Advice building a 90V DC power suply

Posted by: @salp

↑

I have a thread mill motor that I would like to control using an Arduino. need some advice on the power supply, I could buy a AC/DC converter, but I want to learn how the circuits work (just a hobby).

I'm having a hard time identifying the components required to build a switching power supply to output 90V DC ( 10A max), all the tutorials I have found seem to top off at 24 volts, and I have a hard time going through the electric supplier web sites to find the correct components.

Any advice or link to learning resource is appreciated.

Thanks

Sal

You are right, it is hard to find a design for your specs. Hobbyists more often are building 3.3V, 5V, even 12V switching or linear supplies. If you are internet on DIY vs purchasing a solution if it exists then your best bet is likely to be a book on switching power supply design.

Here is a book that seems authoritative https://amz.run/7BSs and here is a web site that seems thorough https://www.electronics-tutorials.ws/power/switch-mode-power-supply.html

Good luck

Hi @salp,

   I hate to be defeatist, but unless you are already proficient in building switch mode power supplies, I personally would be very reluctant to consider designing and building a 900W 90V output supply. In addition to my concerns about the voltage of 90V, which itself could cause a fatality, such a supply would use capacitors and inductors storing a considerable amount of energy, under conditions that put considerable stress on those and other components, and it is not unknown for them to explode, especially when developing a new design, firing parts of the component like shrapnel.

One notable example (from several) I remember is seeing a capacitor embedded in a high ceiling (maybe 15 feet above the bench) in a (very large) lab, following such an incident, and I think that supply would have been lower power than you are suggesting. (I was not personally involved in the development and only heard the bang from afar!)

If you are determined to proceed to develop a supply from scratch, then I would start by looking for manufacturers' application notes on power supply controller chips designed for higher voltage applications like this, and they may refer to develpment boards which can be purchased. But please take adequate safety precautions, including but not limited to, protective screens to prevent you being hit by high speed, hot, components as they are fired like bullets across the room... Unfortunately, it is very likely you will have failures, and the risk of injury is high if you are in the firing line.

-------------

An alternative approach, which again requires similar safety precautions, but may be feasible, is to find a commercial product whose output voltage and current are (modestly) in excess of the values you require, and which it is possible to modify the circuit slightly to adjust to the required values. Sorry, I cannot recommend such a product, but with the help of Google or similar, you may be lucky. Of course, this will probably mean identifying the circuit and designing the necessary modificatons, as you are unlikely to be provided such information.

--------

But please be extremely careful  ... power electronic circuits can, and sometimes are, lethal. I have tried to make some positive suggestions, but I am in no way recommending that you follow any of the suggestions. Any use of the suggestions is entirely at your own risk.

Best wishes and please stay safe, Dave

@salp @davee I second Dave's comments, this is a potentially dangerous project but I wonder if you have the facts straight. Can you tell us a bit more as to why it is 90V as that is not a common voltage. Also why is it 10A, that is a fair amount of current. The more details you can provide the better we are able to help.

@zander Thank you for the references.  I'll start reading.

@davee I appreciate the advice, I'm not rushing into anything, I'm doing the research to see if this is something I want to tackle.

I like to know how things work, but in the end if its too much I'll purchase the power supply.

Thanks

@salp Good idea, but it might be beneficial to tell us more about your project. Why 90V at 10A for instance. There are all kinds of experiences here, you never know what kind of help is available unless you ask.

@zander The thread mill motor I have is rated at 90V and the specs say it has a max draw of 7A.

All the information I have seen uses a potentiameter to control the motor, but that does not give a good control as the load varies. I would like to use an Arduino to get RPM feed back and adjust power to maintain a set speed.

You and Dave are both correct this much more power than I have worked with before, if this gets too risky I'll stop the project.  (It's a hobby, not worth getting hurt over)

Thanks

@salp Thanks for the info. So the first thing I notice is it is 7A not 10. That is a reduction of about 37%, not trivial.

You are correct, the load will affect the motor to some degree.

If you want to maintain constant RPM then yes, there are lots of designs out there that will show you how that is done. I think what you want to look for is information on ESC (Electronic Speed Controller)

Good luck.

Hi @salp,

   I notice you said:

thread mill motor I have is rated at 90V and the specs say it has a max draw of 7A.

