CubeRevo HD PSU repair

[stroker]

No problems - it's just disappointing because I think we were starting to get somewhere.

I think TH81 (thermistor - I think) has burnt out - the black round thing just above the big cap next to the two diodes. Push it over a bit so you can get at the leads and check for continuity. If it's open circuit as I suspect see if you get get any details from it. For the sake of comparison check continuity of smaller blue component next to R82. Having said that that may have burnt out as well - if it is open circuit check for details on that as well. Any colours I mention refer to the resource pics and may be different on your board.

Try to check the two black diodes and one glass diode near IC81. Also check for short from the diode next to R82 and the heatsink.

Please also measure voltage on the striped end of one of the diodes (D81, D82) above big cap, they are connected together, expect around 280V dc (a rough guess).

[jphilweybr]

Spot on stroker. TH61 is open circuit. It is labelled NTC 100-9, albeit I’ve also read it as 110-9 – bit of a smudge there. The adjacent blue C86 is OK – 67kOhms. There is no continuity between D86, next to R82 and the heatsink – OL.
D85, and D86 are OK. D87 is showing about 0.6v in both directions.
While I was about it, I checked D81 and D82, which are ok. The voltage from them to the IC91 heatsink came in at about 2.4 then I.5. However, all voltage readings now seem to have gone haywire. To R82 I’m also getting 2 or less whereas I was getting 19. To R98. I get 0.3, whereas earlier it was 5.1 and last night 11.
I sent you a PM two days ago on another matter, and I’ll send another shortly.
Looking at the picture on the other resource, I noticed there was a fan next to the mains input. Mine does not have one, nor there are header pins on the main board to which one could be attached. So perhaps they discovered early one that their PSU boards tended to overheat?
Many thanks

[stroker]

I have been looking for a replacement for TH61 but cannot find NTC 100-9, is it possible it could be NTC 10D-9. I am not familiar with the markings on these but somehow it seems 10D refers to 10ohms and -9 rates it as 2Amps. If you can get a replacement for TH1 I suggest you get two in case we miss something and blow another.

I made a mistake C86 is a capacitor and not a thermistor so ignore that. D87 not sure why it's 0.6v both ways.
Yes I read your PM it could be useful.

The purpose of TH1 is to reduce the current surge on power up as the large cap charges up.
We really need to bypass TH61 to check for further problems but I don't want to risk blowing the fuse on the board (just more things to replace).
I was thinking if you has some 2A fuse wire you could solder a piece between it's leads but this may blow anyway due to the power on surge.
Failing that we need to get a replacement for TH1. The checks we've done don't suggest any problems but if IC81 or IC91 are faulty the fuse wire would blow.
I will leave it up to you - do you feel lucky?

I am having problems uploading pics for some reason such is life!

[jphilweybr]

Thanks. Yes, I have a 2a capsule fuse. What would I join it to? The legs of TH61 having snipped the top off?
I’ve had difficulty loading photos. My last attempt was to prepare the text and then choose Go Advanced, not Post Quick Reply. Then I clicked on the insert image button, then select files in the pop-up box, and somewhere along the line there was an option to upload files, and that worked.
I’m out for the rest of the day, so it may be tomorrow before you here from me.

[stroker]

I would leave TH61 in position if possible just connect fuse wire across the leads.
This will prevent the leads of TH61 disappearing through the board if you apply enough heat to melt the soldered joints on the board.
If you are connecting with flying leads make sure it's well insulated as we don't want 300V going where it shouldn't.
If it blows don't keep trying there may be a problem we have not noticed or it may just be surge current.
Could check for a short from each end of TH61 to the heatsink but I think we may have already done this - not sure.

[jphilweybr]

Stroker VERY SORRY, I got it wrong two days ago. Having checked resistances from TH81 (and not TH61 which I read at first) and the heatsink – all high, I discovered there IS continuity across the two leads of TH81.
Well I know you like a challenge, but you don’t want them added to you in this way.

[stroker]

OK I'm just about to have some grub so may not post again for a while.
In the mean time could you measure the voltage form the heatsink to each leg of TH81 expect about 300V.
Also the voltage from the heatsink to each end of R85 and R86 which are next to IC81.
I think TH81 is the only thermistor on the board so don't worry if you get the name wrong.

If you do get 300V on both leads of TH81 you could link R98 and R95 again and check if you get 5V, 3V3 and HDD 12V.
I suggest you do this with the mainboard and HDD disconnected and prepare for a possible flash.

