1. Home
  2. Electronics Repair
  3. Today's Odd 'un ...

Today's Odd 'un ...

Fosdex multi-tracker / mixer with HDD and CD drive landed on the bench with the reported fault of "Won't power up. When it last did, smoked"

Once opened, the problem was immediately apparent in that every secondary side electrolytic on its switchmode power supply, was bulging to a greater or lesser degree, with a couple of them gone to the point of the top having split, and spewed electrolyte - the ones behind the reported smoke, I would guess.

Having noted all the values, I removed them all (13 in total !), in readiness to do a blanket replacement. Just for sport, I decided to run the ESR meter over them to see just how bad they were. And this is where it got odd ...

A couple of them were well out of spec, as you would expect when confronted by a bulging can, but for the most part, all of them had an acceptable ESR figure, including the 'exploded' ones. So I got out the capacitance meter and ran that over them all. This time, almost without exception, all were low or very low in value, including the two split ones, which were the better part of open circuit. This was all quite the contrary of what I had expected. In many cases in the past, I have found electrolytics which exhibited a very high ESR, but whose value was just about spot on. I don't recall ever seeing it the other way round though, like this. It's actually hard to see how a cap can exhibit a perfect ESR, and yet have a value that has dropped to less than half its marked nominal.

And aside from this, how on earth can every secondary side cap fail in this way, and all at about the same time? It's not as though most of them are placed in the designers' favourite spot of next to a heatsink or power resistor. Neither is the interior of the unit particularly compact. I wouldn't have thought that the temperature inside the case would get much above ambient. I can only assume that the caps that have been used, are either very badly specced, or of extremely poor quality.

Arfa

Reply to
Arfa Daily
Loading thread data ...

Any chance that an upstream failure exposed these caps to excessive voltage? If the switch regulator's sensing feedback loop were to fail, the switcher might have "stuck on", and ended up pumping the raw incoming voltage into the secondary rails?

--
Dave Platt                                    AE6EO
Friends of Jade Warrior home page:  http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!
Reply to
Dave Platt

One or two caps develop enough ESR to cause a regulation problem, duty cycle increases to compensate for the "low" DC level, the actual DC runs away, caps are exposed to over-voltage, then vent.

Just like old Panasonic VCR's. Often only one line 5 volt for example, is monitored for regulation purposes.

Mark Z.

Reply to
Mark Zacharias

How about loss of a single diode, short or open and the resulting AC applied to the caps.

Reply to
tm

All good points from everyone. I will find out later today if any additional probs, as I did not have enough 470uF caps in stock, so had to order some more. I'm leaning towards Mark's theory, because I've seen it before now I come to think about it, and interestingly, the caps that had failed most badly, were a pair of parallelled 1200uF's on the main rail, which is probably the one used as the reference for the control loop. If that is the case, then here's hoping that the rest of the rails didn't go too wild before it all fell over completely. I have seen downstream circuitry damaged beyond sensible rescue, by this kind of failure in cheapo supermarket DVD players employing really cheap and nasty power supplies ... :-\

I'll let y'all know ...

Arfa

Reply to
Arfa Daily

I'm reminded of the brain dead day I had a modem with a blown bridge in the powersupply. Was four individual diodes, and I grabbed what I thought were four standard Si diodes with the right current ratings.

To teach me to not think, fate substituted 5.1 volt zener diodes for what I thought were rectifiers, and the results were, shall we say, unexpected!

The only good thing was that once I replaced the zeners with regular diodes, replaced the now blown regulator with a new one, it all came back to life--but you can be sure I don't use poorly marked parts without carefully checking to see what they are.

I have to agree, I think something other than just defective capacitors is at work here--likely either an overvoltage or A/C is getting into the caps and killing 'em.

Reply to
PeterD

Thanks for reminding me of those. Well not really thanks but more like DAMMIT! Turned me off from Panasonic electronics for a long time.

--
Live Fast, Die Young and Leave a Pretty Corpse
Reply to
Meat Plow

I recently repaired a couple of powerline LAN adapters with about the same defect, secondary SMPS electrolytics bulging (and SMPS not working anymore however). I always test for low ESR all electrolytics and all tested good, including the bulging ones. Replacing them however revived the SMPS. I didn't test them for capacity, I just remember I was surprised by this result. Indeed the first time a bad electrolytic tests good with the ESR meter with me. I tend to believe the original capacitor were of very low quality (and/or the design of the device was indeed bad) as both the units failed about in the same week in their first summer of life (T > 30C here in the summer). Regards

Frank IZ8DWF

Reply to
messineo

"Arfa Daily" wrote in news:yEKxo.10232$ snipped-for-privacy@newsfe21.ams:

I had a TEK 1700 switcher supply that regulated on the +5V rail,and when the +5V filter cap's ESR went high,the other rails drove up,for example the

+40 went to +63,and then the HV osc xstr running off the +40 began cooking the PCB around it's leads until it because a short and caused the PS to current limit(burst mode).Many of the secondary PS electrolytics were cooked to some degree.

