IIsi odd sound issue with MacSD

[JeffC]

I am having a strange audio issue when I use my MacSD internally in my IIsi. I modified the IIsi to have term power to the internal SCSI connector by bodging from the DB25 external SCSI connector, and I also tried powering the MacSD with an external USB wall adapter.

The machine works fine, however I can hear odd noises out of the computer that correspond with drive access and other functions, for instance moving the cursor around creates a buzzing sound, so it seems there is more than just the SCSI circuit involved. The sound comes out of the internal speaker, and if I plug in a headset the sound comes out of both channels. I have linked a video below to an audio clip taken with my phone. If you turn up the volume you can hear it clearly. When I use a spinning platter HD I do not hear the sound generally, but I did find when using a HD, if I put my ear right up to the speaker, I can hear the same sounds but at maybe 10% of the volume.

The machine is recapped with polymer caps and the board was cleaned as thoroughly as I could with IPA (a soak for an hour followed by a thorough brushing with a toothbrush. I do not have an ultrasonic cleaner. The PSU was recapped as well. The only rework done was bridging two small breaks between the pad and trace on two legs of an ALS245 chip. The board is overclocked to 25mhz, the person that installed the 50mhz crystal in the past did a butcher job, but it functions just fine. Pic of that mild carnage below. I doubt it is involved in the issue, but it never hurts to mention the small stuff.

Another thing I tried was using an 18" SCSI cable that let me move the MacSD far away from the machine in case it was creating some sort of RF interference, and there was no change in behavior.

Does anyone have thoughts on the next troubleshooting steps? Unfortunately I do not currently have another internal solid-state drive solution, so I am unable to see if the same problem exists with a BlueSCSI or ZuluSCSI.

Video:

Photo of crystal hack-job:

 

[Callan]

I'm spit balling since I don't know anything about the iisi, but... if I was to guess I would think I had a ground loop of some sort. I would tap my power off of the same power feeding the HD.

 

[JeffC]

I was originally going to hack up a mini USB cable to power the MacSD through the mini USB plug, pulling power from the Molex connector for the original hard drive. However, I figured since I had the same symptoms whether I was running the MacSD off SCSI term power bodged from the DB25 connector, or if I was powering the MacSD from a wall adapter through the mini USB plug, that would isolate the issue from the power source.

 

[Callan]

I'd pin a spare 4 pin molex with my usb out. I'd also make sure its shielded good and have the ground strap from the shield grounded to the molex ground These are standard ground loop preventative measures. Make sure to recheck you caps to (especially around the sound chips). Caps like c42 and c43. I'd check.c2 and c4 too. It sounds like your picking up bus data somewhere and it's being amplified hope this helps. I haven't gotten my hands on. Iisi yet, so am unable to give you a more independent understanding of what's going on. I will say this. If the guy who capped your crystal capped the rest of the board you definently have to go over it with a fine tooth comb. Could he have severed a ground when he pulled that crystal? Quite possibly. Being it's a multilayer board. Again... just spitballins what case you could use some sort of audio isolation transformer to power it up.

 

[JeffC]

I'd pin a spare 4 pin molex with my usb out. I'd also make sure its shielded good and have the ground strap from the shield grounded to the molex ground These are standard ground loop preventative measures. Make sure to recheck you caps to (especially around the sound chips). Caps like c42 and c43. I'd check.c2 and c4 too. It sounds like your picking up bus data somewhere and it's being amplified hope this helps. I haven't gotten my hands on. Iisi yet, so am unable to give you a more independent understanding of what's going on. I will say this. If the guy who capped your crystal capped the rest of the board you definently have to go over it with a fine tooth comb. Could he have severed a ground when he pulled that crystal? Quite possibly. Being it's a multilayer board. Again... just spitballins what case you could use some sort of audio isolation transformer to power it up.

Thanks for the info. Alternatively to powering through the USB plug, the board also has a pin header I could use to provide power. This might be easier than stripping a mini-USB cable and connecting those wires to a Molex connector. I think I may have a spare Molex cable in storage I could cut and re-purpose. I am not sure though if I can use the "logic ground" to connect to the ground pin of the Molex connector. Unfortunately, I have little knowledge of electrical theory.

I did the recap, and fortunately my soldering skills are solid. I will however take a close look around those caps with a microscope and see if maybe there are damaged traces I missed when I did the recap.

 

[Callan]

I'd pin a spare 4 pin molex with my usb out. I'd also make sure its shielded good and have the ground strap from the shield grounded to the molex ground These are standard ground loop preventative measures. Make sure to recheck you caps to (especially around the sound chips). Caps like c42 and c43. I'd check.c2 and c4 too. It sounds like your picking up bus data somewhere and it's being amplified hope this helps. I haven't gotten my hands on. Iisi yet, so am unable to give you a more independent understanding of what's going on. I will say this. If the guy who capped your crystal capped the rest of the board you definently have to go over it with a fine tooth comb. Could he have severed a ground when he pulled that crystal? Quite possibly. Being it's a multilayer board. Again... just spitballins what case you could use some sort of audio isolation transformer tonpowerm

I'd pin a spare 4 pin molex with my usb out. I'd also make sure its shielded good and have the ground strap from the shield grounded to the molex ground These are standard ground loop preventative measures. Make sure to recheck you caps to (especially around the sound chips). Caps like c42 and c43. I'd check.c2 and c4 too. It sounds like your picking up bus data somewhere and it's being amplified hope this helps. I haven't gotten my hands on. Iisi yet, so am unable to give you a more independent understanding of what's going on. I will say this. If the guy who capped your crystal capped the rest of the board you definently have to go over it with a fine tooth comb. Could he have severed a ground when he pulled that crystal? Quite possibly. Being it's a multilayer board. Again... just spitballins what case you could use some sort of audio isolation transformer to power it up.

 

[Callan]

Lol... and that's why i don't type messages at home when I have the flu. I'd be suspicious of the guy trying to help talking about capping crystals (I've heard it's all the rage ) Honestly though. Good luck! Yeah the header pin would be fine too.

 

[JeffC]

Well I was reading your message before I had coffee, so our mental states probably weren't far off. Thanks for the advice, I'll update after I get some time to test further.

 

[YMK]

I've also followed this issue at 68kmla. Something @JeffC mentioned there is that mouse movements also leak into the audio circuit.

If you have a 25 or 50 pin terminator, you could try installing it and pulling MacSD's resistor blocks.

I think it would only be masking an underlying problem with the audio system, but it could reduce the noise.

Since the noise persists even when you power MacSD from an outside supply, I don't think it has to do with power draw.