j^aws wrote:

fantastic work Fagear - very professional!

Thanks.

j^aws wrote:

, what soldering iron were you using? I think I need to upgrade mine.

goot PX-201 (~230V 70W, temperature control)

j^aws wrote:

You mention daughter boards - what plans/ ideas do you have for those? Something like DSPs for effects?

Well, I have thought about typical power amplifier. Some 1-2W mono IC, pair of capacitors and mini-jack.
There was mentioned SPDIF-out converter on phantom.sannata.ru as possible option.

There are digital power (+5V, GND) pins (marked "PWR_VCC"), analog power (+9 or +12 V) pin (marked "PWR_VDD") and audio output pins (two pins marked "AUDIO OUT").

Fagear wrote:

goot PX-201 (~230V 70W, temperature control)

Cheers, I'll check it out.

Fagear wrote:

Well, I have thought about typical power amplifier. Some 1-2W mono IC, pair of capacitors and mini-jack. There was mentioned SPD […]
Show full quote

Well, I have thought about typical power amplifier. Some 1-2W mono IC, pair of capacitors and mini-jack.
There was mentioned SPDIF-out converter on phantom.sannata.ru as possible option.

There are digital power (+5V, GND) pins (marked "PWR_VCC"), analog power (+9 or +12 V) pin (marked "PWR_VDD") and audio output pins (two pins marked "AUDIO OUT").

Both those are interesting tweaks: The power amp section can add some flavour to the sound, especially if you can easily swap out the components and experiment. And the SPDIF-out sounds like a great way to feed that into an external effects processor!

Shut up and take my money already!

j^aws wrote:

And the SPDIF-out sounds like a great way to feed that into an external effects processor!

I think I'll leave that thing to community. 😀

Great Hierophant wrote:

Shut up and take my money already!

Well, I'm low on some parts now, I have parts for only three boards now (of 30).
Parts are ordered, waiting for them.

I think in some time I'll start to contact everybody about payment.

Now I want to release some "technical" info about available SID clocks.

There is jumper to the left side of the board. And some legend for it above (about selecting 0.89 or 1.02 for different counter ICs).
But you can get some other frequencies.

Full list:

174HC192:
2none:  1,193 MHz (14,31818/12)
3left:  0,895 MHz (14,31818/16), like original SSI-2001
4right: 1,023 MHz (14,31818/14), like C64 NTSC
5both:  0,796 MHz (14,31818/18)
6
7555ИЕ6 (555IE6):
8none:  1,023 MHz (14,31818/14), like C64 NTSC
9left:  0,796 MHz (14,31818/18)
10right: 0,895 MHz (14,31818/16), like original SSI-2001
11both:  0,716 MHz (14,31818/20)

And I have some examples of sound: Lambada.sid and 1988_Top_40_MEGAMIX.sid.
Real hardware, new board with MOS6581R4 in it. Playing *.SID file with bristlehog's alfa-version software.

^^ Nice, sounds of the eighties SID-stylee... all you guys involved in this project rock! Shout-out to bristlehog on the software side to make all this complete. Nostalgia trip...

BTW, 74HC192 looks like the most balanced counter IC; you can go a little faster or a little slower than the two standards, whilst the 555IE6 has a slower low-end setting. Which one will be supplied to us?

j^aws wrote:

BTW, 74HC192 looks like the most balanced counter IC; you can go a little faster or a little slower than the two standards, whilst the 555IE6 has a slower low-end setting. Which one will be supplied to us?

Well, 555's I can easily buy at local store and 74's I have to pre-order. And 74's cost three times more than 555's.
So most ICs I have are 555's.

On the other side SID datasheet says that 1.0 MHz is maximum frequency (even 1.02 is a little bit high).
So 1.19M will be overclock.

Fagear wrote:

Well, 555's I can easily buy at local store and 74's I have to pre-order. And 74's cost three times more than 555's.
So most ICs I have are 555's.

If it's easier for you, and cheaper to build, then go for the 555s. Get a standard base model built and delivered. Of course, I don't know what everyone else thinks, but I personally don't mind the 555s, and can always upgrade later if I ever feel the need to. Although, I don't know how much the 74s would cost as a single unit, rather than buying in bulk.

Fagear wrote:

On the other side SID datasheet says that 1.0 MHz is maximum frequency (even 1.02 is a little bit high).
So 1.19M will be overclock.

I never realised the NTSC SID implementaion was running slightly out of spec... Hmm, the thought of strapping a heatsink to a SID did briefly cross my mind, but a 19% overclock sounds absurd for very little gain and risk of frying it.

I suppose there is no way to implement odd divisors without redesigning the circuit, so that a divide by 15 would be available, which would result in a true C64 PAL frequency.

Great Hierophant wrote:

I suppose there is no way to implement odd divisors without redesigning the circuit, so that a divide by 15 would be available, which would result in a true C64 PAL frequency.

No way.
Divider is made of two ICs: 74HC192 (/8) and 74LS74 (/2 with 50% duty cycle).
74LS74 will always give you even divider.
74HC192 can not count above 9. So /15 is unavailable without re-designing circuit.

I have a small thought about another project... ISA-8 board, blue color, SID, compatible with SSI-2001, without stupid joystick port, all SMD (except of SID), odd divider support, better filtered power rails to SID. But that's just a though. 🤣

How about SSI & HardSID compatibility? That way a bunch of software could be re-used.
http://hardsid.8m.net/

Fagear wrote:

I have a small thought about another project... ISA-8 board, blue color, SID, compatible with SSI-2001, without stupid joystick port, all SMD (except of SID), odd divider support, better filtered power rails to SID. But that's just a though. 🤣

A cool project would be a Tandy 1000 sound compatible ISA card. 😎

carlostex wrote:

A cool project would be a Tandy 1000 sound compatible ISA card. 😎

I'm not really familiar with Tandy 1000... What "wrong" with it? Where can I read something about it? I mean, what's so special about it in the way of sound card?