Some other TEK instruments had similar problems.(all switchers)

--
Jim Yanik
jyanik
at
localnet
dot com
Reply to
Jim Yanik

"tm" wrote in news:ia8267$id5$ snipped-for-privacy@speranza.aioe.org:

usually a shorted diode causes a current limit problem. (burst,"click" or "chirp" mode) I haven't seem many open secondary rect. diodes in switchers.

--
Jim Yanik
jyanik
at
localnet
dot com
Reply to
Jim Yanik

OK. Here's the update ...

Replaced all the caps, and that got the PSU trying to start, which was a step in the right direction. Closer examination then revealed something that I hadn't earlier spotted - a small area of PCB was a bit 'charred' looking on the top side, close to a little TO92 shunt regulator. A check underneath revealed that this had been caused by two 0603 sized sm resistors, both of which were burnt up. They were connected to the shunt regulator IC, which in turn, had the feedback opto connected to it. There was no way to determine the values of the burnt up resistors, so I went online to look for schematics, with no great confidence that I was going to find any. Imagine my surprise then, when I found a complete service manual being offered. Unfortunately, it was a pay site, but one month's access was available for 5 USD so, much as it went against the grain, I reached for my PayPal account, and signed up.

Considering that this is a Chinese manufactured piece of kit, and the general standard of service manuals these days, I have to say that I was pleasantly surprised. The manual is spot on in terms of both content and production quality, with very clear diagrams, layouts, and parts lists.

Having this allowed me to obtain the values of the two resistors, and to see that they are in fact, the voltage set resistors for the shunt regulator. Hmmmm ...

I fitted two new resistors, and ran the psu up again in isolation. This time, the supply started up, and produced a nice steady standby voltage of dead on +12v, as indicated on the schematic. So, with some trepidation, I hooked it back into the unit. Upon hitting the standby / on button, the hard drive coughed a couple of times, and the power supply tripped. Disconnecting the HDD allowed the power supply to come fully up, and the unit then worked otherwise normally, including the DVD writer drive.

Casting around the junkbox, I came up with four IDE HDDs , three very old, and one not so old. I'm not sure of the status of any of them. The one originally fitted to the unit is a 40 GB Western Digital. Of the drives I had, the three very old ones were all much smaller than that, and the newer one, much bigger at 250 GB. If you connect any of them, they run, and the power supply stands up OK. Two of the three old ones produce an error message on the display, and ask to be formatted. However, when you start this process off, it seems to begin to happen, then just hangs. The 250 GB drive never asks to be formatted, nor does the menu allow you to select that option.

So now, I'm in a bit of a quandry. Clearly, the original HDD has been twatted by the failed PSU. But the question now is, does the unit require a specific make and size of drive in it ? Or is there other damage as well to the IDE interface chip ? I'm inclined to think not, as the DVD drive is a slave on the same bus cable, and that seems to work. I've come across other units which employ HDDs, that read the device identifier from the drive, and will only run with the right make and model in them, but again, I'm disinclined to think that this is the case here, as it does at least try to start a format on two of the drives.

So tomorrow, I'm going to ring my mate with the computer shop, to see if he's got a redundant 40 gig WD drive kicking about that he can let me have to try. I seem to recall that the early SKY + sat boxes were fitted with these also. Given the time and trouble that this unit has already had invested in it, and the fact that it is now *nearly* fixed, I'm loathe to give up on it, but it's really not going to stand much more time and effort ...

Arfa

Reply to
Arfa Daily

What model? My guess would be MR8HD.

If so, it's FAT32 and the internal formatter should work. Just plop in a drive, do the format, and you're done. There's probably an upper limit to the size drive it will handle. 250GB is probably too new and too much.

No. Nothing specific. Find the model number on the blown WD drive and see if it's some kind of special.

Keep going, you're doing nicely. However, I hate to see the bill for this repair.

--
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558
Reply to
Jeff Liebermann

Yes Jeff. All my thoughts exactly. The original drive is definitely bad. I unscrewed its board, and a pinhole is blown in the top of what is probably the platen motor drive chip. The foam padding above it has a nice burn mark ...

It's a "Caviar" series, specifically model WD400BB. My mate's shop should be open now. I'll call him in a minute, and see what he's got.

