10. Power Supply Repairs, modifications and Teardowns
INSTRUMENT REPAIRS & MODIFICATIONS (IN ALPHANUMERIC ORDER)
# Action Status Problem Manufacturer Model Type Outputs
1. Improve Ongoing Noisy fan Bio-Rad 200/2.0 Electrophoresis 200W 200V 2A
2. Repair TBD Blew fuse Bio-Rad 1000/500 Electrophoresis 250W 1kV 500mA
3. Mod Ongoing 115V/Disable Life Technologies 4001 Electrophoresis 200W 4kV 300mA
4. Mod Ongoing VMains/Disable EC Apparatus EC-650 Electrophoresis 200W 6kV 100mA/2kV 350mA
5. Repair Ongoing Ch2 stuck in CC American Reliance PPS-1202 Programmable 240W 18V 4A x2
ELECTROPHORESIS POWER SUPPLIES
I have several electrophoresis power supplies as they were quite cheap.
Unfortunately they often have µC controlled membrane keys to set voltage, current and power that is forgotten as soon as the supply is turned off.
Another annoying safety is a cut-off mechanism that disables the output unless a load is present.
This is perfectly understandable for the intended use working with wet DNA gels that could easily electrocute you, but very inconvenient when used as a general purpose bench supply,
1. BIO-RAD 200/0.2
200W / 200V / 2000mA
text
2. BIO-RAD 1000/500
250W / 1000V / 500mA
text
3. Life Technologies Gibco BRL 4001
200W / 4000V / 300mA
I have a BRL 4001 that I bought on eBay USA and looks very similar to the EC600-90.
USA/UK MAINS SELECTION - MOV? present but no spark gapC
currently it only works off 120Vac but on the pcb top left (see first pcb photo below), there is a DIL connector that appears to be wired for 120Vac or 240Vac, accompanied by '240Vac <---> 120Vac' printed on the silkscreen running along its length.
Up until now I have left it alone as there may be more to it than simply rotating the DIL 180°.
Google AI 2026:
The explicit text '240Vac <---> 120Vac' confirms your hypothesis this connector is a hardwired selector designed to alternate between series (240V) and parallel (120V) configurations for the primary windings of the main power transformer. However, before you physically flip the connector and plug it into UK mains, you must complete three critical validation steps to prevent blowing up the input filtering stages or losing your overcurrent protection.
1. Match the Connector Alignment
Look very closely at the plastic DIL connector and the PCB markings before rotating it:
There is usually an indicator (like an arrow, a notch, or a dot) on the jumper block that aligns with a corresponding marker on the board. Ensure that when you rotate it 180°, the orientation marker shifts clearly toward the '240Vac' side of the text label.
2. Replace the Main Fuses (Mandatory)
Operating at double the voltage (240V vs 120V) means the device will draw exactly half the current to achieve the same wattage. You must downsize the fuses to maintain safety:
a) Check the current fuse installed in the rear fuse holder (it is typically a 7A or 8A slow-blow fuse for the 120V model).
b) Replace it with a 3.15A or 4A Slow-Blow (Time-Delay) fuse of the exact same physical dimensions (typically 5x20mm or 1/4" x 1-1/4").
c) Warning: Leaving the original high-amperage fuse inside means the circuit breaker or transformer could experience a catastrophic meltdown before the fuse ever senses a fault at 240V.
3. Check for 120V Metal Oxide Varistors (MOVs)
This is the hidden trap in legacy equipment. Look at the components right next to the AC power inlet and that DIL socket:
a) Look for a disc-shaped component (often blue, red, or yellow, resembling a thick ceramic capacitor) that bridges the Live and Neutral lines.
b) Read the text printed on it. If it says something like 131K, 151K, or explicitly 130V / 150V, it is a surge protector designed for US mains.
c) If you apply 240V to a 130V MOV, it will interpret the UK voltage as a massive power surge and instantly short-circuit, violently exploding the moment you flick the power switch. If you see one, it must be desoldered and replaced with a 250V or 275V rated MOV (e.g., a 431K or 471K part number).
4. To be 100% sure without risking the hardware, run this quick resistance test before plugging it in:
a) Leave the unit completely unplugged.
b) Flip the physical power switch of the unit to ON.
c) Use a digital multimeter to measure the resistance (Ohms) across the pins of the unit's main ac power cord plug while it is in the 120V mode. (Let's assume it reads around 10Ω.
d) Rotate the DIL connector to the 240V mode.
e) Measure the resistance across the plug pins again.
Because the transformer windings are now in series rather than parallel, the total resistance of the primary coils must increase by roughly a factor of four (e.g., if it was 10Ω, it should now read around 40Ω. If the resistance jumps up correctly, the primary side is perfectly configured for UK mains.
Before you switch it on, let me know:
What amperage value is printed on the fuse currently inside the device?
Are there any disc-shaped surge components (MOVs) near the ac inlet, and what numbers are printed on them?
NO LOAD FAIL SAFE FEATURE
This website describes overriding the LM3900 comparator circuit that controls this feature on the
EC Apparatus EC600-90, a 4000V/300mA/300W:
http://www.kerrywong.com/2019/08/17/teardown-of-an-electrophoresis-power-supply/
Below are Kerry Wong's photos of the EC600-90 mainboard, HV and his No Load detection mod.
