UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Television and Video

Notices

Vintage Television and Video Vintage television and video equipment, programmes, VCRs etc.

Closed Thread
 
Thread Tools
Old 5th Jun 2022, 10:13 pm   #1
1100 man
Octode
 
1100 man's Avatar
 
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
Default 3V16/22 Clock display repair

With Mr 1977's 3v22 fully functional again, I just couldn't call it finished until the clock display was sorted.

Failing LED's now seem to be standard on these displays, so I guess that sooner or later they will all need rebuilding. Sure enough, I powered up my 3V16 only to find several bits of digit had gone missing. It's amazing what havoc just three missing bits of seven segment display cause in terms of readability of the display.

The following is a link to a thread on Vrat written by HurtyAC, who is also a member here, describing the full rebuild procedure he perfected for these displays. Full credit goes to him - I'm just trying to follow the same process to hopefully get two working displays!

https://www.radios-tv.co.uk/communit...clock-rebuild/

The plan is to tackle my 3V16 one first as a 'learning curve' before doing '77's one.

To power the clock display separately, rather than having it plugged into the main machine, it is necessary to supply it with 12V on pin 6 of its plug. Pin 5 is ground. It also has 7V AC supplied to pins 2&3. It won't work without this supply, so I rigged up a 15V transformer with a 47R resistor in each arm feeding pins 2&3. That is how the supply is derived in the machine.

The following pictures show just how bad it looks! Believe it or not the first picture should show 1.39!

The first step will be to unsolder the display PCB from the driver PCB and then the fun will begin!

Cheers
Nick
Attached Thumbnails
Click image for larger version

Name:	DSCN0995.jpg
Views:	227
Size:	79.7 KB
ID:	258467   Click image for larger version

Name:	DSCN0996.jpg
Views:	215
Size:	79.8 KB
ID:	258468  

Last edited by 1100 man; 5th Jun 2022 at 10:25 pm.
1100 man is offline  
Old 6th Jun 2022, 1:54 pm   #2
dj_fivos_sak
Heptode
 
Join Date: Mar 2015
Location: Lefkada, Greece
Posts: 969
Default Re: 3V16/22 Clock display repair

Hi. Thanks for sharing this. I have an HR-3330 and oddly enough none of the LEDs have failed on mine yet. Do you know what causes them to fail in the first place?

Regards

Fivos
dj_fivos_sak is offline  
Old 7th Jun 2022, 12:42 pm   #3
Welsh Anorak
Dekatron
 
Welsh Anorak's Avatar
 
Join Date: Apr 2005
Location: North Wales, UK.
Posts: 6,884
Default Re: 3V16/22 Clock display repair

It's very rare to see one that's OK - you're lucky Fivos!
I think that they used quite advanced (for the time) SMD LEDs which perhaps didn't have a very long life. However they did seem to work for years, but when left off for a while and reconnected used to fail.
You'll see the VRAT repair reduces the drive to the new ones which will mean a long life.
__________________
Glyn
www.gdelectronics.wales
Welsh Anorak is offline  
Old 8th Jun 2022, 10:38 am   #4
1100 man
Octode
 
1100 man's Avatar
 
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
Default Re: 3V16/22 Clock display repair

Hi Fivos,
I was completely unaware of the problem until Mr'77 sent me the link to the repair thread by Hurty. He also said his own 3v22 display was faulty. I dug out my 3v16 machine and sure enough, its display had lost several 'bits'. It had worked fine when last powered up about six years ago, so it seems that it's age rather than use that's causing the failures.

Anyway, I've now stripped the display down to reveal the PCB with the LED's on. This involved unsoldering the resistor array both from the driver board and the display board. The driver board end was easy enough as it's single sided, so unsoldering with braid worked well. However, the display board end uses double sided print with plated through holes. In the end, I decided that the resistor array was going to be replaced anyway, so I cut all the legs off. This allowed me to unsolder each leg separately.

The purple plastic cover was then removed. This is held to the board by plastic pins with the ends melted over, so these had to be cut off. It will be reattached with some epoxy.

This revealed the plastic 'lens' that forms the actual segments of the display. This is also held on by plastic pins with the ends melted over.

What surprised me was that between the PCB and the lens was a layer of what looked like clear silicone sealer. This was also between the top of the lens and the purple cover. It peels off easily enough, but what is it there for? Is it to keep out dust or does it have some optical property to improve light transmission between the LED's and the lens?

The PCB with the faulty LED's could now be studied in detail. I was expecting to see tiny surface mount packages much like the new LED's, but it looks like the LED's have been formed directly onto the board. On each pad there is the tiniest pin head of the LED and a just discernible hair like wire joined to the adjacent pad. There is no package of any sort around it, just a covering of clear epoxy over the top. I wonder if it's the connecting wires that are fracturing with the thermal movement of the epoxy? There are hairline cracks in the epoxy covering.

So the next stage is to dissolve the epoxy coating so the LED's can be removed. Acetone was specified as the solvent of choice and I discovered that this can be purchased from Boots the chemist. They sell it as 'household solvent' in 50ml bottles kept hidden away under the counter!

The second picture shows the size of the new LED's compared to a standard size one from the '70's - I wonder how many I'll lose!! They really are insanely small!

Cheers
Nick
Attached Thumbnails
Click image for larger version

Name:	DSCN0998.jpg
Views:	176
Size:	122.6 KB
ID:	258600   Click image for larger version

Name:	DSCN0999.jpg
Views:	171
Size:	49.3 KB
ID:	258601  
1100 man is offline  
Old 8th Jun 2022, 11:48 pm   #5
19Seventy7
Octode
 
19Seventy7's Avatar
 
Join Date: Jun 2018
Location: Essex, UK.
Posts: 1,306
Default Re: 3V16/22 Clock display repair

Bleedin' eck, those LEDs are small! I never ever noticed the new LED in that photo until you pointed it out!

I hope it's not too challenging of a repair that it's not enjoyable for you!

I definitely agree, after revisiting Hurtys post on VRAT the look of a working clock really does round the look of the VCR off nicely! I was happy to live with the broken segments but at the same time I'm glad you offered to do the clock, so a massive thank you again as I can't imagine it's the most fun of jobs out there!

Thanks
'77
19Seventy7 is offline  
Old 9th Jun 2022, 6:50 pm   #6
1100 man
Octode
 
1100 man's Avatar
 
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
Default Re: 3V16/22 Clock display repair

Quote:
Originally Posted by 19Seventy7 View Post
I can't imagine it's the most fun of jobs out there!
Ahh that's where you're wrong!! It's very rewarding because it's a real challenge for someone with very limited surface mount experience like me The SM resistors I used on the Hedghog look like house bricks in comparison to these new LED's!

Before starting to dissolve the glue with Acetone, I thought it would be a good idea to work out which pins did what on the display PCB. Also which way round the new LED's were to be mounted was rather important too! I really didn't want to mount 40 of the things the wrong way round!!

It would seem that if you put the red lead from my digital meter onto the back end of a diode and the black lead on the plate end, it will conduct. I've always used an analogue meter for testing diodes & transistors which works the other way round, just to confuse me!

The new LED's have a very faint green band at one end which marks the front (plate) end. Putting the black lead on this end and the red on the other makes it light up quite brightly.

The picture shows the connections to the display PCB. There are two common lines, A & B. The bits of digits 1 & 2 share the same group of seven pins. Digit 1 will light if common A is used and digit 2 will light if common B is used. The same applies for digits 3 & 4 and for the top row of LED's which indicate the days of the week.

So by putting the red meter lead on common 'A', the bits of digit 1 that still work, will light when the black lead is placed on any of the seven 'digit 1' pins. Each of the 40 LED's can be tested in this way, so once the new ones are installed, they can all be tested before the board is soldered back onto the driver PCB. The glow from the old LED's when tested with my meter was very faint.

I left the board soaking in the Acetone for a few hours. Even though I had a clip on lid on the plastic tray, the evaporation rate was pretty high during the course of the afternoon and I didn't have much left by the end! The Acetone doesn't really dissolve the glue that well, but makes it go crumbly. I found carefully scraping the glue with a small screwdriver every couple of hours removed it quite well. It took about 5 hours in total, but the end result was a board devoid of glue and much less of the green etch resist missing than I expected! Remember the vapour from this stuff is pretty strong so do it in a well ventilated area!

The next step was to 'tin' the pads in preparation for soldering in the new LED's. I applied some surface mount flux and then a quick 'scrubbing' action with the soldering iron tinned the pads and also removed the residue of the old LED's. A quick clean up with desolder braid completed the process followed by a careful check of all the pads with a lens to make sure they were all OK.

Tonight's mission is to mount the new LED's and to see how easy or otherwise it proves to be!

