Lab Power Supply Design / Build

[MJL21193]

Originally Posted by AndrewT

Have you looked at dissipation requirements when high current is required from one or both supplies at low output voltage?

Hi Andrew,
Not sure how closely you looked over the circuit. Typically, there are no large dissipation issues at low voltage and high current. The supplies are limited to a maximum 2A (via R17 and R20) at any voltage.

[MJL21193]

Originally Posted by Workhorse

A Switching regulator is more likely choice for variable output demands.

I will be getting into switching supplies eventually but I'm not ready yet.
So far, all my testing up to this point indicates that this is a pretty good hobby supply. Cheap to build too.

Originally Posted by aMateRasU

sir MJL21193 im just new to pcb making can i ask what is the software you are using for making the pcb layout...

Hi,
I use Ultiboard. Most all of the footprints shown in the layout were drawn or redrawn by me.

[MJL21193]

Originally Posted by megajocke

:P Thanks!

I take it you mean the voltage between the output pin of the LM317 and the negative terminal of the bridge changes. That is inconsequential however as what you need constant really is the voltage between the LM317 output pin and the negative output terminal. You want the current sense resistor R17 to be "outside" the voltage regulating circuit.

It is clear to me now about wat you are saying. I ran the prototype last night both ways and, of course, you are correct...! Again.

My prototype is looking pretty rough. Too many parts swapped out to try various things. I just burned out the LTP when a long resistor lead broke free. I'll need to either clean it up a bit (and replace those T's) or go ahead and make a new board, closer to the final product:

LAB SUPPLY BRD.png

NEW: Changed the ground reference point for the LM317

[AndrewT]

Originally Posted by MJL21193

Not sure how closely you looked over the circuit. Typically, there are no large dissipation issues at low voltage and high current. The supplies are limited to a maximum 2A

I have not studied your circuit, if you use a 30+30Vac transformer for <=30Vdc output then the regulating BJTs can see ~45Vdc.

Two channels putting out 2A 5V will need dissipation of ~160W.
That's a loud fan and an efficient heatsink to allow ~80W through each 21193

[MJL21193]

Originally Posted by AndrewT

I have not studied your circuit, if you use a 30+30Vac transformer for <=30Vdc output then the regulating BJTs can see ~45Vdc.

Two channels putting out 2A 5V will need dissipation of ~160W.
That's a loud fan and an efficient heatsink to allow ~80W through each 21193

Ok, when was the last time you were running something that required 2A @ 5VDC? I never have and most likely never will. Designing to reasonable parameters with the end use of the supply in mind saves time and money. If I thought that I would be powering such a load for an extended period, I'd have kept the second pass transistor (mulitap transformers aren't cheap and would add MUCH complexity to the circuit).
Your worst case figuring above is high in any case, as the transformers I'll use are each 26VAC (not a single primary and CT secondary). That would be less than 60W dissipation. I believe a single MJL21194 can handle 60W with a descent heatsink and no fan required.

[AndrewT]

Originally Posted by MJL21193

the transformers I'll use are each 26VAC (not a single primary and CT secondary).

will they allow 30Vdc and 2A with both current limit circuitry and with regulator?

[MJL21193]

Originally Posted by AndrewT

will they allow 30Vdc and 2A with both current limit circuitry and with regulator?

It has thus far.
Unloaded AC out is ~28.5V and the transformers are rated at 2.8A with 15% regulation @ full rated load. If limited to 2A, that equates to about 10% regulation at full load, still enough to give me 30VDC @ 2A. Current limiter only starts cutting base drive at Q1 at around 2.1A - this still needs some fine tuning but that is the target.

[megajocke]

I'm pretty sure a 26V transformer is a little on the low side for 30V, but if the load is light compared to the current rating of the transformer it should be enough as long as the mains voltage is not low.

I'd use a 30V >=3.6A transformer for a 30V/2A output requirement. Actually, just having a center tap can be useful to reduce power dissipation though so you can use standard transformers with good results for tap-changing circuits.

I used the 36V CT 3.5A transformer from one of those old integrated record player-tape-tuner combos to make a 0-36V, 0-2A bench supply. The winding originally used to power the tape deck and tuner was perfect for powering the "floating" regulator circuit and panel meters with +12V, +5V and -5V referenced to the positive output terminal.

IIRC I made the voltage setting go all the way up to 42V but at full load it will drop out of regulation before reaching that high.

[megajocke]

I tried simulating a 26V 2.8A 15% regulation transformer connected to a bridge-rectifier feeding a 10mF capacitor and finally a DC load of 2A and I get:

transformer peak current 7.0A
transformer RMS current 3.4A
transformer rectified average current 2.0A (same as output current)
(=> crest factor 2.1 and form factor 1.7)

output peak voltage 31.4V
output valley voltage 30.4V (!)

at nominal line and neglecting transformer leakage inductance. Leakage inductance will reduce output voltage further but also decrease input RMS and peak currents. The transformer may or may not be within its current rating depending on this.

Even though the transformer may be overloaded at 3.4A it will still take a long time for it to overheat with "just" 50% extra loss in the windings. It looks like the maximum continuous output current that won't overheat the transformer is about 1.6A. At that current the voltage might be enough to let you have 30V output but my model isn't really accurate enough to say anything about that. 1A 30V shouldn't be a problem though.

[MJL21193]

Originally Posted by megajocke

I'm pretty sure a 26V transformer is a little on the low side for 30V

It is.
Some tinkering this evening (after I cleaned up the prototype) and I have nearly 4VDC drop when drawing 2A at 30V.

Waiting in the weeds is a dual secondary Plitron I bought a year and a half ago. It is 30-0, 30-0 rated for 400VA so it has the umph needed to power this supply and I don't have any other plans for it. I connected it and it has no problem delivering 2A at 30VDC through this regulator. Slight drop at the output (~.7V) but I can certainly live with that.