Tuning 1st Series RX L EFI .

[discopotato03]

This is a spin off from the Turbosmart FPR thread because it goes off the topic .

I got around to setting up the wide band sniffer in the RX L to see what was going on mixture wise now that mad mode is no more .
From that other thread I fitted an adjustable FPR just to make sure that the almost 24 yr old one wasn't letting me down .

With the fuel pressure set to 38 , std is 36.3 , the AFR's were a bit on the rch side like ~ 12.6 at idle and floating around the 12.5 to 13.5 in normal driving mode . The first thing I tried was screwing the vane AFM's bypass screw in and out to see if the mixtures change which they didn't . Next step was to back the FPR off which did make a difference though not enough . It did put the AFM bypass back in its workable range but not by a whole lot .
At that stage I remembered that we altered the AFM's internal wipers spring tension in the past trying to fix a lean out that ended up being an oxygen sensor issue , eventual disconnection fixed that one .
Anyhow we had marked where the nylon gear was inside the AFM so we had a reference point to work from . I reset that to the std position and that put the system pretty much back to factory standard and with a bit of adjusting the hot idle AFR was set to about 14.6 to 1 .

On the road I'm finding that it sort of settles at mostly 13.7 but at very light feathered throttle it creeps into the low 14's .
I'd like to lean it a tad more so that slightly more of the light load running gets in the low 14's but it remains to be seen if these crude old heads will like that . I hoping that low 14's and a bit more advance will make things crisp and clean and we'll see how detonation resistant this BP98 really is .

Also something I've noticed is that except for the 16-1800 rev range it doesn't seem to do a closed throttle fuel cut like most later EFI cars do . I'll have to make sure that my TPS is set properly and that throttle pedal micro switch is not sticking .

More later , cheers A .

[twilightprotege]

14 is pretty high for idle. having tuned microtechs a fair bit, it's hard to hold a good smooth idle that high. i usually tune my idles on the richer side like ligh 12's.

and are those a/f readings with the o2 plugged or unplugged?

[discopotato03]

The std probe is there but disconnected so no closed loop feedback .

I don't find that mid 14 AFR to be a problem at idle , condenser fan cuts in and out (Davies Craig bulb type controller) and the idle doesn't seem to suffer . Same with headlights .

Never used a Microtech only early Haltechs and an Autronic , and an Apexi Power FC .

I reckon in this case the factory would have had a good handle on their matched ECU and injectors so even 14.7 shouldn't be a problem .
All I'm doing ATM is trimming the idle mixture via the AFMs air bypass screw .

Your Microtech no doubt uses a MAP sensor as a load measuring device and its anyones guess how steady manifold pressure is in these old manifolds at idle . In early and late versions Fuji used airflow measuring meters on the atmospheric side of the throttle so probably a more stable airflow signal than MAP . Also with a factory system you get the use of their factory developed temp/density mapping at least with a vane AFM , hotwire is even better again because it measures airflow by mass meaning volume and density in one go .

Most tuners reckon if an engine is capable of idling at a set AFR its capable of doing it in open or closed loop mode . If its making large corrections in closed loop then somethings wrong or the mapping was a long way out to start with .
Ignition timing has a reasonably large say here too , these old jelopies were early in the days of production L Jet (Bosch under licence) EFI and they were struggling with emissions legislation . Often things didn't run real well particularly on the garbage ULP we had here in 85/86 , with better fuel and the want to make things run properly there's a bit more leeway nowdays .
So as I said clean mixtures and the best timing compromise I can come up with may help .
What doesn't help is the "ported" vacuum these early 82T's had from the factory , with a bit of throttle you lose much of the vacuum signal to the distributor and that pulls the timing back just when you don't want it too taking usable torque with it .
Not hard to see why they went with the CAS and transister switched coil ignition system in the later cars , they don't have to rely on manifold pressure and flying bob weights to control ignition timing - prehistoric really .

Cheers A .

[wrxer]

yea it took chrysler until 1979 with the introduction of elb (electronic lean burn) to get rid of those prehistoric bob weights. elb was basically electronic spark advance control, and thats all i know about that, and it would have been one of the first systems available here, certainly much earlier than fyord or holdon. m2c

[discopotato03]

I seem to remember in the past trying different places to get the dizzys vacuum signal source , some upstream of the throttle plate . Its just a juggle to get the 20 odd degrees at idle and still pull enough timing as it climbs into boost .
It would have been interesting to have the early MPFI NA dizzy the play with as its mechanical advance characteristics would have been different , the right ignitor etc in one of those would still allow the nock system to work as normal .

All the hoops the factory went through to have that low 7.7 static CR - and not lose everything off boost . How things have changed .

A .

[steptoe]

You need to refer to the tuning specs guide for countries that had this engine without O2 sensor control then. I seem to recall this being the case for Saudi Arabia.

