A Cheap, High-Compliance Constant Current Source
Introduction
~~~~~~~~~~~~
A problem that has plagued cs aficionados from the beginning is
finding a way to generate constant current through the cell.
The cell resistance is quite high at the start of the process, and
the voltage needed to set the desired current can easily exceed the
40V limitation of ic regulators.
This means the current will start lower than desired until the
conductance reaches the point where the regulator starts operating.
This lengthens the brew time and makes it difficult to calculate the
amount of silver liberated using the Faraday equation.
Here is a simple but effective way to generate a fixed current with
a compliance up to 150V.
Schematic
~~~~~~~~~
Here is the schematic (View in fixed width font)
160VDC
Vin --+-----+----------+-----------+-------------
| / | | |
| \ R1 | _|_ | C1
_|_/ / 33k | ^ D1 --- 1nF
/ ^ | / R3 / \ 1N4007 --- 1kV
/ \----+ \ 6.8k --- |
TL431 --- | / | ---
| | | +------ -
| / R2 | | _|_
| \ 10k | | ^ D2
| / |/E | / \ 1N4007
| V1 | B| | ---
+-----+--/\/---| MJE350 PNP | |
| R4 |\C | ---
/ 100 | | -
\ R5 | |
/ 2 X 47k ------------+---- Iout = 1.544mA
|
|
---
-
Circuit Description
~~~~~~~~~~~~~~~~~~~
The TL431 is a 2.5V adjustable shunt regulator with 50 ppm per
degree C temperature stability. This is excellent performance, and
barely an order of magnitude away from the best regulators (5 ppm/C)
from Analog Devices. It is made by STMicroelectronics, Onsemi,
Fairchild, National Semiconductor, etc:
http://www.onsemi.com/pub/Collateral/TL431-D.pdf
It was made to sell for about $0.17 in quantity from distributors,
but I paid a couple of bucks at a local electronics store.
The MJE350 is a 50MHz, 300V, 0.5A PNP transistor in a TO-126
(SOT-32) plastic package. It has pretty good gain for a PNP,
especially a power transistor, and is made by STMicroelectronics,
Onsemi, Fairchild, etc. I got a package of two for $1.50
http://www.onsemi.com/pub/Collateral/MJE350-D.PDF
The TO-126 plastic case is not too bad either. It doesn't need an
insulator and gives a junction to case thermal resistance of 6.25
degrees C per Watt. It will need a good heatsink and a bit of
thermal grease.
R4 is a parasitic suppresion resistor mounted right at the base of
the MJE350. I didn't notice any instability without it, but it's
always good to have in high gain, wide-bandwidth circuits,
especially when long leads are involved.
D1 and D2 are plain 1N4007, 1000V diodes to provide a bit of
protection against ESD events. C1 is a 1nF, 1kV disk ceramic to
bypass ESD to ground. Good HF bypassing and short leads are needed.
The diodes should be mounted where the wires exit the case, with the
case connected to ground and C1 connected with very short leads.
In any event, never trust circuits that handle lethal voltages and
currents. Treat them as if they were live and ready to kill you at
the first opportunity. Keep this circuit away from inquisitive kids
who like to touch everything.
I added a small NE-2 neon bulb in series with a 150k resistor to
ground to indicate when high voltage is present. Everything fit in a
junked pc power supply case with plenty of room to spare.
Setting the Operating Current
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The reference voltage V1 can be set to any voltage between 2.5V and
36V with two resistors (R1 and R2.) The equation is
Vout = ((1 + (R1 / R2)) * Vref)
where Vref = 2.495V. With the values shown above, I measure 10.928V
from the base of the MJE350 to Vin, and about 10.35V from the
emitter to Vin.
R3 is the current setting resistor. I added a switch to select up to
ten different currents. The approximate value is calculated for an
output current of 1.5mA:
R = E / I
= 10.35 / Iout
= 10.35 / 1.5e-3
= 6900
I used a 6.8k
The actual value will depend on the beta of the MJE350, which varies
from one device to another, and the operating current. So you will
have to trim if you want an exact current.
Functional Tests
~~~~~~~~~~~~~~~~
The Vbe changes with temperature, which causes a slight warmup drift
as the MJE350 heats up. Here is the measured performance into a
short circuit from a cold start:
Cold Current = 1.5400mA
After Warmup = 1.5441mA
Percent change:
1.5441 / 1.5400 = 1.0026623 = 0.26%
The output current changes slightly as the output voltage changes.
Here is the performance driving a 100k resistor:
Vin = 164.75V
Vout = 150.08V
Iout = 1.5385mA
The load resistor was shorted and the output current measured:
Iout = 1.5421mA short circuit
Percent change:
1.5385 / 1.5421 = 0.99766
1 - 0.99766 = 0.00233 = 0.23%
So, the worst case performance looks like about 0.5%. But this will
depend on your operating current, the cell resistance, and the
quality of the heat sink.
Safe Operating Area
~~~~~~~~~~~~~~~~~~~
If you decide to run at higher currents, add plenty of heat sink,
and obey the Safe Operating Area spec of the MJE350.
And don't kill yourself. Never trust this circuit. It can and will
kill you.
Best Regards,
Mike Monett
--
The silver-list is a moderated forum for discussion of colloidal silver.
Instructions for unsubscribing may be found at: http://silverlist.org
To post, address your message to: [email protected]
Silver-list archive: http://escribe.com/health/thesilverlist/index.html
List maintainer: Mike Devour <[email protected]>