[stroker]

None of this is making sense we need to double check we have mains getting through.
Check continuity from live and neutral to the none striped ends of D81 and D82, one should have continuity to live the other to neutral.
If you do not have continuity (probably from D81) you need to double check the on-board fuse.
Live goes through the switch, then the fuse and then through a winding on L1 to D81.
Neutral goes through another winding on L1 to D82 - I would not expect L1 to be a problem.

[jphilweybr]

Thanks stroker.
Well after your 2.42 p.m. post I went to check voltages and got baffled because I couldn’t find any. I was about to report that when I saw your second post – no it didn’t make sense. I was wondering whether unplugging and re-plugging in the mains had damaged something.
Attached is a pic of the underside of the board in that area. What I’m now finding is that there is continuity between one of the input pins in the R80 and R81 area (upper part of the pic) and the hot side of D82. However, no continuity between the other input pin and anything, and not between D81 and anything. (Incidentally, how does the board distinguish between mains L and N, or doesn’t it need to?)
And now I swear to you over the last two or three days I’ve checked main fuse F81 periodically, but tonight it’s showing as blown.
Blast, to put it mildly
IMG_4654.jpg

[stroker]

Brilliant pic and clean looking board.

I spotted what I hope the problem is as soon as I saw it.
L1 is shown as two black lines as if it were links, which being an inductor or choke it is a short to dc.
Anyway if you look at the left hand line in the pic the solder joint at the top end looks fractured.
It could just be a shadow but it looks like there is a break in the solder joint looking like a black ring around the pin.
If this is the case just resolder it adding more solder.
Perhaps you can find a way of bridging the fuse with 3A fuse wire or a fuse on flying leads (make sure it's well insulated.
I will try to find a replacement or perhaps a local TV repairer may have one, 3.15A used to be common in TV's, they may even fit it for you.

The nature of the break could mean the continuity was intermittent which would have caused sparking (at the break) and surges in current (trying to charge the big cap when continuity was present) which would eventually blow the fuse. So perhaps this is an explanation and you (and me) are not going mad!

[stroker]

Forgot to mention - the board doesn't care which is live and neutral.
The mains is filtered then full wave rectified and the used to charge the big cap.

Everything after the rectifier is "half live" the positive half of the cycle goes to the positive connection to the capacitor and the the negative half of the cycle goes to the negative connection. The same is true for the neutral which is opposite it polarity and so supplies negative when live is positive and vice versa. This means that each connection of the capacitor is live for half of each cycle of the mains, or to put it another way live for half the time.

Not easy to explain or understand but trust me you don't want to touch either side of the resulting 300V+ supply.

The 300V is chopped/switched through the transformer primary winding and then the isolated secondary windings are used to produce the dc voltages we need. Feedback signals are passed back to the mains side of the circuit to control the output voltages produced. These feedback signals are passed through opto-isolators to maintain isolation between the low voltage side and hazardous mains side.

[jphilweybr]

Thanks for all the explanations!
First I’d better make sure I understand the lingo.
I don’t have anything marked L1, but I assume your suspect link is the one from near C81 to near C83. If so, the pin protrudes a bit but it looks OK and there is continuity there, and e.g. from R80 to R91. Could be intermittent though, so will add a bit of solder. More pics attached.
While, I was about it I checked for continuity from the screw hole (top ends of C83 and C84) and the GND pins on the output end, and found none. Is that OK? And no continuity across C83 and C84.
My plan to by-pass F81 was to run sheathed wires to a capsule fuse well clear of the board. Is that what you mean by flying leads?
It strikes me I should take some time out to systematically check the diodes and resistors that we’ve looked at before.
I’ve ordered some replacement components which should be with me in two or three days.
IMG_4659.jpgIMG_4658.jpgIMG_4652.jpg

[stroker]

Sorry should have been L81 - looks like a transformer but is two inductors (or chokes) built on one transformer core.
The sort of break I am seeing would still show intermittent continuity.
I would also explain a the lack of continuity from D81 to anywhere as reported earlier.
You should have continuity from D81 back to the fuse and until you do we aren't going anywhere.

The screw holes each end of the board will have continuity when the board is screwed into the chassis.
C83 and C84 are low(ish) value caps and should not show continuity, they are used to filter out electrical noise which we are not worried about at this stage.

You plan for the fuse is fine as long it is reasonably well insulated which I think you understand.

