[Jyri]

I thought that those are only for the analog pins?
At least pin 8 that is for INJ1 goes directly to the teensy pin 8 I didn't check the others, but I think they all go directly.

[PSIG]

I said they should have — not that they do have.

The root of this specific issue is the use of MOSFETs with gates requiring CMOS level signals (typically 10V), rather than Logic-Level gates (5V or less) to turn the FETs fully on.

There are three general responses to this situation. First is to use MOSFETs with LL gates for either Teensy (3V3) or Mega (5V). The others are to raise the drive voltage to the existing FETs with either a level-shifter, or a proper push-pull gate driver much like the IGN circuits have.

This has been an issue with most board designs since the beginning, but they have "worked" at 5V, and with a lower parts count and cost the designs were left as-is. When you aggravate the situation with even lower gate drive (signal) voltages, the issues become magnified. Pick your solution, but the easiest and simplest fix for existing boards is likely to change the INJ FETs to LL type. If designing a new board, then MOSFET drivers would be the most efficient solution, at slightly higher cost.

[Jyri]

Ok, that clears that.
Why would it be designed with the STP75NS04Z, Is the cost or?
Would something like this https://www.infineon.com/dgdl/irl1404pb ... 5b382824f0 (IRL1404) work directly on the board without other modifications?

[dazq]

The output drivers were chosen for their spec and work perfectly, but this was when only the mega2560 (a 5v MCU) was in use!
since then various other MCU using 3v3 have been added this has where the problem came from

[PSIG]

Why would it be designed with the STP75NS04Z, Is the cost or?

I couldn't say. One comment was that the VGS(th) was under 5V. Unfortunately, that spec is for "barely on", where it can't carry any useful current. It is a robust protected and avalanche-rated device though.

Would something like this https://www.infineon.com/dgdl/irl1404pb ... 5b382824f0 (IRL1404) work directly on the board without other modifications?

It could. The datasheets do not show anything below 4.5V,so it's not so easy to back-calculate the performance at 3V. If you have done that and it works-out, then go for it.

There are a number of more common alternatives, and one random type example is the IRLB3036 or the IRLB4030, both of which have performance graphs down to 2.5V (VGS). Note as the gate voltage goes down, the cost goes up. Pick your parts and let us know how it goes.

[Jyri]

Why would it be designed with the STP75NS04Z, Is the cost or?

I couldn't say. One comment was that the VGS(th) was under 5V. Unfortunately, that spec is for "barely on", where it can't carry any useful current. It is a robust protected and avalanche-rated device though.

Would something like this https://www.infineon.com/dgdl/irl1404pb ... 5b382824f0 (IRL1404) work directly on the board without other modifications?

It could. The datasheets do not show anything below 4.5V,so it's not so easy to back-calculate the performance at 3V. If you have done that and it works-out, then go for it.

There are a number of more common alternatives, and one random type example is the IRLB3036 or the IRLB4030, both of which have performance graphs down to 2.5V (VGS). Note as the gate voltage goes down, the cost goes up. Pick your parts and let us know how it goes.

I couldn't find either of those, but I did find irlb3034 locally, they seem to work ok, I managed to get the car running and have run it for approx. 5h or so, so I would say that they are also a notable replacement, if you want to run teensy+adapter with 0.4.3c board.

And thank you all for helping me.