Folderol wrote:Hmmm. Are you soldering chips directly into the board? Unless there are space constraints I always put them in chipholders, so there's no risk of damaging the chips in the soldering process, as you don't insert them until all the work has been done.
We soldered that one in, yes. We just put it back the way the original had been installed. A chip holder sounds like it would have been a good idea though. I didn’t realize you could add one wherever you wanted to - is that an option?
Folderol wrote:Hmmm. Are you soldering chips directly into the board? Unless there are space constraints I always put them in chipholders, so there's no risk of damaging the chips in the soldering process, as you don't insert them until all the work has been done.
We soldered that one in, yes. We just put it back the way the original had been installed. A chip holder sounds like it would have been a good idea though. I didn’t realize you could add one wherever you wanted to - is that an option?
Yes, nearly always. The makers didn't do it partly to save a few pence per chip and also because it's quicker on automated systems. Of course, this isn't possible on modern surface-mounted components
Ok we found pin 22 is shorted to ground. It’s adjacent to 23 which is supposed to be grounded. No physical evidence that they are shorted together. Pin 22 is the d3 output. We are replacing the chip again. As part of that repair we are also installing the chip socket and will use rework wire to make our jumpers run from pin to junction.
It’s possible I mishandled the chip to cause the problem. It’s also new old stock. Either way, lessons have been learned and we’ll try again!
jjlonbass wrote:A short between pins 22 (D3) and 23 (I) of the 4514 could cause the bottom segment of the displays and "ARP ASSIGN" to be stuck off. Similarly a short between pins 10 (S2) and 11 (S0) or between pins 11 (S0) and 12 (Vss) would affect keypad buttons 0 to 7.
Other combinations are possible!
Thanks nickle15; no genius involved, it's just that I've been working with electronics for a long time and have seen many of these sorts of problems.
Folderol's advice is good with regard to fitting an IC socket here, although you should fit a good quality one to avoid puzzling problems in future. Go for a "turned pin" type if you can.
If you're replacing faulty ICs, in order to avoid PCB track damage when removing, it's a good idea to cut all of the IC pins close to the body of the chip, remove the body then desolder all of the pins. If you're lucky, the pins will just drop out of their own accord with the merest touch of the soldering iron. It sounds a bit brutal, but the old chip is useless anyway.
Just a thought, the service manual shows the artwork for each of the PCBs and it can be seen that board 1 is double sided i.e. it has PCB tracks on both sides.
Double sided PCBs usually have through plated holes that join the top PCB pads to the bottom ones. However, if this board doesn't have through plated holes for all of the 4514 pins you won't be able to use an IC socket and also you'll need to solder the IC's pins to both top and bottom side pads.
You can check for this by removing the old chip and measuring the resistance between the top and bottom pad of pin 2 for U104 which only shows a blue PCB trace on the artwork, then repeat for the top and bottom pad of pin 23 which only shows a grey PCB trace. If you see roughly 0 ohms in each case then the board is through plated and you can use your IC socket.
jjlonbass wrote:Just a thought, the service manual shows the artwork for each of the PCBs and it can be seen that board 1 is double sided i.e. it has PCB tracks on both sides.
Double sided PCBs usually have through plated holes that join the top PCB pads to the bottom ones. However, if this board doesn't have through plated holes for all of the 4514 pins you won't be able to use an IC socket and also you'll need to solder the IC's pins to both top and bottom side pads.
You can check for this by removing the old chip and measuring the resistance between the top and bottom pad of pin 2 for U104 which only shows a blue PCB trace on the artwork, then repeat for the top and bottom pad of pin 23 which only shows a grey PCB trace. If you see roughly 0 ohms in each case then the board is through plated and you can use your IC socket.
John
Thanks again - we'll double check that when we get back in there.
We actually did cut the pins of the old chip in order to remove it and I'm not sure how the traces came loose. Unfortunately the pins didn't fall out of their own accord but we stood the board up and worked from both sides to remove the pins. It worked well!
I use a pair of tweezers with the tips lined with thin cardboard to pull the cut pins out from one side while heating the other side. The cardboard doesn't chill the metal, and actually lasts quite a long time before you need to replace it.
Personally I don't like the idea of soldering both sides for chips where there is broken through plating. It puts a lot of extra heat stress on the chip. With a fine tip iron it is usually possible to get in to the pins on sockets, and it doesn't matter if you melt just the edge of the plastic. Where this really is not possible, my usual approach is the thread a single strand of fine wire through the hole and tack it down on both sides with just a tiny bit of solder. This provides a path to wick the solder through when you fix the holder in place.
So, progress. We used a chip holder and got a new 4514 chip installed tonight. It took quite a while to get it all done but when we finished we fired it up and it made sound! The display was back to normal as well. However, now we don't have any sound coming from keys C2-G2. The original keys that didn't sound are working now. So even though it's not perfect it's certainly progress. We're going back in tomorrow night to check for any other damaged traces first and then go from there.
Again - thanks to all for the invaluable tips and advice through this process. It's not over yet but it was an encouraging evening to say the least!
The C2 to G2 keys are driven by output S11 on pin 19 of the 4514. It would be worth checking for continuity from the IC pin to the keyboard busbar connection for those keys.
Do you have an oscilloscope? If you do, you'll be able to observe whether you have activity on pin 19 of the 4514 and whether the logic levels here are correct.
Hopefully you don't have another duff 4514 but at least it will be easy to replace now that you have the IC in a socket.
Indeed, we located a broken trace to pin 19 and rewired it. The keyboard is finally working perfectly!!!!!!!!!!!!
I can't thank you all enough for the time and effort you put into all the thoughtful recommendations. It was a huge help to us and we are grateful!
I work with the guy that owns the 600 and brought the keyboard to him this morning. Needless to say he's very excited to have it up and running. Thanks again everyone!
Now on to the next thing - I'm looking at you Polysix...
Last edited by nickle15 on Thu Apr 15, 2021 2:52 pm, edited 1 time in total.
But ... all that pain and heartache and it's not even yours????
I hope the owner appreciates this. Pizza and beer is the very least he should be offering you!
He was very appreciative. I advised him of the cost of the parts which is all we were asking for and he said there would be more than that coming.
I learned a lot through the process and will pass along any excess to my dad who did the hard work.
My friend is in a nice spot though. Between the original tech bill and the parts we replaced he's out $200. The keyboard was given to him at no cost so if he wanted to sell at this point there's a lot of money to be made!
