awgamer wrote on 2023-08-17, 00:56:

Lots of youtube hardware troubleshooting repair channels. Daunting starting from scratch but maybe able to run though some of […]
Show full quote

Lots of youtube hardware troubleshooting repair channels. Daunting starting from scratch but maybe able to run though some of the initial/basic troubleshooting, like testing voltages on the card with a multimeter.
https://www.youtube.com/watch?v=wkAp5x3Z_gc
https://www.youtube.com/watch?v=p6BJvS3nrb0
https://www.youtube.com/watch?v=2I6Mudx0Rm0
https://www.youtube.com/@adriansdigitalbasement
https://www.youtube.com/@necro_ware

but before that, clean the pins on the card with a pencil eraser, the proverbial turn it off and turn it on again move.

Yeah, I'm familiar with general hardware troubleshooting.

I was more wondering if there is anything specific on this particular card (e.g. likely point-of-failure). Or even if a general schematic is available somewhere.

Another basic thing to do, clean the output jacks, wiggle them to see if there's static or intermittent output from doing so. probe for activity out of and into the dac. if there was output from the dac but no sound that would seem to isolate an issue in the short path from the dac to the stereo/line jack output plug and confirm the whole card is working up to that point. no output from dac, test into dac, if activity to the dac, dead dac, then etc. work your way back. Could probe things from software, card found by gold install/test software? init work? tsr loads okay w/o an error? opl responds to queries? games detect the card? Is there a software mute setting or other such settings that could turn off output? besides physical plugs giving out from physical use breaking solder joints, amps seem to be a common thing to blow.

awgamer wrote on 2023-08-17, 02:21:

Could probe things from software, card found by gold install/test software? init work?

It's not detecting at all, that's the issue. If I run the basic Adlib Gold drivers, it says no card is detected. It's also not being detected by games.

I even attempted a control chip reset (per instructions listed earlier in the thread) multiple times to no avail.

I've also given it a quick inspection under the microscope and no obvious damage, bent pins, or anything else particularly obvious. And tried repeatedly cleaning the contacts just in case, but they're pretty clean to begin with.

If software isn't detecting the card then I'd start with checking that voltage is alive, correct, all around the card, and looking for shorts. No power, no software reporting, a short, locked up, no software reporting. Scratch that, another first basic test before this is trying to identify hot spots, whether by feel, temp sensor or thermal cam(there are phone apps that do this for something quick and dirty.) hot spot = component or ic went caput = no software reporting. Find a hot spot, can test if that's the problem with compressed air blown on the spot or something cold to press on the hot spot to see if that brings the card back alive. Also could stick the card in the freezer for awhile then plug in to see if that brought it back to life before it heats up again. psa watch out for condensation.

Also heard of the opposite, where cards need an inspection for broken solder joints where a reflow would help. I still suggest the youtuber hw diag/fix channels as a guide, they should lead you to a solution faster than my bag of marbles recollection of diagnosis.

Shponglefan wrote on 2023-08-17, 02:46:

It's not detecting at all, that's the issue. If I run the basic Adlib Gold drivers, it says no card is detected. It's also not being detected by games.

I even attempted a control chip reset (per instructions listed earlier in the thread) multiple times to no avail.

I've also given it a quick inspection under the microscope and no obvious damage, bent pins, or anything else particularly obvious. And tried repeatedly cleaning the contacts just in case, but they're pretty clean to begin with.

That is weird... Time to send it back for checking/repair then - send me an email to arrange it !

Shponglefan wrote on 2023-08-17, 00:38:

Unfortunately my GoldLib appears to have died. :( […]
Show full quote

Unfortunately my GoldLib appears to have died. 🙁

I was testing it in a 486 DX-33 build. Initially it worked and I was able to detect and configure it just fine.

I'd been adding and testing other sound cards. At the time I was mainly working with the AWE64 and Yamaha Audician 32. Then went back to try the GoldLib again, and it wasn't working anymore. The Adlib Gold setup program won't detect it.

