Hi all,
Thanks very much to Bob, Kevin, Darcy, Jerry, Mat, Ryan, and Hsin for many
useful comments.
Some more background to the problem: the amplifier in question is a vacuum
tube high-voltage amplifier. Furthermore, the current displayed on the
high-voltage power supply is only 0.009 A, not 0.032 A. According to my
notes, it was at 0.032 A in 2016 and at 0.009 A in 2018, at which point is
was still working. The voltage is and has always been 2.25 kV.
The solution was to clean all visible connectors (cable to probe, cable to
RF front end, cable to amplifier) with compressed air and isopropanol. With
this done, it has stayed on for experiments for the past four days.
However, the current is still at 0.009 A and so I suspect that we are still
heading for trouble in the future. Next steps will be to look for arcing
at the RF front end and inside the tube amplifier. One thing I learned is
that the vacuum tube itself is likely interchangeable at different frequencies,
so the tube (but not the tuning elements!) in our spare 300 MHz high-voltage
amp could be helpful.
More complete comments below.
Robin
****
Some high power amps I've worked with have had safety switches that engage
to protect other downstream components. These safety switches are usually
triggered by high T.
Maybe during one particularly harsh experiment this safety switch flipped,
but has not properly flipped back all the way? So now it doesn't take much
to cause the circuit to open again?
Tune takes hardly any power at all.... So the amp shouldn't be forced to power
down when you tune.
Do you have access to schematics for this amp? Look for safety switches. I seem to recall once having to use some jumpers to force a reset.
****
Your issue maybe a bad internal gating circuitry, a bad low voltage or high
voltage power supply or a bad RF Tube and associated circuitry.
****
what does the Glassman power supply display? Should be something like 35 mA
at 2.25 kV.
Are you sure that the input AMT amplifier is OK (fans running, no overtemp,
front LEDs not blinking)?
If you open the tube amplifier (top lid, switched off obviously), are there
maybe any arcing traces or cracks around the large brown resistors or the
small knob capacitor near the tube or the tube itself?
****
1. Check the power supply for the amp - this could be a source of amp voltage issues
2. We recently had an amplifer controller card give out on us (Herley amp); this
caused duty cycle error messages at the VJ spectrometer software side even though
the duty cycle was well within the realm of safe
3. Finally, amplifier overheating can be caused by impedance matching issues when
dealing with solids / solutions amps.
****
The first thing to check for an intermittent problem is the interconnecting
cables. With the equipment running, twist, bend and push/pull every applicable
cable at the connector to see if the problem becomes apparent. Wiggle them
and watch what happens.
HV circuitry gathers microdust by static attraction, and after years of accumulation
it can lead to leakage currents, distorted coronal effects, and just generally
push operations to the edges of their design tolerances. HV cables and connectors
are especially vulnerable because they become exposed to a less controlled environment.
Since your equipment is operating well in the majority, I would just give it a good cleaning.
Start with outside the cases - disconnect each cable connector in time, and
flush/brush/flush all with isopropanol and air dry. Use a horsehair brush, because
it doesn't accumulate static charges. I'd start at the probe and work back toward
the HV power supply, cleaning every connector. Pay attention to remove metal dust
from the white insulator. If that doesn't clear it up, then I'd go inside the
cases of the HV supply and amp.
Inside, avoid touching anything with your skin, which will leave oil behind. Fashion
a 1M-ohm resistor with a wire and alligator clip, in order to discharge all capacitors
before proceeding further. Using a vacuum cleaner with crevice tool, and a horsehair
brush, disturb and remove the static microdust accumulation on all surfaces. Do NOT
disconnect any connectors on the inside. Avoid using detergent or solvent which can
wick dust into crevices or leave conductive films. Just dry brush it all with the
vacuum tip nearby, including the lid. I'd do both the HV power supply and the amplifier.
****
I can't speak for the amp on your VNMRS, but for our Infinity Plus, this happens
whenever the amp gets too dusty. Pretty much the exact problem you describe, and
it happens regardless of if the fans are working or the intake filter is clean.
Roughly every 6 months, I have to remove the amp from the console rack, open it
up and give it a thorough cleaning. Then it is good to go.
****
Sounds like you're right on the verge of something tripping out. Can you contact
the OEM of the amp? They might have things to trouble shoot/test.
I'm betting a component has just gotten old?
****
I imagine you might have a vacuum tube high band amplifier that needs tuned
periodically. Or perhaps 400 is low enough to be a solid-state unit which
should be more stable. I am not competent enough to comment on these.
Last summer, we accidentally arched one of our probes. It was found later
(with another probe, I think, the logbook is in the labs, I don't have the
details) that CP sometimes (or more often than not) fail. This was traced
to the power supplies in the Front-End (next to the preamps, etc). These
power supplies are small circuit boards, no more than 2"x3", and two of them
were charred by our probe arching. That solved our problem. It is probably
worth looking into. BTW, our system is 600-DD2. The Front-End may be similar to yours.
> From: Robin Stein
> Sent: January 5, 2021 11:42 AM
> To: ammrl_at_ammrl.org
> Subject: Chemagnetics/Varian high-power amplifier (SSNMR) turning itself off randomly
Hi all,
The high-voltage unit on the 1H amp of our Varian VNMRS 400 MHz SSNMR system turns
itself off randomly, with the RF overload and high voltage delay LED lights on
the RF front panel coming on for a few seconds each time. The probe and amp are
tuned well and powers are normal. Once we saw the lights come and high-voltage
turn off simply on starting trtune, but more usually it turns off after between
2 and 20 hours of use in CP experiments (3 ms CP, 20 ms acquisition time). Fans
seem fine and filter has been cleaned. Can you think of any reason why this
might be happening? Thanks!
Robin
Robin Stein
MC2 NMR/EPR Manager
McGill University
801 Sherbrooke St. W
Montreal, QC H3A 0B8
514 398-6219
https://mcgill.ca/mc2
robin.stein_at_mcgill.ca<mailto:robin.stein_at_mcgill.ca>
Received on Tue Jan 12 2021 - 02:11:45 MST