VOGONS


New sound card project: AWE64 Legacy

Topic actions

Reply 121 of 378, by appiah4

User metadata
Rank l33t++
Rank
l33t++

For ISA I think Terratec cards (EWS, Maestro, Gold) are unrivaled but to be fair I have to admit that I'm not familiar with Turtle Beach's ISA offerings first hand. I'll hand them this, though, Turtle Beach's PCI cards are overall very good hardware.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 122 of 378, by MrSmiley381

User metadata
Rank Newbie
Rank
Newbie

Dude, what's with all these DIY sound cards? This is too awesome. Major props for implementing the intelligent mode MPU-401. Every standard one of these cards includes is always appreciated.

So aside from the CSP/ASP chip used by TFX (an old flight sim) and the option for non-PNP support, is there any real advantage to using an AWE32 over this card? I totally want one regardless. Those boards look sleek.

I spend my days fighting with clunky software so I can afford to spend my evenings fighting with clunky hardware.

Reply 124 of 378, by Schyz

User metadata
Rank Newbie
Rank
Newbie
MrSmiley381 wrote:

So aside from the CSP/ASP chip used by TFX (an old flight sim) and the option for non-PNP support, is there any real advantage to using an AWE32 over this card? I totally want one regardless. Those boards look sleek.

Some AWE32, with chip CT1747 that integrates a real OPL3, have the capability of playing the FM sound digitally without any analogue conversion, is the only way I'm aware of getting pure, untouched OPL3 sound. Also it's possible to pass the OPL3 sound through the EMU8000 and apply effects to it, again, digitally.

This cards would be the perfect SB16-compatibles if not for the "single cycle DMA clicking noise" (not present in AWE64). Also, some of them are quite noisy compared to the AWE64.

Reply 125 of 378, by Deksor

User metadata
Rank l33t
Rank
l33t
kolderman wrote:

Do we have to donate our awe64 to get one? I am currently using a Gold and would never destroy that but I have a number of plain ones. How much will it cost?

This, I have a AWE64 value that doesn't work right (one of the two stereo channel is silent for some reasons) so harvesting it's chip shouldn't be an issue since the card's unusable at the moment.

Reply 126 of 378, by truemaster

User metadata
Rank Member
Rank
Member

yes that a good question how much the cost will reduce if we provide our awe64? also do you plan to create awes from other models exept 4520 example ct4380 ct4500 or even the gold ct4390? just before i see this post i orderd a value ct4380 and its good. and with this redisign it will be AWEsome

Reply 127 of 378, by dr_st

User metadata
Rank l33t
Rank
l33t

Good question. CT4520 is the only version, AFAIK, that has the CT8920 single-chip. All others use a two-chip solution - CT8903 DAO and CT1745A DAC/ADC. Integrating multiple chips on a new PCB may be more difficult.

https://cloakedthargoid.wordpress.com/ - Random content on hardware, software, games and toys

Reply 128 of 378, by Eep386

User metadata
Rank Newbie
Rank
Newbie

The YMF289B does not have I2S compatible output data, it's right justified. That said, I've seen a Gateway socket 5 motherboard with built-in ViBRA16S and YMF289B-S use a CS4333 DAC chip instead of the YAC516. The CS4333 has 2dB better dynamic range and slightly better stopband attenuation, but unfortunately the pinout is radically different from the YAC516. Also, it's IIRC out of production just like the YAC516. 🙁

That said, the main benefits of the YMF289(B) over the YMF262 are lower power consumption, lower sample rate (a double-edged sword as while the 44.1 kHz output rate is theoretically easier to mux into SP/DIF, it also means that aliasing might be slightly more noticeable on very high frequency sounds vs. the YMF262) and a smaller overall board footprint. And if a (cheap) source of CS4333's can be secured and made to work, the footprint of the genuine OPL3 part can be made really small in theory. (CS4333 is a teeny 8-pin SOIC the size of an SOIC LM358.) Plus, my first OPL3 FM was a YMF719, which uses the YMF289 FM core internally so to me, the YMF289's output is most familiar to me. 😀

To clarify, I'm not saying use YMF289 _over_ YMF262, just maybe make it an option (have pads for both YMF262 and YMF289) if it isn't too much of a hassle.

