[NickZ]

if they are from Aliexpress then they are highly likely to be fakes, some fakes will probably work OK but there is no easy way to test them. Sourcing small quantities of genuine CJ125s has always been a problem because of the 2400 minimum quantity but they are now on 6 months lead as well. I don't have any sources anymore, I have a long term project to produce my own control circuit using discrete analogue circuitry. I would suggest that you check that the sensor is correctly installed, ensure that it isn't turning on when cold and ensure that you have a genuine Bosch sensor, I only source those from reputable motor factors, no eBay or Amazon sensors.

Have you seen this project?

https://rusefi.com/forum/viewtopic.php? ... 2237ab2aed

these are great, im currently using them, BUT they are CANBUS output only

[dener0987]

Guys, I was just noticing. I am using design 0.0.2. Looking now at the 0.0.3 hardware I noticed that there is a difference. The voltage reading on 0.0.2 is done using the +12v label while on 0.0.3 it is done using UB. Could this affect how the controller heats up? Since the read voltage will be different? That makes sense?

[Alfagta]

Hello,

I dont think it makes a big difference, the voltage with 12V label is max 0.7v lower do to the froesrd voltsge of the diode.

[JHolland]

https://rusefi.com/forum/viewtopic.php? ... 2237ab2aed

I have seen that, there are a few similar projects. I designed a purely analogue controller because I needed a high bandwidth for what I was doing; the bandwidth for the CJ125 isn't stated and I only have a few genuine chips left, I did consider the L9870 which has a 100kHz bw but they were not available at the time.

[dener0987]

Hello,

I dont think it makes a big difference, the voltage with 12V label is max 0.7v lower do to the froesrd voltsge of the diode.

OK. Another thing I was thinking about. Perhaps it is not better to increase the condensation time or maybe increase the USUP_MIN variable? So that the battery voltage stabilizes after starting? Because, for example, if the probe enters the heating phase and the voltage is not stabilized, from what I saw in the code it only takes a reading of the battery voltage at the beginning of the heating phase to calculate the step, but if the battery voltage increase during the warm-up phase due to the alternator, could this not cause problems?

[chaoschris]

I dont think it makes a big difference, the voltage with 12V label is max 0.7v lower do to the froesrd voltsge of the diode.

The forward voltage on this Schottky diode depends mostly on temperature and current. I expect less than 0,5 Volt, more in direction 0,4V

[dener0987]

Another thing guys.
Does anyone have a discrepancy between what is marked on the controller and what is shown on the ECU? I believe this is due to not having a gnd for signal and a gnd for the heater.

[JHolland]

Another thing guys.
Does anyone have a discrepancy between what is marked on the controller and what is shown on the ECU? I believe this is due to not having a gnd for signal and a gnd for the heater.

If you put a scope on it does it change as the heater turns on and off?

[dener0987]

Another thing guys.
Does anyone have a discrepancy between what is marked on the controller and what is shown on the ECU? I believe this is due to not having a gnd for signal and a gnd for the heater.

If you put a scope on it does it change as the heater turns on and off?

I didn't get to test it with an oscilloscope. I was just basing it off the megasquirt grounding scheme.
As speeduino does not have a separate ground for sensors and power. The sensors must be grounded to the gnd of the ecu. However, as the wideband controller does not have a separate gnd for the heater, I believe it is not possible to connect the gnd of the controller directly to the gnd of the speeduino board, as this would interfere with the reading of all the sensors.

Page13:
http://www.msextra.com/doc/pdf/MS3baseV ... re-1.5.pdf

[PSIG]

IMO, this is a potential issue with any system that uses common grounds, rather than isolated grounds. However, the point is that the signal grounds must be treated equally. If you do that, then it will read equitably anywhere in the system.

This is one reason the ECM has multiple ground wires, in order to share the current, change in resistance and voltage signal. But, if the controller and the ECM have the same ground reference then they should read equally, including any real offset. Additionally, any potential offset is minimized by the multiple low-resistance grounds to the same single block ground point. The data is still relative to each component.

I hope that makes sense of one way to deal with common grounding, assuming no odd offsets are created in the component internal circuits.