Hanns g 28" Monitor Power Supply Question

[Steven_C]

I've got the older Hanns g 28 monitor with power supply Phihong psm217-h-r. My power supply quit. I've troubleshot it down to the 12V & 24V transformer failing (i.e. Phihong 26002170001). These boards and the transformer are nowhere to be found.

I've got a couple of questions for those who might know the answer. I don't have 5V or 12V at CN1. I don't have 24V at CN3. I have 5VSB (I've already swapped out the dried out capacitors). Additionally, the PSON functions properly.

(1) Does the 5V derive from either the 12V or 24V transformer? The PWA states that it supplies 7A.

(2) Since these boards are difficult to procure, is it possible to solder a couple of wires from the transformer pinouts which connect the 12V and 24V duo diodes and supply the DC voltage externally from regulated switching power supplies? I'd leave the transformer out and solder 2 wires for the 12V and 24V supplies and 1 wire for the returns respectively. This would cover 12V and 24V DC the transformer is supplying. My assumption is that the transformer output at the duo diodes is DC.

(3) There is a center wire leading from the transformer of which I am uncertain what it is to supply/what it's function is. It is connected to a diode on the PWA.

(4) If the 5V isn't derived from this transformer, where is it supplied from? Perhaps I have an additional problem on the board.

Any help would be appreciated. Thanks much.

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

The PN on the power supply is psm217-404-h-r. I overlooked that mistake in my original post.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

Pictures, pictures, pictures!!!! Not a lot, just one or three - one overall shot of the circuit boards, one top view of the power supply, one bottom view of the power supply. 2000 x 2000 pixels, please.

Virtually every large monitor I've seen has a dual power supply. The 5 volt standby supply is on at all times; the main supply is turned on only when the monitor is active. If you look at the board you should see a broad line zig-zagging across the power supply. One side is labeled Hot, the other Cold. How many transformers straddle that line?

Given the fact that PSON functions properly, it indicates a problem with the main supply.

PlainBill

[retiredcaps]

Re: Hanns g 28" Monitor Power Supply Question

I have 5VSB (I've already swapped out the dried out capacitors).

1) What brand and series did you use for replacements?

2) Did you swap out all the capacitors in the power section?

3) How did you know they were dried out? ESR meter?

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

(1) They are generics purchased at a local electronics shop. Probably not ideal replacements (the capacitance @ the rated voltage is identical to the originals) but that is not my present problem. I'll worry about getting ideal brand name caps once I get 5V, 12V and 24V at CN1 and CN3 respectively.

(2) I swapped out all of the dried out capacitors on the main power supply cca. I used an ESR meter. It ended up being about 6 small capacitors. There were none dried out on the standby cca.

(3) I've read your posts about this card on two other threads and apparently, the replaced caps are the commonly dried out caps.

(4) I've probed the duo diode outputs. These are all shorting to ground on the cca. That's how I know the transformer is bad. I've used a flyback tester on the outputs of the transformer and all combinations are bad. With the transformer removed, the ground shorts go away.

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

I forgot to mention that the input to the 12V/24V transformer tests out fine on the flyback tester. I measured ac rms voltage with a multimeter (not a scope) at the input and it is ~ 92V on each input pin.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

The typical SMPS design has one side of the transformer secondaries tied to ground. The input side of the dual diodes appear 'shorted' to ground because the secondaries have very low resistance. A ring tester isn't much good on these transformers, either. I AM curious how you measured the 'AC' voltage on the primary. If the SMPS controller was working it would have been in the order of 400 volts.

PlainBill

[momaka]

Re: Hanns g 28" Monitor Power Supply Question

Hi Steven_C, did't realize you had this thread.
Here's my replay to the post you made in the Hanns-G HG281D thread:
https://www.badcaps.net/forum/showth...477#post160477
...
From the pictures you posted here, I can definitely confirm that the 5v rail is indeed derived from the 12v rail via Q501 and Q502. If the 12v rail is not present, then you won't get a 5v rail either.
See if the 5v rail is shorted to ground. Also, you do need to have low ESR capacitors on the 5v rail because the MOSFETs (Q501 and Q502) likely operate at a high frequency.

