SMPS Voltage Confusion

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
This should be something simple although I am obviously missing something. I have two power supplies, one good, one bad. I am measuring the DC voltage on the rectified side of the bridge rectifier. I am measuring between the - and + points on the rectifier. On the good power supply, I measure 330VDC and on the bad supply, I measure 336VDC. They are similar but my questions is, I thought they should be in the area of 150-170 VDC. What am I missing.
When I measure Emitter-Collector(Q1 and Q2) on the good one, I read around 155VDC which is what I expected. On the bad one I read 5.5VDC and 320VDC (Q2 and Q1) so I suspect the problem is on the Q2 side of things.
Just curious about rectifier voltage.
 

Attachments

Jony130

Joined Feb 17, 2009
5,487
The half bridge converter will always see 300V DC at his input, no matter the input mains voltage 230V/120V. So, we have a universal circuit that can work almost everywhere.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
Well I thought I was on track, now I have to figure out why I have full voltage on Q1 section and minimal voltage on Q2 section.
I am measuring from the emitter to collector on both transistors and have 330VDC on Q1 and 0.5 VDC on Q2. I guess I need to go back and check the components in those circuits.
Any explanation on how this part of the circuit works or it's purpose? I am guessing that it does the switching to create pulses for the transformer but not sure how it works.
 
Last edited:

ian field

Joined Oct 27, 2012
6,536
Learned something new today. That is pretty slick. Why do that though?
The 2 rectifiers that end up not doing much, are the ones that end up in parallel with the capacitors. The textbook voltage doubling bridge simply replaces 2 of the rectifiers with capacitors - rectifiers left in place don't do any harm and greatly simplify voltage switching.

The reservoir capacitor(S) are a peak detector - under load; your DMM will show pretty close to the AC peak value - with no load; the reservoir gets topped up a little more by various crap on the AC waveform.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
As I am not familiar with this type of circuit, where is the best place to place my ground lead from my meter to measure voltages? Should I go from the center tap and use it as zero reference? Any suggestions?
 

dendad

Joined Feb 20, 2016
4,451
The half bridge converter will always see 300V DC at his input, no matter the input mains voltage 230V/120V. So, we have a universal circuit that can work almost everywhere.
Just as long as you remember to change the voltage selector switch!!!
I can testify to what happens when you don't and run the power supply set to 120V on 240V.
BANG!
And it is particularly embarrassing when it belongs to a costumer. Fortunately, they weren't there when I did it.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
Anybody like to comment on a reference point for measurement? There are no voltages listed on my schematic so I am using a good supply to reference from.
 

dendad

Joined Feb 20, 2016
4,451
What sort of power supply is this?
Point 1. Use an isolation transformer. If you don't have one, use 2 transformers of the same sort back to back to give yourself isolation.
Point 2. Wire a 40W or 60W "normal" lamp in series with the mains input to prevent things exploding, well, limit the likelihood anyway.

Your ref for a start will be the -ve end of the bridge rectifier, or C9. But when it is running, I'd not measure much as there will be big volts here and there.
BUT be careful and really stay away if you do not have an isolation transformer. DONT'T prod about with an oscilloscope without one. A very good idea is to leave these beasties well alone.
Still, with it well discharged and disconnected, check Q1 and Q2, D1 to D4 and measure the resistors.
Check the caps for loss of capacity too as that is the biggest killer of switch mode supplies.

Did I say be careful enough?
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
Thank you dendad for the advice. I hear you loud and clear.
I have used bridge -ve for reference and found the Q2 side very low as if pulled to ground and Q1 high as stated in an earlier post. Thhis makes me think the Q2 side is bad. Both transistors were changed as well as the small caps. Which caps were you referring to as the killers? I replaced c25, c26 and c27 as well which are not in the diagram I provided.
This is a Samlex SEC1235 power supply.
 

ian field

Joined Oct 27, 2012
6,536
As I am not familiar with this type of circuit, where is the best place to place my ground lead from my meter to measure voltages? Should I go from the center tap and use it as zero reference? Any suggestions?
Its dangerous working on the primary side - you should power it through an isolating transformer.

