0

MC43 low side pulled up

Petit 2 months ago in Master modules / MC4x updated 1 month ago 6

Hi all,

I'm currently testing an MC43 on our HIL bench and notice some issue with the low side driver at startup.

On our system, all the low side driver of the MC43 at pulled up to supply voltage with a 10K pull up resistor.

The problem is that when we switch on the ECU, it's going into error with the code 4:2.

We find out that by waiting 1.5 sec at startup before pulling up the low side driver fix the problem. 

The problem is that on the vehicle we will probably end up in this situation as some low side driver are going to be pulled up throw some relay coil. 

Is the unit is going into error if any low side driver is pulled up or only some of them trigger it ?

Thanks for your support,

Grégoire

+1
Under review

Is it the low side outputs used for COUT and Digital out HS+LS ? In that case it sounds as if you are triggering the error detection that the MC4x have in place for protecting against reverse feed. 

Or is it the single LED/relay driver outputs, Digital out LS ? Then I don't know what is going on, these outputs are intended for that type of connection. 


Can you answer the questions below?

We do pull up all low side drivers to detect there state by a voltage measurement, so the (LS) COUT's are pulled up as well.

So yes from error code 4:2 we expect a reverse feed, but we don't understand why.

On the vehicle we have eg. a coolant pump controlled by PWM with C1:39 (Out11) .

The Pump has a separted supply and is measuring the rectangula PWM Voltage signal with a very low load (~1.2mA) current on this PWM pin.

How can we detect the Lowsidedriver state without pullup, we thougt we can do it based on voltage?

Does this mean we can not use the PWM outputs which are at the same time Current outputs as a PWM output without Load or without related High Side connected?

Do we really need to remove all Pullups on the COUT LS drivers at startup?

Why would the coil of a proportional valve not triggering this error when the other side of the coil is connected to the supply?

Thanks for your support,

Grégoire

+1

Yes, you need to use both the highside and the lowside as a combination (either PWM out HS+LS or COUT), with a load between them. 


It sounds as if your application has to control an external device with a PWM command. There is a related discussion on a problem with that type of connection here, this one is about off-state open load detect: 

https://forum.iqan.se/communities/1/topics/2201-mc4x-option-to-disable-open-load-detect-on-pwm-out-hsls


The reason for the check at startup is to prevent reverse feed via the internal flyback diode. 

At startup, the MC4x checks that there is no voltage on the switch supplying the high-side drivers.

As the PWM out HS+LS is always controlled as a pair (same as the COUT), it is easy to check this at startup.  



I didn't understand the part about "detect the Lowsidedriver state" ?

When you control a regular coil with a highside lowside combination (like the COUT channel), the monitoring of the complete function is done by the controller, and reported on a high level as status on the channel in the application. 

+1

Thank you very much this clarifies all my questions.

The Lowside driver state, we want to detect from the outside with our Hardware In the Loop Test Stand.

We wanted to sample the PWM duty cycle and frequency based on the Voltage.

It’s a bit inconvenient but we will remove the Pullups, and try to rearrange the PWM outputs.

I have a similar issue with PWM Drivers on the MC41.

Challenge for me is that a major cooling fan OEM needs a Low Side (open collector) PWM signal to control the fan speed. 

With the way the MC4x series is set up, I am required to have a pull-up/down resistor in the harness to get this voltage in the right range - which is not ideal.

If I connect to the PWM-LS driver (or even any other PWM output) - the pull-up voltage coming from the fan itself is detected as a "short" condition and the device does not start. (if my harness resistor breaks, I would be in the same boat- leading me to an non-diagnose-able failure mode with no system operation).


Can we please have the ability to disable these STB / STG conditions on PWM outputs? That would make life so much easier, and the hardware much more functional in this case.

Thank you!