schrödinger's cat - so which one am I ???

You are getting similar EGT readings to what I had at full load.

About 150 degrees hotter on the inner cylinders.

I'm not convinced that we can get an accurate EGT reading as the thermocouple tip has to be close to the fitting to fit within the port. So a lot of heat soak, but the fact that it reacts that quick suggests that it can't be too far out.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

good stuff Rod, keep up the good work.

Team www.sheepspeed.com Racing