Variable frequency AC power supply design

Well I've found a way to eliminate the class D amp.

While looking for a better amp for a speaker I use at work I found these.

https://www.amazon.com/dp/B0CDWTL57H

Tried a pair bridged and they work great to drive the speaker.

I then thought why not use it to power the motors. The amps don't put out enough voltage to directly drive the motors, however a 120V to 60V transformer might work and if not a 120V to 40V transformer will.

I tried the toroidal transformer and it worked ok. I then tried it with 1/2 the primary connected to the outlet and it still worked ok. The heatsinks got rather warm so I might see if I have a transformer of the ratio I need and try it to be sure it's not the toroidal causing the heatsinks to heat up. If the heatsinks don't get warm then I'll be good. If they do get warm I'll need a larger heatsink for sure.
 
I suppose the toroidal transformer might be better far as that is concerned.

Now what I could try is to connect the output to 1/2 the primary (two 115V windings in series) and take the output from the whole secondary just to see if it works like that. I'll then have se sort of idea whether or not a 1:2 ratio will work.
 
Another variable frequency power supply I built a few years back using a Bogen CHS-60A amplifier did not like a toroidal transformer so I used a regular 15V to 120V several amp transformer which seemed to work fine.

Once I get the real 2050 chips installed I will see what transformers I have laying around the house and try them. I wanted to try the amps driving the load directly, however the maximum output voltage wouldn't be enough.

Using the class D amp I put an ammeter in series with the motors and they draw 120mA at 40Vrms at 30Hz which is about 9.23mA per motor given I have 13 motors. The current being drawn by the DC supply is 360mA.

Far as the power supply goes I could use a 1A DC supply if I still use the class D amp.
 
So here's the results with the TDA2050 boards.

The small heatsinks get warm to the touch but not hot enough to burn me.

The setup.


20250208_191159.jpg



Loaded with the motors.


Loaded.jpg



No load.


Unloaded.jpg



Will also try it with the toroidal transformer.
 
Found a transformer that seems to work fine.


20250208_214417.jpg



The power supply is set to its maximum of 30Vdc for the following measurement with the load of the motors.


20250208_214423.jpg



20250208_214556.jpg


So I will need to up the B+ voltage to 35-40Vdc so that I can get 40Vrms out of the transformer.

The heatsinks heat up about as much as they did with the other transformer only they no longer heat up without any load connected.

I will see if I can find a larger heatsink so that I can mount both amps to it.

For the power supply I'll use this toroidal transformer.


20250208_221616.jpg


A bridge rectifier and a 10,000uF 50V filter cap. Just checked and at the output voltage shown I'm only drawing 310mA from the DC supply. Given the voltage is higher than 24Vdc I'll need a 24V regulator for the meter LEDs and the 5 volt regulator for the function generator.

I could use this one if I keep the unloaded B+ below 40Vdc or if I use a zener in series with the input terminal.

https://www.mouser.com/ProductDetail/STMicroelectronics/L7824CV?qs=GQP/xgjKl96cWFC%2B68wxSg==

That said I do not like the function generator I used due to it having to be manually set each power up. So I will find another that can put out 30Hz and see what dc voltage it runs on before I get any regulators.
 
Last edited: