Re: Modifying Electrophoresis Power Supply for High Voltage Microamperage

SMH wrote:
Winfield wrote:
On Dec 30 2007, SMH <s...@xxxxxxxxxxxxxxxxxx> wrote:

I have an urgent problem. The extent of my knowledge of electrical
engineering is a trimester course in university physics in
electricity and magnetism; I am molecular biologist/biochemist.

We have some old electrophoresis power supplies we are using for a
method called "isoelectric focusing." These power supplies have a
load detection system in which the current must be in the MILLIampere
range (say 5-10 mA). Our electrophoresis strips (used in proteomics)
must have a current NO GREATER than 50 microamperes (50 uA). We have
been using one power supply (a BRL 4000 with 4kV dc output) in which
its load detection system was FORCIBLY overridden by trying to lock
the "DC ON" switch using a wooden stick wedged into it. The power
supply is on the fritz now; we're not sure if it was because it was
forcibly set on.

We had considered setting up some sort of dummy circuit in which we
set up a variable resistor, say from 50,000 to 800,000 ohms, to
produce a constant 5 mA current as the voltage output was varied from
250 to 4000 V.

We would than tap into that circuit somehow to get the 4000 V with a
limiting 50 uA per strip. We use as many as 10 strips, so the
ultimate microamperage limit might be 500 uA.

I need to get something up and working right away, and budget is a
factor. We know that it will cost us $8000 to buy a system that is
supposed to work with the strips (it puts out 10,000 V and is quite
programmable and detects microampere currents...it's called the
Bio-Rad Protean IEF).

What sort of schematic should I be drawing that fills the bill?
Also, if it's not too much trouble to get higher voltages, can I use
a step-up transformer? These strips are able to take on 10,000 V,
and the max d.c. input from our old power supplies could range from
as a low as 500 V to 3000 V.

What sort of box/container will I use to hold the soldered
components...after all, this is very high voltage? Also, I don't
really know how to solder them together...any guides? And what about
heat buildup: will that be a problem for a resistor on a dummy
circuit, which might be putting out heat in the tens of watts range?

I suggest you start with a general-purpose high-voltage
power supply on eBay. For example, a Spellman RHR 15
15kV supply capable of delivering up to 10mA (that's
150 watts max), without any awkward minimum-load stuff.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=120126870617
You may be able to get one of these for less, if you're
willing to wait 6 - 18 months with email search notice.

You can make a high-voltage precision current limit
using 1.2kV MOSFET transistors in series. This is a
risky thing to do at higher currents, where they make
high-power oscillators at RF frequencies (solve with
ferrite beads on the gate and source leads), but it's
not so hard at low currents like 50uA. Can you tell
us your minimum and maximum setpoint currents, and
your minimum and maximum voltage-compliance range?
(There needs to be a way to power a resistor stack).

I have attempted to explain my predicament in a highly illustrated web
page in which I put in a few hours.

http://tinyurl.com/3xvsyu

As to minimum current, I have no minimum current: the procedure is an
isoelectric focusing technique in which 1 mm thick (when rehydrated)
strips of polyacrylamide are loaded with a protein mixture and the high
voltage is used both to establish a pH gradient along the strip, and the
proteins have a net charge at a pH and move to a pH zone where that net
charge is zero (the "isoelectric point" or pI...the pI = pH at which the
net charge on the molecule is zero) and the molecules stop movement...no
current. Theoretically all proteins move to that point in the generated
pH gradient where they have zero movement and thus zero electric
current...no more charged particles move to create a current. In
practice, there is always a net slight current of about 10 microamperes
when the voltage is about 10,000 volts...so the currrent reduction follows
an asymptote...and current reduction to the asymptote can be used to
monitor completion of the procedure.

As to maximum current, each gel strip on the system can carry 50
microamperes maximum. There is an electrophoretic power supply sold by
Bio-Rad in which you enter the number of strips and it automatically
allows the current maximum of the entire system to be the number of strips
times 50 microamperes...of course, this assumes all gel strips will have
approximately equal resistance which is a generally safe assumption; of
course if one strip has little resistance and all the others have a lot,
that gel might take far more than the 50 microampere current it is rated
for and burn up, which is why they tell you not to run strips where the
sample type varies too much...for example, where you have
rehydrated/infused the gel strip with a protein solution having a high
salt content and the other strips do not have it...that is sure to create
problems since conductivity and therefore resistance varies.

The mimimum load feature is even present on these thousands-of-volts-
metering-in-the-microampere-range scientific power supplies, although they
trip in a way I am not aware..perhaps if the current goes down to below 5
uA. The idea is to sense that resistance is so high that there is no
effective focusing (electrophoresis) going on, and you are more likely to
have a very high resistance condition that would cause the strip to burn
up due to Joule heating, and you would lose your probably already focused
sample. The voltage shuts down in that event. It is less of a condition
in which you might have a fire, and more about preventing one's gel strip
from turning to a black or brown ash, which we have seen on occasion.

I see your EC3000P only goes to 3kV, is that sufficient?

I suggested a constant-current sink to keep the supply on,
as a convenience, because, while using a resistor would
work, you'd have to keep changing the value if you change
the voltage very much. If not, a resistor should be fine.
But I take it simple current monitoring is not sufficient,
because you also need a shutoff. A comparator + flipflop
on each strip could drive a 5kV relay to provide shutoff.
These items should cost less than $200, depending on how
many strips you want to run at once. Safety interlocks,
protective covers, etc., should of course be included. A
maximum-current shutoff as well, although perhaps that's
already in the EC3000P supply?

If you have a schematic for your model 4000, it shouldn't
be too hard to get it repaired.
.

Relevant Pages
Quantcast