My synthesizer hierarchy: Novachord > DX-7 > YM2151 > YM3812 > CT1747/YMF262/YMF289 > CQM > PC Speaker > AudioPCI "FM" (retch)

Reply 129 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie

Hello guys!

I guess it's time for the second major update.

First, let me say a big THANKS for all you supporters who have filled in our questionnaire. It has really shed a lot of light on your expectations, that will let us focus on the more important features while discarding the less desired ones. The questionnaire has been just shut down.
The first post of this thread has been just updated with our current status, I'm copying it here:

IMPLEMENTED & TESTED
ALL BASIC FUNCTIONS
CT8920+EMU8011
SIMMCONN+SIMM72
WAVEBLASTER HEADER (IN PARALLEL WITH LINE-IN)
2xRCA OUT
OPL3 (YMF262+YAC512, IN PARALLEL WITH CD-IN)

IMPLEMENTED, BUT NOT TESTED YET
GAMEPORT
S/PDIF OUT

TO IMPLEMENT
INTERNAL MPC HEADER FOR CD-IN
INTERNAL MPC HEADER FOR LINE-OUT
INTERNAL 2-PIN HEADER FOR SPK-IN (PCSPK)
FRONT PANEL CONNECTORS HEADER (AC97 PINOUT, MIC IN AND SPK OUT WILL MOVE THERE FROM THE MAIN BRACKET)
MOVE GAMEPORT BACK TO THE MAIN BRACKET
CHECK MIDI I/O PROTECTION/BUFFERING AND CURRENT LOAD LIMIT
ONBOARD S/PDIF OUT LEVEL CONVERTOR (WHILE ALSO LEAVING THE TTL AS A 2-PIN HEADER ON THE PCB)
METAL BRACKETS (PORTS FROM THE BOTTOM TO THE TOP: GAMEPORT/MIDI, RCA S/PDIF OUT, 2 X RCA LINE-OUT, 3.5 MM TRS LINE-IN)
INTELLIGENT MPU-401 ONBOARD, HARDMPU DESIGN (CREDITS: ab0tj)
SWITCHES/JUMPERS TO CHOOSE THE ROUTING OF THE MIDI OUTPUT FROM BETWEEN BOTH CT8920 / HARDMPU PART AND WAVEBLASTER HEADER / GAMEPORT MIDI PINS
SWITCHES/JUMPERS TO DISABLE THE OPL3, HARDMPU (AND CHOOSE ITS IRQ / IO PORT)
FREE SPACE ON THE PCB FOR SOME ADDITIONAL STAND OFFS FOR THE DAUGHTERBOARD
TB RIO COMPATIBILITY
RETURN THE JP2 (MFBEN) FROM THE ORIGINAL CT4520
REASSIGN THE PARALLEL INPUT PAIRS: CD-IN + WAVEBLASTER; LINE-IN + OPL3

HELP NEEDED
RESOURCE MAP (EEPROM FIRMWARE) FOR CQM AT 0x398H

DISCARDED DUE TO HIGH COMPLEXITY / LACK OF INTEREST / USELESSNESS
NEW S/PDIF IMPLEMENTATION
NEW ROM
ADDITIONAL WAVEBLASTER HEADER
MEDIATRIX 3D-XG STYLE FOR WAVEBLASTER HEADER ANALOG SIGNAL ROUTING (DB60XG COMPATIBLE)
LED

Although I have already answered this question several times, I think I'll put it once again:
Q. Do you need my CT4520 card? Anything else?
A. No, we don't need your CT4520 yet. Please keep it, maybe you'll be entitled for some discount (no idea of how much) when the official sales start. We don't know when they do, nor do we have any estimated, rough, indicative, whatever else timing/pricing for that. No pre-orders, no need for submitting/donating any components, cards, chips, money, etc. Stay tuned and hoard your CT4520, that's all I can suggest at the moment.

