The AM70A has the harmonic analysis and IM. http://www.shibasoku.com/download/avc/am70a_e.pdf
The key difference is using DSP for the analysis, which makes a lot of sense.
You make a good point, but the combined cost of a very good soundcard plus the software plus a soundcard front end will generally be greater than the $450 cost of the QA401. Thats what I meant by bang for the buck. Moreover, don't overlook the potential additional hassle factor with such non-turn-key solutions and the additional amount of bench space taken up.
Cheers,
Bob
Hi, Bob,
Actually the cost is (more or less) the same.
200$ for the card and 200$ for the pro version of Virtins.
The attenuator can be a simple one, some resistors, pots and diodes.
Both solutions, sound card and QA, need a PC, so, the necessity of space over the bench is the same.
Regards,
Respectfully I disagree that the attenuator can be a simple one unless 1) only measurements at one power level are made, 2) the user makes manaul adjustments to the audio interface (amp to soundcard) each time the power level is changed in order to maintain the correct input level to the soundcard or 3) an automatic interface like Jan's AutoRanger is used to automate #2.
It is important to keep the input to the soundcard correct - too much and "magic smoke", too little and soundcard performance suffers (increasing distortion, etc).
#1 is not feasible, #2 is tedious (believe me, I still have my homemade level switch box) and probably doesn't get the input levels realistically
Obviously #3 involves increased cost, either Peter Millet's or Jan's AutoRanger 2 come to mind (since I have the first and am waiting for the second from Jan), but there goes your $400 budget out the window.
I kind of wish I had started with a QA401 instead of my soundcard/ARTA/interface route, but I have learned alot about optimizing measurements doing it "the harder and more expensive way"
Last edited:
My old solution was made the way I said in the last post.
A metallic box, grounded, short cables and linear pots, resistors and some diodes for protection.
The image below is the THD measurement of an active crossover I designed last year. You can see the diference in THD and SNR between the channels, caused by the problem in the Asus line out.
QA401 can´t reach this measurement´s level, but, imho, is a very good "everything in a box" solution. I´m using its hardware with Virtins (ASIO) and the integration is very good. The residual THD is very good. The SNR is a little bit high for me.
Regards,
A metallic box, grounded, short cables and linear pots, resistors and some diodes for protection.
The image below is the THD measurement of an active crossover I designed last year. You can see the diference in THD and SNR between the channels, caused by the problem in the Asus line out.
QA401 can´t reach this measurement´s level, but, imho, is a very good "everything in a box" solution. I´m using its hardware with Virtins (ASIO) and the integration is very good. The residual THD is very good. The SNR is a little bit high for me.
Regards,
Attachments
Last edited:
Oh, I'm not saying it can't be done your way as described. I've tried but gave up quickly and bought interface kits (mostly because I've always loved putting together electronics kits since I was a kid with an old clunky soldering iron). Just saying a solution like the QA401 gets the job done quicker and more certainly, though I'm not sure it is as versatile for a hobbyist as having a soundcard and software.
And I believe we can agree to disagree with each other 😀😀😀.
The measurements I have shown are actual measurements made with the simple attenuator.
In this case, the adjustment is quite simple, just adjust the potentiometer to about -6dBFS in Virtins (lowest THD of Asus board) and measure.
It has to be done the same on QA401 by setting it to -18dBFS (best measurement point).
Using Autoranger would be more expensive but much better for both the sound card solution and QA401.
For my part, I will continue to use the simple attenuator, until I have the opportunity to purchase Autoranger, and the QA401 for faster measurements of power amps.
Regards,
blmn: You spectra are truly impressive for an internal PCI-e soundcard + the LME buffer. 60Hz mains at -170dBFs, 0.0001%THD, 0.0003% THD+N. Also the spectra for your xover output - 60Hz at almost -160dB.
Yep.
The measurements were made at 2kHz because the crossover cutoff frequency was around 900Hz. Since it is a high
-pass side, the rejection of the power supply artifacts is not so impressive.
