Sound Blaster 1.0 or 1.5?

According to some sources, a Sound Blaster model CT-1320C is a Sound Blaster 1.5. But according to Creative’s own reference materials (which do contain many inaccuracies and cannot be taken as gospel), the only difference between a Sound Blaster 1.0 and a 1.5 is the presence of C/MS chips which provide Game Blaster compatibility.

For example this card is clearly a Sound Blaster 1.5:

CT1320C (SB 1.5)

Model CT-1320C, no C/MS chips, case closed. But what then is this one?

CT1320C (SB 1.0)

That’s another model CT-1320C, but with C/MS chips. According to Wikipedia, 1320C makes it a SB 1.5. According to Creative, C/MS chips make it a SB 1.0. How to tell who’s right and who’s wrong? Fortunately, there is a way.

The answer lies in the packaging. The Sound Blaster 1.5 came in this box:

SB 1.5 Box Front

Notice the small “VERSION 1.5” label in the lower left corner of the box. There can be no doubt that this is a Sound Blaster 1.5.

But what about the other one, the one with C/MS chips? The box looks a bit different:

SB 1.0 Box Front

There’s no mention of any version. But look… “STEREO MUSIC”, “24-VOICE” (a nice way to massively exaggerate while not technically outright lying), and “12-Voice C/MS stereo music synthesizer”. Yep, that’s a Sound Blaster 1.0.

That means the 1320C was shipped from the factory with the C/MS chips installed and not upgraded later. The serial number on the box in both cases matches the one on the card itself, so there’s no doubt the boxes match the respective cards.

The disks shipped with the cards also reflect this. The SB 1.0 disks include C/MS software while the SB 1.5 disks do not.

The bottom line is that a Sound Blaster model CT-1320C can be either a Sound Blaster 1.0 or a Sound Blaster 1.5.

Upgrade or Downgrade?

In case it’s not obvious yet, these old Sound Blasters are an incredibly rare case where upgrading the later version meant in fact downgrading to the older model. In other words, SB 1.5 + C/MS chips = SB 1.0.

By late 1991, more or less every game supported AdLib (OPL2) music and the Game Blaster was considered obsolete. Creative removed the C/MS chips and being the marketing-driven company they were, called the reduced-function product “version 1.5”.

SB 1.0 with C/MS Chips

The board layout was the same and because the C/MS chips were socketed, it was trivial to leave them out and offer the chips as a separate upgrade for those few who cared.

For reference, the DSP-1321 chip in the photo is a version 1.05 DSP (some form of an Intel 8051 microcontroller driven by a 12 MHz clock). FM1312 is a Yamaha OPL2 chip and the small FM 1314 chip is Yamaha’s DAC.

DSP Versions

For those who think that the dizzying array of Sound Blaster 16/32/AWE models was insanely confusing, rest assured that Creative had an early start and lots of practice.

In order to further blur the lines between the 8-bit Sound Blasters, some of the 1.x boards were shipped with a DSP (Digital Sound Processor) version 2.00, such as the SB 1.5 seen in the photo at the beginning of this article.

The original Sound Blasters had DSP version 1.x which only supported “single” DMA transfers. At the end of each DMA transfer which moved samples to or from the card, an interrupt was signaled; the interrupt handler had to reprogram the DMA controller and send a new transfer command to the DSP. It was impossible to have entirely smooth playback, although with the low-quality 8-bit samples and slow sampling rates the tiny gaps were not terribly noticeable.

For improved quality, Creative implemented “auto-init” DMA where the DMA buffer was read from or written to in an endless loop, signaling interrupts along the way as directed. This enabled software to implement double buffering and glitch-free playback with no gaps.

The newer Sound Blaster 2.0 model (CT 1350) used DSP version 2.01 or later. Besides auto-init, the 2.01+ DSPs also supported “high-speed” DMA with sampling rates up to 44.1 kHz. The 1.x and 2.00 DSPs were limited to about 23 kHz sampling rates. Whether this made any real difference for noisy 8-bit sound cards is questionable, but Creative clearly thought it worthwhile.

To make matters more interesting, the DSP on the old Sound Blasters was socketed (see photo above) and Creative offered DSP 2.00 upgrades to Sound Blaster 1.x owners. Therefore even an older Sound Blaster might be equipped with a newer DSP. What this meant was that software requiring auto-init DMA might or might not work on a Sound Blaster 1.x, depending on which DSP was installed.

Confusing? Heck yes, and that was just the beginning.

This entry was posted in Creative Labs, PC hardware, PC history, Sound Blaster. Bookmark the permalink.

17 Responses to Sound Blaster 1.0 or 1.5?

  1. John Elliott says:

    Notice the small “VERSION 1.5″ label in the lower right side of the box

    Lower left, surely?

  2. Michal Necasek says:

    Gremlins, surely! Thanks.

  3. Joshua Rodd says:

    What’s the difference in hardware between the 1.x and 2.01 DSPs?

  4. Michal Necasek says:

    The hardware itself was more or less the same (some kind of MCS-51) but the card design was substantially different. The Sound Blaster 1.x (DSP 1.x or 2.0) basically didn’t have any kind of bus interface unit and the DSP directly interfaced with the ISA bus. Not unlike, say, the MCS-48 (aka 8042) keyboard controller. Every sample was lovingly processed by the DSP.

    The Sound Blaster 2.0 had a dedicated bus interface unit sitting between the ISA bus and the DSP. I assume that is what enabled Creative to implement the high-speed DMA transfer modes. I haven’t seen the 2.01 DSP microcode but I’m guessing the bus interface unit shuffled the samples around at least for the high-speed DMA modes.

    Just guesswork really.

  5. Chris M. says:

    I have a Soundblaster CT-1320C with the CMS chips (with 301 stickers), but I don’t have the box. It came with the CMS software though, and a CMS programming guide. I always suspected it was a plain old 1.5 card that got the CMS upgrade. Creative sold a CT-1320A with the CMS chips soldered on and a PCB dated 1989. Some say this is a true 1.0 card.

    http://img40.imageshack.us/img40/7463/slja007.jpg

  6. Michal Necasek says:

    If yours came with C/MS software, it must have been sold as a SB 1.0 (that is, not 1.5). What DSP revision does it have? And is there anything interesting in the C/MS programming guide?

    The box in the photo sure looks exactly the same as mine. Other than the soldered vs. socketed chips, I couldn’t spot any real difference between the 1320A and 1320C. Whatever difference there was must have been very minor. Much less than the difference between some of the later Sound Blasters with the same model number!

  7. Chris M. says:

    Mine appears to be a 1.5. The card has DSP 2.0 on it (chip date code is 1991, no DSP sticker), and the manual states Version 1.5 on the first page. There is also a card with CMS chip installation notes for Sound Blaster 1.5 cards, so it appears this was indeed a 1.5 card with a real CMS chip upgrade. The programming information for CMS just goes over registers and such, basically a copy of the Philips SAA-1099 datasheet.

  8. Michal Necasek says:

    The DSP chip alone means little, because a) it was upgradable, and b) I doubt anyone can say for certain that Creative never shipped any Sound Blasters with C/MS chips in the original box and with the 2.00 DSP.

    But if you have the SAA chip installation notes then that sounds like a very convincing proof the board is an upgraded/downgraded SB 1.5.

    If you could scan the documentation that would be terrific (I’d be happy to do PDF production if needed). At least I don’t think those documents are available online anywhere.

  9. Raúl Gutiérrez says:

    Just curious… did the SB 1.5 sell for a higher or lower price than the 1.0?

  10. Abrahan says:

    I have one of this 😀

  11. Michal Necasek says:

    Excellent question. I don’t know the answer. I suspect the pricing did not change immediately, but the price was dropped significantly when SB Pro came out. I’ll do more digging.

  12. lys says:

    Has someone analyzed what type of ADC version 1.0 and 1.5 SB is using?
    DAC is clearly MC1408 but there’s no separate ADC chip, so ADC function is created by using different components.

    DAC and ADC are independent from the so called “DSP”, since there’s written that sample rate doesn’t change replacing the DSP.

    “Owners of previous revision Sound Blaster boards could upgrade their board by purchasing the V2.00 DSP chip from Creative Labs, and swapping the older DSP V1.0x with the newer replacement. The upgraded board gained the auto-init DMA and new MIDI capabilities of the Sound Blaster 2.0 but not the expanded sampling rates.” – source – wiki SB article.

