That's spot on with no aerial connected, was that the case? Hopefully it corrected to the right time when you connected the aerial. Glad it worked for you.
Yeah, I was fixing it on the kitchen table so plugged it into the mains in the kitchen without an earial just to see what would happen. It's now back in the bedroom working like a charm (I have a 9300t in the living room) for the good wife to tape all that quality programming the Five channels have on offer!
Yeah, I was fixing it on the kitchen table so plugged it into the mains in the kitchen without an earial just to see what would happen. It's now back in the bedroom working like a charm (I have a 9300t in the living room) for the good wife to tape all that quality programming the Five channels have on offer!
This thread should definitely be stickied...
Thanks again Les.
Yes I thought that would be a good idea now people are reporting success, does a moderator do it if you ask nicely.
Thanks, I figured it out, it wasn't hard. Will have to wait ill tomorrow till I get some cleaner though. The back of the clock board seems a bit sticky in patches. I guess that's what you're cleaning.
Success! I did have a couple of goes at it. The first time I powered back up, no clock. I went back and reread the instructions, concentrated my efforts around pin 1 of the RTC device and hey presto, the next time I powered up I had a clock. That's got to be worth a few saved pennies in electricity. The box has been on 24/7 for the last year!
Success! I did have a couple of goes at it. The first time I powered back up, no clock. I went back and reread the instructions, concentrated my efforts around pin 1 of the RTC device and hey presto, the next time I powered up I had a clock. That's got to be worth a few saved pennies in electricity. The box has been on 24/7 for the last year!
Glad you got it going in the end. My instruction to pay particular attention to the area between the chip, super capacitor and crystal focuses on the critical connection between pin 1 of the chip and the crystal, but don't neglect the rest of the area I mention. I mentioned in my initial post that I could see the oscillator getting stronger with each cleaning process.
This is a fantastic piece of work. I dont have an issue with the clock at the moment "touch wood" but this guide will be my first port of call should things go awry - well done!
I just wanted to add my thanks and a report of my experience to this thread.
I have had a failing clock for a long time now but never really knew when it started as I don't use standby.
With v23 I decided I should start using auto off and on at least which is when I noticed the dodgy clock.
I followed Big-les's instructions to clean around the clock circuits but I still had no clock after about 6 iterations. I did notice though that there was some sort of glue holding the crystal to the board which had all but vanished and the legs seemed a little loose so I decided to have a go at resoldering them.
Hey presto, the clock sprung back to life and I can use standby safe in the knowledge that it will still record things.
Thanks so much for the research and detailed information which led me to finding my (related) problem. It also gave me the opportunity to give the insides a thorough cleaning from the years of dust build up and put back the official Humax drive which has now been thoroughly checked and low-level formatted by my PC.
Thanks Big-les, I got a tin of I-P-A from Maplin (alcohol that is, not India Pale Ale) and cleaned the board as you indicated. Noticed that the I-P-A cleaned off the conformal coating from the back of the board - oops. Anyway, after a month or two without the clock, we now have it back again and it has been three days now without fail. Good stuff! Was about to send £30 or so to Humax for a new clock board. Incidentally, do you have any idea why the Humax now shows E-OS every morning? Has it been looking for the firmware upgrade perhaps? Or do I have a new fault? Once again, many thanks for your research and remedy.
Cleaned my clock board today and it's now working fine. I did notice some whitish patches on the clock board (both sides). Hadn't opened the box up before and it's a dead easy job to do, so if anyone's worried about it - don't. Gave the inside a good clean out of all the accumulated dust while I was at it. The fan was quite gunked up.
Just got to wait until tomorrow now for the v23 OTA and hopefully get the new software installed.
Well done Big-les and a big thank you for giving us this.
I know this is probably getting repetitious, but I just wanted to add my enormous thanks to Big-Les for this information.
