Manson hcs 3202

[semerchet]

i Guys,

I hope that someone can help me with the following problem:

I've bought a Manson HCS 3202 broken power supply; no output voltage.
The primary part of the supply is good, all the power rails are good.
On the primary side of Tr2 I have the correct pulses, checked U3, all fine.
Secondary on TR2 I have nothing.
Sec on TR3 I replaced D13 and the output volatge was there.
However after about 20 min -with no load- the power was again 0 volt.
D13, and Q9 and Q10 are good, and so my question is where do I go from here as I've checked capacitors, diode's etc. I think I'm overlooking something...,

Any help would be highly appreciated

[Longbow]

Re: Manson hcs 3202

Not sure what you mean by "nothing" on the secondary of TR2?? No voltages? No signals? Remember that the secondary of TR2 and primary of TR3 have a different "ground". Was D13 A/B actually faulty? The supply might be going into shutdown with no load. I would go back to TR2. Correct drive waveforms at primary side? Please add Sheet 1, the input circuitry.

[semerchet]

Re: Manson hcs 3202

HI longbow, Thanks for the answer; on the drain of Q10 I've 380 V, the source of Q10 start with 190 en then slowly starts dropping.
Tr 2 sec gives about 1,2 V on both sides (D12 and D 11)

Ans I know about the different grounds., but thanks anyway. The same happend when D13 was defective, only now I replaced D13 and the error is back again with a good D13. This Smps doesn't need a load to operate.

Again thanks

[Longbow]

Re: Manson hcs 3202

Thanks for page 1. This looks like a nice supply.

on the drain of Q10 I've 380 V, the source of Q10 start with 190 en then slowly starts dropping.

OK the Q10 Drain is supposed to be connected to the full HDV (high d.c. voltage). But, taking d.c. measurements inside the circuit probably isn't productive. Do you have the correct d.c. at HALF? That would be 380/2 = about 190v.d.c.

Tr 2 sec gives about 1,2 V on both sides (D12 and D 11)

Again, d.c. measurements inside the circuit won't tell you much and might be dangerous to you. 1,2 volts is just telling you that there is no oscillatioin going on. Have you done simple checking of the components between T2 and T3? I'm going to call that circuit a drive circuit - someone correct me if I'm wrong. Simple in-circuit testing for short/open. Does C44 check OK? Nothing will happen without power to that end of the transformer.

I don't understand why you are having problems with D13 if there is no drive to TR3??? What is the resistance accross C55? (Power off, cap discharged). I suppose a direct short at C55 might prevent the drive circuit from operating. But of course you have already checked the supply outputs for shorts?

[zx8401]

Re: Manson hcs 3202

There is a shutdown input into D17 on the secondary diagram, is it possible that the micro controlling the psu has lost controll and takes shutdown high?

[Longbow]

Re: Manson hcs 3202

Sure. It would take about 5 seconds to check this. The OP says that there is normal drive at TR2 primary, so there is no shutdown condition. My question is...will a problem in the output (D13, etc) keep the driver circuit from working? Seems like the output would have to be shorted in that case. Why would it work for 20 minutes, then fail?

[semerchet]

Re: Manson hcs 3202

Hi folks, the input on D17 is low, so that's fine. I also checked U3 fine.
When I bought it, it didn't work checked everything and found that D13 was s/c. Replaced it and it worked for about 20 min. Now D13 is fine, but still no output. Sec on TR2 I have a sawtooth of about 5V (if a pict would help, let me know)
and freq about 60 Khz, so it's oscillating
Btw thanks for your help!

[Longbow]

Re: Manson hcs 3202

When I bought it, it didn't work checked everything and found that D13 was s/c. Replaced it and it worked for about 20 min. Now D13 is fine, but still no output.

This is not making any sense, but for now I'm going to ignore the shorted D13.

Sec on TR2 I have a sawtooth of about 5V (if a pict would help, let me know)
and freq about 60 Khz, so it's oscillating

Does not seem that 5 volts p-p is enough, but regardless, there must be enough voltage at the GATES of Q9 and Q10 to turn them on. Of course, you have already checked the parts between T2 and T3? Transistors, diodes, FET's?