I tried removing everything and just testing the Goldlib, but it still wouldn't detect it. I tried it my other 386 DX-40 system, it won't detect there either. I tried the control chip reset instructions listed here (three times in both systems). Didn't help.

I also confirmed that my original Adlib Gold card works in both setups (386 and 486). So it doesn't appear to be a system specific issue.

I did a cursory inspection for any physical damage, but didn't see anything immediately obvious.

Any suggestions on anything further to try?

Did you change the AdLib Gold port to something other than 388h? If you did, and have the jumper config in single port mod, the AdLib Gold software won’t find the card anymore and it will feel that the card is dead. I had to do one of 2 things:

- Either change the card for dual port mode, where the card will respond on 2 different ports (the one you set the card to and fixed 388h) by changing the jumper to the middle and lower pin;

- Patch the software you want to use for that new port address you set your card to, there’s just no other way;

I always thought the card’s software would look to the SET GOLD environment variable, but that never helped me and seems to happen only when the jumper is set to dual port mode.

I finally decided to have my card on 398h with IRQ 4 and now the card co-exists peacefully with a Pro Audio Spectrum just fine. Main drawback is that i ran out of DMA channels, making me wish i had the MMA also wired to DMA 0 which would require the 16 bit connector part to exist.

Patching the software wasn’t a huge drawback to me since most game support for the GOLD is via sound libraries such as the AIL2, DIGPAK, HMI SOS etc… Cool thing about the AIL2 drivers is that I still have digital sound without a DMA channel set, but alas, other libraries can’t play without DMA. That’s a consequence i was ready to accept.

Dune is an excellent test for the Gold because the game IMO was very well programmed and it follows the instructions you give to it. When you provide the base address of the Gold to the game it really uses THAT range. When i changed my card to 398h and the Gold drivers stopped finding the card i tried Dune with the port I set it to and the card was working fine. That’s when i realised what was going on.

I recommend you set the jumper to the bottom position to set the card to dual port mode and try the AdLib Gold drivers before trying anything else.

carlostex wrote on 2023-08-17, 05:46:

I recommend you set the jumper to the bottom position to set the card to dual port mode and try the AdLib Gold drivers before trying anything else.

Success! Thank you for suggesting this, it did the trick!

For some reason, the port was set to 80 and the SCSI port to 300. I had not changed the port myself (and certainly not to these values). IRQ settings had also been cleared (previously I had it set to IRQ 3).

Mixer settings were also messed up with volume set to zero, and the L-R balance with L set to 100.

In subsequent troubleshooting, I did notice a couple more issues:

1) Initially the mixer would lock up on me when trying to write values. However, after repeated attempts, it seems to be working again.

2) Mixer values don't save the entered values for volume, treble or bass. It always saves a slightly lower value. However, this does appear to be "normal" behavior as my original Adlib Gold on my 386 does the same thing.

I also did another control chip reset for good measure. This worked this time and reset all the internal settings back to defaults.

Did a couple more hours of troubleshooting and everything seems to be working okay. Tested a variety of games including Dune, KGB, Arkanoid 2, Space Quest 3, Settlers, Warcraft...

Only issues is I notice that Dune changes the mixer settings (it sets volume to 97 and changes treble and bass). And Warcraft's FM music seemed a bit quieter than other games.

keropi wrote on 2023-08-17, 04:58:

That is weird... Time to send it back for checking/repair then - send me an email to arrange it !

Good news, it seems to be working again! I tried carlostex's suggestion of changing to dual port mode.

For some reason the port had been changed to 80, along with changes to other settings.

I did get some subsequent mixer lockups when attempts to write changes to memory. However, even that issue appears to have subsided for now. I've probably changed the mixer a couple dozen times in a row. Other than the values not writing correctly (which also happens on my original Adlib Gold), it does seem to be working fine.

