Slow burn project- T1 Speeduino EFI

[Bruce.m]

Engine is back in & running. Oil leak at the bellhousing is gone for now. Oddly, there was a drip from the back edge of the left under tray tin. Have cleaned the tin & will see if it reappears.

[Bruce.m]

This is the car the engine is for. Was doing a few jobs on it today. Still need to do an alignment & fit the shocks.

[Bruce.m]

Couple of learnings from recent experience.

I’ve was adjusting the ITBs side to side, to get them balanced with an air flow meter. This was more trouble than I expected with the cable linkage. I seem to have it close now but the cable has a bit a stretch & I need to preload the tension but without holding the slave side open. There are two types of control cable. There is “wire rope” and “twisted strand”. Wire rope is more flexible as it is a bunch of smaller twisted strand wires, which are twisted as a set (like rope). But wire rope has more stretch. So I’ve ordered a length of 19x1 twisted strand cable which hopefully has less stretch & makes it easier to balance the ITBs.

I had a problem with the fuel pump. It started leaking up through the wire terminal socket. Not much, just a small seep. So I ordered another pump since I’d had it a while & they aren’t expensive. After a few hours of use, the replacement did the same. I’ve now repaired one with epoxy & so far that has sealed it. However the reason seems to be due to using a “dead head” fuel system. I have the pump at the front of the car & next to it is the fuel filter & pressure regulator (one part). The regulator output feeds back into the fuel pump housing as per its design. However because of the short path from the pump, to the regulator & back to the pump housing, only a small amount of fuels is circulating through the pump. This means the pump get warmer than ideal. With hindsight, having a longer circuit to the regulator & back (double pipes to the engine bay & back) would flow a lot more fuel around the circuit & should keep the pump cooler. I will probably do that upgrade over the winter, as I did plan to have a boost sensitive regulator in the future anyway.

[wreck]

when I was using a home grown cable set up on carbs , I needed a light second spring on the slave side , it also took a little while to settle in after the cable was ever removed and refitted , normally had to adjust it a few times to take up a little slack over a few days driving , once that was done it was fine .
With the injection , I made a centre crank with rods , the crack is snail shaped , so I get better throttle feel at light throttle . works well . I've 50mm ITB's on a 2600cc engine .
using LS1 coils .

throttle .jpg

[Bruce.m]

Thanks. I think you are right & it did improve the next day after the cable had some time at preload. I also need to make some improvements to get more pedal progression.

[Bruce.m]

A busy day getting the suspension alignment done & shocks fitted. Also cut the base plate for the left air filter & got that fitted.

Still having issues with the throttle body balancing. I replaced the Bowden cable so it might need to settle. I also tried to add preload to the stock return spring on that side, but it wasn’t possible. Have a few more adjustments to make or I might need to add an extra spring. Hopefully the last issue to solve before tuning.

[Bruce.m]

First drive. A few miles to get a tank of super.
Tune was rough & had to get through tight village streets, up a hill, hill start at lights, before on an open road. Auto tune had made a big improvement by the time I got to the gas station. Took a slightly longer route home & it’s getting better.

I’ve got a big hole if I snap open the throttle at idle. Guess I need some TPS enrichment under 2k. Elsewhere is fine. I did have some poor areas if I loaded it up under 2k but I’d set the min rpm for auto tune at 2k. Have dropped that to 1.5k now.

[Bruce.m]

Tune improving. Used it for various errands today with auto-tune doing it’s thing. Due to the stock cam the torque gradually falls from 3k.

Dialled out the flat spot off idle somewhat but a sharp stab still falls in a hole. I may have to switch to ITB mode below 2k-ish to fix that.

No sound deadening, carpets, tar board etc, at the moment so it’s very loud.

So far the brakes are excellent. (Boxster full set)

[wreck]

thats great you've got it driving , you can tune the AE stationary , just blipping the throttle . What settings did you put in the O2 delay chart .
I'm still playing with that , using the hard rev limiter as a cut off to check in the log . I've good data for higher in the revs but still need to do the lower end .
I'd like to get the open loop working .

[Bruce.m]

Thanks. I might have found a wrong setting in AE which might be my problem for off-idle. Will check later.

My lambda delay settings were a best guess using some basic maths, to calculate the volume of exhaust pipe up to the sensor. Then calculate the amount of time taken to fill that space at various revs / VE, based on cylinder volume * VE & msec of time from rpm [1/ (rpm /60 /2)] . I guessed the settings above 2k would be less critical so put a limit of that to begin with.

Basic tune is fairly good now but getting a great tune is another matter.

For example. The temp sensor on the underside tinware seems to be cooled by the airflow so I’m getting lower temps than I expected on the move. I used the solid metal coolant version. As a test I could swap it for the intake air temp sensor which is exactly the same part except the sensing end is plastic & should be less affected by the metal part cooling. Or I could move the whole part into the hot area so it has no cold air contact.

[wreck]

I'm using the same sensor for "coolant" temp and intake temp . just a generic motor sport open type . It is screwed into the lower cylinder tin near the head . It does read lower than CHT , I have my warm up enrichment fully off at 75C . (about 95C CHT) .
I have made a "coolant sensor" out of a 100K thermistor with a range of -40C to 250C . I soldered to leads to the thermistor , then put a little thermo paste into the bottom of a crimp terminal for 6mm cable with a 6mm id ring , then filled it with epoxy and then a bit of waxed heat shrink to finish it off . I plan to just bolt it to the head where the upper cylinder tin screws up.
https://www.jaycar.com.au/100k-glass-nt ... b736bec8b5

[Bruce.m]

I was pushing it very hard and struggled to hit 60C.
On the positive side, the engine (a 1641) is running well & not overheating

[Bruce.m]

Fixed the cause of the off idle hole. The AE page had the graphed curve set to add 30%+ rising to 150% or so. However underneath the drop-down menu was set to multiplier & not adder. So it was pulling fuel instead of adding it! I’ve got it fairly well dialed in now, at least good enough to not stall / choke pulling away from standstill.

Also noticed when reading logs, that the O2 reading was a continuous 14.6 & not moving around. I pulled the sensor & found it was coated in soot. I gave it a gentle clean with a blowtorch. Just enough heat to burn the carbon off but without over heating the sensor. Re-fitted & went for another tuning run. The tune pulled some fuel from the peak around 2.3k and added more up to 4.5k. The upper range has now opened up and it pulls to 4.5k. Given the stock cam I’m happy so far.

I do need to get the dash screen set up now, because I’m having to guess the rpm by ear currently. That’s likely to be a winter project, using ESP32 microcontrollers.

[Bruce.m]

[Paul H]

If you switch to a run around fuel system you'll end up with a tank of hot fuel with the fuel rails acting as a cooling system for the engine bay