Looking at D13 ANODE would be a safe place to look at your main switching waveform. Is it there, or nothing at all? Some parts that seem important to me would be:
C44
R57, R58
R51, R55
R52, R56
Q7, Q8

I think that your results show that the driver stage is NOT working. The pulses you see are produced by U3, but the driver transistors are not switching. Just a guess, but if the driver stage produces an incorrect waveform, the result might cause diodes like D13 to remain in Linear mode too long and overheat. Just a guess. Right now, find the problem between T2 and T3. Voltages, scope waveforms, please.

I see that U5 is sensing current through S1, the power supply ground leg. What is the meaning of U5 output "CM"?
Since S1 is a current "shunt", check to see if it has become open. Also check continuity through the chokes, L6, L7 and continuity from TR3 center tap to the power supply output connector.

BTW, you did not say anything about your standby supply. Is this section working OK?

[semerchet]

Re: Manson hcs 3202

To start with the last question, standby is working ok.
The "shunt" S! is a wire approx 1mm 0 checked it ok.
L6 L7 both have continuity.
Understanding CM on U5 I've included the schematic from the cpu-unit
To me I also think the problem is in the driver section, although I've checked everything
but are unable to find anything wrong.
C44 is ok
R 51 through R58 also fine and Q7 Q8

Voltage ZD5 0,5 V
Zd 5 180 V and falling slowly.
Pic one showing Tr 3 /D1
Pic two T3 3 /D2

And two pic of S1 // L7

[Longbow]

Re: Manson hcs 3202

Thanks for the information. I see that "CM" probably means Current Monitor and allows the micro to shut down the supply if necessary.

Your driver switching transistors Q9 and Q10 are not working. The small waveform on D13 is just some leakage through the driver circuit. ZD4 and ZD5 protect Q9 and Q10 from being damaged by G-S voltages that are too high. Just a diode check is sufficient. Of course F2 must be in good condition.

Since you have checked the components in the driver, my only other suggestion is to take a close look at T2 and T3 wire connections to the pcb. These parts are heavy and the transformer wire leads can break inside the solder eyelet, cause a small break in the copper trace, or a wire might snap in a place that can't be seen. The driver design is not self-oscillating. Gate terminals of Q9 and Q10 must have a healthy drive signal - I will guess at about 5-10 volts p-p from S-G. If any of the transformer leads have become disconnected, nothing will work.

One trick is to use a small rubber hose to listen to various parts. You can listen for unusual noises like hissing or clicking and locate a possible loose connection or component problem. Often it helps to watch critical spots with the scope at the moment of turn-on.

What is the waveform across C38?

[semerchet]

Re: Manson hcs 3202

Thank you so much for your help. I've repaired a lot of these smps. They are used to charge E-bikes, but this thing is driving me nuts, moreover because D13 was s/c and it worked for about 20 min.

Okay inclosed I've put 2 pict of C38,

Again, thanks for your patience and help

[Longbow]

Re: Manson hcs 3202

Thank you. Now things get interesting. I would say that T2 primary should have a much larger squarewave. Others please correct that assumption if wrong. Transistors Q3/Q4 emitters should switch between 0v and 12v. The same applies to the emitters of transistors Q5/Q6.

My opinion is that we are not looking for a faulty component unless T2 or T3 have failed. Cracked pc traces and loose components make more sense. This is a forward converter with active reset. There are many detailed descriptions of this circuit available.

If something is shorted or faulty on T2 secondary, the primary of T2 may be loaded down. But in that case the 4 switching transistors (above) would be getting hot. Confirm Vcc of +12, and use the correct ground for T2 primary. Confirm emitters of Q3/Q4 have 12v p-p square wave, and emitters of Q5/Q6 also 12v p-p square wave.

Back to the driver. Please show the waveform on each of 2 secondaries of T2. The waveform on both windings should be the same. If Q7/Q8 are shorted, the driver won't work.

Also, Y1 and C56 seem like interesting parts to check. Make sure that those 2 have not become shorted or leaky.

Check capacitors and resistor values in snubber networks R57/C42, R58/C43.
Perhaps other forum members have ideas on this mystery. I think you are close to the answer, anyway.

