i like this post (click again to cancel)
i dont like this post (click again to cancel) remove favorite mark from this question (click again to restore mark)

Hello, I've been working through the probe performance tests and noticed that when I change pfgon='yyy' to pfgon='nyn' lock level drops to about 50% of the previous.

Does it indicate some problem in the PFG amplifier or elsewhere?

Thank you.

P.S. gradtype setting on our system is 'ttt'

asked Jan 15 '10 at 09:26

Evgeny%20Fadeev's gravatar image

Evgeny Fadeev

updated Jan 15 '10 at 10:31

Evgeny, Do you have three axis (x, y, z) gradients? Or z axis only? -Kirk - Kirk Marat (Jan 15 '10 at 10:15)

we have triaxial gradient probe and three channel grad. amplifier, thanks. - - Evgeny Fadeev (Jan 15 '10 at 10:35)

2 Answers:
i like this answer (click again to cancel)
i dont like this answer (click again to cancel)

All of my experience is with the single axis gradient system, but I would guess that the three axis system would be the same.

I don't think you have a problem. When the amplifier is on but not pulsed (pfgon = 'yyy', but no experiment running) there is a tiny leakage current through the gradient coil(s). This current is relatively stable, so you can easily compensate for it with the shim settings. However, if you turn one or more of the amplifier(s) off (e.g. pfgon = 'nyn') then the shims are compensating for a leakage current that isn't there, thus the drop in lock level. You can re-shim for the new pfgon setting, although I think it is probably best to simply leave the pfg setting at 'yyy' or whatever is appropriate for your probe. e.g. if you have a three axis probe then you can probably leave it at 'yyy' all of the time. Is there a reason why you wanted a setting of 'nyn'? If you have experiments that require changing the pfgon setting you could simply create shim files for each of the required settings.

Can anyone from Varian comment on this?


answered Jan 15 '10 at 11:47

Kirk%20Marat's gravatar image

Kirk Marat

Thanks, Kirk. I've now narrowed it down to 'yyy' -> 'nyy', so it's caused by the X gradient. There is no reason for me to use setting other than 'yyy', just discovered this by playing with the parameters. - Evgeny Fadeev (Jan 15 '10 at 12:08)

What I'm wondering is whether that silent current might cause distortions to the field which might be hard to compensate with shims? - Evgeny Fadeev (Jan 15 '10 at 15:45)

Agaim, my experience is with the z gradient only, not x, y, z. I think that as long as the leakage current is not huge, shimming it out will not be a problem. There are many good reasons for the engineers to make the gradients "pure z", pure x etc. We have no problem getting ls spec. with amp on. - Kirk Marat (Jan 19 '10 at 05:30)

If there is a high offset current from a Highland L700 gradient amplifier, you might try adjusting the "Zero" potentiometer within the amplifier. That will reduce or eliminate the problem at the source. - Jerry Hirschinger (Sep 28 '10 at 12:12)

i like this answer (click again to cancel)
i dont like this answer (click again to cancel)

In my experience with the Varian 500 (with xyz gradient probe and amps) I have seen the same thing. It is mainly a change in the shimming due (presumably) to the leaking current from the gradient coils when they are on (as the previous answer says). We used different shim files with different settings but be sure to check that different users don't have different "pfgon" settings or that if they do they use different shim files.

The leaked current seems pretty stable and not a problem except on the z amp where we think some instability caused shimming problems and problems with long experiments. In that case we turned that amp off and if we needed z gradients we used one of the other gradient amps (they all seem to be identical) and changed our pulse sequences accordingly.


answered Sep 16 '10 at 15:11

cdr's gravatar image


Your answer
Please start posting your answer anonymously - your answer will be saved within the current session and published after you log in or create a new account. Please try to give a good answer, for discussions, please use comments and please do remember to vote (login to vote)
toggle preview

powered by CNPROG