This is close the exact component I have.
http://www.maida.com/content/product...6121ZOV181RA04
It shows 320pF. Its marked 04UL is supposed to mean 40 joules, although the
data sheet says 60J.
Tim

I want to conclude that the MOV can short pulling over 25ma thru the 1/8th
watt resistor, heating it up, due to 10khz components on the voltage, and
not an actual high voltage. Would this be a true statement?

Thanks

"tjs" wrote in message ...
Sorry , I crossposted to rec.radio.amateur.homebrew and sci.eng.electrical
sys-protection

I'm having trouble finding data on the MOVs. The markings MDC Z181 04UL
seem
to point to a 175 vac 225vdc 17-19 joule device with 130 pf capacitance,
1200 amps Ipeak, 25mW transient dissipation. I searched Digikey and NTE.
It
is about 10mm diameter disk type, maybe 8.5mm per some datasheets.

Tim



"tjs" wrote in message ...
Not a REC.Radio issue but still appropriate here...


Question: will 1-10khz harmonics riding along on 60 hz power cause
protective MOVs to charge and short out thusly drawing damaging
currents.

The scenario: I have an Allen Bradley PLC digital input card and it
contains
1 MOV across ac input channel terminals (I think they are MDC Z181
spec).
They are there to protect against high voltages at the terminals which I
contend never occurs, never seen it happen. What does happens is that a
232
ohm 1/8th watt (current limiting?) resistor just ahead of the MOV will
slowly heat up and burn to open circuit in about 30 seconds after
connection. Using 0.125W (P=I^2 x R) 232 leads me to exceeding a
current
of ~25ma drawn to burn the resistor.

The IO being monitored is motor run status from a VFD driven motor
(variable
frequency drive, PWM type, naturally using ~50-100khz synthesis
methods).
The control power is riddled with harmonics from 500hz to 10khz, and I
estimate 10vp-p maybe less (as seen on the oscilloscope). As soon as I
close the electrical connector to the IO card the resistors start
heating
up, smoke, then fail.

I know I need to isolate the control power and rid the plc of the
harmonics.
I just want to confirm the high frequency components can cause MOVs to
short
as if there was a high voltage event when there isnt one. I beleive the
capacitor model of a MOV means it should charge up at higher frequency,
and
maybe this is why it takes 30 seconds for smoke to appear.

Regards
Tim KF8XW

tjs wrote:
This is close the exact component I have.
http://www.maida.com/content/product...6121ZOV181RA04
It shows 320pF. Its marked 04UL is supposed to mean 40 joules, although the
data sheet says 60J.
Tim

I want to conclude that the MOV can short pulling over 25ma thru the 1/8th
watt resistor, heating it up, due to 10khz components on the voltage, and
not an actual high voltage. Would this be a true statement?

Without having the exact waveform, it is impossible to say exactly how
much current the MOV capacitance allows through. I suggest you try an
experiment of removing one of the MOVs and see how much cooler the
resistor runs.

Yes I think I will have to test the MOVs to really know. Or as you suggested
take a MOV out and measure resistor temperature.

"John Popelish" wrote in message
...
tjs wrote:
This is close the exact component I have.

http://www.maida.com/content/product...6121ZOV181RA04
It shows 320pF. Its marked 04UL is supposed to mean 40 joules, although
the
data sheet says 60J.
Tim

I want to conclude that the MOV can short pulling over 25ma thru the
1/8th
watt resistor, heating it up, due to 10khz components on the voltage,
and
not an actual high voltage. Would this be a true statement?

Without having the exact waveform, it is impossible to say exactly how
much current the MOV capacitance allows through. I suggest you try an
experiment of removing one of the MOVs and see how much cooler the
resistor runs.

tjs wrote:
Yes I think I will have to test the MOVs to really know. Or as you suggested
take a MOV out and measure resistor temperature.

I was suggesting that you might prove that the MOV current is the
cause of the resistor overheating, by removing an MOV and applying the
input signal. I can't guarantee the input will work, correctly, however.