VOGONS


My Sound Blaster 16 CT2230 Review

Topic actions

First post, by James-F

User metadata
Rank Oldbie
Rank
Oldbie

My SB16 CT2230 Review

CT2230.jpg
Filename
CT2230.jpg
File size
305.52 KiB
Views
13647 views
File license
Fair use/fair dealing exception

I have received my Sound Blaster 16 model CT2230 Rev.49432.
DSP v4.13, CT1745A-S mixer, CT1747 OPL3, CT1703-A DAC.

Test points:
* Distortion in high volume games?
* Oscillation in SBPro high-speed Stereo mode?
* Clicking after sample playback?
* Noise from Wavetable header amplifier?
* Low Harmonic Distortion playing -0.1dbFS 440Hz sinewave in 8bit and 16bit.
* Capable 8-bit to 2/3/4-bit ADPC.
* Present Dynamic Lowpass filtering.
* Working Mixer?
* Authentic OPL3.
* Bug-free MPU-401?
* Low noise, above 70 SNR?

Let me start by saying that this card is absolutely silent.
I get 76dbFS of SNR just like my CT2950 (older CT1745A mixer) or later Vibra16S CT2890 with the proper mixer settings.
I have no clue where the rumor that CT2230 cards are noisy came from but it's absolutely 100% false.
EDIT: Inthis post I think I may found the biggest culprit for noise:
My Sound Blaster 16 CT2230 Review

Proper mixer settings would be: everything muted to minimum but Voice, MIDI, Master, Bass, Treble and all X's unticked,.
Voice and MIDI (FM) set to full and Master set one click below full to avoid distortion, Output Gain-L/R set to x1.
These settings produce outstanding SNR with crystal clean distortion-less audio path.

There is no distortion with high volume games, unlike the Vibra16S card I have which distorts no matter what settings in the mixer.
The CT2230 produces exactly the same sound as the CT2950, measured using VST frequency analyzers and accurate VU meters.
The audio path is very clean and has very low harmonic distortion both in OPL, PCM sound, and Wavetable return.
The Dynamic Lowpass Filter which is unique to the SB16 is working and makes low sampling rate games sound good.
fjsxkm4s6srb2syzg.jpg
27t9af2z9katofmzg.jpg
0aw965am7ay76vfzg.jpg
0fle32034t1jhelzg.jpg

There is a click sound in Prince of Persia after the footsteps sample, other games have similar click sound like Mortal Kombat 1, Skyroads and Supaplex, not present in all games.
This click sound is a residue after low sampling rate PCM playback, it is absent from YMF71x, ES1688, Vibra16S, and DOSBox.
Note that this click is very quiet and some may never notice it, I guess this is just how the stand alone SB16 DSP works.

When the CT2230, CT2950, Vibra16S or any other SB16 card (I believe so) is forced into high-speed mode (SBPro Stereo mode) which they don't support,
they oscillate and make a high pitched noise when sampling rate is less than 44100Hz and if the game plays sound not from the absolute Center (mono).
The oscillation is heard the most when the game/program tries to play sounds not from Center but from Left or Right, this can be tested in Descent setup in SBPro mode.
No game that I know of forces the SB16 into SBPro Stereo mode without having a SB16 option in their setups, if you have the option try to completely avoid SBPro stereo option.
I assume that it is also beneficial/important not to set to T4 (SBP) in the BLASTER variable but the proper T6 for SB16.
I used MPXPlay for DOS, forced to SBPro Stereo Mode, 11kHz sampling rate, with a mono 440Hz.wav panned hard left to produce this screenshot.
5whoiiqnohnglcqzg.jpg

OPL3 sounds great and is balanced with the PCM sound when both set the same level in the mixer, so older SB1/2 games sound as intended without too much hassle.
Functional 8-bit to 2/3/4-bit ADPC for games (a game?) like Duke Nukem 2.
No hanging note bug using the MPU-401 with the CT1747 chip but still stutters in Duke3D and Tie Fighter, I just use the YMF71x for MPU-401 only and be done with it.
I also have my DreamBlaster S1 wavetable plugged into the YMF71x and looped back into the Line-In which I enable only when I want to hear it.