[discopotato03]

I dug out the factory 86 L RX-T EFI section manual to check up on a few specs .
The OE 02 probe is a basic Zirconia single wire narrow band thing so it really only reads with any "accuracy" in the narrow band of 14.7 to 1 , basically cycles on off either side of that theoretical figure .
I reckon if thats all it was designed to see then it must have been trying to trim AFR to that figure in closed loop mode .

"Pressure Switch Turbo" .
Long/short , two switches first being for turbo dash light which illuminates at 6.7 kpa (~ i psi +) and in Australia models sends "heavy load signal" to cancel closed loop mode .
Second switch is an over boost fuel cut set to 62.7 kpa / 8.96 psi+ on Oz models and 69.3 kpa / 9.9 psi+ .
Not going to be real hard to send no pressure to the second one hey ...

"Regulation of Supercharger Pressure " .

Just the specs , Aust model regulated at 49.3 - 57.3 kpa / 7.05 - 8.18 psi+ .
Except Australia models 56-64 kpa / 8 - 9.14 psi+ .

Higher boost than I'd thought these critters ran so a bit more leeway to play with .
Be real interesting to see what boost my poor old girls running and I seem to remember its turbo having an XR6 actuator on it and I think early ones only regulated at 4.5 pounds of boost . LOL in my case just enough to switch the "Turbo" light on !
If only has a blurts worth off boost ATM it should lift its skirts a bit higher with maybe 8-10 pounds .
Not now surprised that the AFRs fall off the Tech Edges richer than 10 to 1 AFR lower limit and sit around 11.7 at full throttle in 5th on a steep enough grade to not pick up speed .

Interesting times ahead , finally time to get some more boost .

Cheers A .

[steptoe]

I for one are waiting for you to get a boost/vacuum gauge

I use the light from the ash tray shoved in the back of my boost gauge so the only other hook up is the vacuum line to boost vac switches there on RHS. Gauge is mounted in a roll of flannelette in where a radio once sat before I got Cheap Grief.

There are just the two adjustments for non Aus models and that is the screw on the TB and the one on AFM. Someone once said the screw on the AFM did nothing as it had nothing on the other end for it to adjust! Just a trick tune device!!

[discopotato03]

No it does have function , which is why the factory puts a er anti tamper aluminium plug abouve it . Strangely mine must have fallen out ...
It just an air bypass around the vane or flap in the AFM pretty much like the throttle body has only it makes a variable leak to set idle AFR .

Boost gauge should arrive next week . A mob I know had a barely used boost T which I got for a barely used price and thats in circuit now .
It does work but I think I'm going to have some turbo issues because of that turbine I had fitted to it . I had an idea that the exhaust flow would be a bit weak compared to a REAL engine like a two liter four valve inline four but I reackon its weaker than I'd guessed . Its not trying hard to make meaningfull boost until about 4000 and thats a bit high most of the time . The fuel cut works and I've bounced off that a couple of times .
Looks like I'll be asking about the other optional turbine which probably means another turbine housing as well .

A .

[nncoolg]

Im gonna chirp in here and tell you about my old 3 plug experiences again, I also played around with the spring tension AND POSITION of the wiper inside the AFM, had good results with that, but you need to keep the o2 sensor hot so it reads correctly, I reckon the sensor ports are WAY too far away from the heads in these things, should be in the merge, under RH head. I've got an old flapper here AND a dizzy if you want to have a play.

[discopotato03]

Well the oxy probe std on an EA82 is in the dump pipe AFAIK on L's and in the up pipe at least on Spider turbo Vortexes .
From memory std EA82T headers and dumps are quite well insulated so would heat up quite quickly . My header is a stainless steel one thats Thermotech wrapped and it warms up fairly quickly as well .

I use a Tech Edge wide band controller and the probe itself is a late pump cell heated wide band one - LSU something or other . Its usable range is from 10:1 up to at least 22:1 which is more than wide enough .
What you have to do with these is connect everything up and power it up before you start the engine , when the flashing orange light goes steady you can start knowing the probe is up to temp and allegedly reading accurately even when the engine is stone cold . They say never to run these sensors unpowered because it damages the probes tip so have to remove it when not connected to its controller box .

If you had an unlimited budget you'd have one in each engine pipe and a pair of EGT probes as well .

I've been thinking of late that I wish there was a way to get a bit more advanced timing with more than light throttle off boost , you can advance the dizzy but that advances the whole range from idle to whatever full mechanical advance engine revs are . Its a pain that subaru set these early 82's up wih "ported vacuum" from a port in the throttlebody to the dizzy , as soon as the throttle is open you lose much of the "vacuum advance" and that makes a low compression engine feel a bit flat . I'm not getting any significant boost in the low/med rev ranges so the retarding timing and lack of boost together gang up on me a bit . Probably the only way to get around this with conventional components would be to have four small vacuum taps at the bases of the inlet manifold and feed them into one line and connecting it to the distributor . This would help a little and still allow boost pressure to retard the timing as it does now .

