[time-nuts] 10638A Degausser Update

John Ackermann N8UR jra at febo.com
Fri Nov 23 19:06:38 UTC 2012


A couple of weeks ago I posted about problems with my HP 10638A 
degausser (that's used with the 5061 Cs units to remove residual 
magnetism from the high-performance option tube).

My troubleshooting indicated that the problem was likely in the op-amp 
that is used to generate an exponentially decreasing current to drive 
the degaussing coil in the tube.  The circuit is basically a 10uF/10Mohm 
network on the input to the op-amp, with the output driving a Darlington 
pair to feed 28V through the coil.  A couple of large resistors are in 
series with the current path.

The op-amp was driving full power, full time, into the coil.  Thus, 
rather than degaussing, it "gaussed"!  It's clear that this had been a 
problem for a while, as *both* sides of the PCB under the current 
limiting resistors (which are mounted a half inch or so above the board) 
were cooked.

The op-amp is an Intech A-233, a potted device with FET input.  You 
don't find those at Digikey, unfortunately.  But a couple of the 
wonderfully helpful people on the list gave me leads.

First, CharlesS pointed me to a message board where one of the former 
Intech engineers had posted about their products.  I contacted "wsj" 
(name deleted for his privacy, since he's not on this list) and had a 
fascinating chat with him about the history of the company and its 
products.  He also kindly reviewed the 10638A schematic for me, and 
suggested that a more modern high-impedance op-amp like the OPA130 would 
likely work with a few bits added (trim circuit and supply bypassing).

Second, RichMcC pointed me to plccenter.com, a parts liquidator which 
showed a couple of A-233 units in stock.  Now, in the past I've seen 
places like this that have inventory and price lists last updated five 
years ago, so I frankly didn't expect much.  But amazingly, they 
actually had two new-in-box A-233 units available, and they are now in 
my hands (they have a $50 minimum order, so there was no point in 
getting just one).

I soldered the new part in, and the degausser is now working just like 
it should.  After a couple of degauss cycles the 5016B/004 appears to be 
within about 1e12 of GPS with the C-field set at the nominal value shown 
on the meter reading log, so all seems to be well.

Thanks to everyone who jumped into the thread, and especially to Charles 
and Rich for giving me incredibly useful leads.  And, if anyone else 
finds themselves in the same situation, I have a spare A-233 a couple of 
OPA130s!

John



More information about the time-nuts mailing list