What Is DXP44Q?

Today I came across a random 1996 vintage ISA sound card (no name, made in China). The only two chips worth noting on the board are an OPTi 82C930A chip and a mysterious DXP44Q—a square chip center left, to the right of the CD-ROM connector.

Sound Card with DXP44QIt’s clear enough what the OPTi chip does—it’s essentially a SB Pro + WSS clone with an integrated DAC/ADC and mixer. But what the heck is the DXP44Q chip? I had never come across it and I can’t find any documentation about it, not even who made the chip!

Unfortunately search results are completely poisoned because someone has DXP44Q on a list and now thousands of Chinese chip peddlers offer “DXP44Q” without having the slightest clue what it is or what it could possibly be good for.

There are a few photos of other sound cards with the same mysterious DXP44Q chip. Is it possible to at least determine what it does given the OPTi 82C930 datasheet? Let’s see…

The OPTi 82C930A chip (the 82C930A is presumably close enough that any differences are not significant) is highly integrated and implements most of the functionality of a typical Sound Blaster clone sound card. It includes bus controller circuitry (including DMA and interrupt logic), Sound Blaster and Windows Sound System register-compatible logic, a game port, a CD-ROM interface, a MPU-401 interface, and a codec/mixer. The only part conspicuously missing is a Yamaha OPL3 implementation (but the 82C930A is designed to communicate with an external chip).

Why would OPTi not integrate an OPL3 FM synthesizer when everything else is there? That’s easy—John M. Chowning’s U.S. Patent 4,018,121, initially assigned to the Stanford University and in 1982 reassigned to Yamaha Corporation (formerly known as Nippon Gakki Co., Ltd.). Because of this patent, others were prevented from implementing OPL3 clones.

The patent expired in 1995 (in some countries possibly in 1994). At that point, it became highly desirable to implement OPL3 clones, since Yamaha was charging relatively high amount of money for their chips. Perhaps that the DXP44Q is such a clone. The DXP44Q on my board is from early 1996 and thus wouldn’t violate the FM patent (not that that would be a proof either way).

With the OPTi 82C930 datasheet and a multimeter, it is easy to confirm that the 82C930 pins dedicated to communication with an external OPL3 chip are, in fact, connected to the DXP44Q.

It is thus fairly certain that the DXP44Q is a functional equivalent of an OPL3 chip and its associated DAC—a clone of an YMF262 plus a YAC512. To be fair, others at least suspected that the DXP44Q might be an alternative OPL3 implementation, as evidenced here.

What’s more, on the audio board’s PCB there are unpopulated locations (U6 and U7) that just happen to exactly match an YMF262-M and a YAC512-M (the SMD versions of those chips). It seems probable that the PCB was designed around 1995 and was intended to accommodate either the genuine Yamaha OPL3 or a third-party replacement (the likelihood that the DXP44Q chip was made by Yamaha is slim).

DXP44Q and Empty SpacesCase closed? Almost, except for one nagging question: Who the heck made this chip? And why is this chip not mentioned in any discussion of OPL3 clones? Any insights are welcome.

This entry was posted in PC hardware, Sound. Bookmark the permalink.

12 Responses to What Is DXP44Q?

  1. Stiletto says:

    I’m not sure why the big fuss, it’s clearly a Chinese bootleg chip? Bootlegs are often not discussed, whether it’s stated directly or not most people won’t give them the attention.

    For the flip side of that, see some old posts on Bunnie’s blog…

  2. Michal Necasek says:

    But a bootleg copy of what? This doesn’t look like any OPL3 variant I know of.

    Also, if it’s a bootleg that makes it more historically interesting, not less…

  3. Joshua Rodd says:

    Well, I’ve got one of these chips, also on a clone SBPro ISA card. Pulled from, I believe, something of a Compaq or Packard Bell vintage in the mid-1990s. No markings on the chip other than the ones on yours.

  4. Michal Necasek says:

    So this mystery chip landed even in reputable OEM systems… out of curiosity, what is the controller chip on yours? Is it an OPTi as well or something else?

  5. Joe says:

    In case you’re interested, I’m hunting around for an OPLL/2/3 and was just handed a Soundblaster compatible card that has a DXP44Q on it – right next to a blank spot labelled OPL3. At least some of the tracks that would connect to the OPL3 also connect to the DXP44Q.

    Incidentally, this page is now the first hit in Google for DXP44Q!

  6. Michal Necasek says:

    Any chance you could provide a high-res photo? I have only one DXP44Q and it’s on a board with OPTi 82C930A which I don’t have a datasheet for. And I only have that one board with 82C930A, too.

    The board I have also clearly has unpopulated spots for OPL3 and its DAC, and the DXP44Q is definitely some kind of OPL3 clone. What I don’t know is whether it’s just the FM synth or also a DAC. It is possible that the 82C930A might have a DAC for the FM chip built in, but I really don’t know.

    It is in theory possible that the DXP44Q would be an OPL3-L (YMF289B) clone, but the OPL3-L came in a 48-pin package while the DXP44Q is 44-pin, so that seems unlikely.

  7. Pingback: More on DXP44Q? | OS/2 Museum

  8. Izlude says:

    Oh man this brings me back… Ever try to play Warcraft 2 with the 82c930a? Nothing but stuttering voices. This chip was onboard my Mobo.. model escapes me. It was part of an NEC tower, Pentium 120mhz.

  9. GL1zdA says:

    The card is a Shuttle HOT-235 card. Not exactly identical to this one https://www.shuttle.eu/_archive/older/de/235.htm , but it looks like a revision of it. On mine, there’s a label on the back side of the PCB “HOT-235” in the top right corner.

  10. Michal Necasek says:

    My understanding of the mid-1990s clone sound card market is that they were produced in Taiwan or wherever and sold to numerous OEMs, perhaps with very minimal changes. Shuttle in the Netherlands, Terratec in Germany, probably Guillemot in France, then there was Addonics and who knows how many others. Who actually designed the cards… I have no idea. Some companies designed their own high-end gear but also sold cheap generic models (e.g. Turtle Beach).

    It’s possible that a lot of the generic card design often came straight from the chip companies (OPTi, Crystal Semi, Avance Logic).

  11. rasz_pl says:

    I just dug out from my hoard a Vibra 16S card with DXP44Q on board. Its full of Korean chips (Samsung and LGS), but claims to be “assemble in Taiwan” without the “d”. QA markings are from 9613. Both IDE and wavetable connectors are vertical, so it doesnt match CT2801 layout from Statson despite sporting one randomly placed “S2801” print.

    Exploiting the fact Google image search uses OCR you can just search for Vibra 16S “assemble in taiwan” and land on https://i.ebayimg.com/images/g/rk4AAOSwLSxe45Ut/s-l1600.jpg https://i.ebayimg.com/images/g/A5UAAOSwkWRe45WS/s-l1600.jpg

    sadly seller put a stupid card over the most interesting part of PCB covering both DXP44Q and mysterious S2801 marking.

  12. Michal Necasek says:

    Nice find. I think the Vibra 16S chips were available to OEMs, right? The card is not pretending to be a Creative Sound Blaster, so it would not be surprising to find shady chips on it.

    Creative probably switched to CQM before the fakes became available so they never used them. I’ve seen the fakes on name brand cards from Turtle Beach and Terratec but those were models built by Taiwanese OEMs.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.