CubeRevo HD PSU repair

[stroker]

D91 and C92 are connected in parallel (across each other) so I suggest you remove C92 and check D91 and C92 separately to see which is shorted, maybe both.

The problem with D91 its it is a zener diode which come in a range of voltage ratings (I think this one will be around 20V) this will be marked on it's glass body and is usually difficult to read so if it is faulty you will need to remove it carefully and try to read the details - it usually ends in "C" followed by digits referring to the voltage rating.

I am not sure where D87 is - is it next to D91? If this is it you reading could be affected by a shorted D91/C92. Also where is D93?

With IC91 I want you to remove the remains of the legs from the board but do not fit a new IC yet. At some stage I will want you to power on to check some voltages but we are not ready to do this yet.

I think R93 and R91 are 3W resistors but not absolutely sure (they could be 5W) - are you sure only 3mm diameter? R91 is 220K not 110K (red=2). Try to remove R91 I am pretty sure is split/broken.

By the way where did you get the ohm character - I want one!

[jphilweybr]

InkedIMG_4669_LI.jpg
Well, what do you know! I can help Stroker. If you’re using Microsoft Word, there is a facility somewhere to Insert/Symbol. In Word 2010, there is an Insert tab, and you click on symbol, and choose what you want to insert. And in fact their symbol for symobl is…Ω This means I usually prepare the text of a posting in Word, as I’m doing now, and then copying it onto the website.
D93 is the black job next to D91. D87 is the other side of the board pointing backwards from the transformer towards IC81 (see pic). I do wonder whether D87 got shorted to R98 during the accident.
I’ve taken off C92 and it is OK – passes the increasing resistance test. There is continuity across D91 though.
Will get on with the other jobs, maybe tomorrow.

[stroker]

Thanks with the ohm symbol I remembered that it is pretty common to use "R" instead so 100R means 100 ohms. I usually just type my replies into the web page but it's good to know these things.

You will have to remove D91 and check it off board remember we need to know what voltage rating it is (I'm guessing 22V). The same goes for D87 which is also a zener diode. I do not think D93 is a zener but you're probably better off doing the same with that.

I have been a bit side-tracked today - looking at your motherboard pics. It seems to me that we should not underestimate the effects of the poor soldering on the 15 pin power connector on the motherboard and I would not be surprised if this is the main cause of our problems. Of the 15 pins 7 need attention one end of the connector and 3 pins at the other end.

By the way you had difficulty soldering the third pin because there will be heavy internal tracks inside the board which will conduct the heat away from your soldering iron. You will need to leave the iron on the joint for a long time before the solder will melt properly and use lots of flux. It is very important to get the solder to flow freely otherwise you will have a high resistance joint which is no good for the high currents involved. Not all the pins are high current but 3V3 and 5V definitely are, you will find it much easier to solder the last 2 pins the other end of the connector which carry +FL and -FL which do need re-soldering and I believe are something to do with the filaments in the front panel display! Of course a more powerful soldering iron would also help.

[jphilweybr]

IMG_4982.jpg
The Satpimps email alert system seems to have stopped working, so have just picked this up.
Meanwhile, I have taken off IC91, R91, R93 and D91. From the pics the resistors are about 3mm at the waist and maybe 4mm at the ends. As you’re about it, how do I clean the iron bits – this having been a problem the last two days? There are YouTube posts which talk about tinning, cleaning with p[lumbers’ flux etc. Is there a simple solution?
I have a reasonably powerful magnifying glass, but have not got anywhere near reading D91. I think I identified which side the letters were on and took about seven pics, but totally failed to hold it in the right position. Will try again!

[stroker]

I got an alert when you posted - click thread tools unsubcribe and subscribe again and see what happens.

I think the resistors may only be 1W but if we had higher wattage it would not matter perhaps you can measure the distance between holes in the board so I know the max length we could use.

As for cleaning iron bit I have always just used a wet sponge, I don't know if they are a special type - try searching for soldering iron sponge. Also look for soldering iron tip cleaner - steel wool type thing, I have only used these a few times but they can be useful. You should only tin the end of the bit otherwise it could collect a large blob of solder which could drop off randomly in places where it's not wanted. To not file you bit - modern ones are iron coated - can't remember why.