I´ll try to find a low-pass figure.
I was very careful with the power supply and used the LME49720 throughout the circuit.
At this noise level, the difference in the placement of grounding points is impressive. Moving a few inches makes a lot of difference.
Regards,
Still I mean even the Asus loopback. I have never seen such noise figure for a soundcard, let alone an internal one.
Neither do I until I used this card.. I repeated this loopback measurements many times and the results were the same. Im using a very short cable connecting the Line IN with the line OUT.
Now the sound card is disassembled because I´ll try to replace the PCM1792 in order to solve the diference of THD problem.
I´m thinking on use a PCI extender and a isolated power supply in order to achieve better noise ratios also. Placing the card far from the PC and near the attenuator pots can improve it.
Regards,
Not sure you need the AutoRanger with the QA401 (?) Have not used a QA401 or even looked at it more than in passing. I would hope not, kind of defeats the purpose if so.
If you work usually with voltage levels up to 56vpp or 20v RMS it would be fine.
In my case the output swing voltage is normally higher, so, the Autoranger would be a nice tool in this case.
Regards,
Last edited:
Thanks Jan, I took your advice and emailed them aobut the Distortion Magnifier and got a reply. As you stated they said they are working on creating a new website. So i am assuming the heath issue statement on the current website is an old one.
Although, i am waiting for clarification, seems the end price to me for a board is more then id feel reasonable (over 150 usd delivered). perhaps he meant that for a completed unit? .. again waiting for conformation that the price was just for a single pcb.
If so, considering i already layed out a version of the original design, and am laying out a new version with additional features, i think i may go with just ordering a minimal batch out of china.
Thanks again Jan
The kit I received from Pilghim Audio was everything but the case and power supply. I ordered a wall wart and metal proejct box from Digi-Key. Works well.
I bet they probably quoted me a kit price then .. I did request just the PCB...waiting for them to confirm what they quoted.
I'm glad someone kicked this thread back open (several times). In getting ready to repair a Krell KSA50s amp that has intermittent high-end distortion in one channel, though I have tons of RF gear, my audio stuff was lacking. So I picked up an older HP DSA 3562a cheap, an HP 8904 generator and probably a mistake, an 8903 analyzer. I envisioned running the DSA or something through this channel having the problem and getting a trigger from a scope mask or something in the collection I have, to see if I can find the failure. It has something to do with the bias circuitry as it was damaged when one of the hv legs opened (unsoldered itself).
Anyway, I'm looking for a low distortion audio generator 'kit' now or the most easily realized set of boards, schematic, etc. I can build Bob's from his analyzer and was reading the thread over in eevblog but was hopping for an 'ebay special' or a pointer to someone that offers a kit or just boards. I don't need components.
Thanks,
Jerry
edit, by the way, when I mean low, I want it to be the last one I build. Also, the problem in the bias circuitry was all replaced, component by component. the amp is an original KSA50S not a knock-off . I use it in a bi-amp setup for the highs with a KSA-160B in the lows.
Anyway, I'm looking for a low distortion audio generator 'kit' now or the most easily realized set of boards, schematic, etc. I can build Bob's from his analyzer and was reading the thread over in eevblog but was hopping for an 'ebay special' or a pointer to someone that offers a kit or just boards. I don't need components.
Thanks,
Jerry
edit, by the way, when I mean low, I want it to be the last one I build. Also, the problem in the bias circuitry was all replaced, component by component. the amp is an original KSA50S not a knock-off . I use it in a bi-amp setup for the highs with a KSA-160B in the lows.
Last edited:
In your position I would get one of Victors oscillators. I don't think you could do better let alone for the money. You still get to build a power supply and housing so it is still diy but getting the performance would take an enormous effort any other way.
I am in the same situation as you with the DSA... sent you PM on how i am going to handle it. that DSA is a pretty amazing piece of kit.