Weird science!

I underestimated the time it takes to prepare so many chips... I had to remove all excess solder so pins are properly flat underneath and that took a while plus all the chips were covered in flux to varying extents. Overnight soak in acetone made things clean and I could start to actually properly align the pins to allow reflowing. I used an old solder paste stencil for that, pushing pins into the openings to get them perfectly aligned. This took better part of today but I am finally done with all the preparative steps ~

Now I need to figure out how I'm going to actually test the chips... I have the test machine roughly usable but things are far from optimum. I'm not sure I'll get to do any testing today though, I have to actually set up some sound playback capability first 🤣

That is truly a labour of love, I hope the majority are working.

I like the legs of of your desk BTW 😅

badmojo wrote on 2026-10-12, 00:44:

That is truly a labour of love, I hope the majority are working.

I like the legs of of your desk BTW 😅

Perfect use for spare cases !

Even better if they are working PCs hooked up to a KVM switch !

They are fully kitted out 286, 386, 486 and a NexGen machine, but they are still wrapped up and waiting for a more proper solution... there is an 8 input KVM in the neighborhood and all the machines will connect to it once there's a better setup ~

Anyway, first YMF704 is tested and it makes pleasant sounds ~

EDIT: I forgot, there are speaker attached to the wall now. I need a subwoofer or something, those little things cannot do what stuff upstairs can 🤣

EDIT2: Testing takes longer than anticipated too, the socket is not meant for anything but new chips so I need to polish up the pins with paper prior to trying. Anyway there are quite a few bad ones :
* 10 bad ones - no sound or garbage noises
* 25 good ones
* 1 weird, most tracks play fine but few do weird things... strange fault
* 52 left to test.

This day is over, I will prepare for sleep now... *yawn*

More good than bad, nice work 👍

Tiido wrote on 2026-10-12, 17:52:

Anyway, first YMF704 is tested and it makes pleasant sounds

Thanks for bringing us along for the ride as you progress on your testing and problem-solving. To me, this is one of the most valuable things on a forum like this. Seeing how others solve peculiar problems and the solutions they create is always a good source of new ideas. It is far more useful and interesting than something like, "I tested x chips; y worked, z were garbage". I definitely appreciate the time you take to share the details.

Today I finished testing all the YMF704 chips. Total is 11 bad ones and 46 good ones, and the previously weird one is actually ok. It is the game that doesn't always update all instruments every track and coming in from a reset when everything is piano, sometimes few things stay a piano unless you go through a few more tracks. My back says no to further tests but I did want to end the day with a YMF721 making sound and I have that, but also a bad one which was the first chip I tested. One with missing pin, and it actually gets hot so it is definitely dead inside 😒

640K!enough wrote on Yesterday, 19:07:

Thanks for bringing us along for the ride as you progress on your testing and problem-solving. To me, this is one of the most valuable things on a forum like this. Seeing how others solve peculiar problems and the solutions they create is always a good source of new ideas. It is far more useful and interesting than something like, "I tested x chips; y worked, z were garbage". I definitely appreciate the time you take to share the details.

Someone possibly getting use out of this is definitely one of the goals, but I also want to ramble lololol. To polish/clean the pins I use a sheet of paper laid over something flat and in my case it was an aluminium heatsink. Irregularities can create bumps that will make pins hang up and get bent on... But some irregularity can be helpful and provide an area to get stronger activity. Sometimes there is a little solder spike under a pin and it starts digging into the paper and may even get the pin bent. In those cases I use gentle force or if I know exactly which pin then use tweezers to hold it down and make it dig so that the spike wears off as a grey mark on the paper. This dramatically increases success with the QFP adaptor, although I still have to align the chip manually in one since there is a bit too much playroom in it... sometimes pins are actually able to short to a neighboring one...

46 working chips is a great result, well done everyone involved 👍

Awesome! Significant progress on bringing more of your Yamaha sound cards into the world.

Top job from Hammy.

And now all the chips are tested. YMF721 were unfortunately lot more dead than YMF704 chips, which is a shame as I wanted them the most...

* 17 good chips that appear to work perfectly, including two I thought were gonna give problems as they had corroded legs...
* 9 bad ones that don't do anything at all
* 1 chip that plays music fine but is hot. There must be an internal short on an IO pin or something...
* 1 chip that plays perfectly for a bit and then starts making weird noises
* 1 chip that plays for a few seconds and then goes silent
* 1 chip that plays with big distortion on the output. If I am lucky I can use the PCM only output and it works ok but that is yet to be tried
* 1 chip that played for a while and then got hot and went silent, and no longer works afterwards unlike all the others that continue to work.

17+46 good chips and 25 bad, out of 88 total. Not bad considering they were loose chips in a plastic that are all salvaged from somewhere. From condition of some of the chips I imagine the hardware they came from was in quite rough shape...

Tiido wrote on Today, 21:29:

And now all the chips are tested. YMF721 were unfortunately lot more dead than YMF704 chips, which is a shame as I wanted them t […]
Show full quote

And now all the chips are tested. YMF721 were unfortunately lot more dead than YMF704 chips, which is a shame as I wanted them the most...

* 17 good chips that appear to work perfectly, including two I thought were gonna give problems as they had corroded legs...
* 9 bad ones that don't do anything at all
* 1 chip that plays music fine but is hot. There must be an internal short on an IO pin or something...
* 1 chip that plays perfectly for a bit and then starts making weird noises
* 1 chip that plays for a few seconds and then goes silent
* 1 chip that plays with big distortion on the output. If I am lucky I can use the PCM only output and it works ok but that is yet to be tried
* 1 chip that played for a while and then got hot and went silent, and no longer works afterwards unlike all the others that continue to work.

17+46 good chips and 25 bad, out of 88 total. Not bad considering they were loose chips in a plastic that are all salvaged from somewhere. From condition of some of the chips I imagine the hardware they came from was in quite rough shape...

Is it possible to provide the nonfunctional chips to someone who can take photos of the YMF704 and YMF721 dies?