On 08/23/2017 09:09 PM, Jon Elson wrote:
On 08/23/2017 01:27 PM, John Kasunich wrote:
Numerical integration is unnecessary. Energy stored in a cap
is 1/2*C*V^2, for 160V and 4700pF that is 60uJ. Times 100K
events per second is 6W.
Thanks so much for the analytical solution and for
confirming my calculation is not
way off the mark!
(The cap energy is completely dissipated in the resistor,
because the RC time constant is only 47nS, far shorter than
the period of the PWM waveform. So the cap is fully charged
and discharged during each cycle.)
You could lower the capacitance to reduce the dissipation.
Of course at some point the snubber will no longer be effective
at damping oscillations.
Right. The failure mode is when the resistor goes open,
the amp trips the fault logic.
So, it seems the snubber is needed. Now, for exactly HOW
much snubbing action is actually required,
I don't really know.
Did you put the snubber in there to deal with a specific
oscillation or ringing problem, or just on general principles?
Actually, for a totally different reason. On the original
brush motor servo amp, I had a destructive event when
a half-bridge was sourcing current to the motor. When the
high-side transistor shut off, the filter inductor
continued to pull current from the half bridge. This
pulled current through the low-side transistor, and the body
diodes are REALLY slow to turn on. When the junction of
the two transistors went to about -7 V, parasitic diodes
in the FET driver started conducting, and they couldn't
handle 20 A. Both transistors turned on and shoot through
blew them both. I added a super fast diode to prevent
that terminal from going negative enough to cause damage.
And, at that time, I added a snubber network to give the
diode time to turn on.
If dealing with a specific problem, there are design techniques
that can be used to minimize snubber dissipation. You need
to know the ringing frequency and the impedance of the
resonant circuit (either knowing the L or the C will do).
I can go into more detail if it would be helpful.
Well, I generally recommend running these drives at no
more than 120 V DC, and everybody has been pretty
happy there. Well, at that voltage, the resistors are
still well above their rating - so piss poor engineering!
Changing up to the 2 W resistor will get them within the
rating, so I will have to do that for any more units
built. When I next revise the PC board, I can think about
other techniques. Also, it is possible the caps could
be reduced. I'd have to set up some experiments to see
what level of snubbing action is actually required to
keep the logic happy, and then make sure I have more.
Thanks,
Jon
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers
