What is it with Yamaha releasing dos drivers via an installshield wizard ? I dont have win98 on that freedos box. Anyhow thanks to vogonsdrivers I found the dos drivers that seem to have a chance to work on freedos, also the adlib disk. I think I'll give the ymf724 a try to see how it sound then install the adlib. Using the eeprom tool I disabled the part of the GUSPnP that use port 388h, anyway, trying to use it on freedos with megamem and sbos crashed the init sequence of any players at best. I'm so eager to load jukebox.exe and listen to parade.rol on the adlib just like in the good old time, hey I did a cover of that tune if anyone cares : http://junkskool.net/audio.aspx?d=/C_WalkingW … ship_parade.mp3

Well the ymf724 didnt play along anyway and I dont really care about that card, it keep complaining that the codec might not be installed. Anyway I plugged the adlib and work since day one. I'm listening to the ROTT soundtrack with dosmid and I'm actually impressed about how well it work! Any reason why dosmid is in monochrome when using an OPL2 ? Seem like a kind touch of retro.

I'm not too found of the potentiometer at the rear of the card, seem like it sound distorted if you give too much level. Any idea how to mod the card to make it achieve good line level ? My goal is to put it on a guspnp cd-in anyway so I dont even have to install another phono jack.

ElBrunzy wrote:

I'm so eager to load jukebox.exe and listen to parade.rol on the adlib just like in the good old time, hey I did a cover of that tune if anyone cares : http://junkskool.net/audio.aspx?d=/C_WalkingW … ship_parade.mp3

Hey, thanks, that's a good recording.

Jo22 wrote:

Hey, thanks, that's a good recording.

Wow! it's me who thank you that comment, that was unexpected 😊.

I came back to the thread as I've start to make some research on how to bypass the amplifier on the adlib card. They are more common here in Quebec but still, I do not want to waste it so I am very careful on how to hook the adlib output to the guspnp onboard cd-in connector. I did some research about the Y3014B chip that seem to be the OPL2's DAC. It is stated that the pin 2 is the analog output. It seem to be linked to the Texas Instrument's RC4136N pin 5 which is called 2IN+ wich is an op amp so that make sense. What's weird (or maybe not) is that pin 6 and 4 are linked together and seem to go to the series of resistances that we can see on the left of the Texas Instrument chip. My guess is that if I follow it I would get to the output of the card, but it's getting very late and dark and my eyes start to ache.

What other obvious thing I can see is that the potentiometer next to the phono jack is just a variable resistance to the output, it dont control the feed or the vcc of any amplifier. So I start to believe that I can tap a wire at the pin 5 of the Y3014B and another somewhere in the ground of the card and have a nice line level out of that card to plug on the cd-in of it's computer case roommate the guspnp, what do you think ?

Btw I have the 1990 version as you can see on the second picture : http://nerdlypleasures.blogspot.ca/2015/07/th … sound-card.html

Ourgh .. I just passed 1 hours with a coworker who is a former electronic guy (electronician?) talking about my project and he suggested that I need the buffer opamp and splitting the wire after the DAC is not a good idea. He suggest I change the potentiometer at the end of the adlib with a fixed resistance that match the impedance of the cd-in of the GUSPNP. I have yet to find that impedance, or the one of a typical cd-out of an internal cd reader. I suspect it is around 10kohms, but anyone can confirm this information?

just found out this site who state 39k for line-in and 50k for cd-in http://support.creative.com/kb/ShowArticle.aspx?sid=5800

Yes, your colleague is right. You can't connect directly to DAC output - it could be very weak output drive capability, and also the amplitude is 2.5Vpp riding over 2.5V DC bias, far too loud for line level input (and there's DC bias that needs to be removed first).

The first op-amp is used as a buffer (which is why pins 6 and 4 are linked), and the second two are used as a 4th order filter, but the gain of the filter is so huge that you get about 6.3Vpp audio riding over 6.3Vpp bias.

This is then fed to some extra audio filtering before the potentiometer, which attenuates the audio before the LM386 amplifier input. So the pot does control the volume level fed to the amplifier.

Your colleague's suggestion about changing the potentiometer with fixed resistance to match impedance of CD-in of GUS PNP is weird. Line level audio does not work like that, it does not use impedance matching at all. To get maximum voltage transfer, the line outputs are small impedance (below 1kohm) so it can drive whatever devices you put there, and line inputs are large impedance (10kohm or above) so it does not load the output too much.

My best suggestion is to get the audio from middle pin of the potentiometer, so then you get exactly the filtered audio with volume control so you can set the level to your liking.