In the mean time I have tested joystick port under Win98SE (had to force install Gameport driver at port 201), everything works fine.
I will get ordered parts tomorrow.

Fagear wrote:

I have a small thought about another project... ISA-8 board, blue color, SID, compatible with SSI-2001, without stupid joystick port, all SMD (except of SID), odd divider support, better filtered power rails to SID. But that's just a though. 🤣

How hard would it be to have mountings for 2 SIDs, both 6581 and 8580, so that the variations in these SIDs can give a pseudo-stereo effect?

Regarding Tandy 1000: IIRC, I don't think there was a discrete card released to the public, although there are plenty of sound options for it in games.

Fagear wrote:

I'm not really familiar with Tandy 1000... What "wrong" with it? Where can I read something about it? I mean, what's so special about it in the way of sound card?

http://en.wikipedia.org/wiki/Tandy_1000#Desig … nd_architecture
http://en.wikipedia.org/wiki/Texas_Instruments_SN76489

The Tandy 1000 was an IBM PC compatible line of computers from a US company called Tandy Radioshack. The first ones were pretty much IBM PCjr clones which included a single Texas Instrument SN sound generator. This chip was included in the motherboard and it is not an expansion sound card. But an ISA card with the same sound generator should be possible. There are some caveats because Tandy 1000's used 8 bit bus and only 1 DMA controller.

Because Tandy 1000 will register at C0H it will conflict with the slave DMA controller. What could be done as ,Cloudschatze suggested is to place the supposed Tandy card at 1E0 because this is where it sits on the Tandy 1000 RSX and then have a TSR that redirects the C0 calls.

This and forcing the games via command line arguments to use EGA instead of the expanded CGA graphics modes of the PCjr and Tandy, would allow PC clones to play older games with the 3 voice generator instead of the PC speaker.

The idea here is not to clone a card but to create one from scratch, as there was never a commercial ISA card that gave 3 voice generator sound on PC compaibles that were not Tandy 1000's. And the value of such a device would be enourmous, as many older games have support for the 3 voice sound generator of the Tandy 1000 and PCjr. In other words this would NOT be a cloning project, but engineering a new card.

However you would probably need some help as you obviously have no knowledge of the Tandy 1000 hardware. But i'm sure that with proper help and your talent this could be done very well. In comparison, the Innovation SSI and the Covox Sound Master are much more complicated circuits than what a Tandy 1000 compatible sound card would be.

carlostex wrote:

Because Tandy 1000 will register at C0H it will conflict with the slave DMA controller. What could be done as ,Cloudschatze suggested is to place the supposed Tandy card at 1E0 because this is where it sits on the Tandy 1000 RSX and then have a TSR that redirects the C0 calls.

NewRisingSun let me know that EMM386 will not allow port trapping in the 00h-FFh range, thereby precluding it from being used for the Tandy sound redirection. Presumably, JEMM and QEMM have the same restriction. 😢

j^aws wrote:

How hard would it be to have mountings for 2 SIDs, both 6581 and 8580, so that the variations in these SIDs can give a pseudo-stereo effect?

AFAIK you can install two replicas, one with 6581 and other with 8580, to use the same I/O address, and they will play simultaneously. Fagear will correct me, if I'm wrong.

Cloudschatze wrote:

NewRisingSun let me know that EMM386 will not allow port trapping in the 00h-FFh range, thereby precluding it from being used for the Tandy sound redirection. Presumably, JEMM and QEMM have the same restriction. 😢

Really? Then does the Tandy 1000 RSX have a limited compatibility with older Tandy 1000's?

Then i'm left wondering how DOSBox does its Tandy emulation. Then there's also the other Tandy emulation TSR called TEMU that i think worked on a Covox Speech Thing. Maybe we could do some better emulation via DOSBox code, but it's way disappointing to know that it might be impossible trying to make a real card with a real sound generator.

bristlehog wrote:

j^aws wrote:

How hard would it be to have mountings for 2 SIDs, both 6581 and 8580, so that the variations in these SIDs can give a pseudo-stereo effect?

AFAIK you can install two replicas, one with 6581 and other with 8580, to use the same I/O address, and they will play simultaneously. Fagear will correct me, if I'm wrong.

Yep.
You can install two (or more) replicas. That's how it sounded with old PCB revision and two 6581R4.

shock__ wrote:

How about SSI & HardSID compatibility? That way a bunch of software could be re-used.

I've downloaded some HardSID-compatible software, I'll try to test it later. My Win98 test machine is down now...
P.S. shock__, I've sent the PM to you.

I've got part for replicas... Starting to assemble boards and preparing kits...

HardSID cannot be compatible with SSI-2001, it's based on more complicated logic and has hardware means for time synchronization.

carlostex wrote:

Then does the Tandy 1000 RSX have a limited compatibility with older Tandy 1000's?

Unfortunately, yes. Today anyhow, Tandy sound support with the 1000 RSX or 2500-series systems requires native use of the 1E0-range addresses. I don't know that anyone has ever compiled a list of such supporting titles, but Sierra's SCI0 through SCI1+ games would be included, as would a number of Brøderbund, Virgin, and Disney games.

...it's way disappointing to know that it might be impossible trying to make a real card with a real sound generator.

I wouldn't say it's impossible; it's just not as simple as leveraging EMM386's port-trapping capabilities in a manner similar to SoftMPU, as I'd suggested earlier. It does seem like TEMU's redirection could be used as a model...