Thursday, October 29, 2009

What have I, what have I, what have I done to deserve this?

When your electrophysiology rig has only a single BNC cable connecting the patch clamp output to the digitizer, you know things aren't good.

Especially when there's still a little 60 Hz ripple visible.

And when you still haven't solved the periodic pipette vibration that makes getting seals hard.

[Then after you write some schlock like this, you worry about what PLS thinks about it.]

In other news, I have discovered a couple things:

When there is 0.5 pA RMS noise on a well shielded headstage, cleaning where the holder attaches can do wonders. Isopropanol + canned air = 0.05 pA RMS.

When ultrasonic cleaners are used correctly, they kick total ass. If you have one, try the glass slide test. HOLYMOLYCOWTHATISCOOLANDSADLYMIGHTBETHEWEEK'SHIGHLIGHT!


Dr. Jekyll and Mrs. Hyde said...

No joke? That would never have occurred to me. So is the inside of the headstage/holder junction like the bellybutton of the rig, lint-wise?

I think that's an impressive week, which tells you how many weeks I've had like yours...Congratulations.

Dr. Jekyll and Mrs. Hyde said...

Also, now I just want to look at your rig to see the periodic vibration issue. Why is it that other people's rig issues are so much more fun than one's own?

Nat Blair said...

I remember reading about cleaningg the headstage input somewhere, maybe the 200B manual or the Axon Guide. But I always filed it away in the "Yeah, whatever, no way that does anything" file.

Nat Blair said...

Actually I did solve the electrical noise problem yesterday, so that was good.

I finally realized that the Sutter MP-285 manipulator on my current setup is different than my previous MP-285. On the older one, the grounding tabs on the 3 piezoelectric motors were connected to power ground, which turned out not to be the case on my current one.

I thought I was grounding the nonscope stuff to those tabs (i.e. table, cage), but that turned out to be incorrect. Actually grounding those to the building ground on an outlet solved the excess noise.

We'll see what the vibration issue turns out to be. I had thought that swapping out the "single metric screw stuck in an English Newport table hole" to attach the arm holding the manipulator would have helped, but alas, it did not.

Did I mention I didn't build this rig myself?

Other people's rig problems are more interesting because it's a chance to learn, in case you have your own issues later. For the person suffering from them, it's just an impediment to data collection.

Dr. Jekyll and Mrs. Hyde said...

The screw...exactly the sort of "ahah!" moment that seems like you've found it...I was losing cells to pipette drift and was happpy to find that one corner of the table wasn't floating...except that didn't solve it. Eventually redid the entire headstage supports.

Yoe said...


I love you. :)

Comrade PhysioProf said...

One of the massive benefits of being a PI is not having to dick around endlessly with douchey rig crap like this.

Comrade PhysioProf said...

A poorly appreciated source of pipette drift is tightening the pipette holder too much. It should only be tightened just enough to allow the red gasket to maintain pressure differential.

Nat Blair said...

This stuff is kinda douchey, but I consider it the necessary downpayment to the later fun of actual recording.

And luckily, due to my general anal retentiveness, things only get to this point maybe once every 3 years.

Nat Blair said...
This comment has been removed by the author.
Anonymous said...
This comment has been removed by a blog administrator.
tideliar said...

And it's moments like this that I remember why I left lab science and electrophysiology! :)

Although, if you have a PI who will leave you figure it out and learn that's cool (and yours is big enough name enough to that I hear).

Mine was an interfering douchemonkey who solved everything himself with liberal doses of relentless sarcasm and scorn. I felt like I knew less - had somehow forgotten things - by the time I left his lab.

Nat Blair said...


I ended up deleting my previous mea culpa comment, as if I were embarrassed by it. I actually meant to delete spam. What a goof!

Here's the previous comment:
Wow, I suck. Turns out one of the table legs was periodically touching its base, hence negating its ability to isolate vibrations. Once I actually looked and saw that, I had it fixed in about 15 minutes (enough time to listen to Phantom Lord, Seek & Destroy, and Metal Militia).

Note to self: check legs first.

At least things are cleaned up and the electrical noise is much better than it ever has been since I took over this rig.

Nat Blair said...

Tideliar - All of my advisors have been pretty hands off, and yet helpful when asked for help about this sorta thing.

And actually, denoising rigs is something my current boss loves- it must be the electrical engineer in him. He also taught me a different way of grounding than anyone else had, resulting in much lower RMS noise values. I'm trying to hammer out a post on it.