Yep, and I know only pci-e internal to pci (external) devices. Nothing with usb input, but the cost would make an usb "soundard" cheaper, especially a second hand 0404.
Still more and more people uses them to measure distortion of sources and amplifiers.
For my needs, all commonly used sigma-delta ADC have problematic HF noise from about 30khz, making it unusable to see extended harmonic spectra from any signal higher than 1khz.
I have yet to check the one that ESS made in 2011.
The power supply problem can be easily resolved using an external 5V regulated psu or batteries. Ideally, both the generator and the soundcard should be battery-powered. The "on" time makes this a non-issue.
Which one? I'm trying to find a 239A in Europe (and later mod it). No luck so far, i'll have to get it overseas.
A stock sound card is plenty good for setting bias, but for measuring distortion of solid state, not even close. They are also fine for measuring distortion of drivers. Even 192K sample rate is not close enough to look for amplifier stability problems. You need a real scope for that.
HP 209A. In it's low distortion mode, it is about .1% so really no better than the PC. Nothing beats it for sweeping around trying to find out what is buzzing in a room. ( I have a spring inside one window jam that resonates.) For doing real amplifier work, you need real laboratory equipment. I just run it through Spice, build it, listen and hope for the best. If someone wants to drop off a few hundred grand, maybe I could build a better test bench. Let's face it, a modern op-amp has about as much distortion as a metal film resistor. This is serious lab measurement range. By serious, I mean go talk to the folks at LBL or CERN.
For those of you who think they can do laboratory measurements with a $95 sound card, let me remind you the basic rule of thumb is that the totaltality of the test system must be ten times better than what you are testing. Go ahead, try and build that ADC or DAC with .00001% distortion. Heck, I want to see the caps you would use! Oh yea, when that measurement reads out a value like Qes of 2.12345, just use 2.1 You are not losing anything real. The air density changes more than that during the day. Remember what you are measuring.
HP 209A. In it's low distortion mode, it is about .1% so really no better than the PC. Nothing beats it for sweeping around trying to find out what is buzzing in a room. ( I have a spring inside one window jam that resonates.) For doing real amplifier work, you need real laboratory equipment. I just run it through Spice, build it, listen and hope for the best. If someone wants to drop off a few hundred grand, maybe I could build a better test bench. Let's face it, a modern op-amp has about as much distortion as a metal film resistor. This is serious lab measurement range. By serious, I mean go talk to the folks at LBL or CERN.
For those of you who think they can do laboratory measurements with a $95 sound card, let me remind you the basic rule of thumb is that the totaltality of the test system must be ten times better than what you are testing. Go ahead, try and build that ADC or DAC with .00001% distortion. Heck, I want to see the caps you would use! Oh yea, when that measurement reads out a value like Qes of 2.12345, just use 2.1 You are not losing anything real. The air density changes more than that during the day. Remember what you are measuring.
You obviously don't build with tubes, you don't need a .00001% distortion ADC to dial in a tube design. I think agood asynchronous USB adc/dac could be built for a few hundred bucks that could esily compete with something like the expensive dScope. Would be a handy tool.
Actually regal, yes I have built with tubes. And yes, they do have so much distortion I could have used my PC for most of it The sound card is really great for setting the bias by measuring the distortion rather than just the current. In this case, balance by voltmeter got the distortion down to about 3%. Using the spectrum analyzer I got it below 2% at 6W. Not much help setting the bias on my FET amps which model in the .0001 range.
Tube amps don't have quite the problems you need to track down like a good fast FET amp will, so looking for stability issues at 200K is not a problem. 90K might be good enough for you. Hey, if you think you can better DScope, have at it.
So, something as simple as my Profire, 192K-24 bit would have the resolution if it had low enough distortion. Just build a box with a high quality gain/attenuation and safety clamps to use as probes. JFet inputs could even be high enough inputs to use conventional scope probes. What you won't get is the kind of resolution, buffering, and of course software as you do buying one. I wish my Zelscope could run at 192. It has the features, delayed sweep, triggering etc. The sound card is not the hard part.
Tube amps don't have quite the problems you need to track down like a good fast FET amp will, so looking for stability issues at 200K is not a problem. 90K might be good enough for you. Hey, if you think you can better DScope, have at it.
