Stuart Yaniger (our own SY) had a six-part (yes he likes to talk ;-) series on soundcard hardware, software and measurement techniques in AudioXpress. He goes through all the issues and helps you select the best stuff for what YOU want to do. It's worthwhile to get the articles.
There's soooo much more out there
Jan
The 'problem' has been accuracy below -100. Sure, you will see all sorts of things down there...... but with several test equipment I have compared against, the accuracy of just using a sound-card is so bad as to be almost useless data below -100dB.
Unless, you use a notch filter on the fundamental, I would be interested in the make/model of what soundcard hardware you use.
THx, RNMarsh
Last edited:
M-Audio Delta 192. Extremely reliable and repeatable well below -120dB. The Lynx card is even better, but I don't have the spare cash for it at the moment.
Of course, part of the trick is understanding what you're doing (especially for choosing measurement parameters), which is why I've been writing that series (now up to 7 parts) for AX.
Thanks for the input.... as you know...reliable and repeatable do not equate to accurate harmonic measurment at below -100. I'll check into your choices though. Thx.
http://sound-cards-review.toptenreviews.com/m-audio-audiophile-192-review.html
As for AX... dont subscribe. How to get just your info in it?
THx-RNMarsh
Last edited:
So after a VERY long delay, I finally wired this up.
I ended up using banks (5 of them) of LIFEPO4 batteries for power. The bank will put out 16-17V, 18V at full charge.
It seems like the DC offset jumps around until it warms up. I need to play and adjust the offset more.
And the one problem with battery power is I don't just leave it on to burn in, as to not drain down the batteries. So burn in may take a while.
But even with just a little time on it, sounds really good.
Randy
I ended up using banks (5 of them) of LIFEPO4 batteries for power. The bank will put out 16-17V, 18V at full charge.
It seems like the DC offset jumps around until it warms up. I need to play and adjust the offset more.
And the one problem with battery power is I don't just leave it on to burn in, as to not drain down the batteries. So burn in may take a while.
But even with just a little time on it, sounds really good.
Randy
SY...look into the work done in Equipment and Tools forum with Davada. I already did the work on these sound cards type compared them to the best commercial test equipment.
Without reading your back issues in AX, I can say you can get best results with notch filter and a sound card. Do you use a notch filter?
THx-RNMarsh
Last edited:
The original devices. No changes.... Just excellent matching and ac and dc trim.
THx-RNMarsh
When needed, which is not often. You can lose a lot of important information doing things that way.
What about DIYing a Twin-T notch filter? Any experiences?
In the sim I assumed to get the caps matched to 0.5% by adding some smaller ones in parallel, starting out with 2% polystyrene for low tempco. The resistors are assumed to be matched to 0.1% by series trimmers added to each leg of the 'T'. Add in a nice dual opamp and a Dual Salas shunt-reg power supply...
In the sim I assumed to get the caps matched to 0.5% by adding some smaller ones in parallel, starting out with 2% polystyrene for low tempco. The resistors are assumed to be matched to 0.1% by series trimmers added to each leg of the 'T'. Add in a nice dual opamp and a Dual Salas shunt-reg power supply...
Attachments
Dick;
that site looks to be focused on different concerns than a card that is optimum(ized) for instrumentation/measurement purposes.
However, I certainly grant that along with all the features (Surround, Kararoke, etc), ASUS certainly has impressive numbers though.
My question has been what card has the whiz bang chips/implementation/performance without all the whiz bang features that are useless "for me"?
... That also won't require me becoming a traitor to the "faith of the soldering iron" and becoming a mush-head (aka software engineer) just to get it to run in Windows nor spend up my hard earned $$$ that i can spend on more useful things, like, say, unobtanium JFETs?
SY's articles are helpful in that regard, I think, but I haven't read them all yet ...
mlloyd1
that site looks to be focused on different concerns than a card that is optimum(ized) for instrumentation/measurement purposes.
However, I certainly grant that along with all the features (Surround, Kararoke, etc), ASUS certainly has impressive numbers though.
My question has been what card has the whiz bang chips/implementation/performance without all the whiz bang features that are useless "for me"?
... That also won't require me becoming a traitor to the "faith of the soldering iron" and becoming a mush-head (aka software engineer) just to get it to run in Windows nor spend up my hard earned $$$ that i can spend on more useful things, like, say, unobtanium JFETs?
SY's articles are helpful in that regard, I think, but I haven't read them all yet ...
mlloyd1
in the test equipment mods thread Dick refers to, someone whose name escapes me at the moment discussed a nice one that is on his website.
Dick, Davada, Demian and several others shared A LOT of interesting stuff in those threads.
definitely worth taking a look ...
mlloyd1
Dick, Davada, Demian and several others shared A LOT of interesting stuff in those threads.
definitely worth taking a look ...
mlloyd1
Exactly right. And a key part of selecting a card, IMO, is having a balanced input with high CMR.
Let me ask this another way.
I finished a Marsh amp build a couple weeks ago. For the complementary pairs, I used KSC2690/KSA1220. The hFE of an NPN-PNP pair can be matched very closely. I used them because, according to my measurements, the hFE of an MJE NPN-PNP pair is not nearly as close.
So I am wondering (and this is my question) why specify the MJE devices when the KSC/KSA devices are much closer in hFE?
Looking at the data sheets, the hFE (gain) of the MJE devices is more linear than the KSC/KSA devices. Because the MJE devices are used as complementary pairs, the hFE differences tend to "cancel out", such that the MJE hFE difference becomes a non-issue, leaving only hFE (gain) as the deciding factor.
If the logic above is correct, then devices which are more linear are preferred over devices that are less linear, regardless of hFE. And if this is correct, then a KSC3503/KSA1381 pair might be suitable substitutes for an MJE pair. Their Ic is limited, but would work in this application, and on paper the hFE looks even more linear than an MJE pair. I found them actually used here:
MJR7-Mk5 Mosfet Power Amplifier
And now the disclaimer: I'm not suggesting that everybody jump up and rebuild their existing amps, and am not suggesting that MJE pairs be discarded in future builds. Just thinking out loud. I do plan to build one up and see how it sounds.
Any thoughts or comments?
PS- arrow.com has been offering free shipping for the last few days. No idea how long it will last. No advertisement, no banners, just checkout and choose "free shipping" before saying "I do".
That site also gives the mfr distortion and other specs.
If you want simple and easy FFT testing just buy the QA400. Designed only for testing.
For more than double that price you can wait for their QA401.
FWIW -- Stereophile has tested just about every DAC and ADC made over the years and J.A, notes that 21 bits is all he has seen -- beyond that it isnt accurate.
THx- RNMarsh
Last edited:
Start with the most linear in the current range to be used and then match them.
I am sure there could be others but I would have to try them and then measure the distortion to know if they are as good or better substitute.
THx-RNMarsh
Thanks, that makes sense. Don't know why I didn't pick up on this earlier. Now I know what I won't be using in the next build.
BTW --- The designer of the QA400 and 401 emailed me that he will send me a QA401 soon for my evaluation. He had sent me a early QA400.
If it is important to you to have bal input capability, the QA401 has that feature.
THx-RNMarsh
I am interested in this also. Is there a way to purchase (if need be) just these articles?
With help from Preamp I was using ARTA to do some tests on a PSU I built. I purchased an ESI Juli@ XTe to do this better but was only using the single-ended input on the card with the other end of the cable clipped to output and GND on my on-board measurement pre-amp. I'd love to get some advice on how to conduct these tests more effectively.