Preregulation of a linear bench PSU

[not1xor1]

Hello,

Sorry - I've been away - I didn't see great spikes at switch-off of the pre-regulator and I also didn't have issues with the transformers getting hot or humming.

Could be the transformers I used were over-kill. I used  160VA ones from RS but never draw more than 4.8A from the output of the supply (at up to 15V 3A at 30V)
https://au.rs-online.com/web/p/toroidal-transformers/6719012/?relevancy-data=636F3D3126696E3D4931384E525353746F636B4E756D626572266C753D656E266D6D3D6D61746368616C6C26706D3D5E2828282872737C5253295B205D3F293F285C647B337D5B5C2D5C735D3F5C647B332C347D5B705061415D3F29297C283235285C647B387D7C5C647B317D5C2D5C647B377D2929292426706F3D3126736E3D592673723D2673743D52535F53544F434B5F4E554D4245522677633D4E4F4E45267573743D36373139303132267374613D3637313930313226&searchHistory=%7B%22enabled%22%3Atrue%7D

I slightly modified the classic circuit and used a comparator. I added a little hysterisis and I made the circuit slow to reduce the pre-regulator voltage but quick to ramp it up. This stops the PSU getting caught out if it briefly goes into current limit.

The biggest issue I had was the bounce in the ground line when the pre-regulator MOSFET turns on. I used a differential amplifier to measure the output voltage fed back to the analog control loop though so this minimized the impact.

I don't have a current probe but if you want me to capture any voltage waveforms I'm happy to crack the unit open and poke a probe into it.

Tom

BTW I disagree with you about the need for fast recovery from current limit. A slow recovery is not a problem and might even be useful to avoid overshots.

Apart that, I tested similar circuits 5-6 years ago and am sure that unless you slow down the MOSFET (with further power dissipation) and add a capacitor (a low ESR one) before the MOSFET (and/or use a relatively large capacitor + resistor across the MOSFET) you get huge voltage spikes (and noise) from the transformer.
It is a classical step-up circuit. You suddenly switch off the current through an inductance (in this case the transformer secondary) and the voltage increases.

But what I did not measure then (and can't do now because my lab needs masonry works) is power efficiency.
I ran lots of simulations on variations of that kind of preregulator (i.e. where the capacitor charge is interrupted as soon as the voltage gets to the desired value) and all of them showed a quite poor efficiency.

I mean I suspect that even a humble center tapped transformer PSU in the worst case (lowest voltage before switching to low voltage tap and maximum load) would provide similar if not better efficiency without all the hassle.
And that may be one of the reasons why no commercial PSU preregulator works this way, but rather starts to charge the capacitor after the AC voltage peak.

So could you please provide any information about that? 
thanks

[tombi]

Hmm I can try. Again a current clamp/probe would be handy. Let me think about it a bit. Maybe a resistor in series that i could measure with my diff probe.

I wasn't aiming for efficiency - it's a linear power supply after all. All I wanted was to reduce the amount of heat-sink I would need on the pass element.

Just for my understanding - and please note I am not an EE but this is a hobby for me - if the transformer voltage spiked after the MOSFET cuts off, why does this matter? I would think resistance of the MOSFET is pretty high in the off state so this will not go anywhere.

Tom

[not1xor1]

Hmm I can try. Again a current clamp/probe would be handy. Let me think about it a bit. Maybe a resistor in series that i could measure with my diff probe.

I wasn't aiming for efficiency - it's a linear power supply after all. All I wanted was to reduce the amount of heat-sink I would need on the pass element.

Just for my understanding - and please note I am not an EE but this is a hobby for me - if the transformer voltage spiked after the MOSFET cuts off, why does this matter? I would think resistance of the MOSFET is pretty high in the off state so this will not go anywhere.

Tom

No doubt that even this kind of pre-regulator saves some energy, but IMHO that is probably much less than one might expect as the transformer and rectifier bridge have to work harder.

In any case MOSFETs are affected by breakdown voltage, and as far as I can remember from the tests I made a few years ago, if you do not take any counter-measure to dump the spiikes, they would get as high as that breakdown value. Then the MOSFET acts like a sort of power zener dumping them.
BTW I did not dare to check how long they can stand that. 