The only differences I have found between the CT2230 and CT2950 is the OPL3 vs CQM and absent Hanging Note Bug on the CT2230, all else is exactly the same in terms of Audio and Noise.
The CT2950 is Full-PnP which requires CTCU whether the CT2230 is Semi-PnP and only requires BLASTER and Diagnose.exe /s to initiate.
Overall, this is a great sound card and I recommend it for High Compatibility, Authentic OPL3, Bug-free MPU-401, balanced PCM to FM, and Low Noise.

A side note:
Line-Out should see a proper load of 10KΩ (10,000Ω) by standard, so only connect to a proper load like a Stereo System, Amplified Computer Speakers, or other Line-In.
Do NOT connect headphones to the line-out, they are 16Ω-32Ω (typical) which will not properly load the SB line-out and you'll lose bass frequencies (confirmed), distort the DAC, or worse.
You can and should connect your headphones to the Speakers jack because it expects low impedance and has the proper current supply by the two big mosfets you see on board.

Last edited by James-F on 2016-09-13, 09:36. Edited 10 times in total.


my important / useful posts are here

Reply 1 of 65, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++

Sweet, I just recently got this card. DonutKing sold me a retro care package 😁

It looks like yours, but instead of the socket next to the large chip, it has a second chip.

YouTube, Facebook, Website

Reply 2 of 65, by badmojo

User metadata
Rank l33t
Rank
l33t

Very nice. I'll just add that the CT2290 has exactly the same fine qualities as the CT2230 in my experience, with the only difference being that the CT2290 is missing the socket for an ASP.

Life? Don't talk to me about life.

Reply 3 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie

There are at least 3 revisions of the CT2230 as pointed here: Duke Nukem 3D stuttering with Dreamblaster X1
I believe all of them are as good and quiet as the next; they just need proper mixer settings and you get the best SB16 model without much lottery.

EDIT: Yes, the CT2290 is an ASP and IDE stripped CT2230, and just as good.


my important / useful posts are here

Reply 4 of 65, by GuyTechie

User metadata
Rank Member
Rank
Member

I'm going to try those mixer settings. Mine has a hiss once it's initialized. But as we chatted in a different thread, mine has a different mixer chip that the 3 known variants.

http://imgur.com/DmQyuxv

It has a CT1745-S (note the missing "A")

Reply 5 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie

I believe the mixer chip has little effect on the noise level because my CT2950 has the CT1745A which is the oldest and considered to be the noisiest yet still achieve excellent SNR.

Untick all the X marks in the Output section and disable AGC, that should get you started.
I also recommend Voice and MIDI (OPL3) at full and Master one click down, for maximum SNR before distortion.

Start like so, add thing later only if you need to.

Mixer.png
Filename
Mixer.png
File size
2.01 MiB
Views
13596 views
File license
Fair use/fair dealing exception
Last edited by James-F on 2016-09-12, 16:33. Edited 3 times in total.


my important / useful posts are here

Reply 6 of 65, by boxpressed

User metadata
Rank Oldbie
Rank
Oldbie

Excellent review. Time to give my CT2230 a whirl. Which set of drivers did you use?

The completist in me wants to know what retail packaging accompanied this card. Does anyone have one CIB? I saw one on eBay sold alongside a Creative CD-ROM drive, so it may have been part of a kit at one point.

Reply 7 of 65, by keropi

User metadata
Rank l33t++
Rank
l33t++

nice review, the CT2230 is my SB16 of choice several years now - I even overkilled it and recapped the one in my main ss7 build 🤣 🤣 🤣

WP_20150619_005_zpsqycdga1b.jpg

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 8 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie
boxpressed wrote:

Excellent review. Time to give my CT2230 a whirl. Which set of drivers did you use?

