Various factors could contribute to that noise. Just a few: 1) be sure there are no bubbles in the tubes, 2) make similar rate of inflow and outflow for ACSF using tube of same diameter, 3) If you using peristaltic pump check if this is a source of noise.

Getting rid of noise is one of the most annoying experience in electrophysiology, so be patient and good 

We also have suction noise problems.  The problem is probably that electrical connection is made and broken and made and broken with the suction system, which causes electrical transients.  It is straightforward to diagnose by just temporarily switching off the suction and seeing if the transients go away.  There are several possible solutions I know of:

1) Use a drip chamber.  In your outflow line, insert a little bottle (e.g. 50 ml Nalgene), which your slice chamber empties into at the top, and the suction pulls into at the bottom.  It is meant to remain mostly empty, so there is no electrical continuity between the first part of the suction line and the vacuum source.  Not everyone in the lab thinks this helps.

2) Use a very small hole for the suction tube.  I use teflon tubing to actually do the suction from the slice chamber, and I mash it down to a narrow opening.  Some people draw a suction tube from pipette glass, with a very small tip.  This also makes suction more consistent, in my experience.

3) Pass the suction line through a grounded metal tube.

4) Don't renew the suction tubes.  I also find that new suction tubes in particular cause problems.  This is probably related to solution #2 above.  Over time, they fill up with gunk, which seems to eliminate the problem, so I avoid replacing them unless I really must.  

Hope this helps.

-Matthew

Another approach that we have used successfully in the past is to pass some dilute warm agar through a new suction tube, before mounting it on the rig.  The agar left behind on the tube is fairly inert, and has the effect of reducing the surface tension between the solution and the tube, to prevent the intermittent loss of electrical continuity. 

Dr. Akopian, Dr. Xu-Friedman and Dr. Lupica

I appreciate your kindness and prompt replies.

I tried Dr. Lupica's advise. It worked very well. I flushed the suction tubes with 100mL of 0.5% warm agar. After the flushing, droplets in the tubes has gone, and the spike noise almost completely disappeared!

To Dr. Akopian

I'm sorry I didn't explain it clearly enough. I don't use a peristaltic pump but a hand-made inflow system which is driven by gravity. For the outflow, I use a vacuum pump.

>Getting rid of noise is one of the most annoying experience in electrophysiology, so be patient and good

What you said is very true, I feel. Thank you for your kind words.

To Dr. Xu-Friedman

Thank you for many ideas.

> The problem is probably that electrical connection is made and broken and made and broken with the suction system, which causes electrical transients.

I agree with you. Maybe, I need to flush a large amount of ACSF to make the tubes slight dirty. Additionally, I throw brain slices into the suction line at the end of experiments. But it takes several days to reduce the noise...

>2) Use a very small hole for the suction tube.

I also use a pipette glass for the outflow. I brake the tip of a patch pipette, polish it and bend the pipette.

>3) Pass the suction line through a grounded metal tube.

I tried this in the past, but it did not help for my setup...

>4) Don't renew the suction tubes.

This is the most easy solution, I think so too. I don't renew the suction line so frequently (about every six months or a year).

But in my case, the tubes pick up other noise as I used it longer. Then I must renew them...

To Dr. Lupica

I was also aware that this is a problem of the surface tension, but such good idea never crossed my mind at all.

Thank you so much for letting me know your great tip.