(MAINS IS SET TO 240VAC) MOV? + spark gap present
Again, the mains swap DIL connector on his pcb appears to be in the 120Vac position but unlike my pcb, spark gap SG1 is present and a fat component looking remarkably like a MOV is the position marked for C52, and there is a hint of a blue wire soldered into one of the holes occupied by what looks like a thin blue MOV on my pcb in the position identified as D4:
Below, my photos of the BRL 4001 mainboard, HV and No Load circuitry section.
It looks like I can copy his mod onto the BRL 4001 to disable the No Load detection feature.
However more investigation is required before I dare rotate the main 2120V/240V DIL connector.
4. EC Apparatus EC-650
200W / 6000V / 100mA or 2000V / 350mA
I bought my EC Apparatus EC-650 before discovering it had the no load safety feature.
Unfortunately the EC-650 does not share the same electronics as the EC-600 and I have not yet been able to identify the section of circuitry that does, let alone obtain a schematic.
OTHER ELECTROPHORESIS PSUs WITHOUT SAFETY CUTOUTS
This link starts with two 500V / 400mA / 200W electrophoresis supplies without this feature:
Pharmacia LKB EPS 500/400; Hoefer PS500XT, and further thread discussions reveal more models.
PROGRAMMABLE PSUs
Text
5. Amrel PPS-1202 dual channel 0-18V 4A Programmable PSU, 3.5 digits, resolution 0.05% 5mV / 0.15% 1.5mA (Amrel is aka American Reliance, bought by Motech)
Problem: Ch2 stuck in CC mode / Noisy fan
This is a nice dual channel 0-18V 4A programmable PSU that I bought without realising one channel was faulty. The seller gave me a 75% refund and allowed me to keep it.
Overview of series here:
a) Ch2 Stuck in CC mode
When the PSU boots it first identifies itself on the LCD and momentarily displays firmware build V8.02A.
Channel 1 works fine but according to the LCD (the little dot to the right of 'CC2') channel 2 is stuck in Constant Current (CC) mode and its output is 29 Volts, way beyond its supposed maximum of 18V:
My first thought was a fault with Ch2's output driver. I looked on the web for a service manual and schematics, and found a post on EEVblog about the 1-ch 18V 4A PPS-1002 that led into a discussion comparing it with a 2-ch 0-30V 3A PPS-1204, which is very similar to my 2-ch 0-18V 4A PPS-1202:
https://www.eevblog.com/forum/testgear/motech-amrel-pps-1002-diagramschematic/
Links are provided for manuals, added below, but full schematics appear unavailable; there is a block diagram in the Service manual, but little else. However the discussion describes an amazing reverse engineering of the address decode logic held in a GAL16V8D PAL on the 1002 which is probably the same in my 1202, and there is a technical description of the 1204 circuitry with a partial schematic.
Popping the lid off my 1202 unsurprisingly reveals a similar design to the 1204: a large mains toroid is mounted behind the front panel and two driver pcbs sit on heatsinks at the rear. The IEC mains inlet feeds into a pcb with rectifiers and large smoothing electrolytics. In the middle two analogue pcbs plug into a small digital motherboard mounted on the base containing a Temic TSC80C31-16CA µC, a TI TMS27C512-10 Eprom, an Atmel 93C46 EEP and a Lattice GAL16V8D-25LP PAL. Chip manf dates are circa 1999.
As with the author, I too removed my Eprom and EEP and saved the contents. Wix won't upload BIN or HEX files, so I uploaded both BIN files below as .TXT files. If you D/L them you need to rename them .BIN.
The analogue boards are identical, and each has an identical connector that leads to one of the output driver blocks. The analogue pcb nearest to the mains toroid at the front is Ch1.
Swapping their positions in the mobo revealed the fault appears to lie with analogue board 2, as the LCD CC indicator swapped from CC2 to CC1, and I was able to correctly program Ch2 and get a valid voltage out but no longer on Ch1 which is now stuck at 21V. This suggests the output driver stages are ok.
However there appear to be a couple more issues as Ch2 LCD voltage is incorrectly shown as fixed at 13.868V with over-voltage protection set, and Ch1 is shown as 'OFF' despite the presence of 21V.
The next move is to try to debug analogue pcb #2. I'll first see if it's possible to build an extender to raise analogue pcb #1 above pcb #2 to facilitate easy comparison in-situ.
b) Noisy fan
The fan is very loud and is on all the time regardless of load. Maybe fit a quieter pwm fan? Noctua?
It's 80mm square x 23mm thick and has only 2 leads. The fan label says: AVE AV825H12B ZP 12V 180mA.
I assume the H means high speed/power as on the net I found a 120mA M model and 90mA L model.
I assume B means it has better Ball bearings than the more common S model which I'm guessing means it has cheaper/noisier/shorter life Sleeve bearings. I can find nothing more about it or the manufacturer. Noctua explains fan bearing differences here: https://noctua.at/en/sso-bearing
This video shows a simple mod to an Amrel PPS-2322 that only turns the fan on when a heatsink is hot:
https://www.youtube.com/watch?v=XUs7ZQLJke0
I'd rather have one that runs it slowly instead of flat out.
The same guy above also has a teardown of an Amrel PPS-2322:
http://www.kerrywong.com/2015/05/06/amrel-pps-2322-programmable-power-supply-teardown/
text
text
µ Ω ± ° ⌠ ⌡ ∫ │ ─ √ φ θ Θ ∂ δ ζ ξ ς λ ψ ω τ µ Ω ∆ Δ ∑ ∏ π Ξ ○ ≠ ³ ² ±