Lock signal/solvent saturation - a follow-up:

Dave Babcook (dbabcook@zeeman.unmc.edu)
Thu, 05 Feb 1998 19:21:50 -0600

Original post:
In the process of tracing an intermittent lock signal problem
on a Varian unity500 (turned out to be a bad BNC connection
between lock converter box J3802 to 20dB attenuator AT1005), I
found a loose connection on the +32 input for the 0-63 dB
attenuator AT1008 (this makes sense as lock signal would drop
out when incrementing lock power from 31 to 32). Before, D2O
would saturate around a lock power of 30, now it saturates at
60. Given the disconnected attenuator, we're running the same
power to the probe. Trouble is, Varian thinks a lock power of
30 was a bit high, let alone the current value of 60!
What do other unity systems run for a lock power vs. solvent
saturation?
Or is this not consistent over a product line?

The follow-up:

Thanks for your replies. It is apparent my lock power is
quite high compared to the consistant values you reported.
Almost all lock powers were in the same range for the various
solvents mentioned. This lab was using these same values you
reported until discovering, then fixing the open connection to
the 32dB relay of the 0 - 63dB attenuator, AT1008.

Your suggestions:
Change probes - all probes show this behavior.
Tune/match H2 - probe tune/match is fine.
This problem is independant of the lose BNC connector.

Might I ramble, please correct me if I'm mistaken:

Looking into this with a scope, the maximum voltage delivered
to the probe via my unity lock pre-amp and bandpass filter is
150 mV so I'm in the ballpark here (Varian says I should
expect 200 mV at lock power of 63db, VXR systems vary at this
point, they are working with 2V max into the probe). My
lineshape (acetone-d6) sample saturated at 12dB before
reconnecting AT1008, now it saturates around 45 dB. I scoped
this signal at the probe and see 19 mV (lock power = 45). A
rough estimate, based on 150mV at lock power = 63, if
acetone-d6 is saturated around lock power of 12 (as some of
you reported) around 0.4mV is getting to the probe. I scoped
this all the way to lock power of 27 (2.2mV) where my scope
could go no more. All appears fine from lock power 63 to 27.

The AT1008 is a series of attenuators (-/0 1dB, 2, 4, 8, 16,
and 32) activated via relays switched by the lock power
parameter. This attenuates the 2V signal from the lock freq.
converter box to J5206 of the lock pre-amp, then J5205, max
output of 200mV, through a bandpass filter to probe. When
lock power is set to 63dB, all AT1008 relays are open, i.e.
no attenuation. Set power to 0dB, all relays close and the
signal is cropped by 63dB. If the 32 dB relay is not
connected, the attenuator becomes a 0 - 31dB. Lock power
settings from 0 to 31 would then correspond to lock power
settings of 32 - 63dB (12dB lock power = 44dB) as the software
thinks the 32dB attenuator is switched on when this parameter
is set below 32. I'm wondering if Varian did this with all
systems using a lock pre-amp with a 2V max input and 200mV max
output.