So, something as simple as my Profire, 192K-24 bit would have the resolution if it had low enough distortion. Just build a box with a high quality gain/attenuation and safety clamps to use as probes. JFet inputs could even be high enough inputs to use conventional scope probes. What you won't get is the kind of resolution, buffering, and of course software as you do buying one. I wish my Zelscope could run at 192. It has the features, delayed sweep, triggering etc. The sound card is not the hard part.
Whoah, nobody wants to replace a (50Mhz+) scope here.
For some SS gear a card like the EMU is enough, and anyway better than nothing at all. Not for DACs and some line-level gear, though.
I wouldn't use it stock. Modded, it can become sub-ppm. Search for 339A in this forum.
This is correct. I'm aiming for -130dB and will settle for -120.
I dumped my EMU 1616m due to driver memory leak problems. If you motherboard does not have an on-board video, you may be OK. Other that that, it was a great box. Very confusing software because it did "everything" for the home recording set. Oh yea, the input impedance on the "line in" was very low. I had to add buffers ( battery powered of course) for the "reference" feed. The M-Audio Profire works fine without them.
My noise floor for my M-Audio Profire is about 120 dB, rising to about 90 at 20K. Battling some power line noise. Turns out in the last few years my trusty old Rotel bench amp has gotten very noisy. Time for a power supply rebuild with modern caps, hexfred diodes, and I am sure I can clean up the dress a bit. I lowered the noise floor in my DH-120 by 18 dB this way. Any way, If I get the whole thing back below 90 dB 20 to 20, I am in good shape for speaker building. Getting a test bench to actually have 130 dB floor is no easy feat. You may need a bank of power line conditioners, a ground plane, shielded power lines within 10 feet or so. 120 may be do-able with only hard work. You may need an opto-isolator between the PC and the acquisition box. Anyone make a USB isolator? Should not be too hard to do. Then feed what you can from batteries. Even things like the phantom power for my mic was too noisy, so it is not battery powered.
It is said the DCX A-2-D is pretty good, and they have a PCM link output, but no idea what software that could feed. Above the emu and M-audio, you are going to real instrumentation or studio stuff. Start adding zeros to the price.
My noise floor for my M-Audio Profire is about 120 dB, rising to about 90 at 20K. Battling some power line noise. Turns out in the last few years my trusty old Rotel bench amp has gotten very noisy. Time for a power supply rebuild with modern caps, hexfred diodes, and I am sure I can clean up the dress a bit. I lowered the noise floor in my DH-120 by 18 dB this way. Any way, If I get the whole thing back below 90 dB 20 to 20, I am in good shape for speaker building. Getting a test bench to actually have 130 dB floor is no easy feat. You may need a bank of power line conditioners, a ground plane, shielded power lines within 10 feet or so. 120 may be do-able with only hard work. You may need an opto-isolator between the PC and the acquisition box. Anyone make a USB isolator? Should not be too hard to do. Then feed what you can from batteries. Even things like the phantom power for my mic was too noisy, so it is not battery powered.
It is said the DCX A-2-D is pretty good, and they have a PCM link output, but no idea what software that could feed. Above the emu and M-audio, you are going to real instrumentation or studio stuff. Start adding zeros to the price.
Creative is coming with a new zxr model. Containing the pcm4220 adc chip. Looks all right. Will buy and test it when it is out.
An interesting video..
Newegg TV: Creative Sound Blaster Z-Series Audio Cards Interview - YouTube
An interesting video..
Newegg TV: Creative Sound Blaster Z-Series Audio Cards Interview - YouTube
FYI.
I asked about a USB interface for audio measurements to Ivo (ARTA author) and he recommended either one of the RME interfaces (expensive) or the EMU 0204 (cheap). He mentioned that neither is good enough to measure very low distortion (<0.0003%) found in modern opamps and DACs. But they are sufficient for most SS amps and definitely loudspeakers.
I asked about a USB interface for audio measurements to Ivo (ARTA author) and he recommended either one of the RME interfaces (expensive) or the EMU 0204 (cheap). He mentioned that neither is good enough to measure very low distortion (<0.0003%) found in modern opamps and DACs. But they are sufficient for most SS amps and definitely loudspeakers.
