And yes sorry 22uF would be best - I miss read your post - still wrestling with my router.
And 470uF for the other but it is more likely to be CE47.
I think the top end of R228 may not soldered properly - you saw 12mV on the bottom end nothing on the top (should be about 5V) - be very careful soldering as these resistors come lose very easily, if it does it may be on the end of your soldering iron bit.
Check for continuity from top of R228 and top of D9.
Last edited by stroker; 23-03-2017 at 03:53 PM.
Too late! I’d already removed the motherboard when you posted.
R228 looks clean through a magnifying glass. I can’t get a better picture though than the one I posted yesterday. If I’m going to get closer, I think I will have to remove both the capacitators next to it.
Shorting D9 didn’t seem to have any effect. I removed the R98/R95 link so SATA stopped working. Shorting D9 didn’t bring it back on produce any light on the front panel.
Resistance between the top end of D9 and the top end of R228 starts off at zero and creeps up to 15Ω. Between the two bottom pins there is continuity. Diagonally, there is 7.7 kΩ. Voltage is 12mV to both bottom pins, 5.7v to the top of D9 and say 2mV to the top of R228.
Attached is a pic of the underside of this area. Neither D9 nor R228 are connected to the underside. In the centre right, there are signs of some damage. I assume that happened during assembly and has been there ever since.
I think I need to look at this again the reset circuit is more complicated than I thought. Looking closer at your earlier pics the top end of R228 is connected to ground and not 5VS which is why you measured 7.6K to GND from the bottom end of D9.
I am convinced we are onto something here but it is not as simple as I thought. I think we can leave the link off R98/R95 now.
It looks like U4 (above the 2 caps) has something to do with the reset circuit. Please check continuity from bottom end of D9 to each of the pins of U4 - pin1 is nearest R41 then count anti-clockwise around the device. Check CE22 is not shorted. Is R42 missing (may not be needed)?
This circuit is really strange and if I understand it correctly could be a very dodgy design - I'll be back!
Thanks for the pic - leave motherboard out for now.
Thanks Stroker. Happy cogitation!
On U4, Pin 7 on the corner near the Syncmos chip is connected to ground. To that pin from D9, resistance starts at zero then climbs to 7.7kΩ. All other measurements are in the MΩ range or OL. Measured on the underside, CE22 reads 270Ω
OK I take back what I said U4 appears to have nothing to do with it. I need some more facts so I need you to check a few things.
1. Remove CE47 and test it just so we know where we are, leave it out for now but keep it safe.
2. Between the top edge of the Syncmos chip and the big resistor R194 there are 2 small resistors, check for continuity between R229 and R228/D9 (bottom ends?)
3. Take another pic of the area above your previous pic, I need to follow the 4 thin tracks in the middle going upwards.
4. Take a pic of the top side of the board showing the area to the right of the Syncmos chip to the far right . It might be easier to take a full width pic of the board below the main CPU as long as the detail is good.
5. From your previous pic the tracks going to the big resistor R194 appear discoloured or possibly burnt, perhaps you could take a closer pic.
Don't expect a quick reply - other things to do today.
IMG_4951.jpgI’m going to have to leave CE47 until tomorrow.
Meanwhile, here are some other pics I took a couple of days ago.
Again difficult to get readings from R229 and R230. Assuming the covering of the Syncmos chip is non-conductive so I am actually measuring the resistors, they both come in at 100Ω. From their bottom ends to the bottom of D9/R228 resistance is 4.2MΩ.
Yes, I had noticed that some areas of the motherboard (green part) look a darker shade, but looking at them now, they all seem clean. Let me know if there is another pic you want me to take.
This is a pretty good pic, I would like a similar pic or the top side of the board please.
From this pic it looks like pin3 of the power connector (third from the right in pic) is not soldered properly but it's right on the edge of your pic. The other pic you posted today shows this problem more clearly. By the way it looks like pin1 of each connector is soldered to a square pad on the board while all other pins have round pads.
I am trying to work out what signal/voltage is on pin3, if it is the same as pin3 on the power supply it is 3V3 which is a very important supply indeed - could this be the cause of our problems?
I could do with a close up pic of this end of the power connector to try to make sure that it is just a fractured solder joint and not more serious damage.
I should point out that this will not be the only problem since the 3V3 (and main 5V) supply only comes on when the box is out of standby. However if we can get the box powered up with the link on R95/R98 on the PSU we may be in a position to flash the firmware. But this is no time to get carried away - one step at a time!
Anyway back to IRLvWAL - come on Wales!
Last edited by stroker; 24-03-2017 at 09:46 PM.
I assume youíre referring to the 11-pin connector to the front panel, so Iíve taken a couple of pics showing these from the side. To me they all seem to have the pin protruding from the solder.
