Group Actions
Engagement
  • Author Author: jw0752
  • Date Created: Date Created
  • Views 1050 views
  • Likes 9 likes
  • Comments 12 comments
Related
Recommended

Experimenting with Auto Tap Selection for a Future Power Supply Build

jw0752

When a linear power supply is in operation the DC output voltage is regulated. The internal regulator circuit of the power supply takes its source voltage and produces a resistive voltage drop using its power control transistor so that the output voltage is at the desired level.  The difference in voltage potential between the source voltage and the output voltage is a critical quantity as it is one of the two factors in the wasted energy that must be dissipated by the power supply.

 

For example lets say that our source voltage is 30 Volts and we have our linear power supply set so that it is supplying 10 volts at 1 Amp to an external load. If we calculate the power that the load is consuming we get Watts = Amps times Volts or 1 Amp times 10 Volts = 10 Watts.  On the other hand the power supply has to waste the internal voltage drop of 20 Volts times 1 Amp or 20 Watts. This wasted energy is dissipated as heat in the heat sink and into the environment around the power supply. Suppose however, if instead of 30 volts for a source we were able to have 15 Volts. In this case we would still have the 10 Volts at 1 Amp output for 10 Watts but now the internal voltage drop is only 5 Volts so the amount of wasted heat is only 1 Amp times 5 Volts or 5 Watts of power.

 

This is the purpose behind having a way to select different transformer taps for the source voltage of a linear supply as the output voltage is turned up. This selection process keeps the source voltage closer to the output voltage and therefore the amount of heat that must be dissipated and wasted is kept to a minimal level. If you have a power supply that has relay controlled transformer tap selection you may hear the relays clicking in and out as the voltage output is increased or decreased.

 

Up until now I have always built power supplies with one source voltage and then used large heat sinks and fans in anticipation of the need to dissipate larger quantities of heat but I knew there was a better way and so tonight I experimented with building a simple circuit that would select between a 12 volt AC Tap and a 24 Volt AC tap. In doing this I would have a power supply where the DC source voltage was never more than 15 volts higher than the output voltage of the power supply.

 

Here is a schematic of the circuit that I built.

 

image

 

One of the first things that I determined was that the selection circuit must be powered with a power source of its own that is electrically isolated from the transformer that drives the Linear Power Supply Board. This can be a separate transformer or a separate winding on the main transformer. The choice of which AC transformer tap to route to the Linear Power Supply board is made based on the DC voltage level that is fed back to the LM393 comparator from the DC output of the PS board. If the DC output voltage is less than 12 Volts the tap circuit chooses the 12 volt AC tap. Once the DC voltage exceeds 12.5 Volts the 24  Volt AC tap is selected and the power supply board has a DC source Voltage closer to 31 Volts.

 

Note that the RMS AC voltage from the transformer will produce a DC voltage that is a bit less than the AC voltage times 1.4. This is why it may seem that I am talking about DC voltages that are higher than the AC voltages. Some of the source DC voltage is lost in the voltage drops of the diodes that rectify the AC voltage.

 

I will attach a video that demonstrates the operation of the circuit as I turn up and down the output voltage of the the Linear Power Supply.

 

 

I anticipated that for my purposes this circuit will have application limited to two or possibly three taps as finding salvage transformers that are designed for this application is difficult and I am not likely to spring the $ for a new transformer that is properly tapped. The number of taps can be expanded by stacking the LM393 sections and having each monitor a different voltage and drive a separate relay that is dedicated to that tap.

 

Before I close out this blog I wanted to mention that recently the vendor of the Chinese power supply board that I have used in a lot of my builds has chosen to down grade the quality of the boards and the components supplied with the boards. They are still being marketed with the name "Original Hiland 0 - 30 Volt Power Supply" but they are not. In one case even the picture of the original was left on the marketing ad but what came in the mail was not good.

 

The original board  on the left and the cheap substitute on the right  looks like this:

 

image

 

I have had nothing but good luck with the Original board as pictured on the left. The new board on the right has poor quality components. The first one I built did not work, defective TL 081 Op Amp as well as the other complaints that I have noted.

 

John

  • in reply to jw0752

    Hi John,

     

    You're right, it is a robust solution, and would not cause any shorts I think!

    It will be fun to try these ideas out sometime, unfortunately Xmas hols are already now getting busy!

  • in reply to three-phase

    Hi Donald.

     

    Sorry just seeing yours and John's message now. An adjustable transformer would be very interesting, I hadn't thought of doing it with a motor.

    Here was the idea I had a while back, it isn't tested, I was hoping to buy a couple of transformers to try it out.

    It's not very elegant, it uses TVS diodes as a crowbar to cause a fuse to blow, although maybe the fuse would blow without it too. The TVS would only work if the voltage taps were such that the TVS was definitely held off ordinarily, but was definitely on if there was a stuck relay and an adjacent relay closed. So, the voltage between taps and the TVS value needs to be selected to work together. I can't recall why I was going with this approach. Maybe it was because I have many SPST relays, or they were more compact. It's been too long to remember : (

    In summary John's solution below is cleaner, requires SPDT relays, and guarantees there cannot be a short, whereas the idea below cannot guarantee short prevention, but tries to protect things if that occurs. To reduce the burden on the relays, there was going to be a single larger relay downstream (before the bridge rectifier and caps) that would have higher switching current, and the plan was for switching that relay briefly off, before switching any of the relays below. That way, these multiple relays could be smaller and cheaper, and hopefully it would extend their life before sticking, since they would never be passing any current while switching. The SPST relays I was planning to use were tiny; 5mm or 7.5mm wide from memory. (EDIT - just checked the CAD file; the relays were 5mm wide, so extremely small!).

     

     

    image

  • in reply to Jan Cumps

    Hi Jan,

    The current sense resistor lies between the boards input ground rail and the output ground.

     

    image

     

    See R-7

     

    John

  • in reply to Jan Cumps

    in hindsight my advice for one transfo may be dodgy, because it connects two points together after the different rectifier bridges. I need to rethink this one...

  • in reply to jw0752

    is the input ground of your power supply standard, on the pcb, directly connected to the output ground, or is there a component in between that is now being shortcutted?