Sanity Check of Microwave DIY Power Supply Design

Thread Starter


Joined May 11, 2016
I am making a 42V DC Unregulated Power Supply from a discarded Microwave Oven. I will use it on a DIY CNC plasma cutter; and, I am hoping for a sanity check to see if my design is correct and welcome any helpful advice.

I have included some inquiries about the driver and stepper capabilities. These are an aside if someone "happens" to know both technologies. If I can ensure my design math is proper, I can eventially solve the driver/stepper questions. My primary concern is that I am applying what I have read and watched online - properly.

I am currently planning on having 42VDC Output and at least 7Amps. I would love to run more volts through the drivers since motors seem to have a VMAX of 59V, but the drivers are rated at 4A 42V each and I will use three of them on the machine. Would a free fix like adding a PC fan to the drivers allow me to safely increase the voltage to say 50V?

The following supply list for purchase:
  1. 1000V 50 Amp Bridge Rectifier (Cheapest one I could find - under $1)
  2. 10,000mF Electrolytic Capacitor, 63V
  3. 22K 1/4W Resistor
On Hand:
1. microwave transformer
2. wire

To determine the max voltage of the steppers I used this formula: 32* √L = VMAX
The Steppers are 3A, 3.5mH inductance, 3V
stepper VMAX = 59 Volts??

The stepper drivers are rated to 4A and 42V, so they are limited to a max voltage of 42 unless I hear differently.

The drivers are the upgraded TB6600. They have very good ratings, unlike the actual TB6600. Oddly, they do not have the TB6600 chip.They come with a TB67S109AFTG chip.

Driver: CNC machine stepper motor driver TB6600 32 segments 4A 9-42VDC | eBay

To figure amperage requirement I used the following formula:
3A each x 3 motors = 9*.68 = 6.1Amps required to run 3 motors.

To calculate the capacitor, I used this formula:
V needs to be at least 20% higher than my 42V output so I went with 42*1.4=59 for V

(80,000*6.1)/59= 8271

The power supply will need a 8271mF 59V+ electrolytic capacitor rated for at least 59V

On ebay I found 63V 10000mF caps for a couple of dollars.

Bridge Rectifier: Found a 1000V 50A for $0.83 (model: KBPC 5010)

I am using a belt drive for X and Y. Each rotation will move 3" since I am using a 15T .2" pitch.

I included pictures of a diode?? bottom pic that came off the microwave 2200v capacitor. Would this work as a bleeder instead of the resistor? If not, I found the 22K 1/4 watt was what a guy used on his power supply. However, I would like to cut a resistor out of a circuit board I have lying around. What is the parameters/formula for selecting the bleeder?

If I need any corrections, or you have some helpful advice, an explanation would be excellent so I can understand how you got there.

Caution: microwave power supplies are very dangerous. An electrical engineer professor got killed repairing one. No room for error.

Thank you for your consideration and advice.


Last edited:


Joined Apr 2, 2009
Are suggesting you wanna use the microwave Tx to make a PSU ?
If yes, it won't work.

May be you can if you can rewind it but I highly doubt it.
You are better off buying a proper toroid or a SMPS

Microwave TX is not suitable. It will heat up and loss is very high.

Moderators note: @R!f@@ , you can edit posts to add more info, I merged them now.
Last edited by a moderator:

Thread Starter


Joined May 11, 2016
Thank you for your response. You have my attention.
What do you men by peril? Safety?
When you say inefficient are you defining the term as using a large input side to extract a small amount output? Or too much work when PSU are inexpensive? or something else?

Thank you,



Joined Apr 2, 2009
If you dunno about rewinding then it is not safe to try that on a microwave TX.
Safety is the issue here.
Regarding efficiency, it will result in power loss and Tx heating up.

You are better of buying a toroid.
It will help in the long run. It is cheaper than your CNC components.
Are you willing to put those at risk ?

And there are Ready made SMPS for CNC's
The SMPS I linked can be adjusted to decrease the out put to 42VDC by turning the trim pot next to the connectors.

Hypatia's Protege

Joined Mar 1, 2015
When you say inefficient are you defining the term as using a large input side to extract a small amount output? Or too much work when PSU are inexpensive? or something else?
Please note that MOTs operate in saturation by design -- Proper re-purposing (i.e. for reasonably efficient and/or continuous operation) will require rewinding of the primary as well - although 'do-able' there are less labor intensive approaches (as @R!f@@ suggests)...

FWIW I sympathize with your desire for application of linear topology to the described purposes -- Please be advised that 'Tape-wound' (line frequency) toroidal forms are all of inexpensive, abundant and available:cool:

Very best regards

PS -- Should you insist upon use of an MOT core - please don't neglect removal of the magnetic shunts...