This morning, I powered on my UPL analyzer only to be greeted with no hard drive detected and CMOS battery failure.
I have the recollection that when I serviced this unit last year, I changed the lithium battery on the computer board. But I'm not 100%, so I pulled the UPL out of service, pulled the cabinet off, removed the speaker and unbolted the motherboard. Took the old battery out, tested it. It shows 3.1 volts. Hmm. I MUST have put this battery in last year.
Needless to say, since I had it apart, I put in a new battery (after checking that it read over 3.1 volts) and put it all back together.
Booted it, and put in all the disk parameters and launched the analyzer software. To check things, I shut off the UPL, waited 15 seconds, then powered it back on, only find out that it lost the CMOS setup for the hard drive and was complaining still about a CMOS battery failure.
This is the only battery I can find on the entire analyzer. What in tarnation is going on here??
I have the recollection that when I serviced this unit last year, I changed the lithium battery on the computer board. But I'm not 100%, so I pulled the UPL out of service, pulled the cabinet off, removed the speaker and unbolted the motherboard. Took the old battery out, tested it. It shows 3.1 volts. Hmm. I MUST have put this battery in last year.
Needless to say, since I had it apart, I put in a new battery (after checking that it read over 3.1 volts) and put it all back together.
Booted it, and put in all the disk parameters and launched the analyzer software. To check things, I shut off the UPL, waited 15 seconds, then powered it back on, only find out that it lost the CMOS setup for the hard drive and was complaining still about a CMOS battery failure.
This is the only battery I can find on the entire analyzer. What in tarnation is going on here??
The interesting thing is that when this happens, the date and time are maintained.
This led me to explore the schematics and see if there were any other batteries and lo and behold, the schematic shows one on the Digital board. Alas, on mine, there was an empty slot where the battery should go--none was ever installed. How it every kept it's config for the past year is a question that nags me.
I happened to have a new, spare 3.6V lithium battery in stock, so I installed it. Curiously, the very next time I booted the UPL, it detected the HDD and booted to the analyzer software without me having to reprogram CMOS again.
This led me to explore the schematics and see if there were any other batteries and lo and behold, the schematic shows one on the Digital board. Alas, on mine, there was an empty slot where the battery should go--none was ever installed. How it every kept it's config for the past year is a question that nags me.
I happened to have a new, spare 3.6V lithium battery in stock, so I installed it. Curiously, the very next time I booted the UPL, it detected the HDD and booted to the analyzer software without me having to reprogram CMOS again.
There are two transistors that provide charging current and power to the VRAM. There is a 47MF 5V capacitor off a 150Ω resistor there. The schematic also shows a battery in series with 2 diodes, which connect to that common supply bus.
Maybe the capacitor was enough to keep the VRAM alive for a few days (two whole weeks last year while I was out of country), but now has dried out to the point where its capacitance is insufficient to keep the VRAM alive for even 15 seconds.
The battery is soldered in. The solder holes in the PCB were unused. The factory never installed a backup battery on this board, though the rectangle and two solder points are there for a battery.
Maybe the capacitor was enough to keep the VRAM alive for a few days (two whole weeks last year while I was out of country), but now has dried out to the point where its capacitance is insufficient to keep the VRAM alive for even 15 seconds.
The battery is soldered in. The solder holes in the PCB were unused. The factory never installed a backup battery on this board, though the rectangle and two solder points are there for a battery.
If those capacitors have little more lifespan than a lithium battery, what's the point of them? Most electrolytics get 30-50 years of good service. That said, I will give it strong consideration.
It is also curious that the schematic shows the battery, connected with the diodes and capacitor, but my unit has never had a battery installed. That part of the PCB has virgin feed throughs that have no solder on them or evidence of anything being soldered there.
This morning I powered up the UPL and it hung during "Verifying DMI pool data....."
I cycled the power and the second time it booted normally.
It is also curious that the schematic shows the battery, connected with the diodes and capacitor, but my unit has never had a battery installed. That part of the PCB has virgin feed throughs that have no solder on them or evidence of anything being soldered there.
This morning I powered up the UPL and it hung during "Verifying DMI pool data....."
I cycled the power and the second time it booted normally.
Started up the UPL this morning and it's lost its configuration again. "CMOS battery failure".
The interesting thing is the date and time were incorrect by about 20 hours. It has yesterday's date at 4:40 as the time.
Then I reprogrammed some of the CMOS back to the required settings and on the next boot was greeting with "Floppy drive failure"
I've replaced the computer backup battery (twice this year) and now also installed a new battery on the Digital board.
I have a new .47MF super cap on order. But I have a hunch there is something deeper going on, like maybe a failure of the CMOS / VRAM or something else.
The interesting thing is the date and time were incorrect by about 20 hours. It has yesterday's date at 4:40 as the time.
