Before we go any further, I do not know if I have the interest in another build, I just want to satisfy my curiosity so I can move on to other, more productive, pursuits.

Some time back, as a lockdown project, I decided to see if I could replicate a piece of obsolete test gear just to say I did it.
(and using a far safer approach - I don't know if the original I had was defective or if the manufacturer had taken a dangerous shortcut in the factory modifications; I got 'bit' at least once trying to get reference readings).
Unfortunately, my replica was completely inert due to a few inaccurate design assumptions (and... one error when trying to troubleshoot it), but that's neither here or there.

The main low-voltage DC power supply in the original unit was apparently a power hog, as it had a heat-sinked LM309K on the 5V bus, but the 12V supply feeding it was nothing more than a bridge rectifier and a filter cap.

Because of that regulator and for ease of assembly, I chose potted SMPS modules for the three voltages my revised design would require, but I now realize I probably went overkill.
I ended up using these sizes in my initial attempt: +24V (10W / 450mA), +12V (12W / 1A), and 5V (6W / 1.2A).

I say revised as I ABSOLUTELY did NOT want to copy the original transmitter design, because it used unsafe voltages and required several custom-wound inductors.

Instead, I'd documented a later design that required only +24V, +12V, and 5V, and only two inductors, both of which I already had at hand.
Unfortunately, it does require an ASIC that I'm not sure is available at this point (which I'd omitted from my initial attempt in error).


Hopefully that should be sufficient background up front.

Now, I have since found specifications for the power transformer used with this later design and the two windings were:

Main 12V/5V supply: 11VAC, 300mA - This winding passes through a bridge, cap, 7812, another cap, and finally a 7805. Neither regulator uses a heatsink.

Transmit B+ supply: 35VAC, 250mA - This winding passes through a bridge rectifier and cap before heading to the output stage (TIP 102 / TIP 107, both with small heatsinks) and two small voltage dividers.


Question 1: It has been TOO long since my power supply lab days - what are the specific formulae that I need to use to convert the AC values given to approximate DC values for the purpose of sizing the power supply?

Question 2: Based on the milliampere rating of the transformer windings, do I assume correctly that the SMPS modules I used were SERIOUS overkill?
I would have been just fine staying with the 7812 / 7805, and something like a LM317 for the 24V supply?

Question 3: If I was to build this, would I be better off using separate supplies for the +24 and 12V/5V or would it be acceptable to cascade all three?
(understanding the cascade approach probably requires I slightly up-size the 24V and maybe 12V to compensate)

 

Hi,

I used to work in the power supply industry and we had an old phrase for companies that wanted to buy power supplies from us or anyone else really. It was based on a lot of observations over time.

The phrase was "Fool load". Yes, that is spelled correctly.

It was a play on words, substituting the word "fool" for "full". It was meant to imply that only a fool would buy a power supply that is to be used constantly at full load. That's because it would be constantly under maximum stress.

There are of course other things to consider too though, such as surge current. Since most supplies will shut down if the load current goes too high, the power supply has to be able to handle that too. With UPS systems nominal run time also becomes very important.

So maybe you did the right thing after all

Combining power supplies in any manner requires good knowledge of the power supplies being combined and of course good knowledge of the load possibilities.

 

1. 35VAC at 300 mA with a bridge rectifier and filter will produce about 47-48VDC. I would not load it to more than the 1/2 the AC current rating, so 150mA.

2. Yes.

3. Separate windings would waste less power and hence produce less heat.

 

So if I go back to the drawing board, it seems the solution would be a pair of small potted style PC mount transformers and the tried and true TO-220 regulator supplies instead!

One 12VAC unit to feed the 12/5V supply and one 25VAC unit to feed the transmit B+ supply (it was actually 28V - I went with 24 due to better parts availability, and the 317 will do nicely for that instead)

 

Dumb question in sizing the transformer - the VA rating is just multiplying the desired voltage and current ratings, right? (I know it's the algebraic sum of the two, with VAr forming the third leg of the triangle)

 

Here's what I ended up laying out...