My Hummy has been failing to switch on for some time and I'd started to blame the children! I saw this info and thought I'd have a go. I was enormously sceptical when I removed the clock board - I could see no deposits on the board which really seemed in very good nick. However, after dutifully following the instructions I re-assembled everything and hey presto the clock is back in stand-by.
You are a star and I think your saving a lot of people here.
Myself and a friend at work both have the same humax box and funny both started showing the --:-- over the same weekend.
Makes you wonder hey ...
Anyway just thought I would drop this forum a line to let you know I too have had success following your instructions.
I did not have any spirits etc, but did find I had some old cd cleaning fluid which contained "ispropyl alcohol".
I gave the clock board a good four / five sprays and wipes with the paint brush and cotton buds on both sides.
Put it all back together and hey presto the hummy droped into standby with the clock showing.
So far all week she's still working and recording again whilst in standby, yipppeeee :-)
Fingers crossed it keeps going.
Thanks again for your extreamly usfull document (which I have save and printed now just in case).
Managed to have a look at my 'donated' PVR9200T today. Following the advice from Big-Les I scrubbed away at the PCB for some time. The cleaning alone wouldn't restart the oscillator, but removing the crystal from the original position and soldering it directly to pins 1&2 of the clock IC then drilling out the two through board links (they look like little blue rivets) that are used to connect the PCB tracks on both sides of the PCB from pins 1&2 of the IC to the crystal did work.
As I (still) have a tendency to suspect the 'super-capacitor' of leaking electrolyte and perhaps causing (or at least contributing to) the contamination problem, I removed this and replaced it with a 2200uF 10V standard electrolytic capacitor which I mounted next to the USB socket on the front panel, holding it in place with cable ties. This provides (in limited testing only...) at least 30 seconds (as long as it took me to screw the front panel and hard disk back into the unit!) of backup power to the clock. This will cover very short interruptions to the power supply, but will obviously not provide much security if the power were to be off for much above one minute.
Without the instructions of Big-Les, I would still be contemplating my first plan of removing the original clock IC, crystal and backup capacitor, then making a little PCB to put a new clock IC, crystal, backup capacitor and associated parts to then connect in place of the originals.
Myself and a friend at work both have the same humax box and funny both started showing the --:-- over the same weekend.
Makes you wonder hey ...
Regards
Sean
I agree that the timing seems suspicious. My 00:00 problem occured immediately after performing the v23 update. My OTA update failed and I had the E... error onscreen each morning, so run the update by download to the box. After the update, the clock showed 00:00 in standby. I've been on holiday since then and having missed all scheduled recordings while away, I was raging, so googled it expecting a v23 related glitch.
Anyway thanks very much to Big Les - i had to try a few times but got there eventually.
So does the time taken to dirty enough from purchase just happen to equal the time it took Humax to release this software update?
I have to say though, the whole box was heading for the bin prior to this update - it really was SLOOOOOWWWW. The fix does appear to have made things a lot better.
This year my clock has been intermittently changing to --:-- sometimes after a few minutes, earlier in the year it took a few hours. Gradually though the year it has been getting worse and worse. I tried correlating it with what channel I left the Humax on when I put it in standby, whether the program guide was fully populated when I has put it in standby and so on to no avail. Who knows, it now probably turns out it was something as weird as how hot the box was or the humidity or something!
After the software upgrade the clock displayed --:-- immediately after the program guide had been copied to disk on going to standby. Every time without fail. Drat.
Anyway, a couple of weeks ago I tilted the front panel back, but didn't have the disassembly instructions to pull out the ribbon cable so I cleaned round the left hand third of the back of the board with a few swabs of surgical spirit. The first swab definitely had some kind of brown/yellow deposit on it. After this the clock showed the time for about 18 hours after switch off, but then it reverted to --:--. A definite improvement, but no solution.
