Bench power supply upgrade

[Sidecar Bob]

OK, that matches my calculations.

Getting the case painted is my highest priority right now because I want to let the paint cure for at least a week before assembly. In winter that means I have to do it in the garage shop and pick a day that's mild enough out that the electric heater can warm the space enough and it looks like Monday will be good. My wife is out for the day with the car so I've put the heater on in her parking space to take the chill off so I can sand the case.
And I still need to find another fuse holder, decide where to put it and drill the hole before I get the sander out.....

I might not get around to shovelling the back deck today either

 

[Sidecar Bob]

Idea: If I mount a normally closed thermal switch next to the transistor and connect it in series with the power cord it would turn the whole thing off if the transistor gets too hot.
The D1047 datasheet says it can handle 150`C so a 100`C switch should be provide a good margin of safety.

BTW: I cut a heat sink to fit the space on the back behind the transistor.
The sanding is done. I want good thermal coupling between the transistor and the back of the case so I sanded all the paint off of that part of the back (inside & out) and the paint didn't want to sand out to a feather edge so I sanded it all off.
The paint on the top & bottom had a wrinkle texture and was in good shape other than some stains and about 50 years of dirt in the wrinkles so I just sanded it smooth.

 

[Audioguru]

In my fix of the old circuit, I used two power transistors at the output to share the heat. Some power supplies and linear amplifiers use many transistors to share the heat.

The large 2SD1047 Oriental transistor can handle 100W when it has a "perfect" heatsink so that its case does not even get warm, but inside it is so hot that it is almost destroyed.
You do not want to operate a transistor at it maximum allowed heat and you do not have a perfect heatsink. The temperature might be even higher in summer.

 

[Sidecar Bob]

I think I'll add the thermal switch anyway. They aren't expensive and it won't hurt anything.

 

[bertus]

Hello,

The following pages on heatsinks may interest you:
https://sound-au.com/heatsinks.htm
https://sound-au.com/articles/diy-heatsink.htm

Bertus

 

[Sidecar Bob]

Very interesting.
I had a basic understanding of how heatsinks work from what I learned in school but they didn't get into a lot of detail about insulating materials. The silicone pad I was planning to use was salvaged from something I scrapped and saved the heatsinks from. Not having seen them before I figured it must be the modern way of doing things but according to that it has a fairly high thermal resistance (I should have expected that - we have silicone oven mitts).
So I guess I'll use kapton tape instead. I don't suppose I can just stick it to the transistor and apply compound to the side against the inside of the case can I? Or would I need to clean the adhesive off and just use the bare kapton (plus heat sink compound)?

And now I'm wondering whether it would be a good idea to add another small heatsink to the front of the transistor, something like this


And I guess with 170mm of fan and meter I'd better add a heatsink to the 7824 too.....

 

[Sidecar Bob]

Maybe your supply voltage is 12V then the series resistor has (12V - 5.1V=) 6.9V across it. The circuit needs 6mA and maybe the datasheet for the zener diode says it needs 20mA. Then the total current in the resistor is 26mA and Ohm's Law calculates the resistor to be 6.9V/26mA= 265 ohms but use 270 ohms which is a standard value. It heats with 6.9V x 26mA= 180mW so a 1/4W resistor is fine.

The time has come for me to calculate the value for the resistor. The zeners are marked "C5V1". From what I can find online I think the current the zener needs in order to regulate is called the "reverse current" or "zener current". I have just spent most of an hour looking at datasheets for zeners that have C5V1 in their names (usually preceded by 4 letters beginning with a B) and the most common values I can find are 2μA and 5 mA
Does this mean I should calculate the value based on 11mA?

 

[kellys_eye]

Heatsink - does it even require an insulating pad? Not all devices do (some with isolated bodies, some where the heatsink itself is 'live' but isolated from the rest of the equipment). Either way, if insulation is required it should ideally be a new washer ('dry' if it's a silicon pad, compound if it's a 'solid' pad).

Adding an additional heatsink is unnecessary if the equipment was designed that way anyway and the thermal path is 'secure'.

 

[bertus]

Hello,

Most times I have used silicon pads on TO-220 transistors:

Bertus

 

[Sidecar Bob]

Heatsink - does it even require an insulating pad? Not all devices do (some with isolated bodies, some where the heatsink itself is 'live' but isolated from the rest of the equipment).