The bill for this repair is a bit of a tricky one. The phone number that the shop took down for the owner is wrong, so they can't get hold of him, and he hasn't rung chasing it yet. Based on the value of the unit (not huge, but enough that you wouldn't want to have to replace it from new again if you could help it) and also what appeared on the surface to be wrong with it, the shop have already made an 'executive' decision and given me an 'up to' price. If I can pick up a second hand drive that works in it from my mate, then I can probably still scrape into that figure, but it's gonna be close. I'm sure with all your experience in the service field, that like me, you come across situations like this all the time where it's a case of gritting your teeth and carrying on for 'just a few more minutes ... ' The trick is to determine when you really are swimming against a retreating tide. This one is getting close to that :-|

Arfa

Reply to
Arfa Daily

Just guessing of course, but I would think a hard drive limitation of

137GB (Greater than this requires LBA48 support) might be where the ceiling is on the piece you're working on. If you can't get the exact model you need, try using any drive between 20GB and 120GB. Also, see if the 250GB drive you have has a capacity limit jumper.
Reply to
JW

The IDE controller and the BIOS of the unit probably doesn't do 48 bit LBA so you might try to jumper a drive to limit its size to 137 GB or less. My Fostex D90 won't see anything past 4.3GB even with a 10GB installed but it still operates at 4.3GB. But then there is no limitation of 48 BIT translation as the old drives less that 137 GB use 28 bit logical block addressing. Just a thought.

--
Live Fast, Die Young and Leave a Pretty Corpse
Reply to
Meat Plow

Well, it's all a bit academic now ... My mate had just had a huge clearout, and dumped all his 'recovered' drives, but he rang another friend who is a PC recycler, and he said that he had 40 gig drives coming out of his ears, and I was welcome to one for free, if I wanted to go over to his place to pick one up. So I made the 20 mile round trip, and picked a drive up. I hung it in the unit, got the error message with a request to format the drive, started that going, and Lo ! - it did it in about 2 seconds flat. Result ! I thought ...

However, with an input on any channel, there was no response - no signal presence light, no peak light, no dancing segments in the display. A few voltage measurements determined that the PSU was not outputting any +15v to the analogue input board, so back out it came. 4R7 sm resistor in the feed to the +15v regulator transistor was open. The transistor itself was also short circuit, and the zener. After replacing these, the +15v was back, but still no response from the analogue board, so I guess that the opamps on there had probably been twatted as the psu failed. This brought me to the point of serious debate as to whether it was worth carrying on. This issue was promptly resolved by a slip of the meter probe when the phone rang and startled me. The +48v phantom feed goes up to the analogue board, along with

  • / - 15v and +5v. Guess which pin the 48v is right next to ...? Yep, the
5v, and the very next place it goes is the hard drive. "Squeal - squeal - squeal " goes the replacement HDD, as the mouse that runs it inside, is electrocuted ...

With a sigh, I reached for the pot of screws, and put it all back together. Some faults are just not intended to be repaired. To be honest, I had reservations about this right from the start, but when work is quiet, and you are presented with what looks like a straightforward problem, it's easy to get sucked in. Note to myself. Trust your instincts, boy ... :-(

Arfa

Reply to
Arfa Daily

Wow. Like you said, it seems some repairs are just not meant to be...

I once traced down a bad resistor in a Yamaha protection circuit voltage divider, which had resulted in falsely triggering the protect circuit. Lots of time and labor expense.

A couple weeks later, the customer's house had a lightning strike, the receiver was severely damaged.

Just not meant to be...

Mark Z.

Reply to
Mark Zacharias

I'm sure we've all experienced the same situation. Multiple times. One in recent past for me was a Yaesu FT847 HF/VHF/UHF amateur transceiver. Suffered a proximity lightning strike down the antenna. The diplexer taking the single downlead combo vertical VHF/UHF antenna splitting it to a UHF N and SO239 on the set took most of the charge. However the set had no VHF RX. I spent what seemed like weeks trying to repair all the damaged components. I did manage to get about half of the RX sensitivity back though before giving up and putting the mess back together. We ended up sending it back to Yaesu-Vertex where obviously they could just replace the whole VHF RX innards. Pissed me off to spend all that time and only get half the receive back. And back then I had some decent test equipment including a Marconi 2955. I was pretty popular with the local amateur radio community repair-wise. Just renewed my FCC Tech ticket. Into year 21 now :)

--
Live Fast, Die Young and Leave a Pretty Corpse
Reply to
Meat Plow

In article , Jim Yanik writes

Had the same problem with a Compaq DLT7000 drive. The PSU's attempts to maintain the 5V line with a leaky cap resulted in the 12V line reaching

22V and cooked the drive. 
--
Mike Tomlinson
Reply to
Mike Tomlinson

Yep. Just as you think that you've got the measure of switchers, and the downstream problems that faults on them can cause, along comes one like this to slap you back down ...

Still, the rest of the week wasn't actually too bad in terms of number of items fixed, the repair values of them all, and cash in as a result. And to get next week off to a good start, yesterday a box of boards from a company that I do repairs for, turned up out of the blue, so that's a half day's work for about a day and a half's money, and then today, I picked up enough repairs from the shops that I do work for, to fill the back of the car, plus the back seat, plus the front passenger seat, and the foot wells in front of both. Now all I've gotta do, is unload it all, and book it all in ... d:~}

Arfa

Reply to
Arfa Daily

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.