Then I reprogrammed some of the CMOS back to the required settings and on the next boot was greeting with "Floppy drive failure"
I've replaced the computer backup battery (twice this year) and now also installed a new battery on the Digital board.
I have a new .47MF super cap on order. But I have a hunch there is something deeper going on, like maybe a failure of the CMOS / VRAM or something else.
Their standard service fee is over twice what I paid for the instrument.
I've attempted to disassemble to the point where changing the super capacitor would be possible, but decided the risk would be too great for an item that I'm doubtful is the cause of the problem, so I abandoned efforts to disassemble. Don't want to risk breaking another display ribbon cable ($800 to fix that last time it happened) and too many other cables that appear not detachable, yet have no slack. I already had the keyboard wires break 3 times in the past 3 disassembly attempts.
What I did check was the batter voltage of the battery I installed on the Digital board. It's holding at 3.6V.
I reseated the CMOS battery on the PC motherboard, too.
I reseated the floppy ribbon cable, because of the "Floppy drive failure" error on bootup.
I had to turn off floppy seek at boot in order to boot the machine.
I find that the machine hangs, such that CTRL - ALT - DEL won't restart it, when I attempt to access or write to the floppy drive. So there goes my ability to make and print reports.
I think there is something deeper going on here. Sometimes the UPL retains it's disc configuration settings overnight, sometimes not. It's inconsistent. What IS consistent is that the floppy drive no longer works.
Could it be that the floppy drive had some sort of electronic failure that is preventing CMOS from storing settings? I suppose I could try to find another floppy drive and swap it out. But if that doesn't solve it, then I've got problems on the PC motherboard most likely. An oddball, being a Cyrix WinChip 86-based one.
If the problems are a motherboard issue, I'll have to just leave it powered on 24/7 on a UPS, to maintain the setup. But I'll need to find an alternative means of writing files. Perhaps I can install some sort of USB port that I can stick a USB thumb drive for access?
What I did check was the batter voltage of the battery I installed on the Digital board. It's holding at 3.6V.
I reseated the CMOS battery on the PC motherboard, too.
I reseated the floppy ribbon cable, because of the "Floppy drive failure" error on bootup.
I had to turn off floppy seek at boot in order to boot the machine.
I find that the machine hangs, such that CTRL - ALT - DEL won't restart it, when I attempt to access or write to the floppy drive. So there goes my ability to make and print reports.
I think there is something deeper going on here. Sometimes the UPL retains it's disc configuration settings overnight, sometimes not. It's inconsistent. What IS consistent is that the floppy drive no longer works.
Could it be that the floppy drive had some sort of electronic failure that is preventing CMOS from storing settings? I suppose I could try to find another floppy drive and swap it out. But if that doesn't solve it, then I've got problems on the PC motherboard most likely. An oddball, being a Cyrix WinChip 86-based one.
If the problems are a motherboard issue, I'll have to just leave it powered on 24/7 on a UPS, to maintain the setup. But I'll need to find an alternative means of writing files. Perhaps I can install some sort of USB port that I can stick a USB thumb drive for access?
You may want to try powering the floppy drive externally if possible (make sure you get a good ground connection) - if that works, chances are the internal PSU needs recapping (maybe with a pinch of dried-up grease in the stepper motor bearings thrown in). Yeah, it's dodgy and may require some extra loading, but some people successfully use this method to power hungry graphics cards.
As for your CMOS problem, I kind of suspect a corroded trace. Maybe one or more electrolytics nearby have started spilling their guts. Supercaps sometimes do the same in old age, actually. That would mean it's only going to get worse.
As for your CMOS problem, I kind of suspect a corroded trace. Maybe one or more electrolytics nearby have started spilling their guts. Supercaps sometimes do the same in old age, actually. That would mean it's only going to get worse.
Visual inspection showed no indications of leaking caps.
I do think recapping the power supply would be prudent though. I had a musical instrument (sampler, rackmounted) have it's caps go bad after just a dozen years of 24/7 power up.
I'm not convinced that the settings for the computer are stored on the digital board. I suspect that cap is for maintaining calibration constants. I could be wrong. Just guessing there.
All traces look good. This unit came out of a medical environment (hearing aid developer that went bust), so very clean and free of corrosion.
I'll try to get this thing on the bench and put a scope to the power supply outputs looking for ripple.
I have a spare floppy drive to try out. Just need to get time to do it.
I do think recapping the power supply would be prudent though. I had a musical instrument (sampler, rackmounted) have it's caps go bad after just a dozen years of 24/7 power up.
I'm not convinced that the settings for the computer are stored on the digital board. I suspect that cap is for maintaining calibration constants. I could be wrong. Just guessing there.
