• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Series diode voltage controller.. maybe?

Not open for further replies.
Given diodes and mosfets of appropriate current handling, heat sinks etc., will this circuit in effect work as a staged voltage contol using the forward voltage drop of the diodes?

*Yeah, I drew that on my phone, sorry 'bout that :oops:
*The transistors should be mosfet but there was no symbol available to use :meh:
*I have no idea how many diodes need to be there yet, the pic is just for reference :)

My idea is to switch a mosfet/diode pair off to attain lower voltage and vice versa.
Without having (and not wanting) a ~10 position switch I'd need a sequential latching circuit of sorts to control the gates via two tactile momentary switchs, one to go 'up' voltage and one to go 'down'. I've looked for an appropriate circuit but currently drawing a blank..

I suspect it might, but if by some miracle my circuit doesnt get shot down in flames.. could anyone point me in the direction of a switching circuit, or maybe tell me what it'd be called so I can search it out please. My search words and phrases so far haven't yielded anything close to what I'm looking for.

Other questions if this idea is actually conceivable; if I paralleled two diodes, I'll get double the current handling with the same voltage drop right?
If all the mosfet/diode pairs were on and conducting, I'd be sharing the thermal load between them?

I've tried to convey as best as I can but it's almost like electronics engineer is another language to me haha. If anythings too ambiguous there I'll do my best to clarify it.

Thanks in advance


Well-Known Member
Most Helpful Member
A diode makes a poor voltage regulator because its voltage changes when its temperature or current changes. It is 0.5V with low current or high temperature or 1.5V with high current and low temperature.
Your idea simply shorts the battery with the diodes and the resistor.
Is the resistor just a heater?
If more than one transistor is turned on then the one nearest the battery controls the voltage and current and the other ones do nothing.

Most voltage regulators use an accurate voltage reference that is fed a constant low current, an error amplifier and a series transistor to pass the current. The amplifier compares the output voltage with the reference voltage and adjusts the transistor so the output voltage is correct by using negative feedback.
Currently the resistor is just various heater elements yes. I do have plans for a motor and high power led's too but that won't be any time soon, and likley not with this circuit.

The trouble with most voltage regulators is I can't make one small enough to to fit in the package that can handle the current I'm using. This idea is me clutching at straws if I'm honest.

I was hoping it'd be better news.. but, it could be worse.. as long as the voltage is lower than the source and doesn't fluctuate rapidly I've not lost all hope in it yet. Parallel diodes and mosfet for current handling and taking a feed after every other diode instead of after each one? I guess as the live one gets hot the next one is cold so I could end up with a situation where when switching to the next in series the voltage goes up before it goes down. Maybe I could stack the diodes so the live one warms the next one in the series, there'd be less of a step between the Vout's?


Well-Known Member
Most Helpful Member
If all you want is to control power to a heating element then the most efficient method would be pulse-width modulation (PWM).


Well-Known Member
Most Helpful Member
Why so complicated? In theory it would sort of work, but not very well. You will be still dissipating huge amounts of power on the diodes, so you might as well use a single transistor as a linear regulator.
If you want to avoid the huge power dissipation, use PWM. The 8.4V figure suggests this is battery powered, so I would avoid any linear solution.
Below 65% duty cycle my pwm test circuit doesnt work well with my all my applications. Among other issues it has an audible rattle snake sound to it and the element its self or anything its contacting isn't capible of storing heat.
If I could double the frequency, or maybe have two timers operating sequentially on a bank of mosfets it might improve it, but varying the voltage would be the ideal for me.

Yes, exactly that flatfootskier


Well-Known Member
Most Helpful Member
Among other issues it has an audible rattle snake sound to it and the element its self or anything its contacting isn't capible of storing heat.
Increase the PWM frequency. If you use a PWM frequency > 20kHz you shouldn't hear anything.
I can't think of any heating element or object which won't store heat (albeit briefly).
Put a pulsed heater element in water and you'll see what I mean Alec_t.. one of my applications isn't too dissimilar. Another application is a Niobium pyrography tip, it needs a good stable red glow to work in the niche way I desire. A soldering iron application wouldn't care about pulses because it has abundant room for lho, problem is I can't get every application I intend to function as desired without varying the voltage.

Hey Ci139. I've looked at/for smps for months now and I'm convinced I cannot fit an inductor and capacitors that can handle 50a+ into my available space. There's certainly no 100a smps circuit schematics posted anywhere on the Internet that will even remotely fit in your pocket when built, most are car battery size! There are some insanely expensive 100a+ ic smps type chips around, most have a max output of 1.2v though, and have I mentioned how expensive they are??

I don't want to abandon the project just yet, hence the clutching at straws circuits. Itll never be a money maker for sure, but I wouldn't mind the extra helping of soul food y'know?

You guys haven't totally ripped my circuit to bits so it gives me confidence my understanding of electronics is improving and hope for future inspiration :) I'd test out this circuit but I've no clue how to switch between mosfets.


Well-Known Member
Most Helpful Member


Well-Known Member
Most Helpful Member
The circuit as shown has a peak output current of approx. 75 amps. AMPS. That's a lot of power, very large diodes, large diameter wires, fat connector pins, etc. Plus heatsinks and a cooling fan. Peak power dissipation in the regulator and the load would be approx. 176 W each. The final assembly would be smaller than a soccer ball...

And, the PNP transistors are upside down in the schematic.

Hey AnalogKid. I was thinking of this type of diode as they're easy to make series http://www.mouser.co.uk/ProductDetail/Vishay-Semiconductors/VS-100BGQ030/?qs=sGAEpiMZZMtQ8nqTKtFS/PAFbn7Ht/GGBuBTLt4C%2bx8=
paired with these mosfet
I have 2.5mm pure copper plate for the conductors, narrowest point would be 10mm wide ~50mm long. Heat sink would be a 'D' shape finned aluminium 25mm at the deepest point ~100ml long and a micro fan internal. I'm aware of the destructive power of amps, it's the main reason I'm running my ideas by you fine folk :)

By and large its just know how Alec_t, there are those who do.. and then there's me. I don't know all the physical effects of changing frequency either, I'd need time to learn how then build and test. Your fairly insistent its an appropriate solution and obviously know vastly more than I do about pwm so I will endeavour to learn further, I already have some pwm thoughts I'd like to ask about.

Ci139 I take my hat off to you sir, you obviously put some thought and effort into that! Unfortunately I've absolutely no idea what I'm looking at, how to interpret it as physical dimensions or how to turn it into a real circuit.. but it sure looks cool! o_O
The 15a circuit you linked is too big, Even if I reorientated the components the caps are too large. I've been eyeballing things like these
but I'd need all my appendage loads constructed and tested before I dropped that kind of capital on a circuit!
Last edited:

Nigel Goodwin

Super Moderator
Most Helpful Member
I've looked at/for smps for months now and I'm convinced I cannot fit an inductor and capacitors that can handle 50a+ into my available space.
As that's likely to by far the smallest solution, then how do you expect to fit a linear solution in the same space? - the whole point of switch-mode is that's it's smaller, cheaper, and more efficient. A linear solution is likely to be many times larger, and produce huge quantities of waste heat (requiring huge heatsinks).


Active Member
i donno whitch you keep in mind but they both can't be converted into anything useful - the 2-nd one i just run to get the apx. values for such setup (because i donno *anything about these things - in a sense of build . . . mostly*)
-- the IRF1010N can't be used coz there's no DC plot on SOA graph the min 3x to 10x in parallel with sufficient heat shields could be considered 24/7 op. . . .
-- the synchronous buck likely will do better in a sense of waste heat -- we have a parallel thread near to subj. here
other than that -- for 100kHz flyback type the page http://www.precision-inc.com/power-switchmode-transformers-p-1-l-en.html?cPath=3 gives the last in the list 1.3 by 1.3 in
this is only the space that the "inductor" requires if anything requires cooling there goes extra space
Last edited:
I've looked at/for smps for months now and I'm convinced I cannot..
Emphasis on the 'I' there I'm afraid Nigel :sorry: .. the best I could do with my limited knowledge right now would be way too large and I doubt I'd get 50% efficiency without significant help.

It sure can be done, one of my links has one such case.. But as I said, I can't find a single suitable 50a-100a buck converter circuit diagram to copy, and as you can tell, I can't design one myself.
Last edited:
Not open for further replies.

EE World Online Articles