Whilst I looked at the voltage and current, and noted I didn't know what 'thread mill' was, I first assumed it was some kind of milling machine (as in a kind of router, typically cutting solid metal).

I am now wondering if you meant 'treadmill'?

-------

That aside, the idea of using a variable resistor for something running at 90 V, 7-10A, is somewhat mind boggling in physical terms ... I know early electrical machinery (perhaps a 120 years ago?) did use such scary approaches, but it is beyond my direct experience!! It is also long before present 'Health and Safety' reguations were effectively applied in factories, etc., so that injuries and deaths were a commonplace event.

--------------

Using my imagination, perhaps a relatively simple controller, such as a thyristor using phase control, could have been intended, connected straight to a 110 Vac supply. The thyristor would only allow 'positive' half cycles through, hence representing DC, albeit pulsed DC.  The lower voltage rating of the motor would be 'offset' against the effective reduced average voltage, by the voltage being zero for 0% of the time. (I am not proposing an exact mathematics of this 'equivalence', but rather that this is an empirical solution that might have been persuaded to work.)

As someone who lives in a country with 230/240 Vac supply, I have never met such a fearsome beast, so I have absolutely no experience to know if it would work. Hence, this is NOT a suggestion of how to do it, merely a conjecture as to how such a motor might have come into existence.

Even if it such a system could be made to work, I suspect the emc (radio) emissions would preclude it from being legal to sell in most countries today. How you view such limitations is your choice, but for obvious reasons, I personally would not be willing to support it. Ironically, such a system would be easier to design and build, as there are no major energy storage components, but that does not mean it is a wise choice.

Sorry, this is not a positive contribution to your project, but perhaps it helps to understand some of the context.

Best wishes and stay safe, Dave

I would be inclined to use a linear power supply design, all cautions still apply. With FET regulation and a FET speed control. Does the motor have a “pulse” output for monitoring rpm? You could use opto isolation between the motor and Arduino to keep it safe and drive the speed control from the MC. Finally take a look at some treadmill schematics to see how it’s done for that application. Run it through chat-gpt to see what it spits out. Good luck fuse everything and be safe.

Hi @kmreed55,

re: I would be inclined to use a linear power supply design,

  From certain viewpoints, including simplicity and reduction of stored power that may cause an explosion, I tend to agree. The comments about opto-couplers and so on are also sensible.

By all means try one or more of the AI systems, but be careful to check their outputs with reality. A discussion in another thread on this forum showed a contrast for a question about wire sizing for a particular current. The same question to Google's and Microsoft's machines apparently produced utter nonsense from Google, but a half-decent answer from Microsoft.  That was only one question, so should not be considered representative, beyond saying that you must carefully check suggestions.

------------

Unfortunately, it seems probable that the dissipation could be considerably in excess of 100 W. A linear power supply reduces the voltage by using transistors as 'magic resistors' and those 'magic resistors' must dissipate the same amount of heat as 'real resistor', dropping the same voltage whilst passing the same current.

I don't have any accurate curves of voltage and current, nor do I know the incoming DC voltage, or even the motor usage, in terms of only be used for a brief burst at a time, or continuous usage, so this is a psuedo-random guess of a particular situation and setting:

e.g. Voltage drop = 40 V Current = 7 A, Dissipation = 40 (V) * 7 (A) = 280 W

Dissipating 280 W, or even 100 W, whilst ensuring the transistors etc. are within their operating range, is not trivial.

Perhaps this calculation is over pessimistic, but it illustrates the principle of one of the first checks that should be applied when more details of a proposed solution are known.

Best wishes and stay safe, Dave

@davee You are correct its a treadmill for waling exercise .

I'm in the US with 120V power, this is just a hobby of mine  to see if I can run my drill press with it and control its RPM.

@kmreed55  Yes the motor has a nice photo interrupt sensor and quadrature wheel built into it.

I'll post my results once I decide.

Thanks to all for the advice.

Hi @salp,

  I don't know how you acquired the motor, but if you removed it from a treadmill, and still have the old control board, then a couple (or so) hours with a magnifying glass, meter for checking continuity, + pencil and paper, maybe a few close up photos, noting the main details, part numbers, and so far as possible, tracing out the circuit can be very educational, not to reproduce it exactly, but as hints.

Of course, you may not have the board, or it may be difficult to decipher, but I usually reckon to learn something!

Best wishes, and good luck. Dave