Hi, the 8515A seems to be a simple unit to
calibrate, as it has not many adjustments inside I can see and the
Keysight calibration page it cost only 329USD, does any one have
any idea how to calibrate that unit? I see some adjustments in the
samplers. There are some labels on some voltage points. Any ideas?
Mike
KE6OJE
Mike,
I was project manager on the development of the HP 8510 test sets
and I designed the sampler preamp board. You are right, the only
adjustments are on the sampler preamp which is housed in those
hexagonal extrusions surrounding the VTO/driver microcircuit. It's
been almost 35 years since I did this work, so I'm going on
memory. You should consult the 8510 service manual too.
The adjustments are a variable inductor which is part of the 20
MHz tuned circuit at the preamp's input and trimpot which slightly
dings the Q of this tuned circuit. The inductors are adjusted
while looking at the ratio of two channels to make the phase
tracking as close as possible. The trimpots are adjusted to make
the magnitude tracking as close as possible among all four
samplers. If I recall right, the phase and magnitude imbalance is
most obvious where the VTO harmonic number changes. I recommend
that you leave the adjustments alone unless you're sure there is a
problem with tracking between sampler assemblies. Error correction
will remove most systematic tracking errors and unless you're
making NIST type measurements, that will be good enough.
There are also adjustments for phase and magnitude balance for
each sampler but I recommend you leave those alone unless you've
replaced the sampler microcircuits.
Regarding the ripple problems you described in earlier posts, go
through and re-torque all the semi-rigid cable connections if you
haven't already done so. They tend to come loose over the years,
especially with rough handling. Also, make sure you are using a
very low loss cable between the source and the test set. The
available power from the source is marginal at 26.5 GHz and we had
to specify a short, low loss cable to connect the REAR output of
the source to the test set to have enough margin.
Regards,
Bob Dildine
