i am trying to make a power supply for electrophoresis, which requires 100V DC. The project is a funded non-profit for some schools, where there are strict requirements when it comes to the individual price.

One way to make the power supply is as posted in this instructable, where they use a boost converter from a 15v DC supply: http://www.instructables.com/id/Gel-electrophoresis-power-supply/. I have found that this in components will cost about 10$ plus the power supply.

Another way, would be to use the mains directly with a bridge rectifier trough a series resistor to a zener diode with breakdown of 100 volt. I know that it should supply between 200mA and 350mA of current when it is connected.


simulate this circuit – Schematic created using CircuitLab

The top circuit is the one i have tried and the only problem is that the resistor has to dissipate quite a lot of heat. But is has been tested to work.

The bottom circuit is one i found somewhere, which has next to no power dissipation, except for in the load.

The question stands, is there a better way to do it or is the proposed circuit safe. And will the alternative circuit work when you put 230V RMS on it? The circuit was found for lower voltages around 12V, so can i find a transistor that can handle my specs? The benefit of the zener versions is that they can be connected directly to the mains and doesn't need a separate dc power supply, which saves money.

  • \$\begingroup\$ DO NOT use the circuits shown - as others have said. The combination of current required and safety issues make them a poor choice. You can very likely get supplies from ebay/AliExpress (Chinese regardless usually) that produce the desired voltage from a low voltage input and at a cost below the parts cost of doing it properly yourself. | If you can spec the actual need better we can make better suggestions. \$\endgroup\$
    – Russell McMahon
    Aug 18, 2017 at 2:38
  • \$\begingroup\$ @RussellMcMahon , the specs are as described in the post. We need to deliver 100V at around 200-250mA. If a low voltage device is used, it should preferably be from 12V, as we have to buy 12V adapters in bulk anyways. When it comes to output ripple, we can easily tolerate 10%, if there is any. I do not know what other specs would be required, but I am grateful for any advice on what to get. \$\endgroup\$
    – Nothin
    Aug 18, 2017 at 7:29
  • \$\begingroup\$ Electrophoresis almost certainly requires a variable supply voltage as results depend on V/cm and bed size and material and ... . There are MANY boost converters on ebay but most do not supply the voltage that you want. This one is probably dearer than you wish but seems very well priced for what it CLAIMS to do. | 350 mA or even 250 mA seems higher than usually needed - BUT if you do want that spec then 100V x 350 mA = 35 watts - ... \$\endgroup\$
    – Russell McMahon
    Aug 18, 2017 at 8:32
  • \$\begingroup\$ ... if supplied from a 12V pack and boost converter that needs a I = Watts/V/efficincy = say 35/12/80% = 12V x 4A supply. A common power supply to operate several boost converters or a eg 12V car battery may help. \$\endgroup\$
    – Russell McMahon
    Aug 18, 2017 at 8:32
  • \$\begingroup\$ A POSSIBLE "easy" solution is a suitably rated 230 VAC to 110VAC transformer with a phase controlled "dimmer" followed by a rectifier and smoothing. The DC output would still be potentially lethal but less so than unisolated mains AND and supply with your spec has the ability to kill. \$\endgroup\$
    – Russell McMahon
    Aug 18, 2017 at 8:38

2 Answers 2


The circuits you propose are dangerous ones. The reason for this is that your circuits are directly mains connected and therefore live.

That means that if you (or any other person) touches part of the circuit or the liquid you use for the electrophoresis might get an electric shock.

What is needed to fix this (I hope that you agree that a mains live setup is not desirable) is mains isolation. That's why the instructable example uses a 15 V adapter. This provides the much needed mains isolation. That 15 V output voltage is safe to touch.

Of course one can still get a shock from the 100 V but then you'd need to touch at two points (with 100 V between them) (and assuming the setup is not grounded) instead of getting a shock from touching it at any point.

Also the 100 V for the electrophoresis will be low-current so you might get a shock but since the current cannot be maintained, it will not be lethal. In your circuits the current can keep flowing so they are lethal !

Since the 100 V you need for the electrophoresis needs very little current, only a relatively simple circuit is needed.