Your pic of D91 I can see two characters when zoomed in - look a "9" or "6" and "0" or "C", really need to see it from different angles. I would prefer to get the right value rather than guess. The ones I used to use years ago were BZY88Cvv and BZX61Cvv where vv is the voltage but not sure "61" is the right number - it was a very long time ago. So if they are similar to those you're looking for "C" followed by 2 digits something like 22.

[jphilweybr]

IMG_5004C.JPGIMG_5003.JPGIMG_5002.JPGIMG_5001B.JPGIMG_5000.JPGIMG_4999A.jpg
R91 and R93 hole centres are 15mm.
Some better pictures of D91 which you can blow up; look in particular at those I labelled A, B and C. I took them in that order turning the cylinder upwards a bit each time.. So they seem to read:
IN or 1N with possibly another character hidden by the white stripe
52
51
Hope that makes some sense.

[stroker]

That took a bit of finding, I wasn't sure which came first the 51 or 52. Anyway 1N5251B is a 22V 500mW zener diode, a pack of 5 for £1.50 that should last a while.
I'll email you a shopping list later.

Do we need a replacement for D93?
Please remove D87 and have a look at it if you think it's faulty.

We also need R91 and R93 is there anything else?

[jphilweybr]

So you guessed right about D91, but glad we’ve been able to check.
D93 is OK , 6.7v and I. D87 gives a voltage both ways. However, I think it’s identical to D91 – I can see the 52, so will have a spare anyway. (I need to take a few days off, and avoid getting soldering kit out again, so a good time to do it would be while waiting for components to arrive.)
My only thoughts are C93 next to the burnt out resistors and C88 across the board, next to D87, which was bent over. Are these caps? If so C88 passes the increasing resistance test. C93 registers a resistance for a split second then goes straight to OL. If you think there might be a problem I’d prefer to order a replacement now, even if it turns out to be a waste of money. It, and C88, are marked:
E
471
I’ve checked all the resistors and diodes and caps on the hot side of the board, but not the blue jobs near the mains supply, nor IC81 or PC41. I assumed that nothing would have got through to the cold side.
Ah, just noticed another brown two legged feller near PC41. It show OL. It’s markings are:
B, possibly with a + to the right of the B or just a black spot
102
Is there anything else I should check?

[jphilweybr]

New components have been ordered, so there will be a delay of at least a few days. In fact, for one resistor, they're only promising delivery on April 10th.

[jphilweybr]

New components have arrived, though looks too nice to be indoors, so maybe Monday before I get round to thing.
Stroker, do you want me to install everything, including IC91?

[stroker]

I agree it's way too nice to be indoors! I would recommend leaving IC91 out for the time being and replace everything else. Then power on with the mainboard disconnected and check the standby supplies, that is everything apart from the main 5V and 3V3. I'd do it this way just in case we have missed something and we have another incident although I think this is unlikely.

If all is ok fit IC91 but do not use the link on R95/98 anymore. I recently found that Q101 on the motherboard (next to the power connector where we found the fuses) is used to supply 5VS to the P/S connection on the power supply board, the collector and emitter of Q101 effectively connect 5VS to P/S so you should be able to short these together to turn on the main supplies. So I suggest you use a piece of solid wire to short 5VS and P/S in the connector on the PSU. Initially power on with the motherboard disconnected just in case.

You haven't said if you have received your high power (100W?) soldering iron yet. If/when you do I want you to touch up the joints on the 15 pin connector on the motherboard, use plenty of flux and add more solder if needed. Keep the iron on each joint long enough to get the solder to flow freely, I cannot emphasis this enough we need really good solid connections on here. Having said that do not go mad and leave the iron on way too long as you risk overheating the board, the low voltage pins like 5V and 3V3 will need more heat than the likes of FL- and FL+

If you are worried about this solder one or two joints and post a pic so I can see before you do the others. I am hoping that these bad joints are the real reason we cannot get the box to power up properly, bad power supplies to the motherboard circuitry may not be sufficient to be passed on to the front panel and prevent the display and switches from working.

[jphilweybr]

OK, back to work Stroker! Monday promises to be a bit of a dull day. Actually, I didn’t order a 100w iron, but I did find my old one, so I do have two 30w ones and four bits. If you think 100w is essential for the connector on the motherboard I’ll get one. Or perhaps try with what I’ve got and take some pics. Another novice question, what is the technique? Do I apply some flux to the joint, melt it, then melt the solder and expect to see it fill the hole? Sorry – basic question!

