Wanted: AdLib Sound Card

[NobodyIsHere]

Hi! I will write up a KiCAD schematic with your notes and we can discuss implementation features, changes, etc. I'll make a folder on the N8VEM wiki to keep it in.

I am thinking a 74LS682 for IO decoder with 8 position DIP switch. 74LS245 bus transceiver, YM3812, 74LS04, 74LS74, 74LS08, YM3012, some jumpers, and a stereo headphones jack. Probably some bypass capacitors, filter cap, etc. Use a pair of generic LM386s for audio amplification (one each channel).

What else is needed?

Thanks and have a nice day!

Andrew Lynch

 

[per]

What else is needed?

Some components for the synth chips are also needed (The master clock source and everything needed for the DAC. See the datasheets for details).

We should also try to decide if we want to use the general output ports of the OPM to anything spesific, or if we should just keep then NC.

 

[jmetal88]

You guys have fun with that.

I'll just stick to making the YM3812 work on the PCjr. My progress will be posted at Mike Brutman's forum.

 

[NobodyIsHere]

Some components for the synth chips are also needed (The master clock source and everything needed for the DAC. See the datasheets for details).

We should also try to decide if we want to use the general output ports of the OPM to anything spesific, or if we should just keep then NC.

Hi! Do you have example schematics for the DAC? I read the datasheet and it is kind of sparse. Actually a schematic for an end to end YM3812 & YM3012 would be good or at least something close. Audio chips are more sensitive to the analog effects than the normal pure digital circuits I work with. A good snippet from a known working circuit would be very helpful.

Thanks and have a nice day!

Andrew Lynch

 

[per]

Hi! Do you have example schematics for the DAC? I read the datasheet and it is kind of sparse. Actually a schematic for an end to end YM3812 & YM3012 would be good or at least something close. Audio chips are more sensitive to the analog effects than the normal pure digital circuits I work with. A good snippet from a known working circuit would be very helpful.

Thanks and have a nice day!

Andrew Lynch

This is taken directly from the Yamaha FB-01 service manual, which contains a schematic how the OPP and the YM3012 is connected, along with the master clock. Just note that the master clock should be slightly slower for the OPM and OPL2. I don't know why they use 4MHz, but it may be that the OPP operates at this rather instead of 3.58MHz. It shouldn't do much a difference anyways.

 

[NobodyIsHere]

Hi! Thanks! Good, that is something I can work with. I'll draft a schematic tonight and post. Hava a Happy New Year!

Andrew Lynch

 

[TheLazy1]

Just as a quick aside, would this project be possible with the YM3812 hooked up to the parallel port?
It would be a nice, easy way to add an OPL chip to newer systems as well as older laptops that had no sound card.

Also, are you all going to be gutting cards for their OPL chips?
It seems like such a waste

 

[NobodyIsHere]

Hi! The YM3812 and Y3012 both appear to be available on Area51ESG.com probably in quantity buys though. I don't see them on findchips.com although I suppose with some digging around they'd turn up somewhere. However you raise an excellent point we'll need to know before this project moves to prototype boards is parts availability. If it requires bulk purchases of parts that'll require upfront cash to buy a store of parts. Also we'll have to find out how many builders would be interested in getting one or more of these boards, etc. There are many questions that'll need answering before the project is ready to go but for now we can make at least a notional design so we can discuss the plan for the rest.

The ISA board is a lot more appealling to me than a parallel port interface dongle. An ISA card does not need additional external enclosure, power supplies, etc. It is a much easier design.

Thanks and have a nice day!

Andrew Lynch

 

[jmetal88]

Just as a quick aside, would this project be possible with the YM3812 hooked up to the parallel port?
It would be a nice, easy way to add an OPL chip to newer systems as well as older laptops that had no sound card.

Also, are you all going to be gutting cards for their OPL chips?
It seems like such a waste

I don't know what these guys are doing, but I'm getting mine from a Sound Blaster 2.0 with iffy analog hardware. Seriously, the amount of static and cutting in and out on this card is awful. It's no waste at all, to me.

 

[per]

I found one who sells OPM-s and DAC-s at:
http://mysite.verizon.net/res8aiig/ICparts/ICparts.htm#YM2151
However, it seems that he sells them for a quite expensive price.

Here is another store who sells them, but they are out of stock.
http://andysarcade.net/store/index.php?act=viewProd&productId=1366
http://andysarcade.net/store/index.php?act=viewProd&productId=1367
http://andysarcade.net/store/index.php?act=viewProd&productId=1385

One can always try one of the many generic recycled-IC stores, whic I am quite certain would have some. I have never personally done any business with these before, so I don't know how much they are to trust.

