Dick, (not W0ID)
If the gold balls theory were valid in this case, then a good high potting
is the proof. Try it. I wouldn't
thump your tubes yet, as it still could be that op amp. Sounds like you are
on to the right folks at AlphaPower.
>Yes, that's what I've seen on most op amps, but this one is in fact designed
>for a higher voltage supply. The op amp in question is a Motorola MC1536,
>which I found on a chip directory web site listed "High Voltage Op Amp +/-
>40V". Motorola's chip literature web site was unavailable, so I couldn't get
>more info.
>The voltages are quite benign: 312 mV on the - input pin and 309 mV on the +
>input pin. That makes some sense to me. Remember that the shunt resistor has
>30 mV across it, which is sampled by what looks like a high impedance input
>network to the op amp. Each side of the shunt resistor is connected to a 1M
>resistor, which is in turn connected to both a 10K resistor to ground and a
>10K resistor to one of the input pins. As a result, I think the 30mV voltage
>drop across the shunt resistor is dropped down to 3mV (312mV - 309mV) at the
>op amp input. Of course, I have no idea how much amplification the op amp is
>set to do. The output is 189 mV, which is cut in half by a divider to 94 mV
>before going on to the microprocessor board. My guess is that the op amp
>output voltage is supposed to directly represent the plate current, perhaps
>plus a small offset to account for losses through the A/D and D/A chips. In
>any event, I think the op amp output should be closer 40 mV.
I looked up the MC1536 and 1436 in Linear/Interface Data book DL128, Rev.
4, Vol I, page 2-91.
Wow, a HV op amp. It has a 7 mV DC max input offset voltage over the temp
range. It has 100 uV/V power supply rejection. Offset null is applied
across pins 1 to 5, with a pot wiper connected to the - supply. Not that
you'd want to do that as a retrofit. The voltage dividers appear to be used
to drop the common mode voltage on both sides of that shunt (33 volt bias
present above ground) to 0.3 volt range, using the 1 Meg and first 10 K
shunt R. Then a 10 K series to the amp, and there should be a feedback
resistor to set the diff gain of this thing.
Since it is bringing the 3 mV difference voltage up to some level which can
be used by the A/D within it's usable range, it is also able to amplify
whatever offset voltage your IC has, which can be up to 7 mV according to
Motorola - for a good device. That appears to completely swamp your desired
input. Was that 30 mV across the shunt R when it was drawing plate current,
or the measured values during cutoff?
Without further scrambling in the dark, I defer to AlphaPower for the final
solution! I suppose it is easier to pop in another MC1536 than to sit here
scratching ideas on email.
By the way, these days there are some real nice IC's made to sense current
and provide a ground referenced DC voltage to a microprocessor. They are
used for 'fuel gage' monitoring in laptop PCs. The Maxim MAX472 is one I
used recently, to monitor collector current in a solid state bipolar RF
amplifier at 800 MHz. I sample 8 Amps through an external shunt R on the
high side of the power supply, and it comes out with a 0 to 5 Volt DC
range, ground referenced. This works while the collector line is at 22 VDC.
Similarly, these little chips should work on a cathode resistor or the
ground return of a plate power supply. Hmmm.
John
K5PRO
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|