0
\$\begingroup\$

I have an existing motor drive circuit that uses an Allegro 4935 to drive a brushless DC motor. The circuit has been working fine. We've decided to change motors, now with some of the motors the 4935 stops driving prematurely and indicates a fault condition. The majority of the motors can be driven fine however. I can read out a 10 bit word from the fault register and it is 0x42, indicating "VDS exceeded on A phase low-side FET" and "Bootstrap undervoltage detected on phase B". I am pretty new to the circuit but am supposing that the Vds exceeded is indicating the current draw across the MOSFET is too high? Are there any basic tests I can do on the problem motors to try and determine what the exact problem is?

Here's the motor drive section enter image description here

\$\endgroup\$
  • \$\begingroup\$ Isn't this just like a question we had here a week or two ago? \$\endgroup\$ – Olin Lathrop Dec 24 '11 at 17:56
  • \$\begingroup\$ @OlinLathrop - it appears to be a significantly new question based on a common problem set relating to installing a new BLDCM in a system. AFAIR I have asked several questiins about sensor signals and timing which went unsnaswered. May be my memory :-). But I do't think we have had a cct diagram before so we are going in the right direction. Or it may be my memory ... :-). \$\endgroup\$ – Russell McMahon Dec 24 '11 at 19:15
  • \$\begingroup\$ A continuation of bring-up problems after switching to a new motor vendor. I have new data on what may be a new problem. \$\endgroup\$ – fred basset Dec 24 '11 at 20:43
1
\$\begingroup\$

Starter for 10:

Your explanation of what you think the error messages mean does not match what I would understand from them. You may have some reason to interpret them as you do but, if so, you should probably tell us.

High side FET gate drive voltages, which I take to be what they mean by "Bootstrap Voltages" are provided by diode pumps driven via C18, C19, C20 which are provided as as the relevant low side FET switches from on to off, carrying the low side of each cap from near ground to near VIN (VN?) = VBB and the top of the cap to about VDD higher. I take the lack of bootsrap voltage to mean that there is a failure in this system in some manner. This can be due to lack of switching speed, lack of enough duty cycle, dead cap, dead IC and a few more things that Murphy can think of.

Being built as your circuit diagram actually shows is one more such thing!!! (or I need more sleep, which I do, so I may have missed a connection, but I don't think so).

NB - if your diagram is to be believed, bootstrap cap B appears to be wrongly connected and so do Q2 and Q5. This is such a severe error that it seems more likely to be a "drafting/draughting/drawing/where did the dot go on the interscetion with Cage-V" error. ie an example of naughty drawing practices. If it's really wired like this then the last motor did well to run.


Vds exceeded on A side low FET means just that - I'd have thought. It says that pin 34 = SA on the IC is being overvoltaged. How it knows this or if it is correct is TBD. If your motor phases produce double voltage on phases which are off under certain switching conditions then this may be what it means.

This may be the datasheet for your IC - if not then a link to the correct one would be agood idea [tm].

http://www.allegromicro.com/en/Products/Part_Numbers/4935/4935.pdf

Vbb_max = 50V so I'd imagine that exceeding that may generate the message. Or not. If the motor acts as a transformer then thsi MAY occur for Vbb > 25V or so.

What is the Vbb/ motor drive voltage?

\$\endgroup\$
  • \$\begingroup\$ Russel, thanks a great deal for your excellent comments. The datasheet you reference is correct. Vbb is 36V in my setup. The Allegro chip uses the voltage at Vdsth as a threshold voltage for determining the overvoltage condition. It comes from a 1K to 4.99K resistor divider to the left of the chip, that's connected to 3.3V Vcc. So the threshold voltage should be about .55V. \$\endgroup\$ – fred basset Dec 24 '11 at 20:53
  • \$\begingroup\$ RE interpreting the 0x42 hex value I am reading back from the error register and referring to page 14 of the data sheet the way I read it is that the AH bit is clocked out first, and bits are left shifted in the software as they are clocked out, so it ends up becoming the MSB in the word. I don't really know enough about motor driving to understand your comment on the motor acting as a transformer, are you saying the back EMF could be double the supply voltage? I will retest my setup with a 24V PSU next week when I'm back in the lab - thanks again. \$\endgroup\$ – fred basset Dec 24 '11 at 20:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.