All traces look good. This unit came out of a medical environment (hearing aid developer that went bust), so very clean and free of corrosion.
I'll try to get this thing on the bench and put a scope to the power supply outputs looking for ripple.
I have a spare floppy drive to try out. Just need to get time to do it.
All winter, the UPL booted up normally. It just started working again, for no apparent reason. For five months, it started faithfully every day. Then, this morning, lost configuration again. Baffling as to why it 'fixed itself' for those months and is back to malfunction.
I think I bought some time by installing a battery where a battery was missing, but maybe the super capacitor that used to store the info has not only failed to hold a charge, but perhaps is shorting out and taking down the battery power as well.
Last fall, I attempted to get to the solder side of that analog board where the capacitor is. Removing the PC motherboard was easy enough, but the analog board proved to have too many delicate cables attached, so I aborted the whole attempt, for concern of breaking something and rendering the whole instrument useless.
I was able to boot up the unit by manually entering the settings in CMOS (disable on board controllers, basically) and the unit would function. But having to do this every morning would be tedious. I can't just leave the thing running 24/7 as we already have a $700+ monthly electric bill.
I wonder if there is a safe way to change that super cap without breaking the delicate ribbon cables and other wires attached to the analog board?
I think I bought some time by installing a battery where a battery was missing, but maybe the super capacitor that used to store the info has not only failed to hold a charge, but perhaps is shorting out and taking down the battery power as well.
Last fall, I attempted to get to the solder side of that analog board where the capacitor is. Removing the PC motherboard was easy enough, but the analog board proved to have too many delicate cables attached, so I aborted the whole attempt, for concern of breaking something and rendering the whole instrument useless.
I was able to boot up the unit by manually entering the settings in CMOS (disable on board controllers, basically) and the unit would function. But having to do this every morning would be tedious. I can't just leave the thing running 24/7 as we already have a $700+ monthly electric bill.
I wonder if there is a safe way to change that super cap without breaking the delicate ribbon cables and other wires attached to the analog board?
I contacted them last August when the unit first lost its CMOS settings. They thought perhaps the PC motherboard was failing, but the particular settings that seem to be lost may be stored in the Analog PCB. I twice changed the lithium battery on the PC motherboard already. The original was reading 3V and the replacement was reading 3.25V after 4 months installed. I replaced it again in November.
Shortly after I added the battery to the Analog PCB, the unit started to remember its settings between power downs. Eventually, it became reliably able to boot up every morning, so I forgot about it. Then today, bam. No rhyme or reason. The problem started in August 2018.
It DOES seem like the PC motherboard may be forgetting SOME settings, but the clock and date are correct. What it's doing is resetting it's config to defaults, so that it can't detect hard drive or floppy drive. When on board controllers enable, the analog PCB controllers aren't detected and the unit can't boot.
But the PC motherboard battery is new. Only other thing I can think of may be a solder joint somewhere that's flaky. Could explain why it stopped working in August and started working a month later up until now.
Shortly after I added the battery to the Analog PCB, the unit started to remember its settings between power downs. Eventually, it became reliably able to boot up every morning, so I forgot about it. Then today, bam. No rhyme or reason. The problem started in August 2018.
It DOES seem like the PC motherboard may be forgetting SOME settings, but the clock and date are correct. What it's doing is resetting it's config to defaults, so that it can't detect hard drive or floppy drive. When on board controllers enable, the analog PCB controllers aren't detected and the unit can't boot.
But the PC motherboard battery is new. Only other thing I can think of may be a solder joint somewhere that's flaky. Could explain why it stopped working in August and started working a month later up until now.
I had a thought just now, thinking back to another piece of equipment that failed last summer..
I had a Kurzweil K2600RS that was taking longer and longer to self test at power on, until last August it would not get past that step in the startup process.
I troubleshot it and found the processor was receiving a reset 120 times a second. I traced that back to a bad capacitor in the power supply. The rippled was just over the limit that a raw DC sample (pre-regulator) feeding a logic gate, determined to be unusable and thus sent the pulses through as resets.
Now in the case of my UPL, the unit is functional once I put in CMOS parameters, but given its age, perhaps a capacitor somewhere has dried out enough to cause a problem that manifests in unexpected ways?
I had a Kurzweil K2600RS that was taking longer and longer to self test at power on, until last August it would not get past that step in the startup process.
I troubleshot it and found the processor was receiving a reset 120 times a second. I traced that back to a bad capacitor in the power supply. The rippled was just over the limit that a raw DC sample (pre-regulator) feeding a logic gate, determined to be unusable and thus sent the pulses through as resets.
Now in the case of my UPL, the unit is functional once I put in CMOS parameters, but given its age, perhaps a capacitor somewhere has dried out enough to cause a problem that manifests in unexpected ways?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.