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

That's very helpful Bill. So my problem probably isn't in the main transformer given that a ring tester isn't so useful at the output ... hmm. I took a straight multimeter ac rms voltage reading with respect to ground on each transformer input. I could put a scope on each input to see if that's useful as 92V rms is significantly lower than 400V. So it's quite possible that my problem is on the input to the transformer ... but even so Bill, I get 0V on both the 12V and 24V output of the transformer. Regardless of the ratio, I should get something higher than 0V given a 92V ac input.

Alright, I'm going to take a scope reading of the input and output of the transformer as it sounds like this would be more useful than a mm reading.

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

momaka, the transformer was the last thing I checked as I figured that they rarely fail. I'll post scope pics of the input and output of this transformer with respect to ground and see if that is helpful. Since I have the transformer out, I'll first take scope pics without it in to see what the input is like without a load.

I was going to leave out the transformer if I couldn't make the card serviceable standalone when soldering in wires for separate 12V and 24V inputs. From Bill's post, it sounds like maybe my problem lies elsewhere. I'll take measurements tonight and post pics. I'd like to get to root cause asap.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

That's very helpful Bill. So my problem probably isn't in the main transformer given that a ring tester isn't so useful at the output ... hmm. I took a straight multimeter ac rms voltage reading with respect to ground on each transformer input. I could put a scope on each input to see if that's useful as 92V rms is significantly lower than 400V. So it's quite possible that my problem is on the input to the transformer ... but even so Bill, I get 0V on both the 12V and 24V output of the transformer. Regardless of the ratio, I should get something higher than 0V given a 92V ac input.

Alright, I'm going to take a scope reading of the input and output of the transformer as it sounds like this would be more useful than a mm reading.

WARNING!!! You have a hot ground on the primary side of that circuit. If you hook up a scope to the primary of the transformer without using an isolation transformer you will do serious damage.

Again, this is a Switching Mode Power Supply. The AC input goes through the bridge rectifier, and is boosted to ~385 volts by the PFC front end. That boosted voltage goes to the primary of the transformer, and is switched by the power FET.

Right now I'm trying to remember where I saw an example of this circuit...

Got it!!! The attached schematic uses the same concept; DETAILS are different - ICs, part number, etc.

PlainBill

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

I made the measurements before that warning Bill so I'm uncertain if the measurements are valid or would shed any light on my board's problem. I'll have to get an isolation transformer as I don't have one on hand. The ground lead of my probe was tied directly to the neutral pin of the outlet for each measurement.

One of the images shows a funny waveform on input 1 (i.e. 24V) to the transformer with the transformer removed. On input 2 (i.e. 12V), the semi square waveform is constant under load or unloaded. Both transformer input 1 and 2 look the same (i.e. semi square wave 150V p_p) with the transformer on the board. The scaling on the unloaded waveform @ pin 1 is 20V/div @ 6 1/2 divisions ~ 130V peak to peak. The square waveforms are 50V/div @ 3 divisions ~ 150V peak to peak which would be more meaningful. The third photo shows All of the outputs as being less than 1 V referenced to chassis ground.

It seems to me that I should have some voltage at the transformer output.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

I don't know how I can make this much simpler. You are dealing with a hot ground. The waveforms you are seeing do not reflect the actual voltage present ACROSS the primary windings of the transformer.

Here's a suggestion. One of the simplest SMPS controllers is the NCP1203D. Here's a link to download a datasheet. Take a look at figure 1. The EMI filter are the pair of torroids and the poly caps under the heat sink, near the AC input. For simplicity, simply imagine each line from the 'Universal Input' runs straight through to the 4 diodes. Of course, one side (let's assume it's the lower side) of that universal input is connected to ground at the circuit breaker box; the other side varies from +165 to -165 volts with respect to ground (if the line voltage is 120 VAC). Now sit down and draw the waveform of the voltage at either side of the filter capacitor. NEITHER side is at 0 volts.

PlainBill

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

I think I understand Bill. That datasheet I believe helped me grasp this concept. I could be measuring what I think is a 150V sinusoid peak to peak when in fact a fluctuating reference gives me a meaningless reading. Moreover, a chassis ground reference is only useful with respect to a secondary measurement because the ground is the reference for that side whereas it is not on the primary.

So if I understand correctly, the isolation transformer kills any DC which would affect a scope reading via the common neutrals to the monitor power supply and scope. Moreover, the reading across the primary must be directly across the transformer (i.e. from figure 1 on that datasheet, it would be the dotted side of the transformer with respect to the non-dotted side on the return. Any other reading is useless.).