So after appealing for the disassembly instructions here, I fully removed the clock board and gave the front of it a clean, concentrating round the crystal and clock chip. Again, the first swab definitely had a smudge of gunk on it. Five swabs later I ressembled it. Success. The clock has held the time all week.
So well done big-les for spotting this and, yes, the more thoroughly you clean the better the result.
This is very interesting, my clock has started indicating --:-- in the last 3 or 4 days or so. Like in an earlier post, I wonder if our recent spell of hot weather has been the catalyst for failure
I will get into the unit over the next few days, take some high quality macro photos and report on proceedings as per big Les' instructions
With my 30+ years of electronic engineering behind me I opened up my 9200 for not the first time. Must say, it's well put together.
Pulled forward the front panel and immediately noticed patches of white dusty like corrosion on the back of the display panel pcb. Standing back from this it looks like these patches may correspond with the fingerprints of the person who installed the board! Maybe the fingerprints over time and humidity have reacted with the conformal coating on the board and the copper pcb strips beneath. There were no such patches on the upper component side of the board.
As a test, and to absolutely prove Les's theory, Using Isopropyl alcohol, I only initially cleaned the area of tracks and pads from where the crystal connects to the RTC chip. Hey presto on reassembly, the clock started working again!
Looking in the Dynamic Characteristics section, the required parameters for the Crystal oscillator are shown. It is my belief, In supporting Les's findings that the contamination present on the board is presenting additional capacitance to the crystal oscillator and damping and ceasing the oscillations, thus stopping the clock chip from operating.
I have taken some nice macro photograps that show the contamination as a white furry substance, and will hopefully post these up in the not too distant future.
Front panel pulled down, here's the back of the RTC Display board. Are they the ghosts of fingerprints I see before me?
Contamination at it's worst... looks like a bacterial growth!
Here's the topside of the area of the circuit we're really interested in, The large capacitor, Real Time Clock IC, Crystal, Infrared Receiver and Power on button can all be seen
And here's the bottom side where the crystal legs are soldered in showing the contamination that needs most attention.
.....
Looking in the Dynamic Characteristics section, the required parameters for the Crystal oscillator are shown. It is my belief, In supporting Les's findings that the contamination present on the board is presenting additional capacitance to the crystal oscillator and damping and ceasing the oscillations, thus stopping the clock chip from operating.
I have taken some nice macro photograps that show the contamination as a white furry substance, and will hopefully post these up in the not too distant future.
Vic
This was certainly not the case with the two boards I looked at, in fact pin 2 of the chip is already loaded with 10pF (well over 100pF is needed to stop the oscillator) and you can load pin 1 with additional capacitance to 'pull' the crystal frequency. As I said in my instructional post, there were measurable voltages on the surface of the boards that I looked at and it was a negative voltage around pin 1 that was stopping the oscillator. I confirmed this in two ways, firstly I simulated the problem on a bench circuit and secondly I biased pin 1 on Coulrophobe's board back to what it should be before I removed the contamination, the results were consistent and conclusive. The only theory I have is how the rogue voltages are being generated, everything else is fact. The protective coating on the boards remained intact and there was no conduction to the copper tracks below.
Your pictures show considerable contamination compared to the boards I looked at, in fact Martin's board gave no visual clues at all.
My 4 year old machine started showing the odd occasional manifestation of this problem around 2 years ago now which probably explains some missed recordings over that time. I guess the conformal coating is becoming conductive before visual signs of breakdown appear. At least, tanks to your work, Les we know whats going on here!
Comments
Yeah, I was fixing it on the kitchen table so plugged it into the mains in the kitchen without an earial just to see what would happen. It's now back in the bedroom working like a charm (I have a 9300t in the living room) for the good wife to tape all that quality programming the Five channels have on offer!
This thread should definitely be stickied...
Thanks again Les.
Yes I thought that would be a good idea now people are reporting success, does a moderator do it if you ask nicely.
Success!
Click on the alert button and ask nicely.