The square pads on the PSU connector and on the front panel 11-pin connector are both GND.
For the PSU connector, Pin 1, defined as the one with the black wire, is on the left of the picture and carries +FL. So the third from the right, call it Pin 13, carries 3VS, and Pin 12 5V.
For the 11-pin connector, Pin 1 with the red wire is on the right of the pic and is connected to GND. Pin 2 carries +5VS.
Now hereís an odd story. You may even regard it as good news, but more likely bad news.
Prompted by your query about the solder, I set about checking the voltages on the front panel.
At this point things are normal. The failing SATA disk is clicking away. To remind you, Iíve always had all the voltages on the PSU output connector, except +5V and +3Vs. Now with the R98/R95 link in, I have those as well.
So checking the front panel I measure in the following order:
Pin 2 5.1
Pin 11 Ė
Pin 10 17.3
Pin 9 24.0
Pin 8 26.5 (could be minus Ė didnít check)
Pin 7 3.5
This seemed significant, so I stopped to write this down and went to re-check in case Iíd missed a pin. Now, I couldnít find a voltage at all in any of them.
So I check the PSU output at the motherboard pins. Apart from two, no voltages there either Ė no 8v for instance and no +12V though the SATA HDD is clicking away. The voltages I do have are 3.4v at+3VS and 4.4v at +5V. I also have 336v between the rear heatsink and TH81.
It may be that just powering down and back on again may restore things. But Iíll leave it on in case thereís something else you want me to check.
Sorry for the delay in answering I took advantage of the good weather to catch up on a job I needed to do on the car.
Thanks again for the good quality pics.
Sorry that some of this is getting confusing. Perhaps I should have mentioned that it was the 15 pin connector with the badly soldered pin. This is best shown in your pic img_4960.img if you look at the third pin from the right you can clearly see the (empty?) hole that the pin goes into below the solder. It is not the solder further up the pin that is important but the (lack of) solder joining the pin to the pad on the board. The solder needs to connect the pin to the pad. The hole in the board is probably what they call "through hole plated" and the pin may be touching the plating but for good connectivity the pin has to soldered to the plating in the hole. The quality of the connection is even more important with low voltages because the currents involved are much higher.
Please re-solder the pin I have high-lighted leaving the soldering iron on the joint for at least a few seconds to get the solder to flow freely through the hole. The pin on the far right could also do with similar treatment and also the 3 pins on the far left - these are not as bad as the one mentioned above but are not the best.
I am not sure what to deduce from your voltage measurements but please continue to mention such things as I am still looking for clues. If the hard drive you are using is a 2.5" one it will not use 12V only 5V.
I am still concerned with the D9/CE47/R228 situation - all I have seen so far are things that discharge CE47 but there must be something that charges it up. It could be helpful if you were to remove CE47 and take a (top side) pic of the area so I can see tracks going to it if any.
Thanks – amazing how you can spot these things from pics, but I can’t see them in real life.
Tomorrow will be soldering/desoldering day.
So the pin you noted, Im going to call it 13, is definitely +3VS. Fingers crossed.
And SATA does not use 12V? I learn something new again – yes some websites show it, but not obvious in Wikipedia!
I think 12V is present on the SATA power connector. The point I was trying to make was that 3.5" drives use 12V and 5V but 2.5" (laptop) drives only use 5V. This is also true for IDE drives. If you look at the label on some drives they state the current consumption on 5V and 12V I don't think I have ever seen a 2.5" drive mention 12V.
Anyway an hours less sleep tonight usually takes me week to adjust! Good luck!
Sorry, I meant SATA 2.5Ē, which is what Iím using here.
Anyway, a couple of steps forward I think.
I could actually see with the ***** eye what Stroker meant. Pin 13 protrudse further from the board than the others and its cloak of solder sits a little high, whereas all the others sit nicely on the board. Iíve filled round with a bit more solder. I canít promise that I was actually getting it down into the hole. In fact it kept moving around, forming joints with the adjoining pins, but finally seems to have sealed where the gap was. Pics attached. Actually I see there is a blob of solder worryingly near Pin 14 so Iíll have to fix that.
Secondly, have taken CE47 off and it displays resistance climbing from modest to OL. So I assume OK.
BUT another big step backwards. Before powering down with the mysterious loss of voltages in place, I thought I should check voltages at R98 and R95, which were in fact as before. However in doing so, I must have disturbed the link and trying to restore it, I did something which produced a bang, a burning smell and a soot- covered R93 at the other side of the board. The main fuse had also gone. I have replaced the fuse, and spent an hour (I timed it) trying to release R93, and am not yet there by some margin. Mercifully, we do have a pic of it in its younger days. Itís blue and from the hot side its stripes are brown, white, white, bigger gap and grey. Obviously I canít power up until itís replaced, and I suppose I should take a little time out to see if anything else is damaged.