Good card? Well, my Audigy USB had high distortion on one channel. As it was more than a month old, Creative told me to buzz off. I bought an emu-1616u and the first one flat failed. By the time I was set up and running the second, the software has a memory leak that would crash the on-board video ram after a minute or two of acquisition. They admitted to the problem but refused to fix it. Dumped it and bought a M-Audio Profire 610. It has a severe problem with leakage/delay/magic causing several dB of ripple in loopback which makes all the measurement tools that use a reference, SE, HOLM etc, useless. They won't answer. It also will cause distortion on the analog out if the mic in is approaching clipping. I have a M-Audio FastTrack, and it mixes too much input to the output. Put your mic close to do a nearfield measurement and you get feedback. Again, M-Audio has no response. I am beginning to think semi-pro recording equipment makers are all trash and that is better than their products. What the heck out there actually works?
I tend to agree with your view. I had an M-audio Transit and it worked really well, until it flat out failed. No computer will recognize it anymore. I would've bought another coz they are cheap, but it turns out they've discontinued it. There are some cards that are esoteric and then there are those that are expensive. Reliable, available, cheap and high performance don't go together, apparently.
I bought the 0204. More than a few people have reported success using it. Let's see how it goes.
I bought the 0204. More than a few people have reported success using it. Let's see how it goes.
Been searching all the music store sites. The FocusRite 2i2 looks promising due to it's lack of features. Like no direct monitor mix. That may be fine for a musician, but it kills us. Drops me to USB and 24/96K which is fine for speakers I guess. 192 was nifty with the Firewire for looking at amp issues. The Akai EIE Pro also looks promising. Of course, Firewire is dead and no one has come out with the new HD interface or USB3 yet.
The 0204 looks like it has the monitor mix that is so problematic. My FastTack has that problem. Of course, emu admitting they have defective software and telling me to stick it makes me never want to deal with them again. I am in the software business, and memory leaks are inexcusable.
The next step up seems to be $2K, which is out of the question for us. Everything in the middle is just features. Quality? Know what the input amp is in the Profire? TLO72.
I am going to call up the Sweetwater folks and talk to them. They have been very helpful in the past. ZZsounds happens to be doing web maintenance today, so I don't know what they carry.
The 0204 looks like it has the monitor mix that is so problematic. My FastTack has that problem. Of course, emu admitting they have defective software and telling me to stick it makes me never want to deal with them again. I am in the software business, and memory leaks are inexcusable.
The next step up seems to be $2K, which is out of the question for us. Everything in the middle is just features. Quality? Know what the input amp is in the Profire? TLO72.
I am going to call up the Sweetwater folks and talk to them. They have been very helpful in the past. ZZsounds happens to be doing web maintenance today, so I don't know what they carry.
Had a good talk with Sweetwater. They really recommended an RME, but at a bit under a grand, is not exactly hobby speaker building. The message is: "it's the drivers!" So, what next? He agreed the Focusrite 2i2 sounds like the ticket at chump change. $ I am going to order one tomorrow.
Last edited:
2i2 is USB. Scarlett 2i2 | Focusrite
I only need 2 in, 2 out. Super basic. From what I gather, the difference is features, not performance until you move up to external clocks etc. I would not pass up a Pro 24 if I did not have to pay for it. It is within thinking about price wise. If you were also going to run something like Bodzio UE, then you need the additional outputs. To get the 196K back, you are talking four digits again.
I only need 2 in, 2 out. Super basic. From what I gather, the difference is features, not performance until you move up to external clocks etc. I would not pass up a Pro 24 if I did not have to pay for it. It is within thinking about price wise. If you were also going to run something like Bodzio UE, then you need the additional outputs. To get the 196K back, you are talking four digits again.
I'm looking only for a good AD and decent drivers to run ARTA or Spectraplus as a spectrum analyser, nothing else, I prefer to use an external generator (battery powered).
On a different computer I have a 1616m pci and I didnt have any issue with its drivers, but I never tried the AD.
"96 KHz, 24-bit conversion" - wonder where the HF noise starts :/
On a different computer I have a 1616m pci and I didnt have any issue with its drivers, but I never tried the AD.
"96 KHz, 24-bit conversion" - wonder where the HF noise starts :/
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.