Thanks!
Driver are from Vogons Drivers, Sound Blaster 16 Pre-Installed for DOS 7.
Autoexec.bat:

SET BLASTER=A220 I7 D1 H5 T6
DIAGNOSE.EXE /S

I got the MPU-401 disabled by un-jumpering MPUEN.
No fussing around, it just works. 😎

boxpressed wrote:

The completist in me wants to know what retail packaging accompanied this card. Does anyone have one CIB? I saw one on eBay sold alongside a Creative CD-ROM drive, so it may have been part of a kit at one point.

5.jpg
Filename
5.jpg
File size
269.7 KiB
Views
13592 views
File license
Fair use/fair dealing exception
4.jpg
Filename
4.jpg
File size
346.61 KiB
Views
13592 views
File license
Fair use/fair dealing exception

And a small CD audio cable in a nylon bag (never opened), that fits the small white connectors on both sides
The guy I bought it from kept it really good.

keropi wrote:

nice review, the CT2230 is my SB16 of choice several years now - I even overkilled it and recapped the one in my main ss7 build 🤣 🤣 🤣

Thanks!
The capacitor change is an overkill though. 🤣

Last edited by James-F on 2016-09-12, 13:25. Edited 1 time in total.


my important / useful posts are here

Reply 9 of 65, by keropi

User metadata
Rank l33t++
Rank
l33t++
James-F wrote:

[...]
The capacitor change is an overkill though. 🤣

I regret nothing 🤣
Here is a better pic of my card, I don't think it matches any of the other revisions you linked (mine had C25 populated) - unless old age finally caught up with me 😵

http://i.imgur.com/J86EarS.jpg

Last edited by keropi on 2016-09-12, 13:25. Edited 1 time in total.

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 12 of 65, by boxpressed

User metadata
Rank Oldbie
Rank
Oldbie

It looks like keropi's has only the interface for the Panasonic/Creative CD-ROM, while James-F's has all three interfaces. Probably no difference in performance, I'm guessing.

Reply 13 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie
keropi wrote:

Here is a better pic of my card, I don't think it matches any of the other revisions you linked (mine had C25 populated) - unless old age finally caught up with me 😵

This is one of the latest CT2230 revisions (49432) with some IDE controllers stripped, and the CT1745-S mixer (no A) along with C25 and L1 next to it.
The mixer timeline goes like so: 1745A (92), 1745-S (93), 1745A-S (94).
I should say again that I believe the mixer chip has nothing to do with the noise level, they simply used what came to hand at the factory, the mixer chips are interchangeable.
A great card you got there keropi.

boxpressed wrote:

It looks like keropi's has only the interface for the Panasonic/Creative CD-ROM, while James-F's has all three interfaces. Probably no difference in performance, I'm guessing.

No difference whatsoever.


my important / useful posts are here

Reply 14 of 65, by jesolo

User metadata
Rank l33t
Rank
l33t

To minimize the "clicking after sample playback", you can try by disabling your high DMA channel (usually channel 5) and only use the low DMA channel (use the Diagnose utility for that).

I experienced the same symptoms on my AWE32 CT3980 (which also has the CT1747 chip) and disabling the high DMA channel solved the problem.

Reply 15 of 65, by clueless1

User metadata
Rank l33t
Rank
l33t

Thank you, James-F. Really enjoyed that review.

James-F wrote:

Untick all the X marks in the Output section and disable AGC, that should get you started.
I also recommend Voice and MIDI (OPL3) at full and Master one click down, for maximum SNR before distortion.
]

On my CT2800 (Vibra), if I have the Master that high, there is very noticeable hiss coming from my speakers with their volume knob set roughly in the middle. I think I ended up with my Master volume down around 35-40%. There is a step in there where the hiss just "goes away" and that's where I keep it, adjusting thereafter from the speaker volume.

I will need to try your setup with the expectation that my speaker volume will be much closer to zero.

