jackinnj said:
My TM503 just came in yesterday, slipped in the SG505 and hooked it to my MAudio 192 card. Right off the bat, after optimizing the input level, it had about 0.00015% THD
and 0.00026% THD + N sampling 1kHz at 48K. With no notch, i.e the full fundamental showing. That was using a 131K FFT and 100 power spectrum averages. You may be able to do better, but I'm happy for now
that's orange juice BTW...
Just purchased a HP-334 distortion analyzer off ebay to qualify DIY toys, not a very good one lets just call it a starter one it is on par with my signal generator. Is there a decent publication on using test equipment on audio circuits? I have a Tecktronixs 2235 100mhz scope and use it in the most rudiment manner.
Bill
Bill
Pjotr said:Nice, I presume you looked at the oscillator of the 505? And what do you get if you use a notch filter? Then you have really the distortion of the oscillator.
Yes, the SG505 feeds the M-Audio Audiophile 192 via a short BNC cable.
http://www.m-audio.com/products/en_us/Audiophile192.html
I have not yet tried a notch. With a notch you eliminate the fundamental, but introduce the notch filter electronics - noise, etc. At this level, everything is important. Perhaps a passive notch would work. Then impedance matching is not to be ignored. Not sure the notched output from distortion analyzers would get THD much lower than 0.00015%, but only one way to know.
Ultimately, I think low level, low frequency non-harmonic noise in the M-Audio card is the limiting factor, at about -115 to -120dB. Another card may do better here, or perhaps using it in an outboard PCI card carrier (Magma) to get out of the PC backplane noise environment may also help. I'll post the plot referenced in my above post later.
" ... or perhaps using it in an outboard PCI card carrier (Magma) to get out of the PC backplane noise environment may also help. ..."
No question about that!
Or maybe M-Audio's pure external devices = they do have better specs than the plugin cards, right outta the box.
Check it out: http://www.m-audio.com/products/en_us/FireWireSolo.html
No question about that!
Or maybe M-Audio's pure external devices = they do have better specs than the plugin cards, right outta the box.
Check it out: http://www.m-audio.com/products/en_us/FireWireSolo.html
Ehh, have a look at the plot I posted at post #261. That is a PCI card straight inside the PC!
Anyway USB/FW devices are also powered from the noisy PC psu, unless you use an external powered hub. But the ones I tried (E-MU, M-Audio Transit and M-Audio Duo) didn’t perform any better even powered external.
Anyway USB/FW devices are also powered from the noisy PC psu, unless you use an external powered hub. But the ones I tried (E-MU, M-Audio Transit and M-Audio Duo) didn’t perform any better even powered external.
FastEddy said:" ... or perhaps using it in an outboard PCI card carrier (Magma) to get out of the PC backplane noise environment may also help. ..."
No question about that!
Or maybe M-Audio's pure external devices = they do have better specs than the plugin cards, right outta the box.
Check it out: http://www.m-audio.com/products/en_us/FireWireSolo.html
Maybe you meant another one, or I'm missing it? Not the one in your link, the 192 easily beats it according to their specs. Look in the line input only, as we are using an external source.
Pjotr: " ... Anyway USB/FW devices are also powered from the noisy PC psu, unless you use an external powered hub. ..." from #261
Working on it. I'm working with a European company to build electronic isolation for FW in the cable (an isolation dongle, if you will). It may not be possible and still maintain FW400 performance.
(The stand alones from M-Audio, Roland/Edirol and MoU stuff all have their own external power brick and all internal filtration specifically designed to combat switching power supply noise = all relatively successfully. If your FW hub or confuser has a 4-pin port, then, with proper cabling, the switching PS noise isolation becomes more manageable, but still not perfect.)
=====
BFNY: " ... Maybe you meant another one, or I'm missing it? Not the one in your link, the 192 easily beats it according to their specs. ..."
Yes, you are correct. FW M-Audio Audiophile = 96k instead of 192k on the analog ports.
Check out: http://rolandus.com/products/productdetails.aspx?ObjectId=731&ParentId=114 ... and ... http://www.motu.com/products/motuaudio/ultralite-mk3/
Wondering if you have access to any of these others to make a comparison of your graphs (above)?
