Nothing gets past the fuse which is connected to the switch.

Which is the first time I've had an answer to the first question a service engineer (like me) should ask.

So we are proceeding in a systematic manner.

To proceed further you need to separate a few components and proceed with the check in the order I outlined.

 

The way I see, by the pictures, your power supply is of low quality. Only four transistors for 35Amps, and the bulk capacitance of the filter is very low, maybe 60000uF (I estimate). For 35A, that's very low.

Sorry to say that, but it is just my opinion. You should definitely buy another PSU, a better one.

 

The way I see, by the pictures, your power supply is of low quality. Only four transistors for 35Amps, and the bulk capacitance of the filter is very low, maybe 60000uF (I estimate). For 35A, that's very low.

Sorry to say that, but it is just my opinion. You should definitely buy another PSU, a better one.

Is there a way to bulk this one up?
each cap is 4700uf 25v and there are 6 of them.

 

Is there a way to bulk this one up?
each cap is 4700uf 25v and there are 6 of them.

Worse yet. You need a lot more caps, or bigger caps. Considering about 18V before the regulator and that you might want a line regulation of 10%, you will need at least 70000uF. And that is not a conservative figure at all. In a real situation, you might need the double of that figure. And the problem doesn't end there. You might need to replace the diode bridge if it doesn't hold the current peak demanded by the new set of caps. Just problems after problems.

Of course the caps aren't the problem (unless one is shorted). Start by measuring the resistance at the primary of the transformer. You should check the double pole switch as well, since it has a neon bulb inside that may short, but that is very unlikely, since those are protected by a series resistor.

I see a lot of people having problems with pyramid PSUs, and not just inside this forum. Your PSU is not worth of such investment. Buy a better one, and you might have a reliable PSU for some extra money for the entire life.

 

Worse yet. You need a lot more caps, or bigger caps. Considering about 18V before the regulator and that you might want a line regulation of 10%, you will need at least 70000uF. And that is not a conservative figure at all. In a real situation, you might need the double of that figure. And the problem doesn't end there. You might need to replace the diode bridge if it doesn't hold the current peak demanded by the new set of caps. Just problems after problems.

so 140000uF in a number of caps at 25v and if I use that then what would be the current peak demanded by them for the diode bridge?

Of course the caps aren't the problem (unless one is shorted). Start by measuring the resistance at the primary of the transformer. You should check the double pole switch as well, since it has a neon bulb inside that may short, but that is very unlikely, since those are protected by a series resistor.

I measured the resistance of the transformer. (not sure if I did it right though) See pic for measurements. Meter set on 200m.

I also checked the switch and found something odd. I tested the switch on and off and got the same results. Between the red and blk wire (bottom wire) = no connection when off and connection when on. Thats Ok. But there is a connection of the red and the black & white wires whether or not the switch is on or off. The blk wire (that is with the white wire) connects to the transformer as well as the red wire. Transformer BAD? or Switch BAD?

 

I found this on another site, if correct can I use different items for my PS-36KX? This info is for a PS-53KX.

The power transformer sure appears to be something that would deliver the goods. Heavy iron. I don't know if it would get to 50A continuous or not, but there's at least hope. Grade: B

The rectifiers are 35-ampere bridge rectifiers. Diodes in parallel. If there's any mismatch in the diodes, one winds up taking all the current, and it fails. That's what had happened in mine. Grade: F

The pass transistors are mounted on good heatsinks, cooled with a small fan. Collectors are bonded together with metal straps. Unclear if they can dissipate the heat under full load, but there's hope. Individual emitter connections to a common point. Grade: B

NO CURRENT-BALANCING EMITTER RESISTORS. Worst possible design decision. Grade: F-----...

Electrolytics are small, PC-board mount devices, with small-gauge wire connecting rectifiers, caps, and rest of circuit. Grade: D-

Crowbar SCR is small, 16-ampere unit. Could not possibly handle crowbar current more than once! SCR is PC-board mounted (HOW did they expect it to handle the current????). Grade: F-----...

If you own one of these, you could bring it to being a first-rate supply as follows:

a. Replace the rectifiers with stud-mount, 85-ampere units (Mouser has them). I did this; took a few minutes with a drill to enlarge the mounting holes for the original rectifiers.

b. Replace the electrolytics with large, PC-grade caps, and run heavy-gauge (10 AWG or better) from the two diodes to the caps, and from the caps to the pass transistor collectors. You need the large caps to be able to land heavy-gauge wire.

c. Install 0.1 ohm emitter resistors on each of the 8 pass transistors. I was going to use cased wirewounds that would mount against the heat sinks (again from Mouser). Install 16 AWG wire from each transistor to a common point.

d. Install the largest stud-mount SCR you can put in. I was aiming at a 135-ampere device. Bond it to the output terminals, and reference it back to the PC board.

e. Install a real meter shunt (another wirewound resistor mounted to the bottom of the case, suitably sized to drive the ammeter with an adjustment pot).

f. Complete installation of heavy-gauge wire everywhere in the DC power circuits between rectifiers, capacitor, pass transistors, SCR, and output terminals. Consider putting larger output terminals on rear of case.

 

I just got back and started checking things out again using my multimeter and noticed while just looking at the 4 pack of diodes and the numbers on the diodes are 1N5393 and the schematic says they are supposed to be 1N4002.
BTW going clockwise (starting from the upper left) in checking them I get a very low resistance on all but the last one DA(d) which show a 1 which I am assuming to be Zero or no result. In fact all of the diodes look like they are 1N5393.