I'd be really curious as to what caused this in the first place. I do have four other sound cards in the same system (Yamaha Audician 32, AWE64, MPU-401AT and a Primax Soundstorm). Not sure if it's possible for something else to have written to the GoldLib and corrupted its settings?

carlostex wrote on 2023-08-17, 05:46:

I finally decided to have my card on 398h with IRQ 4 and now the card co-exists peacefully with a Pro Audio Spectrum just fine. Main drawback is that i ran out of DMA channels, making me wish i had the MMA also wired to DMA 0 which would require the 16 bit connector part to exist.

Patching the software wasn’t a huge drawback to me since most game support for the GOLD is via sound libraries such as the AIL2, DIGPAK, HMI SOS etc… Cool thing about the AIL2 drivers is that I still have digital sound without a DMA channel set, but alas, other libraries can’t play without DMA. That’s a consequence i was ready to accept.

I've found that similarly DMA channels (or lack thereof) to be the real limiting factor in these multi-sound card setups.

I wonder if there is a possibility in some scenario that DMA sharing might work, but everything I've read on it is that DMA sharing seems to be more problematic than useful.

Most cards drive the signals in push-pull fashion so sharing is not going to be possible. In theory if cards used open-drain or source signalling, and didn't try to use same resources simultaneously, it can work. DMA sharing jumper on some of the SB cards probably does just that but if you have just one card this way, it isn't really enough...

Shponglefan wrote on 2023-08-18, 00:52:

Good news, it seems to be working again! I tried carlostex's suggestion of changing to dual port mode. […]
Show full quote

keropi wrote on 2023-08-17, 04:58:

That is weird... Time to send it back for checking/repair then - send me an email to arrange it !

Good news, it seems to be working again! I tried carlostex's suggestion of changing to dual port mode.

For some reason the port had been changed to 80, along with changes to other settings.

I did get some subsequent mixer lockups when attempts to write changes to memory. However, even that issue appears to have subsided for now. I've probably changed the mixer a couple dozen times in a row. Other than the values not writing correctly (which also happens on my original Adlib Gold), it does seem to be working fine.

I'd be really curious as to what caused this in the first place. I do have four other sound cards in the same system (Yamaha Audician 32, AWE64, MPU-401AT and a Primax Soundstorm). Not sure if it's possible for something else to have written to the GoldLib and corrupted its settings?

oh that is good to know
so basically the configuration eeprom data became corrupted somehow but it is good to know the way to come back from this
no idea what caused it, perhaps the hardware combo triggered something.
I have also noticed the mixer app saving a bit lower values... another oddity/bug I guess

Shponglefan wrote on 2023-08-18, 00:48:

Success! Thank you for suggesting this, it did the trick!

Happy to help!

Remember that it's important to keep the AdLib Gold on a range that nothing else writes to it. In my case a PAS and the Gold at 388h was immediate disaster with the MVSOUND.SYS driver writing on that range screwing up the mixer values big time.

With the Gold on its own, without any conflicts, it's a great sounding card. I really like it. Too bad i have no more DMA available for it.

Another thing that surprised me recently was that i couldn't find any MSS Miles driver (DIG) driver for the Gold. The library does have the Adlib Gold MDI file but i can't find the DIG driver. And i fear the reason is that one was never done, which is definitely a shame.

carlostex wrote on 2023-08-20, 22:04:

Remember that it's important to keep the AdLib Gold on a range that nothing else writes to it. In my case a PAS and the Gold at 388h was immediate disaster with the MVSOUND.SYS driver writing on that range screwing up the mixer values big time.

I had admittedly been testing the AWE64 and Yamaha cards with their FM ports set to the default 388h.

I guess at some point something must have tried accessing/writing to that port and messed up the Goldlib's EEPROM settings in the process.

Currently have those two cards set to 3A8h instead, so hopefully that will avoid further issues.

keropi wrote on 2023-08-18, 05:15:

I have also noticed the mixer app saving a bit lower values... another oddity/bug I guess

At least we know that the GoldLib delivers the authentic Adlib Gold experience in that regard. 😉

Hi there!