=====
Again, I have a mercenary interest in this and have for years = firewirestuff.com
Working on it. I'm working with a European company to build electronic isolation for FW in the cable (an isolation dongle, if you will). It may not be possible and still maintain FW400 performance.
(The stand alones from M-Audio, Roland/Edirol and MoU stuff all have their own external power brick and all internal filtration specifically designed to combat switching power supply noise = all relatively successfully. If your FW hub or confuser has a 4-pin port, then, with proper cabling, the switching PS noise isolation becomes more manageable, but still not perfect.)
=====
BFNY: " ... Maybe you meant another one, or I'm missing it? Not the one in your link, the 192 easily beats it according to their specs. ..."
Yes, you are correct. FW M-Audio Audiophile = 96k instead of 192k on the analog ports.
Check out: http://rolandus.com/products/productdetails.aspx?ObjectId=731&ParentId=114 ... and ... http://www.motu.com/products/motuaudio/ultralite-mk3/
Wondering if you have access to any of these others to make a comparison of your graphs (above)?
=====
Again, I have a mercenary interest in this and have for years = firewirestuff.com
FastEddy said:
Check out: http://rolandus.com/products/productdetails.aspx?ObjectId=731&ParentId=114 ... and ... http://www.motu.com/products/motuaudio/ultralite-mk3/
Wondering if you have access to any of these others to make a comparison of your graphs (above)?
No, I don't, but the specs on the 1st one don't look that great. I could not find any specs on the 2nd one. On almost all of these, the input dynamic range and distortion is better than the source. Almost any 24 bit card will give you 0.001%, or 100dB. Getting lower than that by 10-15 dB is more rare, and the exception. There's more data on the other thread I referenced with the EMU Tracker, which is USB based. It does OK at .0006 THD +N.
It appears one of the best is the EMU1212, which is also a PCI card. This apparently betters the M-Audio I have by a slight amount. That one may the best best you can get, at least today. As before, if it must be hooked to a laptop, use a Magma PCI to laptop interface.
All the Magma outboard PCI boxes have their own power. But they are not cheap, used is the lowest cost way.
http://www.magma.com/products/pci/1PCI/
" ... No, I don't, but the specs on the 1st one don't look that great. I could not find any specs on the 2nd one. ..."
Mmmmm... the first spec I looked for is the 192k stereo bandwidth. I did not look for the details. My favorite is the EchoAudio.com AudioFire 4 = which is (only) 4 channel X 96k ... but then I don't use it much anymore, preferring to listen to DVD-A and DVD-Video straight from the "universal" DVD player and the odd SqueezeBox Internet radio sessions ...
Mmmmm... the first spec I looked for is the 192k stereo bandwidth. I did not look for the details. My favorite is the EchoAudio.com AudioFire 4 = which is (only) 4 channel X 96k ... but then I don't use it much anymore, preferring to listen to DVD-A and DVD-Video straight from the "universal" DVD player and the odd SqueezeBox Internet radio sessions ...
FastEddy said:
I have an USB2.0 isolator at work that uses an optical fibre link intended for safety purposes, it’s quite expensive. Tried that one on my USB audio cards but it does not give any benefit over a simple external psu.
IMO these are just theoretical benefits. Practice shows that good PCI cards don’t suffer from noise from inside the PC.
" ... I have an USB2.0 isolator at work that uses an optical fibre link intended for safety purposes, it’s quite expensive. ..."
Which one, may I ask?? We have one in our catalogs, but it is USB 1.1, alas.
" ... practice shows that good PCI cards don’t suffer from noise from inside the PC. ..."
Mmmmmmm ... sometimes, maybe. The VIA chip set PCI and motherboard sound cards are all quite noisy ... some others, not as bad, some quite good in other's opinion. I'm of the old school: switching noise comes through whenever there are cabling / impedance matching problems = IMOP = and the source is cpu power supplies, invariably. I can see it on the 'scope.
Which one, may I ask?? We have one in our catalogs, but it is USB 1.1, alas.
" ... practice shows that good PCI cards don’t suffer from noise from inside the PC. ..."