Just as I sat down to ask for this thread to be pinned, it was pinned. Thank you moderators.
Glad you got it going in the end. My instruction to pay particular attention to the area between the chip, super capacitor and crystal focuses on the critical connection between pin 1 of the chip and the crystal, but don't neglect the rest of the area I mention. I mentioned in my initial post that I could see the oscillator getting stronger with each cleaning process.
I have had a failing clock for a long time now but never really knew when it started as I don't use standby.
With v23 I decided I should start using auto off and on at least which is when I noticed the dodgy clock.
I followed Big-les's instructions to clean around the clock circuits but I still had no clock after about 6 iterations. I did notice though that there was some sort of glue holding the crystal to the board which had all but vanished and the legs seemed a little loose so I decided to have a go at resoldering them.
Hey presto, the clock sprung back to life and I can use standby safe in the knowledge that it will still record things.
Thanks so much for the research and detailed information which led me to finding my (related) problem. It also gave me the opportunity to give the insides a thorough cleaning from the years of dust build up and put back the official Humax drive which has now been thoroughly checked and low-level formatted by my PC.
Here Here!
I'm off to my local tip to get some free supposed write offs!
Cleaned my clock board today and it's now working fine. I did notice some whitish patches on the clock board (both sides). Hadn't opened the box up before and it's a dead easy job to do, so if anyone's worried about it - don't. Gave the inside a good clean out of all the accumulated dust while I was at it. The fan was quite gunked up.
Just got to wait until tomorrow now for the v23 OTA and hopefully get the new software installed.
Well done Big-les and a big thank you for giving us this.
David.
My Hummy has been failing to switch on for some time and I'd started to blame the children! I saw this info and thought I'd have a go. I was enormously sceptical when I removed the clock board - I could see no deposits on the board which really seemed in very good nick. However, after dutifully following the instructions I re-assembled everything and hey presto the clock is back in stand-by.
Once again, massive thanks Big-Les.
I think at this stage all reports of success or failure are valuable. Keep them coming.
You are a star and I think your saving a lot of people here.
Myself and a friend at work both have the same humax box and funny both started showing the --:-- over the same weekend.
Makes you wonder hey ...
Anyway just thought I would drop this forum a line to let you know I too have had success following your instructions.
I did not have any spirits etc, but did find I had some old cd cleaning fluid which contained "ispropyl alcohol".
I gave the clock board a good four / five sprays and wipes with the paint brush and cotton buds on both sides.
Put it all back together and hey presto the hummy droped into standby with the clock showing.
So far all week she's still working and recording again whilst in standby, yipppeeee :-)
Fingers crossed it keeps going.
Thanks again for your extreamly usfull document (which I have save and printed now just in case).
Regards
Sean
As I (still) have a tendency to suspect the 'super-capacitor' of leaking electrolyte and perhaps causing (or at least contributing to) the contamination problem, I removed this and replaced it with a 2200uF 10V standard electrolytic capacitor which I mounted next to the USB socket on the front panel, holding it in place with cable ties. This provides (in limited testing only...) at least 30 seconds (as long as it took me to screw the front panel and hard disk back into the unit!) of backup power to the clock. This will cover very short interruptions to the power supply, but will obviously not provide much security if the power were to be off for much above one minute.
Without the instructions of Big-Les, I would still be contemplating my first plan of removing the original clock IC, crystal and backup capacitor, then making a little PCB to put a new clock IC, crystal, backup capacitor and associated parts to then connect in place of the originals.
Many thanks to Big-Les.
Big-Les you are a genius.
(I used EXPO white board cleaner which contains Propan-2-oL & 2-Butoxyethanol)
I agree that the timing seems suspicious. My 00:00 problem occured immediately after performing the v23 update. My OTA update failed and I had the E... error onscreen each morning, so run the update by download to the box. After the update, the clock showed 00:00 in standby. I've been on holiday since then and having missed all scheduled recordings while away, I was raging, so googled it expecting a v23 related glitch.