[stroker]

I gave you quite a few things to do in the last post, I did this because I thought we may be too busy enjoying the weather to be posting messages. So I wasn't really expecting you to do it all on the same day.

I suggest you concentrate on the the PSU to begin with although if you find you have the time try soldering the FL+ and FL- end of the 15 pin connector on the main board. These I expect to not be so heavy as the 5V and 3V3 pins and so should be easier to solder and my thinking is that if the FL supplies are used on the display you may produce signs of life (this is probably just wishful thinking).

When you come to the 5V and 3V3 pins I am not sure how heavy the power tracks in the board will be. So don't expect these to heat up quickly when you apply the soldering iron, it may take 30 seconds or more before you see any sign of the solder melting. If the tracks are really heavy you may find that the iron cannot supply enough heat to raise the temperature enough, you will have to suck it and see.

I am not an expert at soldering but I remember being told by a soldering trainer that you should heat the joint first and apply solder once it has heated up. However in practice I tend to put a little fresh solder on the iron first then heat the joint and add more solder (if needed) when I see the old solder melting freely. Using flux should also help to spread the heat and in any case shouldn't do any harm.

I'll try to explain why a badly soldered joint is such a big deal on power supply connections. For example you have a 5V supply connected through a bad joint which adds 0.1 ohms into the supply path and the current along the path is 5A, there will be a voltage drop across the bad joint of 0.5V (ohms law V=I*R) so only 4.5V will arrive at the load. Most 5V devices only work reliably down to 4.75V so 4.5V is not enough for them to work properly.

[DCG]

did you recive my pm requesting which account you wanted deleting as ive had no reply

[stroker]

I'm sorry DCG but I have made no such request and have not received any pm from you. Do you have your wires crossed?

[DCG]

that wasn't directed at you m8

[jphilweybr]

I’ve finally got round to fixing all the replacement components other than IC91. Checked diodes, resistances and capacitors. Diodes fitted with black stripe at the arrow end. Inclined to pause before powering up, perhaps even sleep on it. Or shall we go for it Stroker?
There is no continuity between and pins in the new IC91. Can I take it that’s normal until it gets powered up and some switching takes place inside?

[stroker]

Good to see you back.

Power up whenever you feel up to it - I suggest you do these tests with the mainboard disconnected. Check the standby voltages ie. everything except 5V and 3V3 (no link required). Also measure voltage from the mains end of R98 and IC91 heatsink which should see about 22V - this is the standby supply to IC91. If you get anything other than 22V we need to investigate before fitting IC91.

If all is well fit IC91 - you may need heatsink paste or they sometimes use a silicone type washer - you should see this when removing the old IC91 if you haven't already done this. Link 5VS and P/S at the DC output connector with a solid core wire then power on and check you have 5V and 3V3 as well as the standby supplies. Linking this is hopefully safer than the R95/98 link we used before but to work this needs a few more components to be working as well - hopefully this should not be a problem.

If everything is ok you can try connecting the mainboard (obviously power off first) and we should be back to square one (or is it two) again.

The next step is to sort out the bad solder joints on the 15 pin power connector on the mainboard. As I said before attempt to sort out the low current pins first as these should not sink too much heat from the soldering iron. Post some pics if you want me to check what you have done. If you attempt the high current pins you iron may not be able to supply enough heat to get good results. Check my comments in recent posts and let me know how you get on.

[stroker]

I forgot to say - don't worry about the new IC91's lack of continuity most likely the continuity you saw previously was caused by components external to IC91. It might be worth double checking you have the right replacement part.

[jphilweybr]

Some good news. We have the expected voltages at the output end of the PSU, albeit the +8v came in at 7.1v. +5VS is the normal 5.1V. As expected, no +5v nor +3VS and no P/S.
At R98 and R95, I’m getting 1.3v. On re-checking, I find I’m not getting reliable continuity between the cold end of R91 and the hot end of R98, so will try re-soldering that. I’m getting 113.6v/108v at either end of R91 and 403v/332c across R92 near the transformers. Is that OK? If you’re happy, shall I go ahead and fit IC91? Or do we need to get 22v to R98 (which is what we had before)?