[Longbow]

Re: Manson hcs 3202

Now is a good time to find another Manson supply and make comparisons. This should be a rather interesting repair, since the design is complex with all the protection circuits, PFC, different grounds and so on. Given a solid +12 Vcc at U3 pin 12, the primary winding of T2 should have at least 12v p-p or greater squarewave.

It is always a good idea to check VREF which should be +6v d.c. at U3 pin 2, 13 and 14.

In post #11, why are the 2 waveforms different? Use X10 probe and short scope ground lead!

[semerchet]

Re: Manson hcs 3202

Hi There,
Strange.., U3 pin 2 2,5 Volt and 13 and 14 are 5,05 Volt.I'm getting the same waveforms as in post 11 with the probe on x10
12 Vcc is present, and finding another Manson is just the problem.
R57 /c42 and R58 I'll check that later, and being close to an answer, well I hope so
Inclose are the waveforms T2 sec

[Longbow]

Re: Manson hcs 3202

This is all very interesting. First, I believe I am mistaken about the design. This seems to be a half-bridge style supply, not a forward. I get them mixed up.

Anyway, VREF is almost always = 1/2 x Vcc. In this case VREF should output on pin 13 = 5 volts d.c. You can see at pin 2 U5, R36, R37 form a voltage divider that should place pin 2 at about 2,5v.

Is there a problem with isolation between different grounds? I keep looking for a good healthy drive signal at T2 primary. More mysteries. Let's see the drive waveforms on U5 pins 11, and 8. Thank you!

[semerchet]

Re: Manson hcs 3202

This thing is driving me nuts...,

Allright first one is pin 8 the other pin 11

[Longbow]

Re: Manson hcs 3202

That looks correct to me if the scale is 5v/div. These pins drive the output switches, so there should be a 10-12 volt square wave accross C38. U5 has a regulated +5 volts that sets the main output voltage of the supply. Feedback starts from the L6/L7 junction and ends up at U5 1,2 and 15,16. Pin 8 and pin 11 are switching, so U5 is telling the supply to operate. But, T2 has no output waveform. I think this summary is correct, but it isn't leading us anywhere. Is C41 a special Y safety capacitor? These parts tend to have rather short lifetimes. If T2 secondary does not have a nice square wave, something in the driver is loading down the circuit. I'll think about it tomorrow!

[semerchet]

Re: Manson hcs 3202

Hi Longbow
I've checked C41 2 Nf and C 56 95 Nf
Both the grounds are isolated from each other.
C 55 and C57 both fine
Have a nice day

[Longbow]

Re: Manson hcs 3202

Thank you. Just for added interest, here is a similar model with almost the same circuit.

Please note the parts change section at the end of the manual, in case similar parts in your supply have also failed due to heat.

I am running out of ideas because I did not design this circuit! The best thing now would be to locate a similar model to compare the two. Most forum members are very familiar with computer supplies, but this one is quite different. U5 provides PWM error correction by the timing of the pulses on pin 8 and pin 11. Hypothetically, if the timing of pulses on these pins were the same (no time difference), you would see output waveforms on both pins, but no drive to T2.

So, there may be some trouble with the feedback loop from the output L6/L7 connecting back to pin 1 of U5. If so, an over-voltage or over-current failure might account for a shorted D13.

If U5 thinks that there is an over-voltage or over-current condition, obviously it would shut down the output of T2. Let's say U5 thinks it is over-voltage. That feedback comes through R63, and network C45/R61. C45 is my favorite value...1uF. The d.c. voltage on U5 pin 1 is compared with the reference voltage on pin 2 which is 2,5 volts. Over-current has its own dedicated feedback path. Also the microcontroller can control the supply via U4 A/B. Maybe this attack will provide some clues.

[semerchet]

Re: Manson hcs 3202

Ok, As I said I repaired a lot of these power supplies, but This one..,
If I disconnect ie r49 / d11 and d12 r53 would I then have a normal waveform at T2 sec?
And the slowly putting all parts back in order to find the failure? I just came up with that idea, I only don't know if I would damage more parts.
(And using instead of 220 about 80 to 100 V)
Anyway much thanks for your time and help!