• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

SE/30 Sound repair issues

hellslinger

Well-known member
Hi Guys, just going to put it out there that I'm interested in getting an ethernet card for an SE/30 if anyone has one.

I recently recapped an SE/30 and got the right sound channel working. Something to keep in mind is that I didn't have enough 47uF caps to recap the entire board, so the problems I describe later might be a result of remaining bad caps. I have already cleaned the local radio shack out of these caps, so there's no more to be had until my ebay shipment arrives.

The left channel was still out, though, so I checked all the pins on the sound chips to see if the appropriate signal was getting to them. Everything appeared to be OK except for the Left Channel  Sony Snd chip. It's getting power and PWM signal from the ASC chip, but isn't producing any analog signal on the SND pin. I checked to make sure this pin wasn't shorted.

I think that if a CMOS part has died, the power supply of the circuit is suspect. It is possible that this chip was damaged due to static while connecting headphones. I decided to continue anyway.

I have a dead SE/30 Motherboard due to corrosion of the RTC battery. Fortunately there was no corrosion near these sound chips, so I was able to desolder and remove these through-hole chips with some considerable effort. Using tinning flux and a soldering gun, I was able to get most of the solder off the pins for removal, but it still needed a bit of prying and heat. The tinning flux worked, but boy is it messy. I bought a solder sucker and cleaner flux from ebay as a result of the work.

I soldered a DIP socket to the good board so that I could try all 3 Sony chips (2 from bad motherboard). I also wanted to be able to verify that the old chip was actually bad.

I messed up by severing the trace connecting pin 3 of the internal speaker op amp (TLO71) on the negative input terminal. I patched this trace easily with some wire-wrap wire.

So here's where it gets interesting. The old Sony chip was bad. The two chips from the bad board work -- but not exactly right. When the machine is booted the sound works perfectly with both channels. I checked full volume range to see if there was distortion and there isn't. However, the boot chime is distorted, like it is overdriven and clipping. This happens on reboot or power off just the same.

The older damaged board has a slightly different layout than the good one, theres a few surface mount parts in slighty different spots. The good board has a socketed 68030RC and the bad board has it soldered in.

Accordingly, the Sony sound chips have slightly different numbers on them between the board revisions. The resistor values in the sound circuit on the good board are a bit different than the schematic. R1 and R2 appear to be 150 ohms on this board and are supposed to be 50 in the schematic.

Could it be the Sony Snd chips are a different revision between these boards?

 

Elfen

Well-known member
There is usually 1 Sony Sound chip in Macs and the other two are digital/analog amplifiers. So chances are the amp for the left side is out. This sounds like rotted traces to it from the cap that was nearby.

 

hellslinger

Well-known member
Yes, the left channel Snd chip was out. Replacing it fixed the sound after boot. The distorted chime is still puzzling.

In case anyone comes across similar problems, I'll go into some detail. There is 1 ASC (surface mount, 44 pin sound processor) and 2 smaller chips (DIP16), one for each channel. The silkscreen next to these two DIP chips says Sony, so perhaps Elfen is referring to the ASC as the sony chip. These smaller Sony snd chips convert PWM to Analog signals (low pass filter), and probably amplify them slightly. These DIP chips also have 3 (SV 1:3) lines connecting to the ASC, which must be the volume control lines (there are 8 levels of volume on the Mac, so this would make sense).

I connected an oscilloscope to the PWM pin for both and observed working signals from each while the machine was running.

I've also verified all the connections to these PWM->Analog chips, and all the passive components nearby.

I'll report back when I've finished recapping. Thanks for the responses and interest.

 
Last edited by a moderator:

hellslinger

Well-known member
Thanks techknight, I think there must be more wrong with this than just bad caps as you are suggesting.

Another symptom is the machine will freeze for a short amount of time after playing a few sounds. If I click the alert sounds in the sound control panel a few times in a row, it will play the first few just fine and then will hang, sometimes for a few seconds or completely freeze the machine.

On the other hand, if I run The PlayerPro 4.06, the sound works perfectly through both channels and for as long as I need. This program gains exclusive access to the sound somehow, though I don't really know what I'm talking about here.

Again, any suggestions are appreciated. Thanks for the feedback so far.

 

techknight

Well-known member
Again, check your address/data lines feeding to your ASC. 

If the chime sounds "out of tune" or a different pitch/scratchy than it should, the address/data lines feeding the sound IC are broken, well at least one is. 

 

hellslinger

Well-known member
Techknight, I've been looking at the schematics and noticed that the ASC address lines are A(0:11), and since the glue chip doesn't have connections to A(2:11), I assume that the CPU is the thing that directly talks to the ASC. Is this accurate? In other words, should I only be concerned about the connections from the CPU to the ASC?

 

hellslinger

Well-known member
Techknight, got it working!

Thanks for your urging to check into the address and data lines. Apparently, only the first 8 address lines are actually connected to the ASC. I took out the 68030 and checked for continuity on the address lines and saw that A6 was not making any contact. 

http://imgur.com/a/Ck4Zh

I jumpered all of the address lines including A8:A11 and the machine wouldn't boot, so I began removing each of these until it booted. I had a feeling that there couldn't be 5 bad lines especially knowing that PlayerPro worked fine. 

Even the chime works and sounds right. No more hangs. Finally got this little SE/30 to working order!

 

techknight

Well-known member
I figured. 

Glad it works. 

Also, regarding that long ass wire, You could just jump over to the SWIM, or the SCC/SCSI IC and it would shorten your run. 

Also, A0/A1 go back to the Math co-pro. A0/A1 are only used for bus width negotiation/selection. for 8/16/32wide, and the FPU decoding

 
Last edited by a moderator:

uniserver

Well-known member
ok i have toned out address lines many many times….   A7( for me this time)… Murphy's law….

i was pressing down to hard or something...  with my leads… i don't know… when i was checking the address lines they were all good...

i did the same thing as the op did… removed that IC…  swapped it out…    pulled a big chunk of hair out.

everything was working….  great…  just the start up sound was always distorted... robotic sounding.

so i decided to check them one more time… clean up all the flux…. nice and clean and start with my lead on the video rom legs. and just bump each leg from of the ASC starting at pin #41 of the ASC and going up … and when i got to pin #4…   NO TONE… it was like 1280 ohms or something like that …  ran a wire… and now she is good to go…    when dealing with this Crusted up ,  Cap goo'd ….  Partially rotted board fixes….  it can get to even the best of us…    black magic :)

 
Last edited by a moderator:
Top