All the best
Nick
Attached Thumbnails
Click image for larger version

Name:	DSCN1000.jpg
Views:	131
Size:	86.2 KB
ID:	258727  
1100 man is offline  
Old 11th Jun 2022, 10:07 pm   #7
1100 man
Octode
 
1100 man's Avatar
 
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
Default Re: 3V16/22 Clock display repair

Mounting the new LED's went surprisingly well! I only lost two which pinged out of their packaging when I peeled back the plastic film.

Once soldered, they were examined under a strong lens to check the quality of the soldering. Each LED was tested using the meter and it was a good job I did, as I had managed to bridge two tracks near to an LED.

The board was then cleaned with IPA to remove the sticky flux residue and then reattached to the driver board with its 20 x 560 Ohm resistors and two solid links.

Then the moment of truth - would it work?? - Well, actually, yes it did!

The only issue, which can be seen in the picture, is the amount of spurious light within the display which makes the bits that are not lit glow brighter than I would like. The picture does make it look slightly worse than it really is, but I wondered if I could improve it.

The LED's are very bright, even in high ambient light, so I experimented with higher value series resistors. You have to increase the value by a large amount to reduce the brightness by a small amount. 1K2 works quite well, so I'll use that value on the next one and see if it looks any better.

One other thing I realised is that on the first digit, it only ever displays a '1' or it's blank. It's a 12hr clock with an AM & PM light. Therefore, it only needs the two right hand bits that make the '1', the other 5 LED's can be left out!

Overall though, I have to consider it a success, so thanks to Adrian (Hurty) for coming up with the method in the first place.

Cheers
Nick
Attached Thumbnails
Click image for larger version

Name:	DSCN1002.jpg
Views:	134
Size:	68.6 KB
ID:	258891  
1100 man is offline  
Old 12th Jun 2022, 9:16 am   #8
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
Default Re: 3V16/22 Clock display repair

I did a similar repair to an LED display on someone's much loved old clock/radio/alarm - it worked OK but the LEDs I chose to use were much more efficient, therefore much brighter than the original ones so I had to compensate by replacing all the series resistors. The whole job was a much bigger headache than I had naively expected, as I had hoped just to source a complete pin compatible display.

Nice job on this one, well done.
SiriusHardware is offline  
Old 2nd Jul 2022, 9:06 am   #9
1100 man
Octode
 
1100 man's Avatar
 
Join Date: Apr 2016
Location: Ventnor, Isle of Wight, & Great Dunmow, Essex, UK.
Posts: 1,377
Default Re: 3V16/22 Clock display repair

I have now rebuilt the second of the two displays, incorporating some refinements in my technique.

Firstly, to remove the resistor array I cut all the legs off to leave as much 'stump' poking out of the pcb's as possible. The stumps were then easily removed using a pair of tweezers and a quick dab with the soldering iron. The residue was then cleaned up with de-solder braid.

Secondly, I was much more careful removing the display - carefully separating the layer of silicon sealer between the white plastic 'lens' and the PCB, so as to leave it as intact as possible. It would seem the lens itself is filled with silicon which acts as a diffuser. Be particularly careful to ensure the silicon doesn't come out of the two dots in the middle of the display - it tends to want to do that as there is not much to hold it in. If it does, carefully poke it back in or the display will glare where the silicon is missing.

I used 1K2 resistors this time rather than 560R, which reduces the brightness and consequently the spurious light within the display. Comparing the two displays side by side, the overall effect with the 1K2 is much more pleasing. In fact, the display is still more than bright enough, even in high ambient light, so I might even use 1K5 in any future ones I do. Whether the light output between various makes/batches of LED's varies, I don't know.

Overall though, now the machine is back together, the end result is very good and I am more than pleased with it. It was certainly worth all the effort and was an interesting learning curve!

Cheers
Nick
Attached Thumbnails
Click image for larger version

Name:	DSCN1003.jpg
Views:	107
Size:	117.3 KB
ID:	260240   Click image for larger version

Name:	DSCN1007.jpg
Views:	111
Size:	84.3 KB
ID:	260241  
1100 man is offline  
Old 4th Jul 2022, 3:25 am   #10
19Seventy7
Octode
 
19Seventy7's Avatar
 
Join Date: Jun 2018
Location: Essex, UK.
Posts: 1,306
Default Re: 3V16/22 Clock display repair

That looks brilliant! I'm really pleased with how it's turned out too, it really does complete the look of the VCR. I hope it was enjoyable for you as well and it wasn't too hard to do.

I really can't wait to use it!

Thank you again, I really really appreciate this
'77
19Seventy7 is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 2:44 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.