I'll get this test configuration and post scope pics.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

I think I understand Bill. That datasheet I believe helped me grasp this concept. I could be measuring what I think is a 150V sinusoid peak to peak when in fact a fluctuating reference gives me a meaningless reading. Moreover, a chassis ground reference is only useful with respect to a secondary measurement because the ground is the reference for that side whereas it is not on the primary.

So if I understand correctly, the isolation transformer kills any DC which would affect a scope reading via the common neutrals to the monitor power supply and scope. Moreover, the reading across the primary must be directly across the transformer (i.e. from figure 1 on that datasheet, it would be the dotted side of the transformer with respect to the non-dotted side on the return. Any other reading is useless.).

I'll get this test configuration and post scope pics.

Bingo!!!

PlainBill

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

Measurements:

(1) 12V/24V transformer input (one with respect to the other) is less than a volt with a very poor sinusoid.

(2) The rectifier (D10XB60) input is +/-160V peak to peak.

(3) The rectifier output (i.e. pin 1 referenced to pin 4) is 160V DC.

(4) D1 and Q101 have no shorts.

This rectified input leads to a yellow transformer which sits underneath the heatsink. (see 1st pic) Assuming this rectified input is the dotted input to the transformer, the non-dotted return has a diode D1 (anode side) and transistor Q101 (pin 2) electrically connected (see 2nd pic). Measuring this non-dotted side with respect to pin 4 of the rectifier is difficult because (a) it's close to zero and (b) has what looks like aperiodic switching. I can't get my scope to consistently trigger on it. I'm guessing an ic is switching Q101.

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

Measurements:

(1) 12V/24V transformer input (one with respect to the other) is less than a volt with a very poor sinusoid.

(2) The rectifier (D10XB60) input is +/-160V peak to peak.

(3) The rectifier output (i.e. pin 1 referenced to pin 4) is 160V DC.

(4) D1 and Q101 have no shorts.

This rectified input leads to a yellow transformer which sits underneath the heatsink. (see 1st pic) Assuming this rectified input is the dotted input to the transformer, the non-dotted return has a diode D1 (anode side) and transistor Q101 (pin 2) electrically connected (see 2nd pic). Measuring this non-dotted side with respect to pin 4 of the rectifier is difficult because (a) it's close to zero and (b) has what looks like aperiodic switching. I can't get my scope to consistently trigger on it. I'm guessing an ic is switching Q101.

One correction. That transformer under the heatsink is part of the PFC correction circuit. It corresponds to LP801 on the Samsung BN96-03775A schematic I attached earlier.

Two points about the PFC correction circuit. It is enabled only when the main power supply is active. When it is active, the voltage across CP812 (in the BN96-03775A schematic) is about 385 volts. When it is inactive the voltage is about 165 volts.

There are two concepts for the activation. The older controllers activate PFC correction when the PSON signal goes high. Newer controller designs activate PFC correction when the current demand reaches a certain threshold.

PlainBill

[Steven_C]

Re: Hanns g 28" Monitor Power Supply Question

CP812 in your schematic corresponds to the large cap on this board underneath the heatsink. When PSON is enabled, the voltage is in fact ~ 380V DC. When PSON is off, it drops very slowly down to ~ 165 V DC. It jumps to 380V very quickly, but takes a couple of minutes to drop down to 165V DC.

Any suggestions up to this point? It looks like the PFC correction circuit is operating properly. I'm following your schematic as my board appears to be very similar.

[momaka]

Re: Hanns g 28" Monitor Power Supply Question

Yes, seems like PFC is working correctly. Not sure if this is how long it should take for the primary cap to discharge, though. If the primary supply is either detecting a fault on secondary or just faulty on the primary, this could well be why the primary cap is discharging so slowly.

I think a good point to start would be to identify the PWM controller and see if is getting power. There are 3 ICs - U102, U201, and U202. What numbers are written on them?

[PlainBill]

Re: Hanns g 28" Monitor Power Supply Question

Good work so far, and you are making progress. The behavior of the voltage across the large cap sounds normal. The testing you did shows a large part of the circuit is working properly. Identifying the SMPS controller is the next step. U102 would appear to be the likely candidate.

PlainBill