Ad Lib Music Synthesizer Card Clone :
There was nothing proprietary about the Ad Lib MSC card, the OPL2 YM-3812 sound chip and its YM3014 DAC were readily available from Yamaha. The rest of the board was populated with standard logic chips, a pre-amplifier and an amplifier. Ad Lib Inc. scratched off the part number of the Yamaha chips, but once their identity was known, anyone with some spare time and a multimeter could design a clone PCB. Several clones were made back-in-the-day, but they are now as hard to find as a genuine Ad Lib Inc. made card. In 2012, VCF forum contributor Sergey Malinov did just that, he designed a smaller 8-bit ISA card and it works just like the original. It also sounds like the original because he used the original amplifier circuit. It even has a volume dial. The 1987 revision of the real Adlib card used a 1/4" phono jack, but the later 1990 Adlib revision and this board use a 3/8" mini-jack. You can find all the information you need, and a link to the board design, here :
Another individual has also cloned the Ad Lib card, but this individual sought not just to recreate the functionality but to recreate the board itself, including the Ad Lib logo. True Ad Lib cards are valuable, but this design has the potential to defraud a collector given how close it is :
Roland MIF-IPC-A Clones :
A Roland Midi Processing Unit (MPU) 401 was originally contained in a small metal box. It was designed to connect with any 8-bit or better home computer which could provide the addressing and data lines and IRQ it required. Roland sold interface cards for various computers, including the Apple II, Commodore 64 (cartridge) and IBM PC. All the electronics necessary for the interface was contained in the external box including a microcontroller, ROM and RAM, a bus interface chip and DIN connectors for MIDI cables. Replicating the MPU-401 circuitry would be no mean feat. The interface cards consisted only of a few standard logic chips to route the necessary signals from the computer.
The MPU-401 boxes are often found without an interface card, and without an interface card they are useless. Roland originally offered the MIF-IPC interface card for IBM PCs and XTs. The original MIF-IPC card proved unreliable in AT and faster machines, so Roland issued the cost-reduced MIF-IPC-A, which works in any PC. It consists of four standard logic chips and a DB-25 connector on an 8-bit ISA card. Two cards were made, one from a U.S./German source (Bryce) and another from a Dutch source (n1mr0d), have been made. The details are here :
I am not sure whether the U.S./German clone is currently available, but it allows you to easily change the IRQ and the I/O address lines. The Dutch clone is currently available on ebay and allows for easier changing of the I/O port.
James Pearce from VCF and lo-tech.co.uk has designed a clone card called the MIF-IPC-B. This card can be built cheaply as it uses only through-hole parts. It adds useful features to the MIF-IPC-A design, namely selectable IRQ and I/O port selections. It is also compatible in an IBM PC/XT's Slot 8. It fixes the design flaw on the MIF-IPC-A card that leaves unused gates on the 74LS04 floating. It even replicates the logic differences between the MIF-IPC/IF-MIDI and the MIF-IPC-A cards, although I have never read or experienced anything to suggest that the MIF-IPC-A is unreliable in an IBM PC or XT. Still, the :
Music Quest PC MIDI Card Clone :
There has been an even more promising development on the MPU-401 clone front. Recently (as of September, 2015) VOGONS user Keropi made a full clone of the Music Quest PC MIDI 100% compatible MPU-401 card. This card contains all the essential intelligent MIDI capabilities of the Roland MPU-401, but does not require a breakout box. In other words, unlike the MIF-IPC-A clones, you do not need to supply anything to connect it to your MT-32 or other MIDI module. All circuitry is on the card, only an DE-9 to MIDI adapter is required for MIDI In and MIDI Out. The adapter is included with the card. The clone is fully compatible with games requiring intelligent (normal) MPU-401 capabilities. Details here :
Innovation SSI-2001 Clone :
I have already discussed this remarkable board elsewhere. I own one and can heartily recommend it, assuming batches are still being made : http://nerdlypleasures.blogspot.com/2014/01/sid-and-dos-unlikely-but-true-bedfellows.html
You can probably get a card with an 8580 SID, but if you want a 6581 you will have to provide it yourself. Information on how to obtain a card can be found here :
Creative Game Blaster Clone Prototype :
All the clones that have been made available for purchase or replication have been discussed above. For my next discussions, I will identify cards which have had working prototypes made. The principal difficulty in implementing a Game Blaster clone is that for a long time, the CT-1302 chip on the real board acted like a black box. Now that we know what it does and that all it really does is implement a simple autodetection scheme, the scheme has been replicated with standard logic chips. A work in progress Russian clone has come up with some very impressive results so far :
True Game Blaster cards sell for $200-$300 these days, but are desirable especially for the nine or so games that require a true Game Blaster card.
I should also mention here for completeness sake the successful efforts to replicate the PAL of the Sound Blaster 2.0 to allow for Game Blaster functionality. The details can be found here : http://nerdlypleasures.blogspot.com/2012/10/all-you-ever-wanted-to-know-about.html
Gravis Ultrasound Plug-N-Play (GUS PnP) Clone Discussion :
This is a very ambitious project because the main AMD/Interwave chip is a surface mount chip with many pins and small spacing between them. While the Interwave chip does not require RAM to function, its GUS emulation does not work without at least 512KB of RAM. The more advanced features of the Interwave chip can utilize up to 16MB of RAM. The real GUS PnP only supported 8MB of RAM, but that was more than sufficient for GUS Classic compatibility.
As of this writing, the prototype is still in the paper stage, so I do not deem it a real prototype until someone gets a real custom board working. The Interwave-based GUS PnP can play virtually all classic GUS-supporting games, although you may need a utility or special driver for some. This card should be able to offer the same level of compatibility. On the other hand, the advanced GUS PNP features went unsupported in DOS and the card was treated somewhat generically by Windows 95 outside Gravis' applications. GUS cards of any sort have always been pricey and hard to find, so this board should be welcomed by many gamers and fans of old-skool demos that used the GUS.
Tandy 1000 3-Voice Sound Card Discussion :
The idea behind this board is to implement a Tandy 3-voice sound chip at I/O C0-C7 for 8-bit XT systems and and probably 1E0-1E7 for 16-bit AT systems. Unfortunately, AT systems have a 2nd DMA controller at C0-C7, so the chip may not work there.
Games must write directly to the I/O ports to make sound come out of this chip. Some games will refuse to play Tandy music or sound effects unless they detect a Tandy through its BIOS signature or its unique graphics adapter. This card will not work with every game, but it will work with many, many classics that do not care about the BIOS signature or the graphics. James Pearce has committed to making a prototype, but none has yet been completed. Unfortunately, the DAC added in the Tandy 1000 TL and SL and later computers used a Tandy-proprietary PSSJ chip, so without a source for those this board will be stuck with a TI SN76489 (the extra feature of the TI SN 76496 is not required). The NCR 7496 clone is preferable for Tandy 1000s but is virtually impossible to source.
This is the discussion I support the most because it is the only one of these projects for which I do not own the original card or a reasonable substitute. Like the Innovation SSI-2001, there are only two original cards known to still exist. Cloning this card is more difficult than the Innovation for two reasons. First, the main sound chip is more difficult to source. Second, this card has a PAL chip on it that needs to be decoded. Decoding PAL chips has been done before, but it requires a lot of work and a fair amount of deduction.