The D1047's collector is connected to the tab and I want the back of the case to be part of the heatsink so it definitely needs to be isolated from ground.

Either way, if insulation is required it should ideally be a new washer ('dry' if it's a silicon pad, compound if it's a 'solid' pad).

The only washers/insulators I can find are a couple of mica ones for for TO-3 and a silicon one for a TO-220 but this is a TO-3P. I want to use kapton tape because it was recommended in the first link bertus posted above.

Adding an additional heatsink is unnecessary if the equipment was designed that way anyway and the thermal path is 'secure'.

I'm not sure what you mean by "secure".
This is seat of the pants engineering. Heathkit had the power transistor mounted on the back of the case and that's what I was planning to do with the kit's power transistor until Audioguru got me concerned about overheating it.
Fins are good so I started by adding a heatsink to the outside of the case behind the transistor. When I cut that heatsink to the longest that will fit the space available there was a piece left over that is about fit the right size for the front of the transistor so I thought it might be a good idea to add it.

Any ideas about the zener?

 

[Sidecar Bob]

I spent some time yesterday afternoon at the buffing wheel putting a mirror finish on the back of the heatsink and the parts of the case that the heatsink, transistor and thermal switch will be mounted to (I figure that will provide maximum contact between them per bertus' page about heatsinks). After that I applied masking tape where the heatsink will go, sat it in place and cut around it with a sharp knife. I peeled off the part outside where the sink will be, burnished the edges down and lightly sanded the area outside the mask to help the paint stick. Next I wiped the back down with lacquer thinner to remove any buffing compound and get it ready for paint (I wiped the top & bottom of the case with mineral spirits).

This afternoon I heated the garage shop and sprayed, black enamel on the back (as thin a coat as would cover it) and blue on the top & bottom.

BTW: I'm still waiting for someone to confirm whether 11mA sounds right for the zener resistor calculation so I can finish the changes to the schematic.

 

[Sidecar Bob]

Here is version #3 of the modified schematic (if the current I need to calculate for is only 11mA I'll have to re-test the "9V" supply and see what the voltage is at that). I think everything is correct except for the value of the zener resistor so I'm going to start soldering

 

[Sidecar Bob]

The 3300μF cap is smoothing for the rectified DC. Do I still need it with the switching PSU?

 

[Sidecar Bob]

I did some more research today and figured out that the zener's current (Izt) that I needed to calculate for. Izt is 5mA so the current through the resistor is indeed 11mA. The "9v" supply puts out 11.9V @ 11mA so the zener resistor (Rz) needs to be 11.9V/0.011A=618Ω. As it happens the 3rd of the 680Ω resistors I checked is 618Ω.

I think this is the final revision of the schematic for my power supply. I would really appreciate it if Audioguru would look it over and confirm that I didn't miss something.

 

[Audioguru]

I am back.
Measure the voltage across the 5.1V zener diode because it is probably a Chinese one, not a European one.
It might need 20mA or more for it to reach 5.1V.

 

[Sidecar Bob]

OK, I'll breadboard it and test. Would it be safer to set the current a few mA higher than it needs anyway?

Are all the rest of the changes OK?

BTW: I just remembered the last time I worked with zeners. My major lab project in college was building a 4 channel stereo amp and it needed a regulated power supply. 45 years later all I remember is that a basic transformer, bridge & capacitor wasn't good enough so I added 4 big stud mount zeners on a heatsink and they got so hot I had to add a vent on the top of the case.
I spent so much time on that and I don't remember actually using it because it isn't exactly compact and we've never lived anywhere with enough room. It's been in the garage attic since we moved here; I'll have to bring it down and look at it again one of these days. I came across the info from building it when I was digging for the manual for the Heathkit power supply too. Hmmm...

 

[Audioguru]

A 20mA zener diode barely works at 5mA but a 5mA zener works fine at 20mA.
R8 still wrongly connects to 0V instead of to the negative supply. this is from the TL081 datasheet:

 

[Sidecar Bob]

It does. The emitter of VT3 is connected to the negative supply and I drew it connected to that. (it was connected after R6 originally)

 

[Audioguru]

It does. The emitter of VT3 is connected to the negative supply and I drew it connected to that. (it was connected after R6 originally)

No. The emitter of VT3 connects to 0V, not to -5.1V.
The original European circuit got it wrong and the Chinese copied it.

 

[Audioguru]

PS2 must be 30VDC for a 24V at 2.4A output.