Searching \ for '[EE]:Reversing DC Motor Circuit Problems' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page:
Search entire site for: 'Reversing DC Motor Circuit Problems'.

Exact match. Not showing close matches.
PICList Thread
'[EE]:Reversing DC Motor Circuit Problems'
2002\09\22@150230 by DFansler

picon face
Try it again with a subject line!

I started building an observatory with a roll off roof about 3 years ago.
The structure itself is finished.  The roof probably weighs about 2000 lbs
and rolls on steel wheels and while it can be pushed by hand, it takes a far
amount of force.  Not being fond of being cold, I constructed the mechanics
to roll the roof back using a couple of 24v dc motors.  The controller box
uses a 16F877 as the brains.  I am using (trying to at least) to use the PWM
to control the motors with.  The output of the PWM goes to a 220 ohm
resistor and then to the gate of an IRLZ44 (N-channel 60v 50A FET) which
between the motors and ground.  There are a pair of relays that reverse the
leads of the motors to change direction.  I am aware of back EMF and the
need for a blocking diode.  I have a blocking on each relay coil and another
diode between the +24v and the FET.  The FET works the first time, ramping
the motors up in speed, but after the relays switch the first time, the FET
becomes a dead short.  At first I thought the motors were drawing too much
current, so I tried two of the IRLZ44 paralleled - same result.  Anyone got
any suggestions?  I would really like to get this finished.

If the word pictured is not clear enough, I can place the schematic on my
web site.

Thanks to all,

David V. Fansler

-- Going offline? Don't AutoReply us!
email with SET PICList DIGEST in the body

2002\09\22@175831 by Brent Brown

picon face

Had a go at reversing a DC motor with a double pole changeover relay
once and it was a dismal failure. My failure might give a clue (and a

The double pole double throw relay was arranged like a full H bridge
(you work out the circuit , it's not that hard). The motor field
winding was seperate from the armature and had relatively large
inductance and drew about 6 or 7A @ 24VDC. This was much much lower
than the armature current (about 400A) so we decided to reverse the
motor by reversing the field only (this is a valid method).

The intsant the relay tried to changover an arc was dragged between
the opening relay contacts, and carried over to the new contact
position. If you've sketched the circuit for this you will see that
we now have an arc established between the +ve and -ve battery rails!
Even though the contact separation should have been enough to handle
24V (or even 240V) the arc is sustained because once it's there the
impedance is low and voltage is low. Poof. There goes the relay. And
all the circuit board traces!

The problem is you need an inverse parallel diode across the motor,
to get rid of the back emf, but that only applies in one direction.
AC voltage clamping devices like a VDR or TVS applied accros your
motor may work, but would have to be at least 35V or more for a 24V
circuit, and I suspect that would be "too little too late" in
preventing arcing in your relay. An RC sunbber may do the trick, but
values may not work out practical.

The solution we settled on was a full H-bridge of power MOSFETs, each
with a built in diode. The diodes form a bridge rectrifier circuit
which clamps the motor back emf to the supply, regardless of
direction. Worked well for us. (No more smoke!)
Brent Brown, Electronic Design Solutions
16 English Street, Hamilton, New Zealand
Ph/fax: +64 7 849 0069
Mobile/txt: 025 334 069

-- Going offline? Don't AutoReply us!
email with SET PICList DIGEST in the body

2002\09\22@232916 by DFansler

picon face
My thanks to both Tom and Brent for their comments.

Tom, you are right I do have a mounting pile of clown parts and the joy of
the project is some what diminished at the moment!  Most of the software is
already written and ready for real life debugging - which is what I was
doing when I discovered the problem.  So unfortunately I seem to be ahead of
the curve you suggested.

Brent - I feel that your relay H-Bridge is probably similar to what I am
doing.  There are some interlocks that keep my relays from being active at
the same time - the coil of each relay goes through a pair of normally
closed contacts on the other relay, so one can not be energized until the
other releases.  If  I were dealing with high voltage, I would expect arcing
of the contacts, but I guess I would be surprised to have it at 25vdc.  But
there is no doubt that something dreadful is going on to keep blowing my

Looking over a SGS Motion Handbook from the 90's, they have an application
note using a pair of 30DB040D Transpack NPN Power Darlington Module (which I
have 10 of) to form a H-Bridge.  My skills in the analog world are almost
non-existent when it comes to biasing transistors and the like.  The app is
not complete, so I can not just copy what they have and use it in my
application.  So I guess I am going to keep looking and hope to find a
circuit to use.

Thanks again

David V. Fansler

-- Going offline? Don't AutoReply us!
email with SET PICList DIGEST in the body

2002\09\23@165310 by Peter L. Peres

picon face
It is impossible to tell what kills your fet from your description.
Perhaps unfortunate ground setups allow the kickback to propagate to the


-- hint: To leave the PICList

2002\09\23@230045 by DFansler

picon face
My thanks to all who have responded with helpful suggestions.  The motors
probably draw ~10A each so typical H-bridge drivers are not stout enough to
handle the situation.

I was in a Barns and Noble today and was looking around in the electronics
section and ran across one of the many build your own robot for competition
(i.e. Battlebots).  They had a section on powering large dc motors and had a
number of commercial esc for such, but then they noted the existence of a
group that had an Open Source Motor Controller capable of handling dc motors
with continuous ratings of 300A at up to 50vdc.  This group can be found at <> .  Long and short of it I
ordered an OSMC kit (PC board and all parts) for $164.  The OSMC has inputs
for each of the 4 legs of an H-bridge, plus an inhibit line.  The on board
controlling IC accepts these signals and output the same to the bridge of
FETs - the chip has a charge pump to provide a signal 10 volts above the
drive voltage to the upper legs of the h-bridge.  This same chip will not
allow both legs of one side to be active together.  BTW it is a Intersil

So now all I have to do is make a few hardware modifications to provide PWM
signal to the correct inputs, re-write the section of the software that
drives the PWM to include direction, and move everything into a larger
enclosure.  I will soon be able to start enjoying my telescope again.

Again, my thanks to those with helpful suggestions.

David V. Fansler

-- hint: To leave the PICList

More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...