    I noticed there’s comparator chip LM393, however there’s different ADC types (schematics) using comparator. Early prototype is also using the same comparator.

    http://www.vgmpf.com/Wiki/index.php/Sound_Blaster

    Have someone analyzed the ADC part so far?

  13. lys says:

    ps. regarding the differences between ver 1.0 and 1.5 vgmpf states (and those guys know a lot) that CT-1320A is ver 1.0 and CT-1320B is ver 1.5. The difference is how Intelligent Organ is implemented – ver 1.0 uses card CM/S part, ver 1.5 uses OPL2.

    http://www.vgmpf.com/Wiki/index.php/Sound_Blaster

  14. Michal Necasek says:

    I stand by what I wrote: The difference between Sound Blaster 1.0 and 1.5 is the C/MS chips. That means the CT-1320A PCB can only be SB 1.0 (soldered-on C/MS chips, can’t be SB 1.5). However, the article provides evidence that SB 1.0 was also sold with the CT-1320C PCB, and I see no reason why CT-1320B also couldn’t be sold as SB 1.0. In short, CT-1320B/C without C/MS chips is SB 1.5. CT-1320B/C with C/MS chips is SB 1.0, and at least some of them were delivered that way from the factory (the serial number proves that).

    Summary: CT-1320A is SB 1.0 but CT-1320B/C can be either SB 1.5 or 1.0.

  15. Michal Necasek says:

    I’m not an expert at all. But I would guess that yes, the LM393 is part of the discrete ADC on the old Sound Blasters. Maybe it’s some successive approximation or pipelined ADC. The resistors (R10-R18 or so) are likely involved, and I’d guess the Harris CA324E op-amp is part of the circuit as well. There may well be fewer than 8 bits of precision in the DAC. It’s a toy, not even a consumer-grade device.

    Having a firmware dump would help a lot, but I’m not aware of one.

    Question: Why does upgrading the DSP chip not increase the sampling rates? How does the DSP know to limit them? Or is it only a practical limitation of the ADC circuit?

  16. lys says:

    SB Series Hardware Programming Guide [1] states that one of the function of the DSP chip is controlling the sample rate (page 15), however i think the limit is coming from the DSP chip itself and probably also surrounding A/D and D/A implementation. The DSP should output and input samples at appropriate rate, maybe the new 44.1kHz/15 kHz rate introduced with SB 2.0 along with the new DSP 2.0 code is just too much for that poor old intel 8051 which had to do other things too?

    [1] https://pdos.csail.mit.edu/6.828/2014/readings/hardware/SoundBlaster.pdf

    There’s also one A/D converter design that uses feedback from the D/A converter to the comparator. Unfortunately i don’t own any SB 1.x cards, it would be interesting to analyze and recreate the analog path of the signal on those cards on the paper. Using just pictures it’s quite hard. The first op-amp is clearly microphone pre-amp though. There’s several AdLib projects around but none of the old SB 1.x projects…

  17. Michal Necasek says:

    But the DSP is the 8051. If you replace the DSP, you also replace the microcontroller. So that can’t be the reason. The conversion circuits could be. It now also occurred to me that the old boards might have trouble with faster DMA.

    Now I re-read what you quoted and there’s actually no mystery. The replacement DSP was version 2.00 (my SB 1.5 has it), but SB 2.0 uses DSP 2.01 or later (mine has 2.02). DSP 2.00 includes auto-init DMA and MIDI UART commands, but does not support high-speed transfers like 2.01+ does. I’m guessing the faster DMA needed the CT1366 bus interface chip which SB 2.0 (and before that, SB Pro) used.

    About the ADC. With just two comparators, we know it’s not a flash ADC. It could be a successive approximation ADC or it could be a digital ramp ADC, or perhaps something else. At least those two types need a comparator and a D/A converter.

    What I also just noticed is that the analog audio section on the Sound Blaster 2.0 is rather similar to the 1.x boards. A 393 comparator, MC3403 op-amp (two actually), MC1408 DAC. The biggest difference is a TDA7284 (op-amp with automatic level control) instead of the Harris CA324E.

    I could do some measurements on my SB 1.x if you wanted to know what’s connected where, although my electronics skills are very limited.

Leave a Reply

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