Anyway thanks very much to Big Les - i had to try a few times but got there eventually.
So does the time taken to dirty enough from purchase just happen to equal the time it took Humax to release this software update?
I have to say though, the whole box was heading for the bin prior to this update - it really was SLOOOOOWWWW. The fix does appear to have made things a lot better.
Cheers,
Grazer
Did it show 00:00 or --:-- ?
After the software upgrade the clock displayed --:-- immediately after the program guide had been copied to disk on going to standby. Every time without fail. Drat.
Anyway, a couple of weeks ago I tilted the front panel back, but didn't have the disassembly instructions to pull out the ribbon cable so I cleaned round the left hand third of the back of the board with a few swabs of surgical spirit. The first swab definitely had some kind of brown/yellow deposit on it. After this the clock showed the time for about 18 hours after switch off, but then it reverted to --:--. A definite improvement, but no solution.
So after appealing for the disassembly instructions here, I fully removed the clock board and gave the front of it a clean, concentrating round the crystal and clock chip. Again, the first swab definitely had a smudge of gunk on it. Five swabs later I ressembled it. Success. The clock has held the time all week.
So well done big-les for spotting this and, yes, the more thoroughly you clean the better the result.
With the clock just on the point of failing then this is the explanation I would put forward.
I will get into the unit over the next few days, take some high quality macro photos and report on proceedings as per big Les' instructions
Vic
With my 30+ years of electronic engineering behind me I opened up my 9200 for not the first time. Must say, it's well put together.
Pulled forward the front panel and immediately noticed patches of white dusty like corrosion on the back of the display panel pcb. Standing back from this it looks like these patches may correspond with the fingerprints of the person who installed the board! Maybe the fingerprints over time and humidity have reacted with the conformal coating on the board and the copper pcb strips beneath. There were no such patches on the upper component side of the board.
As a test, and to absolutely prove Les's theory, Using Isopropyl alcohol, I only initially cleaned the area of tracks and pads from where the crystal connects to the RTC chip. Hey presto on reassembly, the clock started working again!
Lets go deeply technical....
Here's the datasheet for the Real Time Clock chip on the circuit. http://http://www.farnell.com/datasheets/60445.pdf
Looking in the Dynamic Characteristics section, the required parameters for the Crystal oscillator are shown. It is my belief, In supporting Les's findings that the contamination present on the board is presenting additional capacitance to the crystal oscillator and damping and ceasing the oscillations, thus stopping the clock chip from operating.
I have taken some nice macro photograps that show the contamination as a white furry substance, and will hopefully post these up in the not too distant future.
Vic
Front panel pulled down, here's the back of the RTC Display board. Are they the ghosts of fingerprints I see before me?
Contamination at it's worst... looks like a bacterial growth!
Here's the topside of the area of the circuit we're really interested in, The large capacitor, Real Time Clock IC, Crystal, Infrared Receiver and Power on button can all be seen
And here's the bottom side where the crystal legs are soldered in showing the contamination that needs most attention.
Vic
This was certainly not the case with the two boards I looked at, in fact pin 2 of the chip is already loaded with 10pF (well over 100pF is needed to stop the oscillator) and you can load pin 1 with additional capacitance to 'pull' the crystal frequency. As I said in my instructional post, there were measurable voltages on the surface of the boards that I looked at and it was a negative voltage around pin 1 that was stopping the oscillator. I confirmed this in two ways, firstly I simulated the problem on a bench circuit and secondly I biased pin 1 on Coulrophobe's board back to what it should be before I removed the contamination, the results were consistent and conclusive. The only theory I have is how the rogue voltages are being generated, everything else is fact. The protective coating on the boards remained intact and there was no conduction to the copper tracks below.
Your pictures show considerable contamination compared to the boards I looked at, in fact Martin's board gave no visual clues at all.
Vic