Mmmmmmm ... sometimes, maybe. The VIA chip set PCI and motherboard sound cards are all quite noisy ... some others, not as bad, some quite good in other's opinion. I'm of the old school: switching noise comes through whenever there are cabling / impedance matching problems = IMOP = and the source is cpu power supplies, invariably. I can see it on the 'scope.
FastEddy said:
Have a look at Bob’s results. My PCI card gives comparable results --> spurs well below –140 dB and that with the soundcard itself used as generator.
Being of the old school is not a bad thing (maybe my school is even older
) but that is no excuse to believe in general statements that doesn’t hold in practice.Cheers
" ... My PCI card gives comparable results --> spurs well below –140 dB ..."
Oh, I really liked your results = better than most, as good as it get's without a full tech support staff ...
(Currently I'm home brewing a distortion analyzer ... building the power supply as I write. If my numbers and graphs come out as good as your's I'll be happy ... too.)
Oh, I really liked your results = better than most, as good as it get's without a full tech support staff ...
(Currently I'm home brewing a distortion analyzer ... building the power supply as I write. If my numbers and graphs come out as good as your's I'll be happy ... too.)
bench upgrade
Just purchased an HP 339A option1 (THD down to .00018%) from Valutronics /$600 (only one they had). Been looking for awhile,
not a great price but decided to avoid the risk of an Ebay purchase.
A little off topic:
Also jumped on what seemed to be a great price on a digital OS:
Tek TDS360 for $450( they had 14 of them-Gino). Liked the FFT Analysis and the VGA & Rs232 outs; but this scope is still an unknown for me, we'll see. Should be an improvement from by old 'boat anchor' Tek 5440 w/5A48(2) & 5B42.
Any comments on this TDS 360? Jackinnj?
http://www.valuetronics.com/Details.aspx?Model=Tektronix_TDS360&ProdID=3112
The TI ADS1271 Eval sure looks interesting; but I think have enough to figure out at this point.
Charles
Just purchased an HP 339A option1 (THD down to .00018%) from Valutronics /$600 (only one they had). Been looking for awhile,
not a great price but decided to avoid the risk of an Ebay purchase.
A little off topic:
Also jumped on what seemed to be a great price on a digital OS:
Tek TDS360 for $450( they had 14 of them-Gino). Liked the FFT Analysis and the VGA & Rs232 outs; but this scope is still an unknown for me, we'll see. Should be an improvement from by old 'boat anchor' Tek 5440 w/5A48(2) & 5B42.
Any comments on this TDS 360? Jackinnj?
http://www.valuetronics.com/Details.aspx?Model=Tektronix_TDS360&ProdID=3112
The TI ADS1271 Eval sure looks interesting; but I think have enough to figure out at this point.
Charles
Re: bench upgrade
I don't know how much utility you'll get out of the FFT on this series of TEK scopes (earlier DSO's) -- you should sign up to the TEK group on Yahoo -- a lot of really knowledgeable folks there. If I didn't have a network analyzer, I would probably build a sound-card based VNA to display gain and phase out to a couple megahertz.
I have a TEK 5xxx series scope -- although the bandwidth is low, they are very quiet. I use the differential amplifiers which are just as good as the 7a22n. The scopes I turn on every morning are a 2465B and a 3012B. I would hold on to the 5440 and look for a 5a22n plug-in (assuming that the CRT is still in good shape.)
The FFT on the 3012B is almost useless.
cowboy99 said:
I don't know how much utility you'll get out of the FFT on this series of TEK scopes (earlier DSO's) -- you should sign up to the TEK group on Yahoo -- a lot of really knowledgeable folks there. If I didn't have a network analyzer, I would probably build a sound-card based VNA to display gain and phase out to a couple megahertz.
I have a TEK 5xxx series scope -- although the bandwidth is low, they are very quiet. I use the differential amplifiers which are just as good as the 7a22n. The scopes I turn on every morning are a 2465B and a 3012B. I would hold on to the 5440 and look for a 5a22n plug-in (assuming that the CRT is still in good shape.)
The FFT on the 3012B is almost useless.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.