What do you think the TVS diodes on the drains of the two transistor are doing?

Hi Dennis, they are to (hopefully) protect the mosfets from spikes & BEMF from the motor being run by Q1. We duplicated it for Q2 but it only runs some lights so no inductive spikes on that one.

 

Also, I don't understand what R1 is doing... is it there to limit current in case of a short circuit?

Hi, R1 allows zener D4 to pull down the 12 VDC supply voltage to 5 volts that the micro needs. Its a cheap voltage regulator, I guess.

 

I just want to say a very big thanks to all of you for the great and very helpful advice so far!

 

I have ordered some parts from Mouser and will try all the suggestions you guys kindly made today. As soon as I have tested them all I will re-post and let you know how I got on, and which worked the best,

Again, many thanks!

Peter

 

Thank you, this makes a lot of sense. So, if I drop the pull-up (R4 and R5) to 1 k each, do I also change the value of R2 and R3? Do you feel a transistor will work fine, as opposed to changing Q3 and Q4 to mosfets as another very helpful member suggested?

Yes, you might want to change R2 and R3 also.
For good saturation of a BJT the base current be about 1/10 of the collector current. For 5V drive to the base and 12V to the collector this means R2 and R2 should be about 4kΩ.

At a 1kHz PWM I see no reason to change Q3 and Q4 to MOSFETs as the BJTs should switch more than fast enough at that frequency.
As I noted, the limitation in the switching speed is largely due to the MOSFET gate charge.

 

Yes, you might want to change R2 and R3 also.
For good saturation of a BJT the base current be about 1/10 of the collector current. For 5V drive to the base and 12V to the collector this means R2 and R2 should be about 4kΩ.

At a 1kHz PWM I see no reason to change Q3 and Q4 to MOSFETs as the BJTs should switch more than fast enough at that frequency.
As I noted, the limitation in the switching speed is largely due to the MOSFET gate charge.

Hi Crutschow,
Thanks so much, I'm trying to get out of the office and into the workshop to modify the test circuit with your suggestions. I'll post my results as soon as I'm done. Hopefully I don't "let the smoke out" - lol

 

Firstly changing R2 to 470 ohms and R4 to 1000 ohms completely handled the Mosfet Q1 heating up – so that’s a huge relief! Many thanks to Crutschow and others for supplying such easy solutions!

A Couple of more questions if you or anyone has a couple of minutes;

1) The original TVS was part # SA36A (Vishay). An engineer we used a little while back specified a surface mount part to replace it saying it was identical in specification. It is part # SMF16A-TP by Micro Commercial. I looked at both on Digi-Key and see that the spec is actually quite different. Should I use this as a surface mount replacement? Or is there something better? The working voltage is 12 – 14 VDC.

2) From my testing it is evident that ground plane diode D5 has to go, as it renders the TVS’s D7 and D6 ineffective. The original reason for installing D5 was reverse polarity protection. Is there any other way we could protect from reverse polarity without adversely affecting the ability of D7 and D6 to find a solid ground?

Thanks in advance,

Peter

 

The original reason for installing D5 was reverse polarity protection. Is there any other way we could protect from reverse polarity without adversely affecting the ability of D7 and D6 to find a solid ground?

You may use a p-mosfet as a polarity protector, watch the explanation on this video.

 

You may use a p-mosfet as a polarity protector, watch the explanation on this video.

Hi cmartinez, thanks for this, very interesting video! My circuit can draw up to 35 amps + a margin for the inductive load when the motor first energizes. I guess the mosfet would need to be rated at this amount minimum?

Peter

 

Hi cmartinez, thanks for this, very interesting video! My circuit can draw up to 35 amps + a margin for the inductive load when the motor first energizes. I guess the mosfet would need to be rated at this amount minimum?

Peter

Your fet should be rated for at least twice that much, plus properly protected with TVS between its pins. I also suggest you take a look at inrush current limiting resistors, since a motor can draw up to ten times its rated current when starting.

 

Your fet should be rated for at least twice that much, plus properly protected with TVS between its pins. I also suggest you take a look at inrush current limiting resistors, since a motor can draw up to ten times its rated current when starting.

Thanks again for your help, regarding the inrush current limiting resistors - we have about 3000 units out in the field running happily so hopefully we don't have to install these. We are now "soft-starting" the motor by running the MOSFET at 1000 hertz on a 80% ON and 20% OFF cycle for 0.15 seconds. We are not doing this to protect the circuit, its functionality driven, but I wonder if this is providing a "soft start" effect as far as protecting the circuit overall?

Peter

 

I wonder if this is providing a "soft start" effect as far as protecting the circuit overall?

It definitely is... also, I've just thought that maybe you could configure the p-fet that you're using so that you can PWM it and make it protect your circuit against reverse polarity.

 

It definitely is... also, I've just thought that maybe you could configure the p-fet that you're using so that you can PWM it and make it protect your circuit against reverse polarity.

That would be the absolute best solution - how would I do this?
Thanks again!
Peter
====================
Another idea I had would be to put something like a SMB8J16A-E3/5B across the power input if it will protect against it. It's a 16V 30A uni-directional TVS 800W.

Any one have an opinion on doing this to protect against reverse polarity?

 

Oh, one LAST question - The resistor / zener network (D2 & R1) that supply the micro with its 5 volts draw about 11 milliamps all the time. THe micro only draws 1 milliamp on standby though.

To try to get the standby current draw down is there a nice little voltage regulator that we could use instead of the resistor / zener network? My tests show the micro draws about 8 milliamps max, so it can be a very small regulator. A SMD part would be best for this design.

Any ideas?

Thanks again,

Peter

 

..............also, I've just thought that maybe you could configure the p-fet that you're using so that you can PWM it and make it protect your circuit against reverse polarity.

Not likely you can use a single transistor to do both, since the source and drain are interchanged in the reverse polarity circuit as compared to using it as a PWM switch.

 

Not likely you can use a single transistor to do both, since the source and drain are interchanged in the reverse polarity circuit as compared to using it as a PWM switch.

Thanks for the clarification... coming from you I'll take that statement as a fact, and not just as an opinion.
I'm glad that I used the word "maybe" in my previous post...

 

....................
To try to get the standby current draw down is there a nice little voltage regulator that we could use instead of the resistor / zener network? My tests show the micro draws about 8 milliamps max, so it can be a very small regulator. A SMD part would be best for this design.
..................

One listed here should do what you want.

 

One listed here should do what you want.

Thanks Cmartinez and Crutschow!

I found a bunch of 5 volt regulators through the link you supplied. The challenge seems to be to get one small enough, we only need about 10 milliamps, so a 50-100 milliamp regulator would be perfect.Then most of them only tolerate about 6 volts maximum input where this is an automotive product that could see up to 16 volts input if the vehicle alternator is set on the high side. Then we need super low standby current draw, <1 milliamp would be ideal but maybe not attainable?

If anyone knows a regulator please let me know.

Thanks again,

Peter

 

This search turned up this, for example.

 

This search turned up this, for example.

Thank you, I'll check it out and let you know what we decide on