Anticipating another question:
Q. Why have you chosen the HardMPU design instead of the MusicQuest one?
A. HardMPU is a bit less PCB space consuming, and that was our major requirement for this additional feature. And its lack of MIDI recording feature does not matter much as the Creative built-in MPU-401 should do the job. But if you can prove the opposite, please drop me a line, and we'll think it over once again.

And, to put some cherry on the cake...
All you guys who still believed that Creative was a smart and/or caring company, we have to disappoint you once again 😀
While we were testing the analog part of our prototype card, we've come across some problems with the mixer. We started to debug it, compare our schematics against the original CT4520, etc. etc. etc., and finally discovered some strange behaviour of the Line In jack on CT4520. In fact, it did mono while it was clearly supposed to be a stereo input. CT4390 acted the same. We've become curious, traced the signal, and had to desolder the jack, and...

CT4520 Line In.jpg
Filename
CT4520 Line In.jpg
File size
111.8 KiB
Views
538 views
File license
Fair use/fair dealing exception

...we've found a jumper trace there, shortening L and R channels of the stereo input! Meanwhile, when we desoldered the Mic In, in has turned out to be "stereo":

CT4520 ports.jpg
Filename
CT4520 ports.jpg
File size
103.3 KiB
Views
538 views
File license
Fair use/fair dealing exception

So, Creative has clearly confused the Line-In and Mic-In jacks on the board, despite the silkscreened markings! What a shame! And all you users who tried to feed the output of your GUS/TB/Rolands/etc. into your AWE64 (no matter if it was a Value or a Gold), in fact got pseudo-stereo due to this error. Moreover, when this trace is cut, the Line-In functions perfectly as a true stereo input.[/i]

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)

Reply 131 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie

At least CT4520 and CT4390 are confirmed.

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)

Reply 133 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie

Unfortunately no. It does only EMU8000+CQM (however we haven't tested that ourselves yet, I'm telling the experience of other users here).

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)

Reply 134 of 378, by derSammler

User metadata
Rank l33t
Rank
l33t

So, Creative has clearly confused the Line-In and Mic-In jacks on the board, despite the silkscreened markings! What a shame! And all you users who tried to feed the output of your GUS/TB/Rolands/etc. into your AWE64 (no matter if it was a Value or a Gold), in fact got pseudo-stereo due to this error. Moreover, when this trace is cut, the Line-In functions perfectly as a true stereo input.

"L" and "R" refer to both solder points each, that's why the silkscreen is centered. And as you can clearly see from the other connectors, the outer two pins are the stereo channels, the inner ones are GND. Connecting the two GND lines does not turn stereo into mono. How can you even think Creative was that dumb and no one ever noticed that line-in is mono only in the past 23 years?

Reply 135 of 378, by kolderman

User metadata
Rank Oldbie
Rank
Oldbie
hard1k wrote:

Unfortunately no. It does only EMU8000+CQM (however we haven't tested that ourselves yet, I'm telling the experience of other users here).

Will the DACs on the board at least be very high quality ones?

Reply 136 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie
derSammler wrote:

How can you even think Creative was that dumb and no one ever noticed that line-in is mono only in the past 23 years?

Oh yes, they were. We couldn't believe our expectations too. The inner and outer pins are shortened in pairs (on all 4 jacks), and this jumper on the Line-In make both L and R pairs shortened too. Feel free to take a card and test it yourself.

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)

Reply 137 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie
kolderman wrote:

Will the DACs on the board at least be very high quality ones?

The DAC is integrated into the CT8920 chip, there is no known way to get the digital signal out of it.

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)

Reply 138 of 378, by kolderman

User metadata
Rank Oldbie
Rank
Oldbie
hard1k wrote:
kolderman wrote:

Will the DACs on the board at least be very high quality ones?

The DAC is integrated into the CT8920 chip, there is no known way to get the digital signal out of it.

But isn't that how the spdif gets out for EMU/CQM? Or can you only DAC the I2S or something probably...

Reply 139 of 378, by hard1k

User metadata
Rank Oldbie
Rank
Oldbie

That's where the problem is - you get the digital signal only for EMU/CQM, but not the waveout. Waveout leaves CT8920 only in analog.

New sound card project: AWE64 Legacy
Please have a look at my wishlist (hosted at Amibay)