I was playing a bit of Monkey Island "Ultimate Talkie" and i was running into crashes everytime i tried opening a door on the game. Being completely puzzled, and thinking it could be an oddity of my setup, i removed the GoldLib card, because i simply never had any issues with this game prior to installing it. After removing the card, i just walk to the SCUMM Bar and open the door and this time there's no lockup or crashes at all. This ONLY happens when opening a door, which was weird.

So i'm wondering what the hell is going on, i have my card on a range that shouldn't have any conflicts with a sound blaster (Orpheus II in my case). Since this behavior only happens when opening the SCUMM bar door, or any other door in the game i'm also wondering what the game does at that point that would make this behaviour happen.

Can anyone around with a GoldLib validate this for me? To test this, one has obviously need a Sound Blaster and a GoldLib / Adlib Gold installed in the system. Just start the game normally with voices running and head over to the SCUMM Bar and open the door. Post your findings, please!

carlostex wrote on 2023-09-09, 16:07:

Can anyone around with a GoldLib validate this for me? To test this, one has obviously need a Sound Blaster and a GoldLib / Adlib Gold installed in the system. Just start the game normally with voices running and head over to the SCUMM Bar and open the door. Post your findings, please!

I don't have the Ultimate Talkie version, just the original floppy edition.

FWIW, I tested the floppy version on two different setups: 386 DX-40 with a real Adlib Gold and 486 DX-33 with a GoldLib. Didn't have a crash or other issues in either case.

Does it only do this with the Ultimate Talkie version on your end? Or is the floppy edition also affected?

carlostex wrote on 2023-09-09, 16:07:

Hi there!

I was playing a bit of Monkey Island "Ultimate Talkie" and i was running into crashes everytime i tried opening a door on the game. Being completely puzzled, and thinking it could be an oddity of my setup, i removed the GoldLib card, because i simply never had any issues with this game prior to installing it. After removing the card, i just walk to the SCUMM Bar and open the door and this time there's no lockup or crashes at all. This ONLY happens when opening a door, which was weird.

So i'm wondering what the hell is going on, i have my card on a range that shouldn't have any conflicts with a sound blaster (Orpheus II in my case). Since this behavior only happens when opening the SCUMM bar door, or any other door in the game i'm also wondering what the game does at that point that would make this behaviour happen.

Your issue reminded me of this Mister fpga's AO486 core's issue, that sounded similar: https://misterfpga.org/viewtopic.php?f=13&t=3889

So the game would freeze when starting a new sound before the previous one finished. Is there a lot of sounds in that scene when you open the door?

Shponglefan wrote on 2023-09-09, 16:38:

I don't have the Ultimate Talkie version, just the original floppy edition.

FWIW, I tested the floppy version on two different setups: 386 DX-40 with a real Adlib Gold and 486 DX-33 with a GoldLib. Didn't have a crash or other issues in either case.

Does it only do this with the Ultimate Talkie version on your end? Or is the floppy edition also affected?

No the floppy version is absolutely fine. Only the Ultimate Talkie and it happened only when opening doors.

janih wrote on 2023-09-09, 17:42:

Your issue reminded me of this Mister fpga's AO486 core's issue, that sounded similar: https://misterfpga.org/viewtopic.php?f=13&t=3889

So the game would freeze when starting a new sound before the previous one finished. Is there a lot of sounds in that scene when you open the door?

I never get to see the new screen after opening the door because it locks up immediately and the door sound isn't even played.

Hmmm i think those are unrelated with what i'm experiencing. For some reason it only happens when the door sound is triggered. It happened to me on the SCUMM Bar door and the Voodoo Lady door. Taking the GoldLib card out of the system solves the issue though, so i'm wondering what the game does when the door sound is played that could be interfering with the card.

Any slightest hope on a new batch?