You do not need (to buy) that specific 15 V adapter, almost any mains adapter you already have for charging a phone, laptop or some other household appliance, might suffice. The circuit used to make the 100 V DC from the adapter's output voltage might need some modifications though but these will be minor.

But please forget about the circuits you proposed, they're dangerous and also very inefficient ways of getting the voltage you need. Sure, going down to 15 V and then up again to 100 V might sound silly as well but it does not consume much power and is much safer.

  • \$\begingroup\$ The setup would be enclosed, but i see your point. The problem is that we have to make 200 units and therefore need to include a dc-power supply. What if i added an isolation transformer before the bridge rectifier, would it then be safe enough? \$\endgroup\$
    – Nothin
    Aug 17, 2017 at 8:23
  • \$\begingroup\$ Yes, that is another solution. You could also take care of the voltage that way as well. Just choose a transformer with for example 240 V in and 70 V out. Then after the bridge rectifier + smoothing capacitor, you'd get 70 V AC * 1.41 = 100 V DC, exactly what you need. The 47uF caps in your proposal is overkill, I'd use only a few uF like 1 uF or 5 uF. For small currents that is more than enough. \$\endgroup\$ Aug 17, 2017 at 8:27
  • \$\begingroup\$ You're talking about 200 - 350 mA, that is a lot of power. The instructable circuit will only supply a few mA at most. Why do you think you need that much current ? If you do then power needs to be considered as well and then indeed the smoothing caps may need to be about 47 uF and not smaller as I suggested above. \$\endgroup\$ Aug 17, 2017 at 8:32
  • \$\begingroup\$ I would guess that equipment supplied to schools will need to be certified for product safety? That's an awful lot easier if it's powered from a 15V DC supply that's bought in ready-certified. \$\endgroup\$
    – Finbarr
    Aug 17, 2017 at 10:22
  • \$\begingroup\$ I have found that other equipment delivers between 100 and 250mA, so 300mA was to be sure it had enough. I may go for the 3:1 transformer and then only a capacitor for charging, as this seems safer and when an isolating transformer is used, might as well make the other half of the circuit simpler. If this doesn't go through the certification that Finbarr speaks of, I would have to revise and use the boost converter instead. \$\endgroup\$
    – Nothin
    Aug 17, 2017 at 15:30

1) You need to talk to somebody who is competent in these matters to find out what your load requirements really are. 300 mA for electrophoresis is an outrageous current level. Put it this way - at a 100 volt level this implies that the medium will be dissipating 30 watts. To which I can only say, oh really? Are you trying to cook the sample as well as analyze it? So go find a second opinion on your current level. This will greatly impact your design requirements.

2) Your tested design is, well, I don't think you've thought it through. Not only does the 600 ohm resistor dissipate "quite a lot of heat", like 30 watts or more, but it will have to do so in an enclosed volume, and I don't think you realize just how hot it will get unless you include provisions for forced air flow to cool it. This, of course, will make it difficult to make your enclosure safe for school use. You will have three safety issues, rather than two. Shock, heat AND fan blades. Oh yes, and if the load is disconnected for any reason the zener diode must dissipate the entire 30 watts, and a zener of that size (plus the heat sink required) is going to set you back a pretty penny.

3) Your proposed circuit will give better performance than your first, but at the least you need a power transistor mounted on a heat sink. A 3904 will fail almost immediately. Additionally, your 40k zener resistor will only allow about 3 mA of current to the zener (less when you consider the transistor base requirements) and this will not be enough to drive the zener properly.

4) In either case, your 47 uF capacitors are entirely inadequate for the job, unless you don't mind major voltage changes at a 100 Hz rate. Neither circuit is up to the job of providing 100 volts DC at 300 mA.

  • \$\begingroup\$ I have found from testing and looking at other equipment that I need to deliver between 100mA and 250mA, so 300mA was to be on the safe side. And yes, it is quite a bit, but that is what it takes. You can see that there is condensed vapor on the lids on the professional machinery as well. And as a note, the 40k Ohm resistor and the exact values of the capacitors were estimated, so these would have had to be calculated afterwards. But I think I have to go with an isolation transformer and then chose a 3:1, as described by Bimpelrekkie. \$\endgroup\$
    – Nothin
    Aug 17, 2017 at 15:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.