Raining a lot these days so nothing much going on aside from driving it .

Actually I made a new discovery in the Garrett GT28 ball bearing family , I'll try and keep this short so's not to bore people .

Garrett changed there numbering system a while back to differentiate their GT25BB and GT28BB turbo ranges . Basically the GT25 turbine based units get an 11 bladed turbine thats 53.0 mm major diameter in 62 trim .
The GT28's have 53.8 mm diameter turbines that are 9 bladed in either 62 or 76 trims . Trim size is the percentage difference between the outlet of the wheels cross sectional area and the major diameters cross sectional area so 62T is ~ 62 % . You could liken it to a rims face area divided into a whole wheels face area .

Anyhow upshot is that I'd always thought that the 62T GT28 turbine was an upsized version of the GT25's 11 bladed one . Well its not so I've ended up with the GT28 76T turbine thinking there was no smaller trim 9 bladed turbine available . So with the largest GT28 turbine option I can't get it to spin the compressor wheel fast enough early enough in the engines rev range to make much more than a pound or so of boost , no boost gauge till next week but at least the dash light illuminates so it does do something - just not much .
The options are to use the GT25 62T 11 bladed turbine or the GT28 62T 9 bladed one . Either one is going to need another turbine housing because mines been machined too big for them .

Everyone fallen asleep ?

A .

[Gannon]

Isnt there a vac line just below the throttle plate that see's vacuum at partial throttle because of the pressure drop across the throttle?

[discopotato03]

The pressure drop across the throttle is just the difference between atmospheric and whatever the plenum chambers pulsing pressure is .

My wording above should have read when the throttle opens more than a small amount . When you get into the upper gears you use more throttle even at low medium revs because you're asking the engine for more torque to pull the higher gearing . The result is that the manifold pressure is higher - as in closer to atmospheric and you lose the ability for the dizzy to advance the timing other than mechanically .

Actually thinking about it I should have a look at a spare throttle body to see if theres enough meat to have another tapping point further around the throttle plate and closer to its shaft so that the ported vacuum stays up a little longer in the engines rev range . It'd still see boost pressure and retard as per std .

Also been onto the MX5 people in the US who turbocharge their BP16 and 1800cc engines and the feedback is quite good on the 1600cc ones .
The Mazda BP's are a DOHC 16v engines with quite reasonable porting , they can get 160-190 wheel Hp out of these turbos in 9.5 CR intercooled cars and 10 pounds of boost .
Probably still breathe better than an EA82s wheezebag heads but they still have to spool the turbo with four 400cc pots .

Jees I miss 16v inline fours , cheers A .

[nncoolg]

Disco, something I noticed the other day was that the spare Flapper I have here has a code on the side, 'I' which looks like the codes that are on the ECU, I am wondering if there are different spec AFMs throughout the different models, different makes of that time even, like CA20's, E15's and such...

[Gannon]

Could you use a vacuum canister to store enough vacuum to keep ignition advanced before boost came on?

[steptoe]

and dump the vacuum with a solenoid from the boost light? ah, it'd be storing vacuum too often though, more than when it is wanted, surely

[discopotato03]

Well I believe MPFI EA82's were only made for two years with that vane AFM system and I've never read anything about any being different . For the AFM to be different the ECU would have to have different mapping , I have an 85 ECU in an 86 build RX L and it runs ok .

Gannon I thought of that too but getting acceptable transisions from below to above atmospheric pressure wouldn't be easy .

Really electronically controlled ignition timing is the go and I reckon thats where any gains from 3 to 4 ECU plug cars would have been made . The mass flow sensor isn't necessary to have timing control and some early turbo Nissans were like that eg FJ20 turbo engines . MAF is a little faster and more accurate BUT heaps cheaper to mass produce .

Cheers A .

Oh BTW .

I plotted out the turbine maps for the three turbines available for GT25/GT28 BB turbos and the middle one , mines the largest , looks to be the best overall . Buyable too so will get those balls rolling in that direction tomorrow .

A .

[Gannon]

Didnt i suggest you go with the 4 plug system when you started this project?

[discopotato03]

Yes and I think I mentioned that theres no scope for fuel adjustment unless you use intercepters to "bend" the load signal , often doing this to add fuel makes the computer retard the timing because the signals its getting make it think the engines under higher load and closer to the detonation threshold .
Hard to have a win with OE gear but its fun trying .

Cheers A .

[steptoe]

Disco had bigger plans than a 4 plug ECU. Just give him time - or give him the money. I like my 3 plug ECU system as it runs now - the Karman Vortex air flow sensor out in the shed, fuel pump and injector plug disco nnected leaving the ECU to control EGR function and let the green dash light glow happy with my (slightly) explosive LPG system