SMPS Troubleshooting steps

Guys

I'm back after some more troubleshooting of my plasma smps.
This has three outputs

1. 80V regulated by the switcher
2. 60V which has its own DC to DC converter teed of the 80V line
3. Vcc +5V which tees of the 60V line

If I switch on the unit with all three rails connected, then the unit goes into protection almost as soon as I hear the soft start relay kick in. You can see a momentary square wave on the gate of the switching transistor then nada. Measurement of of the outputs shows zero volts.

Resistance checks of the outputs of the three rails against ground using the diode check setting of my DMM show

1. 80V line =1Meg Ohm
2. 60V line =1Meg Ohm
3. 5V line =1Meg Ohm

I then wired a 250V halogen bulb in series with the 80V rail to see if I have too much of a current draw, and interestingly enough the unit has gone into a power cycle every 4 seconds or so.

After turn on, the switching transistor kicks in as expected, and I get a nice square wave on the gate of the switching transistor ahead of the fly-back transformer. The Halogen bulb lights up.

After about four seconds I hear a tingly sort of click and the square wave disappears. and the Halogen bulb goes out.

on observing the 80V line with the DMM, I see that it starts of at 81.4V and then when it gets to 81.7V the click happens and the power cycles the volts drop immediately to zero

on observing the 60 line with the DMM, I see that it sits at 60.7 V until the click happens and the power cycles and drops immediately to zero

The 5V line stays at 5V constantly

On observing the square-wave on the switching transistor it start of with a large mark space ratio, and then progressively gets smaller (higher frequency) until the click.

Am I right in the following assumptions ?

1. Output caps on the 80V line are bad as it drops off to Zero too quickly.

2. No problem with the circuit board the smps is connecting to, rather an issue with the SMPS unable to hold a load.

3. A potential Over-voltage issue with the SMPS

Thanks for any thoughts

Mike C

 

Here is a good site which may help you:

http://www.nxp.com/acrobat_download/applicationnotes/APPCHP2.pdf

The fact that the control circuit varies the frequency is a good point

Try to measure the frequency of the signal driving the control transistor at the point where the trip occurs with one bulb and two bulbs as loads on the 80V output. The frequency reduces as the voltage rises across the capacitor as expected but I believe that the bulb load on the 80V output is not enough to discharge the capacitor and keep the voltage below 80.7V. For this reason the frequency reduces to a minimum value but it wants to reduce more as the voltage to be kept below 80.7V. However, it cant reduce more, the voltage across the capacitor builds up and finally it trips. The fact the voltage reduces almost instantly is maybe due to a transistor shorting the capacitor out if the overvoltage trips. Thats what my intuition tells me now

 

Thanks mik3

I had the same feeling as you that I was not loading the 80V line enough, but remember that the bulb is in series with the PCB that the smps would normally be in contact with unless the resistance of the bulb goes so high when it turns on it is limiting the current flow.

I will try with two bulbs in later today.

on the tripping, I cant see any transistors on the output of the 80V line, though lots on the 60V and 5V lines which are teed off them.

Mike C

 

It would be helpful if you try to draw a schematic of the PCB by looking at the tracks.

 

Cheers

Here is it for the 80V and 60V lines so far, may not be 100% correct but you get the shape of it

http://farm4.static.flickr.com/3056/3043601561_cc55d1868a_o.png

Mike C

 

At last

I have confirmed that the SMPS is functioning and I actually have a fault elsewhere.

first i disconnected the halogen bulb connected in series with the 80v line and the circuit board that the smps powers.
i reconnected it directly to ground so that it acts as a dummy load.
on power on i get the same cycling behaviour, but the input to the switching transistor is a constant square wave. ie no frequency change. the 80v rail does not ramp up before it cuts out.

The SMPS has three inputs all of which are 3V, which come of the main Processor logic card in the plasma TV. All of them were at 3 V apart from one which seemed to go from 3 V to zero in sync with the the power cycling.

I disconnected this line, and hey presto the supply would not come up at all.
I then shorted this line to one of the other 3V lines and before you could say shissh kebab I have a steady and constant 80V and 60V with no more cycling. the smps just stays up

here is the big warning.

i reconnected the bulb in series with the circuit board that the smps connects to and can now see the 80v rail ramp up constantly. i powered off the plasma once it got to 90v. i assume that at some point the smps would have tripped, but was not going to find out

Thanks for all your help guys, I'm now off to find out why my smps is being pushed into overvoltage. probaby a short
on the circuit board the smps connects to.

for anyone interested this board is an x-sustain board that drives the plasma panel.


MikeC