The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks

Reply 16 of 65, by ZanQuance

User metadata
Rank Member
Rank
Member
jesolo wrote:

To minimize the "clicking after sample playback", you can try by disabling your high DMA channel (usually channel 5) and only use the low DMA channel (use the Diagnose utility for that).
I experienced the same symptoms on my AWE32 CT3980 (which also has the CT1747 chip) and disabling the high DMA channel solved the problem.

From what I've gathered is that games unaware of the DMA Auto-Load mode supported in the SB16's will click as new single 8-bit samples are loaded. The high DMA "should" only be used for loading 16-bit samples, but the drivers might be have been using it instead.

Reply 17 of 65, by GuyTechie

User metadata
Rank Member
Rank
Member
keropi wrote:
I regret nothing :lol: Here is a better pic of my card, I don't think it matches any of the other revisions you linked (mine h […]
Show full quote
James-F wrote:

[...]
The capacitor change is an overkill though. 🤣

I regret nothing 🤣
Here is a better pic of my card, I don't think it matches any of the other revisions you linked (mine had C25 populated) - unless old age finally caught up with me 😵

http://i.imgur.com/J86EarS.jpg

Your's is similar to mine where I have the C25 populated. I guess the CT2230 isn't just a single model, as there seems to be 5 different variations I've seen so far. 3 from James-F's photos, yours, and mine.

Reply 18 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie
jesolo wrote:

To minimize the "clicking after sample playback", you can try by disabling your high DMA channel (usually channel 5) and only use the low DMA channel (use the Diagnose utility for that).
I experienced the same symptoms on my AWE32 CT3980 (which also has the CT1747 chip) and disabling the high DMA channel solved the problem.

ZanQuance wrote:

From what I've gathered is that games unaware of the DMA Auto-Load mode supported in the SB16's will click as new single 8-bit samples are loaded. The high DMA "should" only be used for loading 16-bit samples, but the drivers might be have been using it instead.

Using the same DMA (1) for High and Low did not eliminate the clicking.
I have noticed many older games have this clicking, AFTER the sample has been played.
On the Vibra16S which uses integrated DSP v4.13 there is no clicking.

GuyTechie wrote:

There seems to be 5 different variations of CT2230 I've seen so far. 3 from James-F's photos, yours, and mine.

Actually, there are only 3 PCB revisions, not looking at the Mixer chips (and its peripherals), amplifiers chips, CD-ROM IDEs, ASP/CSP, etc..


my important / useful posts are here

Reply 19 of 65, by James-F

User metadata
Rank Oldbie
Rank
Oldbie

We have a parallel discussion about noise in this thread: Duke Nukem 3D stuttering with Dreamblaster X1
But I'll post my latest findings about noise here.

I have noticed the difference between the quiet CT2230, CT2950, AWE32, AWE64 vs the noisy ones is the CT1703 = 8/16-bit ADC/DAC Codec chip.
CT1703-A would be the latest and quietest which started to appear on the later CT2230 (49432) SB16 up to the AWE64 Gold.
CT1703-TBS is on the CT2230, CT2290, AWE32 and AWE64, I think it is as quiet as the CT1703-A.
CT1703-T is the noisy DAC appearing only the CT2230 (19413) and (29417) but not on the CT2290, also on the first revision AWE32 CT2760.
CT1701 is the oldest and noisiest appearing on the first SB16 CT1740 cards.

The early AWE32 CT2760 which came in 1994 use the noisy CT1703-T DAC.
Later AWE32 had the quiet CT1703-A but also omitted the real OPL3 chip.
All AWE64 have the quiet CT1703-A, but not a single one with a real OPL3.

This has to exclude the possibility that the wavetable amplifier actually adds the most noise, but someone would have to risk jumpering the wavetable on a card with CT1703-T to find out.

And so, this ends an era where is was believed that the Mixer chip CT1745 had anything to do with noise.
It is not, since it appears on all model randomly like my CT2950 with the older CR1745A and newer CT1703-A which is dead quiet.

Last edited by James-F on 2016-09-13, 14:09. Edited 1 time in total.


my important / useful posts are here