No bubble?

Experimental procedures often have lots of superstitions.
Even in electrophysiological techniques there are plenty of those.
Superstitions often make one comfortable while doing stressful experiments; however, it looks silly when you see someone following a superstition that you don't believe in.

Here's an example: the bubbles in a glass capillary microelectrode.
I know some people seriously worry about those bubbles. They let the electrode sit in a tube filled with 3M KCl solution for a few minutes. It looks as if some sort of Buddhist ceremony with an insence standing up in front of an altar.

I don't believe that religion and here's why.
Here is a microelectrodes with ugly bubbles.

I let the silver wire not to penetrate through these bubbles.

And the resistance was...

The voltage drop was -38mV with no bridge when -1nA was passed. 

So the electrode resistance was 38 MΩ.

Then, I let the bubble go out. 

Yes it took me a few minutes to get rid of all tiny bubbles.

And the resistance was...

It was reduced by 1 MΩ. 
So, those ugly stupid bubbles costed me 1 MΩ!

1 MΩ, Oh well...
Tell you what, when you play with 6 electrodes simultaneously poking around neurons looking for Si2 or Si3 or whatever cells, 1 MΩ drift is nothing.  The electrode resistance will change anyway by tens of MΩ when you poke around the brain looking for cells. It is no worth spending a good few minutes just to get rid of those stupid 1 MΩ bubbles. Just go for a poke with it and replace it when clogged. Think about the efficiency of your labor. Don't worry about the bubbles. 

AxoClamp 900A problems

After using the new amplifier in several experiments, I decided not to use AxoClamp 900A any more. The reasons are as follows:

1) I could never get used to control the amplifier with mouse and keyboard. Especially when I needed to push and buzz at the same time. I often "clear" the electrode by giving a long buzz, but this cannot be done quickly in the new amplifier.

2) The bath-ground of AxoClamp 900A and that of AxoClamp 2B don't like each other, when they share the same bath ground. When 900A ground was used, then 2B shows 60Hz hum, and vice versa. This is a critical problem. I contacted the customer service and this is what they said:

I am not sure I can give you a solution to this. But perhaps it is how the two instruments are plugged into the wall socket. Try keeping a single headstage grounded, but vary the way the instruments are plugged into their power sockets on the wall or power strip.

Who could provide two different AC sources for a rig? I cannot believe they did not even consider for those who use two amplifies at the same time for one dish.

3) The software (or my PC) has some problem. Something in the USB system in my PC reset periodically. When it happens, the Ch1 tab comes to the top even when I am working with Ch2 tab. I often buzzed Ch1 electrode by accident when trying to buzz Ch2. Those tabs are small and I have to really gaze at the monitor to check which one is in front.

Well, it should work fine if you use it alone by itself and know what cell you are poking. Good enough for doing e-phys while imaging. I installed my old 2B amplifier back to my rig.

AxoClamp 900A

This week I installed a brand new Axoclamp 900A (Molecular Devices) in my rig.
The new amplifier has several neat features. One can save the setting of the amplifier as a configuration. This may be good for doing different kinds of experiments routinely in one rig.

They changed the headstages. This is really annoying.
The new headstage somewhat resembles that of Axopatch, but shorter. There is no rod and it is designed to be attached directly onto the manipulator. And, of course, it won’t fit on our manipulator. We use Siskyou’s micromanipulator. Siskyou provides a “dove-tail” headstage connector, which costs like $175 for each. This is a rip off.
So, I constructed two headstage holders by myself from scratch.


Photo: The old headstage (blue) and the new headstage (black).
The “dovetail” slide connectors were made from plastic plates cut out from the microelectrode box (Sutter). A wooden rod ($0.87) was bought at Home Depot. It was cut, carved, grooved, and epoxy-glued upon the dovetail plate. The tightness of the dove tail can be adusted by putting something under the headstage. I use a small amount of dental wax. The total cost is less than a dollar + my labor.

Another feature I don't like is that there are no knobs on the front face of the amplifier. You need to control the capacitance compensation using mouse, or type in, like 8 μF through your computer. It has to be all controlled by PC. Bridge balancing is automatic. This sounds good, but actually it is not. In a real E-phys experiment, you don't really balance the bridge that accurately, because I don't trust bridge at all for measuring input resistances. I mostly use the bridge to keep the output within the voltage range of your AD converter.

This was my first day of using it. I might change my mind after getting used to this machine. I don't like it so far, though.

Patch-clamp amplifier vs. intracellular amplifier

People often ask me what the differences are between Axopatch and Axoclamp (both are made by Axon Instruments).

The answer is simple: Axopatch is a highly-sensitive I-V converter (ammeter). Axoclamp is a highly-stable voltmeter.


Axopatch or so-called patch-clamp amplifiers are designed for recording the membrane current, but it can also measure the membrane potential under the "current clamp mode".
On the other hand, Axoclamp is an amplifier designed for recording voltage with high-resistance microelectrodes. You can also measure membrane current by using "voltage-clamp mode." One may think that the continuous single-electrode voltage clamp mode in Axoclamp sounds very similar or almost identical to the whole-cell patch clamp by Axopatch.

Having either one being able do both voltage and current recordings, why do we need both?



The headstage of Axopatch contains a highly sensitive I-V converter, so that it can measure current with better s/n ratio. It can indirectly measure the voltage by calculating from the current being passed and membrane input resistance, but very susceptible to errors due to changes in the electrode resistance. This is why the patch pipette has to be big, with a resistance as low as possible.

Ironically, the large electrode tip makes he recorded potential less accurate, because of the junctional current. Junctional current is produced by the difference between the electrode solution and the cytoplasmic fluid. This current can be minimized by adjusting the ionic content of the electrode solution and reducing the size of the electrode tip (but this will increase the electrode resistance).


The headstage of Axoclamp contains the "voltage follower" circuit. This voltage follower circuit is the "magic" that makes it possible to record voltage with a sharp mega-ohm electrode. A voltage follower produce virtually infinite input resistance, which makes the electrode resistance almost negligible. Thus, the voltage recorded by Axoclamp is more accurate and relieble than Axopatch.
The current measured under voltage clamp mode in Axoclamp is actually the current generated artificially in the feedback circuit. Such indirect measurement is bad in s/n ratio and slower than Axopatch.