Besides that, energy cannot just disappear like magic so, given that energy of an inductance is 1/2*L*I² where I is usually tenths of Amps (on charging capacitors), those voltage spikes at the end will make something ... hotter  .

BTW to have an idea of the efficiency you would need a power analyzer or at least one of those energy meters (although most of the cheap ones aren't probably able to deal with odd current waveforms). You cannot just multiply RMS voltage * RMS current, you must integrate the instantaneous product of V * I...

[Kleinstein]

Even the cheap power meters are usually not that bad. Their limitation is more in limited BW and resolution at low currents.

For a lab supply it's usually less about the efficiency but more about keeping the heating low - it is kind of related, as energy is conserved. Still loss distributed over more parts can be less of a problem.

[not1xor1]

Even the cheap power meters are usually not that bad. Their limitation is more in limited BW and resolution at low currents.

For a lab supply it's usually less about the efficiency but more about keeping the heating low - it is kind of related, as energy is conserved. Still loss distributed over more parts can be less of a problem.

Yes, but the question is (3rd time I write that  ) : is it really any better than a center tapped PSU?

To have at least a quite rough idea one would need:
- at least one of those cheap energy meter
- a pre-regulated PSU
- a constant current load
- beeing disposed to unmount (desolder/unconnect the wires) the pre-regulator + post-regulator for the test sake 

Since an ordinary linear regulator is just a variable electronic load from the perspective of the transformer+bridge+capacitor circuit, to get a good approximation of the power wasted at various output voltages and maximum load one needs just one measure with an electronic load set to the proper value.
The power wasted at a given voltage by a linear regulator would then be just : total_measured_power - (given_output_voltage * I_load).

Once known that, one could just measure the power wasted by the pre-regulated PSU at various output voltages and maximum load current.

Of course for power measurement I mean measurement taken at the 230VAC supply.

[tombi]

I actually did both tap switching (parallel vs series) and a pre-regulator.

I have a cheap lab PSU that I bought from a local electronics shop (Jaycar) and if you set it to say 25V then when it goes into current limit (and the output voltage drops) it will switch transformer taps.

This is fine unless the load is transient in which case it makes a horrible tap switching racket as it flicks between the taps.

I didn't want that. So if you set the target voltage on my homebrew PSU to 25V it keeps the transformer tap configuration the same if it goes into current limit. Instead the pre-regulator reduces the capacitor voltage to prevent the pass element going nuclear.

Apart from this the pre-regulator does help a bit but not a lot. You need a few volts of headroom above the output depending on the size of your capacitors/current so it never goes that low even if you set the output to say 1.8V.

Not sure if this helps but I also reduce the overhead if the output current is low. That is the pre-regulator voltage is a function of both the output voltage AND current. If the current increase then the pre-regulator increase the capacitor voltage.

Tom

[Atom]

Hi seems there is an intresting discussion here  , by the way i manged to find some power mos 80V 75A 15mohm rdson to220 package.

the circuit from blackdog the tap switcher works flawlessly and i would reccomend it to anyone that needs a similar solution, the to220 package didn't even get hot, it was extremly cool to the touch.

Thanks to every one that helped!!

do you have any suggestions when it comes to books? on my to buy list there are:

The art of electronics 3rd ed
jim williams books
Power Supply Cookbook by Marty Brown (i've read online that it's good and i have to get into switchmode as fast as possible (5th year thesis).

In the past i've tryied to get into smps but was limited by the lack of various other things, now i feel ready in the end the basics are easy to understand, the calcs could be more challenging but i have professors that could help me.

 

[not1xor1]

I actually did both tap switching (parallel vs series) and a pre-regulator.

Then it is not bad.

Actually I've been really dumb. 
Obviously, in any case Blackdog's style pre-regulator would be more efficient than a center tapped one from about half max Vout up to 90% where the efficiency should be about the same.

It is when the output voltage is below half of maximum that the pre-regulator offers little if any advantage (vs. center tapped one) according to my simulations.

Anyway, since you used it both with the top and the mid tap of the transformer then that is not your case.

So in case somebody needs to use a single winding transformer, e.g. a salvaged one, a Blackdog's style pre-regulator might still be useful although much less efficient than a buck pre-regulator.

I think anyway there is still much room for improvement, especially regarding efficiency at low output voltages.