DA(a) DA(b)


DA(d) DA(c)

I can build a new circuit board to house the larger caps. cumesoftware suggestion.

I am also thinking of changing out the BR for these.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MP5010W-BPMS-ND

 

Disconnect the transformer wires, marking where they were plugged in.

It looks like Hot is connected to ground on the back, don't know if that is the power switch or power input.

Measure again. Mark the wires on the PCB, desolder one end of each winding, measure again.

-- I agree with cumesoftware on the quality of the unit. Half of those modifications could be done to your supply, but in the end, it is still a 10A supply with a bigger number printed on the front of it. Depending on what your time is worth, a Good supply will last a long time, have REAL output protection, less ripple, etc. The "Protection" light looks neat, but it is simply something to look at while the fuse blows.

 

Disconnect the transformer wires, marking where they were plugged in.

It looks like Hot is connected to ground on the back, don't know if that is the power switch or power input.

Measure again. Mark the wires on the PCB, desolder one end of each winding, measure again.

OK measured again and the switch is fine but red and black make a connection on the transformer.

 

Then the transformer is shot.

That happens to be the most expensive part of the power supply, not even worth thinking about replacing.

Pull the panel meters and save them, or you could get a 10A Secondary transformer and make a 8A Supply out of this one.

Buy a new, non-Pyramid supply. "You get what you pay for" is a very true axiom. Sometimes you are paying extra for the nametag, but some nametags mean a LOT (Fluke comes to mind).

 

So we are proceeding in a systematic manner.

To proceed further you need to separate a few components and proceed with the check in the order I outlined.

It looks like the transformer is shot. You would have determined this quite a bit earlier following a systematic approach.

There's a lesson there somewhere.

 

There are two lessons here, one you mentioned, the other is: "Do not buy Pyramid power supplies"!

I like Astron's supplies - even Motorola likes them. So much, in fact, that they sell Astrons when a "12 volt" supply is needed.

(I must admit, I have used their linear-regulated line a lot, but don't think I have sold more than maybe 2 or 3 of their switching supplies in 25 years).

 

Then the transformer is shot.

That happens to be the most expensive part of the power supply, not even worth thinking about replacing.

Pull the panel meters and save them, or you could get a 10A Secondary transformer and make a 8A Supply out of this one.

Buy a new, non-Pyramid supply. "You get what you pay for" is a very true axiom. Sometimes you are paying extra for the nametag, but some nametags mean a LOT (Fluke comes to mind).

Ok just for S&G (sh**s & giggles) what are the ratings on this transformer in case a buddy of mine has one?

 

I also checked the switch and found something odd. I tested the switch on and off and got the same results. Between the red and blk wire (bottom wire) = no connection when off and connection when on. Thats Ok. But there is a connection of the red and the black & white wires whether or not the switch is on or off. The blk wire (that is with the white wire) connects to the transformer as well as the red wire. Transformer BAD? or Switch BAD?

I think your switch might be shorted between the two poles, but then again might be a short in the transformer primary. If the switch is shorted between the poles, probably the neon bulb is the cause. You should measure the switch alone. You should measure the resistance between poles, and it should be close to infinite. Also, you should check the resistance of the primary of the transformer with the same disconnected.

-- I agree with cumesoftware on the quality of the unit. Half of those modifications could be done to your supply, but in the end, it is still a 10A supply with a bigger number printed on the front of it. Depending on what your time is worth, a Good supply will last a long time, have REAL output protection, less ripple, etc. The "Protection" light looks neat, but it is simply something to look at while the fuse blows.

Just recently bought one power supply from Kiotto, a KPS3005. It is a 30V, 5A adjustable and regulated power supply, and I've payed 70Eur for it. It has true current limiting, and both voltage and current can be regulated. Besides, it has thermal protection and over-current protection both on line and load sides. And the transformer is as big as the one from the pyramid shown. But the difference is that mine delivers half the power, and thus, is more properly dimensioned.

I've tested the current limiting at the maximum of 5A for 10min, and it never complained. It heated to about 40ºC.

 

Ok just for S&G (sh**s & giggles) what are the ratings on this transformer in case a buddy of mine has one?

If you want to build a working lower power one out of this one, get a 12V 20A secondary, that will give enough to regulate to 13.8V without overheating much. Replace the output caps, and.. (erased a bunch here). Skip that. Just use the case to put in a non-adjustable 12V/15A regulator from IC's with support components.

Essentially, the case is the only good part, as long as the meters are accurate, you will also want to re-shunt the meter as mentioned in your post above.

Otherwise, replace transformer after checking power switch and cable, a Slow-Blow 15A Resettable circuit breaker inline with the 12V Output, and use a 2 Amp fuse in the main input transformer fuse, in addition to a 185° Thermal Fuse on the transformer itself. The circuit design should have everything mentioned in your "mod" post, and a bit more. Look for schematics of Decent power supplies to get an idea of what all was skimped on in yours. Example, the 15A SCR "overvoltage protection" is a bad joke, that alone prevents "better" protection from being used, so it would need to be removed.

 

Look for schematics of Decent power supplies to get an idea of what all was skimped on in yours. Example, the 15A SCR "overvoltage protection" is a bad joke, that alone prevents "better" protection from being used, so it would need to be removed.

Can you post a couple of sites where I can get the schematics of better power supplies. I might just build my own.