OK - Robot Wars and Top Gear finished - back to business. I wouldn't worry too much about R93 just yet - check for a short across the biggest capacitor C85, if there is a short I would think you have blown IC91. R93 may have burnt out as well, it is a low value so should measure about 0.4 ohms - your problem removing it is probably because the legs are bent over on the bottom of the board - you need to melt the solder and straighten the legs using a small screwdriver of the soldering iron bit, then melt the solder while pulling the resistor out. If you need to replace IC91 check the short is gone before fitting the new one. In fact I would suggest leaving IC91 out until you check the standby voltages are present first.
I was going to mention before that your linking method was a bit dangerous - from the pics I got the impression you had just twisted a wire around the resistor legs, I'm sorry if I have got this wrong. I really wanted you to put a blob of solder on each resistor leg and solder the wire to these without any loose strands sticking out - however I assumed you would be very careful using your method so I decided not to mention it.
Your latest pics of the 15 pin connector show more problems - the first 5 pins at the dodgy end are not good and the first 3 pins the other end I would not be happy with. I think you have a flux pen - put lots of flux on the pins and re-melt every pin on the connector to be sure. Flux is used to help the solder flow freely, don't be afraid of using too much, it should also prevent solder bridging to adjacent pins. The green coating on the board is called solder-resist and is also used to prevent/minimise solder bridging.
I am a little worried that the third pin we were looking at is protruding further than the others - is it still properly attached to wire (leading to the PSU)? Try to check continuity between the PSU and motherboard while wiggling the wire (with your third hand).
I hope I have covered the important points but if in doubt about anything ask questions first.
First, picture of sad event and one taken of that area in the early days of this adventure. Yes, I decided not to solder that link, because we had cause to take it on and off. And yes, I was very careful Ė I always powered down before fiddling with it. Except two days ago when I was missing those voltages, so I wanted to check what was on there. So I was very careful, but Ö.
I havenít had much time today, but I tried to check what else might be damaged. Iíve confined myself to the hot side of the PSU board. Tomorrow, Iíll try and upload a CSV file or pic thereof, of my measurements. Meanwhile I have identified the following problems/queries:
1 Sooty R93 Ė should be 0.39Ω; my little multimeter shows 0.00 Ω, the more sophisticated one shows 0.6 Ω. Iíd be best replacing do you think?
2 R84 looks shorted.
3 R80 and R91 (the latter next door to the sooty R93) are not giving me stable readings, albeit they home in and flicker around a given figure. In fact on my more sophisticated multimeter, R91 shows 2.78 with the setting at 20k and 17.8 when set at 200k. All the other resistors Iíve checked (81,82,83,85,94,95,96,97) give rational readings but I have not checked them against the expected nominal values.
4 No bulging capacitors but sooty C92 definitely shorted. I have a replacement. C85, C86, C87 have been tested in situ and all pass the increasing resistance test.
5 D91 shorted.
6 D83, D84 and D87 show 6.7v with the red probe on the arrow, and 5.7v or thereabouts the other way. D85 and D86 look OK (4.6v and I) as does D93 (2.2v and I).
7 I reported on IC91 back last year (did we really have nothing better to do on Dec 27th?), and found continuity between three pairs. I now only have one Ė 2 to 3 in that post, but neither of the others. I have a replacement.
Perhaps erring on the safe side, I should replace C92 and IC91, which I have and R80, R84, R91 , R93 and D91. If you agree, Iíll probably need help to source these.
Wow that's a bit of a mess but probably looks worse due to the soot. If you zoom in on the pic you can see IC91 has a crack in it on it's top edge near the screw and the face of the device has blown off so that's fit for scrap for sure.
Your best bet is to remove IC91 to get it out of the way as some of the shorts you are seeing may be due to internal shorts in IC91. Don't fit the new one yet and be careful not to damage the tracks on the board when removing it - I doubt it if we could ever get another board. It might be better to leave it screwed to the heatsink and clear the solder first to avoid stressing the tracks on the board. Although sometimes it's easier to cut the legs off the IC before clearing the solder if you can get cutters in there to do it - then melt the solder on each leg and pull them like teeth. You will have to decide which is best for you and the board.
After removing IC91 check diodes D81,82,83,84 and check for a short across the big cap.
I need to digest your info from today and possibly tomorrow as well before deciding what we need to order.
OK, but just to warn you my experience of electronics as you know is virtually non-existent. My experience of pulling teeth is totally so.