[Supras] SAFT, AFM signal add-ons

[berniek at technicaldevelop.com]

Dear Jim and list:

    Thanks for the insight about the TPS.  The last car I really examined in any detail was the 84 300 ZX which only had a switch for idle for the TPS, no pot from which to gain rate of throttle opening information.  Many years ago when monolithic opamps first appeared I did an opamp based version of the very early Bosch speed-density system used on VW type II's.  The ECU then, if you want to believe it, was made from discrete transistors, with provisions for base-emitter temperature compensation (injector drive transistors were Germanium, if you want to believe it).  The MAP sensor was a variable reluctance, bellows operated monstrosity, excited from an oscillator in the ECU (forget the frequency).  It had an accelerator pump-like function which operated by taking the rate of change of manifold pressure, with an exponential decay time, emulating the spring extension in an accelerator pump.  

    Regarding the 7M, what I know about tip-in problems with air flow tuning add-ons is from what I've read, since I still need to complete the replacement JDM engine assembly and installation (BHG in the original '90 at 156K), let alone the upgrades.  Time is so precious during business hours that I want to do what I can myself, nights and weekends, especially since I'm not looking for tons of power, and since almost all working hours are taken up with necessary other activities (i.e., serving the needs of clients to put food on the table).  Here in northern NJ, the one place I knew of which had a chassis dyno and catered to the performance crowd has long vanished (tire dealer in Verona, NJ).  I took my '65 GTO there once, and made 240 RWHP, good for the time, with cam, carb, and shorty header mods.  The session did not last long, as the overpressure plug blew out of the dyno driven pump.  

    Your statements about the 7M TPS providing accelerator pump-like functioning is comforting, as is your contention that the stock richness under boost is not that bad (will follow suit with the Lex and 550 injectors).  I've searched out and saved a couple of charts illustrating A/F ratio vs. relative power, and I agree with you, on the assumption that the A/F is not richer than about 10:1.  Speaking of which, I'm curious about your statement that your engine did not run well at 10:1.  Any leaner, and EGTs would be higher.  Any ideas?  

    I've looked at the Apexi SAFT information and a writeup intended for Mitsubishi 3000 installation (not that different from the 7M) and now understand it pretty completely.  It still rubs me the wrong way, though, that a near-step change in throttle position at off-idle is not compensated for due to frequency recalculation latency (I did not see any TPS pot input for the SAFC for a stronger "accelerator pump", but if it was there, I could have missed it).  

    Yeah, a pot in series with the air temperature sensor would work, but a paralleled thermistor with the right curve should be better, since it will track in percentage terms.  Common thermistors exhibit a fixed and predicable percentage change per temperature increment, much like the way compound interest works.  There are many sources, like Fenwal, Yellow Springs, and VECO.  

    From what I've read, I agree with you completely about the stock CT-26.  After the engine installation, 550 injectors and the Lex, an upgraded trubo will be next, but again I don't have the time to flirt with not being on the safe side regarding NJ inspection, so both cats will remain in place.  The standards are strict here.  With some cars here, I've had to retard the spark 10 degrees or so to pass HC and CO, despite good running engines, with new injectors and spark plugs (particularly for our salt-sacrificial winter "mule" car, a '92 Maxima with the SOHC V6).  

    I'd like nothing better than a WB O2 sensor, but the prospect of having to set up A/F, then remove it to avoid temperature damage is not appealing.  I do have several thermal conductivy exhaust gas analyzers here including two Sun units from some time ago, and a Marquette unit.  I could use those on the road for WOT tuning also.  There is a section of westbound I-80 here where I can hold WOT for 10-15 seconds in 4th gear, an uphill climb, where cops are scarce.  

    IC piping:  I do not understand the point in using PVC pipe, especially since most of it is Schedule 40 (relatively thick wall as compared with steel exhaust tubing or even copper tubing, M grade being thinnest wall).  When I do the upgradies you mention it will be with thin wall steel or M grade DWV copper.  

    Again, thanks for your informative post. 

        Bernie
--------------------------------------------------------------------------------------------------
Jim Jobe had written:

The TPS is used for throttle tip-in, called TPS based acceleration or
compensation, much like an accel pump in a carburator.  Think about how
far away the 7MGTE afm is from the throttle and intake valve.  It's needed
even on my car (AEM EMS) with the map sensor plumbed directly in to the
plenum.  In my experience with piggybacks, only the VPC makes the throttle
tip-in hesitation worse.  Most of the 7MGTE's I've setup respond the same
after installation of the s-afc, s-afr, maft, and/or lexus afm.  There are
many reasons as to why the 7MGTE throttle response sucks, however, the afm
is not one of them.

I personally don't feel the factory map is as rich as the bandwagon
suggests.  Not many on the bandwagon have put a wbo2 on a completely
stock car (nor have I), but I do know how shitty my car runs in the 10:1
range and the factory setup never ran that bad.

A mechanical solution will not work very well as the change will not be
linear and most likely on/off.  Although some have gotten away using one
of the openings on the accordian pipe and literally putting a small K&N
filter on it, then waiting for the ECU to trim it back out.

You could in fact add a potentiometer inline with the air temp sensor
or the coolant sensor.  Adjust until mixture is where you want it when
fully warmed up, but keep in mind the compensation will plateau then drop
again, adding fuel to cool the cylinders back down.

Skip the fancy lightshow with the narrowband sensor.  There are plenty
of charts available on the internet that show how inaccurate, no, how
ineffective the sensors are outside of the narrow band around lambda.
Wideband O2's can be had for $200-300.  Innovate LM-1 is cheap and works
well.  NGK/NTK is nice.  FJO is very good, about as good as you'll get
for the money under 4 figures.  Most of a simulated nbo2 output so you
can replace the factory sensor completely solving your emissions inspection
issue.

Personally, if you have the walbro, 550's, and other supporting fuel
system mods (proper feed/return lines and afpr), get a MAFT, MAFTv2,
or MAFT-PRO.  The cars I've seen with them run great and it's a cheap
solution.  The stock ct26 can handle random blasts to 15-16psi as long as
it's in good shape to begin with.  I can recall many a race we got in
to where we turned the boost controllers all the way up, CT26's spiking
to 16-18 psi and rolling off to 14psi.  I'd also recommend you fix the
bottlenecks in the IC piping (in particular the aluminum elbow and the
plastic pipe over the turbo, and the end of the 3k pipe).