Did you order a solder sucker? I can't live without one!

By the way I sent you an email.

[jphilweybr]

Continuity.jpg

Thanks, stroker, and thanks for the pic showing where I should have continuity. I’m uploading it now in case anyone else is interested.

My only worry about the GND continuity was that most of the time I have the PSU Board out of the box, so do we still have a 0V point? Hope you’re not getting tired of these elementary questions!

Well do you know, I’m getting on a bit and I’ve managed to live all those years without a solder sucker. In fact, until a fortnight ago, I would have thought that was a misprint, perhaps missing an f somewhere and a b or something somewhere else. Anyway, my life will soon be complete because I have ordered one. As well as some braid, pencil flux, and while I was about it, a new soldering iron. My existing one would probably be OK for the fault you spotted, because there’s nothing else near it, but I wasn’t sure about the more delicate work I expect you’ll be getting me on to soon. I’ve also ordered two replacement fuses (I’m sure to ruin one), and replacements for IC 91 and A3651. In fact, I think I might take a break and wait for that lot to come rather than mess about with temporary fixes. I’m out all day Tuesday anyway, so maybe I’ll get on to some of this Wed or Thurs.

[stroker]

Thanks for uploading the pic, I notice it is now about 1mB whereas when I sent it to you it was more like 4mB.
Perhaps I should send it to myself before trying to upload it on here and then it might work!

I really wouldn't worry about those GND's but when we get the PSU working again you can screw it in the box and you won't have to worry anymore.

After ordering all that we won't have any excuses for not fixing at least the PSU.
It's good to see you've still got a sense of humour after the recent setbacks!

I agree we should have a break from this while your parts arrive.
Whoever breaks radio-silence first is a sucker!

After we fix the PSU it will be back to the real fun and games!

[jphilweybr]

Well would you believe it, the shoulder shucker was due for delivery Monday, but still hasn’t come. Don’t tell me I am to be denied one of life’s rarer pleasures after all!
Actually, a few other things due yesterday or today haven’t come either. Royal Mail 2nd Class must be busy. So there’ll probably be radio silence from me for a few more days.

[jphilweybr]

Well would you believe it, the shoulder shucker was due for delivery Monday, but still hasn’t come. Don’t tell me I am to be denied one of life’s rarer pleasures after all!
Actually, a few other things due yesterday or today haven’t come either. Royal Mail 2nd Class must be busy. So there’ll probably be radio silence from me for a few more days.

[stroker]

I had an item delivered 2nd Class recently and it turned up about a week later than expected.
Alright so I said I couldn't live without a solder sucker but to be honest I cannot remember the last time I used it!

If your sucker still hasn't arrived you could try removing the fuse by just pulling it out while heating the two pins alternately and then clear the holes with desolder braid.
You need to be careful not to apply too much heat for too long or you may risk the copper track lifting from the board. Applying flux or fresh solder will help the solder to be more fluid and make the job easier.

I would like to see what results you get after changing the fuse and touching up the solder joints on L81.
I would not bother with IC81 and IC81 just yet as I suspect one or both of them are still OK.

[jphilweybr]

I did the folder tuckers an injustice – it was next Monday they promised and it came today. However, I’d like to wait for my new soldering iron before getting on with this.

I’ve ordered a replacement IC91 – due last Thurs but not here yet – but not an IC81.

[stroker]

I thinks it's time for a short progress report since we have not posted much over the past few days.

We had a bit of set back recently which turned out hopefully to just be a blown fuse on the PSU board but until this is replaced we will not know if there are any other related problems. I believe the blown fuse was caused by a fractured solder joint on L81 so we will touch up the solder joints on all 4 pins of L81.

Jphilweybr has had to order some parts and soldering equipment and will soon be practicing his soldering skills before turning his attention to the fuse which is soldered on the PSU board between L81 and the white 2 pin connector. We will then look to see if there is any other damage and continue fault finding.

We had previously discovered that the 5V standby supply (5VS) was apparently being overloaded when the mainboard and front-panel were connected. With the front panel disconnected 5VS was ok which seemed to point to the front-panel being the problem.

However jphilweybr managed to find more information on the net which suggested the problem may still infact be with the PSU itself. The finger of suspicion being pointed towards C21 but after looking at the pics of the PSU board I could see that C22 could also be a problem as they are connected together via inductor L21. So when we get the PSU up and running again to the same extent as before we will be checking C21 and C22.