Besides, I would not want to break/modify an original Adlib card myself 😀

Thank you so much for taking time to explain all that. So, I wanted to know the resistance value of the potentiometer when I noticed it did not worked at all as I tough so. I figured that it control another amp, I could not get it more wrong, I mistook bridges from front to back of cards with pot connectors. All that start to make sense to me now about how it work. Your explanation confirmed and put missings pieces of the puzzles in places.

Of course I want to be extra-careful when I work with that card (hence that thread, instead of blindfolded experimentations). So after that exploration I have regrouped my mind and realise that my fear was that I dont want a wild wire to bulge that big potentiometer knob on the back of the card, then be constantly re-adjusting it. On another hand they are cheap potentiometer and it start to gritch, so it is just normal maintenance to change that pot, and I found that a way to kill two birds with one stone is to change it with a cerment potentiometer that will stand into the case, next to the pcboard of the card, just like that http://www.hobbytronics.co.uk/image/cache/dat … pot-500x500.jpg (candle drop sealed 😵). I'm not sure that's the real name for those kind of pots or where to buy them, but I'll check tomorrow. I tested the adlib pot, it is often skipping marks. It is written 10k on the pot, but my voltmeter say 6.8k and it has no shame being a logarithmic one, yet that is being kept secret from the inscriptions on the pot.

Of course your suggestion make alot of sense too. If I understand it well, you say that I could pick the signal after the buffered amp and skip the amp that the potentiometer is controling? Knowing that this value is at most 10k and that the guspnp's cd-in should accept as much as 50kohms, I think I will try your idea. You seem to know that card quite well, so may I ask you if I get audio from your superb idea middle pin of the pot, do you have any suggestion as to where I should ground it ?

If you did not remove the pot but measured it in the circuit, the other connected circuits affect the reading, as you are not only measuring the pot itself but the whole thing. It should be some Bourns 10K logarithmic audio pot, the exact code has been discussed on Vogons.

If you take the audio from middle pin of the pot, you have ground in one of the other two pins and the other of the two is the audio in to the pot. But actually, anywhere where it is easy to put the ground is fine.

I did try your idea, I tested with a pc speaker the middle pin of the pot and ground it, there was never nothing, maybe I grounded it wrong ? But if I sample the left or right pin I got audio, and if I ground the left pin with the right pin and put an amplified speaker in between, I can hear the music playing and the pot has no effect at all. I hooked it to an portable amplified speaker and I could hear clean audio. On a small unamplified speaker I couldnt hear it tho, I should had suspected there was something wrong. With my overwhelming enthusiasm I took a picture:
I was happy... back then !

That was before I knew the guspnp conflict with the adlib and if I init the gus, then the adlib wouldnt init anymore and I need to init the guspnp to change it's mixer settings. I have read about deactivating the sb or mpu emu using pnpcfg.exe and/or iwavecfg.exe hack (talking about hack, does a legit guspnp really say "plug and pray" at bios startup?) I found that the only way to do it effectively is to comment out AdlibBase=388 in the %iwdir%\iw.ini. Anyway after that I realised that I deactivated the atapi line in jumper connector of the guspnp using gus classic drivers and I could not re-activate them with anything from the guspnp drivers... when it actually worked, I realised that the output of the adlib's buffer amp seem to be not powerfull enough to feed the atapi-line-in of the guspnp, or is it that I hooked it wrong ?

So I made a mono phono 1/8 connector to a cd-in connector of the guspnp and adjusting the pot seem to make the work. I'll order one of those internal plastic potentiometer that arent prone to hook on anything. Listening to Miracle Warrior soundtrack on the adlib opl2 using dosmid.exe routed thru the guspnp as I wrote that entry. But I feel that the ultimate entry to listen on that card is kq4 or silpheed.

Well that's it, project closed, I finally found a 10k linear (4 turns) trimpot and swapped the old huge knob. Found a good value and dont really need to touch it anymore, it add less noise on the guspnp output than I expected. The old pot was starting to fade anyway. So I'm happy with that solution and learned alot in the process and luckily havent broke anything 😀

If you mind here is a picture with the blue cylindric trimpot standing on his bed of heat glue. We can appreciate that the green adlib card contrast with the red of the guspnp card, reminiscent of the artificial computer-generated waveforms and the sample-based synthesis difference of that time.

I was expecting homie saying stuff like "Hey I see a guspnp an adlib, an usb controler and a 100 network card. Then a matrox g200 video card with it's line-out addon, can you really make all that work altogether ?" and I would say "yeah, I have an fdconfig.sys that load them all"

Jepael wrote:

Am I crazy or was this the name of someone in the movie "Golden Child" ? Where he was at the custody.. I'm sure somewhere in that movie he say "Japeal" ... or was it another Murphy movie?