All primary side voltages are referenced to the negative end of the bridge rectifier - it has half wave pulses equal to the negative peak of the AC waveform.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
I have been doing it the correct way then. I was just thrown off a little bit by the center tap. It would seem if that was the half way point, it should be ground zero but if you guys say -ve side of bridge, I am good with that. Thanks for the feedback. And yes, I am very careful when taking measurements.
 

dendad

Joined Feb 20, 2016
4,451
It is not a bad idea to replace all the caps if any have failed.
It will be worth looking at all the caps, on the secondary side too. A capacitance and an ESR check too.
The caps in a switch mode supply work pretty hard and the replacements must be low ESR types, and I'd go for the 105 degree ones, not 85 degree.
I found this older thread that may help...
https://forum.allaboutcircuits.com/threads/testing-switching-transistors.101483/
Is this one the same circuit? (From http://w4kaz.com/yaesufp1023.html )
http://w4kaz.com/images/fp1023/samlex_1223.pdf
It is a different model but the part you published looks almost the same.

Have you tested all the diodes and the transistors?
There are not that many bits in the primary and I'd have a go at measuring the resistance across every part, one at a time after making sure all is disconnected and discharged. After all, you have 4 almost identical sections to compare, being as you have 2 supplies. Measure the resistance across each component both ways and see if you can find something that is different and that could be a clue to where to look further. Don't forget to test the transformer windings.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
That first thread that you showed was one from three years ago that I started. As you can see I am just getting back to these now and although I understand them much better, I still am missing something on these particular units. It is very similar to the SEC 1223 and I have a schematic for this unit so I can just follow it along. I started to ohm everything out earlier today and then my first grandchild arrived so this took a back seat. I will let you know what I find and see if we can't figure this out.
Cheers,
B
 

ian field

Joined Oct 27, 2012
6,536
It is not a bad idea to replace all the caps if any have failed.
It will be worth looking at all the caps, on the secondary side too. A capacitance and an ESR check too.
The caps in a switch mode supply work pretty hard and the replacements must be low ESR types, and I'd go for the 105 degree ones, not 85 degree.
I found this older thread that may help...
https://forum.allaboutcircuits.com/threads/testing-switching-transistors.101483/
Is this one the same circuit? (From http://w4kaz.com/yaesufp1023.html )
http://w4kaz.com/images/fp1023/samlex_1223.pdf
It is a different model but the part you published looks almost the same.

Have you tested all the diodes and the transistors?
.
Primary side transistors usually go bang when they fail, most of the important rectifiers are on the secondary side - usually the PSU just stops when they fail - but not always!!!

Electrolytics often show signs, like any deformity of the top of the can, failing caps run hot and often visibly degrade the plastic sleeve - if you take them out of the board, they can feel unusually light.

An ESR meter is a must if you mess with SMPSUs much - I like the Peak Atlas.
 

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
I have an ESR meter and these were measuring good although I replaced them anyways. They are a fail point on this unit.
To recap, the main power fuse had blown and Q1 had literally let the smoke out and showed a visible crack. It was shorted, all three legs. Same with Q2 although no crack and no smoke. R9 and R5 were both way too high when removed from the board and C13 and its counterpart were off the chart for capacity so those parts all replaced. Dim bulb test came on okay and settled in, but no output. Measured voltages as noted above with one side getting all and other side getting pretty much no voltage at less than 1 volt.
Measuring resistance showed difference between good unit and bad but both sides measured the same which isnt what I expected to see. The biggest difference came across R9 and its counterpart at 190 K ohms where the good supply measured 118 K ohms.
I am thinking of stripping out both sides and measuring each component as I put it back on and testing both transformer windings. Any other thoughts?
 
Last edited:

Thread Starter

bwilliams60

Joined Nov 18, 2012
1,442
Well I took everything off the board in those two circuits and only found one bad resistor. I guess that is all it takes. The R10 resistor had gone open so I am going to replace it and all the diodes and caps in the area just to be on the safe side. Is there any concrete method for testing these small transformers?
 

recklessrog

Joined May 23, 2013
985
Something that I do which never seems to be mentioned.
All my meter probes have long bare tips which can easily accidentally short out to adjacent parts when working in crowded spaces. I have put two layers of heatshrink sleeving over them just leaving the tips exposed.
ALLWAYS discharge the hV capacitors before you start to work on the psu, I have a 2.5k 7watt resistor with two old meter probes that I use for this.
Don't short them with your screwdriver!
As I used to do this sort of repair in my business, I have a 5kVA isolation transformer (with the output earth disconnected because your house earth is connected to the neutral) which leaves the secondary fully isolated, a Variac variable transformer, and various arrangements of 100W bulbs.
(two in series across the supply will act as a potential divider with the centre connection giving around half supply volts and one limiting the current in case of a fault.
I know someone will shoot me down for saying this, but years of experience have proven to me to use an analogue meter. The digital meters can give strange results. with this set up, I am still here, not blown up my 'scopes, and repaired hundreds of power supplies.
 
Last edited:
Top