 

[jmetal88]

Hi! I will write up a KiCAD schematic with your notes and we can discuss implementation features, changes, etc. I'll make a folder on the N8VEM wiki to keep it in.

I am thinking a 74LS682 for IO decoder with 8 position DIP switch. 74LS245 bus transceiver, YM3812, 74LS04, 74LS74, 74LS08, YM3012, some jumpers, and a stereo headphones jack. Probably some bypass capacitors, filter cap, etc. Use a pair of generic LM386s for audio amplification (one each channel).

What else is needed?

Thanks and have a nice day!

Andrew Lynch

Ah, useful post, man!

I have an old ISA serial card I can take a 74ls688 off of to do the decoding for my project. It hadn't occurred to me before now. It ought to eliminate at least two chips from the design I was previously contemplating.

 

[NobodyIsHere]

That's just the "Ye Olde FORTRAN" talkin...

 

[NobodyIsHere]

Hi! I uploaded the OPL2 folder on the N8VEM wiki with draft schematics, PCB layout, parts list, and a 3D rendering of the prototype board.

http://n8vem-sbc.pbworks.com/w/browse/#view=ViewFolder&param=OPL2

Please review closely and send me your comments/corrections/changes etc. Please lets discuss.

We'll need at least 5 volunteer builders to get prototype boards so we can see and debug with some real hardware. Typically those are $150 for 5 at www.33each.com so that would be $30 a piece. You don't have to build the board if you buy one or more -- just send it to someone who will.

Thanks and have a nice day!

Andrew Lynch

 

[jmetal88]

I think you uploaded the schematic for the wrong device. Looks like the XT-IDE rather than the OPL2 board.

 

[per]

I would suggest that there should also be a volume controll knob, like on the early soundcards without any mixing capabilities.

In addition, the power-lines going between the jumpers and the Synth chip should be thicker, and "YM2151" and "YM2164" should also be written besides "YM3812", along with jumper configurations.

 

[NobodyIsHere]

Hi! I found an error in the IO decoding logic so I updated all the files and uploaded the new ones. This should include the proper schematic.

Unfortunately there is a quirk in the IO decoder resulting in the board requiring 4 IO addresses rather than just 2 as it should. What I need to fix it is another OR gate but I do not want to add another chip. I have 2 AND gates, an inverter, and a flip flop as spares. If anyone can figure out how to convert some combination of those into an OR gate that would be great.

Thanks and have a nice day!

Andrew Lynch

 

[per]

Hi! I found an error in the IO decoding logic so I updated all the files and uploaded the new ones. This should include the proper schematic.

Unfortunately there is a quirk in the IO decoder resulting in the board requiring 4 IO addresses rather than just 2 as it should. What I need to fix it is another OR gate but I do not want to add another chip. I have 2 AND gates, an inverter, and a flip flop as spares. If anyone can figure out how to convert some combination of those into an OR gate that would be great.

Thanks and have a nice day!

Andrew Lynch

Whatever I try, it's impossible to make an OR gate without an extra inverter. Perhaps we can replace the oscillator with one that uses fewer components?

By the way, it doesn't seems like a huge problem. Most cards can only be configured to modulo-8 port addresses anyways.

 

[per]

I got a solution!
Replace the inverter U7D with an AND-gate where the inputs are tied together. The clock will be "inverted" in terms of what the original design provided, but it doesn't matter due to 50%-duty cycles.

Now we got two inverters, one and-gate, and a flip-flop. The flip-flop can easily be converted into an inverter, which means that we got a maximum of 3 inverters.

Then we can place one inverter on each of the inputs and the output of the remainin AND-gate, and that should make it an OR gate.

 

[NobodyIsHere]

Thanks Per! Can we some how manipulate the 74LS688 with the flip flop or AND gate to squeeze out another address line of decoding? A1 is hanging loose so the YM3812 decodes to a pair of mirrored IO ports. I tried screwing around with the inverter and an AND gate and just made a mess of it. I was also thinking since this is an IO only decode can we *safely* ignore AEN (DMA active) and use A1 to activate the output of the 74LS688? Maybe I am over thinking this.

I settled on just accepting the 4 IO addresses but would love to see an innovative way to fix this without adding more chips. Actually if anyone has suggestions on how to *reduce* part count that would be even better.

Thanks and have a nice day!

Andrew Lynch

 

[per]

Seems like you replied before my latest reply...

But, nope, AEN must be in there due to the nature of how the DMA controller works, but this wouldn't be a problem if we go for the solution I suggest.

I'd still love to see volume-knobs which you can alter without opening the chassis, though...