The maximum output current at 400V(in) 100Hz is 1.2uA. This is way less then safety regulation specifies as the maximum safe leakage current. Extreme caution should be taken on the mains side of it, the output side is safe.
Nothing
I just want to get some impression of what is going on (on the mains) and what the filter is doing (especially HF). The two transformer filter is generating an isolated output voltage, it is safe to look at that directly with a simple resistor divider. It is also safe to look at the intermediate voltage (e.g. the secondary of the input transformer and the primary of the output transformer). The problem with looking at the 50Hz signal directly is that the 50Hz must be filtered out, and I do not have a filter to do that. For now I have to do with this probe
Yes, and no, the Kelvin Varley divider has a common ground connection between input and output voltage. I did use this specific design to avoid a common ground between input and output, this way the current on either side (with the input plug reversible into the mains both sides can be phase or zero) can be no larger than 1.3uA at 100Hz (21nA at 50Hz).
Also, the maximum current from mains phase to output ground (when either output shorted to mains ground) is no larger than 400uA.
Last edited:
Will the QA400 work with a Windows virtual machine?
This is a diversion from the other discussions on the QA400 that I've seen - and enjoyed. I was interested in purchasing the QA400 to do some power amp distortion measurement, so I guess I'll need some sort of interface device, if not just a voltage divider, to connect it to the amp output.
But my main issue is that I use a MacBook. I could run the QA400 software in a virtual machine but I'm not sure it would work given the unique way they are handling sound in the Windows software. Does anyone know for sure if the QA400 software will or won't work? I posted a note on the QA400 forum, but I'd like to hear from some real world users. Thanks.
This is a diversion from the other discussions on the QA400 that I've seen - and enjoyed. I was interested in purchasing the QA400 to do some power amp distortion measurement, so I guess I'll need some sort of interface device, if not just a voltage divider, to connect it to the amp output.
But my main issue is that I use a MacBook. I could run the QA400 software in a virtual machine but I'm not sure it would work given the unique way they are handling sound in the Windows software. Does anyone know for sure if the QA400 software will or won't work? I posted a note on the QA400 forum, but I'd like to hear from some real world users. Thanks.
I think you will have a problem with the Mac not knowing what driver to load to talk to the device.
I would use bootcap and just reboot the system into Windows when you need the QA400. You will need an interface to connect. The output of the box is about 1.5V max (not enough for many power amps) and the input will get fried wit more than about 2V. A simple divide by 100 probe would work, only if the amp output is referenced to ground. otherwise it will need a differential to single end converter (the point behind the box I built). I'm redoing it into a different form to fit a nice box I found while still a miniboard. It should be ready soon. I'm changing to surface mount for most of the parts to fit easier (unless I hear some pushback).
I would use bootcap and just reboot the system into Windows when you need the QA400. You will need an interface to connect. The output of the box is about 1.5V max (not enough for many power amps) and the input will get fried wit more than about 2V. A simple divide by 100 probe would work, only if the amp output is referenced to ground. otherwise it will need a differential to single end converter (the point behind the box I built). I'm redoing it into a different form to fit a nice box I found while still a miniboard. It should be ready soon. I'm changing to surface mount for most of the parts to fit easier (unless I hear some pushback).
Has the schematic changed at all? If it has, then will the sm pcb come fully loaded?
Thx-RM
Only a PCB design and layout. If there is enough interest I might work out getting a run made but it needs to be enough to set up the process.
The circuit is the same. I'm ditching the switches (too confusing to use) and setting fixed gain.
I think I'll see if I can persuade my tech to assemble up a few and offer those. I try to negotiate a price that works. I'll publish the complete DIY instructions anyway, but I do understand the desire to just buy it assembled. I'll include the cables (probably more work than the box) just so its complete.
At -5dBv input level and no notch filter in use on the input --
I did a comparison test against a known Lab Standard (ShibaSoku AD725D) and it looks like the 2H indicates 3dB low from the stock QA400. Will compare/verify with A-Prec 2722 later. I would say their spec of 105dB range is accurate.
Thx-RNMarsh
I did a comparison test against a known Lab Standard (ShibaSoku AD725D) and it looks like the 2H indicates 3dB low from the stock QA400. Will compare/verify with A-Prec 2722 later. I would say their spec of 105dB range is accurate.
Thx-RNMarsh
Last edited:
I asked this question to Demian in a PM. He thought this would be a great question to ask in open forum for all to benefit. I agree.
I need help understanding how to properly interface my QA400 with all the precursor equipment between the QA400 and a power amplifier I am testing. It confuses me accounting for all the level changes from the input oscillator and what is presented to me on my laptop.
I built Pete Millett's "Soundcard Interface" 6 months ago. I am just now getting around to testing it with the QA400. I am confused about adjusting the input levels on the QA400 to accurately display correct readings. I have also noticed a discrepancy in THD between the QA400 and my HP339 Distortion Analyzer.
For this setup I am testing a Carver TFM-15 Power Amplifier.
Here is my setup. "HP239 Oscillator" to "Carver TFM-15 Amp" to "Pete Millett's Sound Card Interface" to "QA400" to "laptop".
I adjusted my oscillator to reflect .775 vrms (0dB) on the output of Pete's Interface. The amp measures 14.6 vrms (25.5dB) on my 8 ohm load. I picked 0dB to make for easy math.
I am confused. How do I take into account the difference between the amplifiers high output and the level reduction of the Soundcard Interface that is presented to the QA4000?
I tried it both ways in the QA400. Input gain set to 0 and then -25.5 dB. I just don't know if I have it right.
Please see pictures.
Second question regards THD. My HP339 measured .012% THD (with the setup above). The QA400 is showing .005%. Why the difference? Do I have my setup wrong?
I think I am on the right track but just not there yet.
I need help understanding how to properly interface my QA400 with all the precursor equipment between the QA400 and a power amplifier I am testing. It confuses me accounting for all the level changes from the input oscillator and what is presented to me on my laptop.
I built Pete Millett's "Soundcard Interface" 6 months ago. I am just now getting around to testing it with the QA400. I am confused about adjusting the input levels on the QA400 to accurately display correct readings. I have also noticed a discrepancy in THD between the QA400 and my HP339 Distortion Analyzer.
For this setup I am testing a Carver TFM-15 Power Amplifier.
Here is my setup. "HP239 Oscillator" to "Carver TFM-15 Amp" to "Pete Millett's Sound Card Interface" to "QA400" to "laptop".
I adjusted my oscillator to reflect .775 vrms (0dB) on the output of Pete's Interface. The amp measures 14.6 vrms (25.5dB) on my 8 ohm load. I picked 0dB to make for easy math.
I am confused. How do I take into account the difference between the amplifiers high output and the level reduction of the Soundcard Interface that is presented to the QA4000?
I tried it both ways in the QA400. Input gain set to 0 and then -25.5 dB. I just don't know if I have it right.
Please see pictures.
Second question regards THD. My HP339 measured .012% THD (with the setup above). The QA400 is showing .005%. Why the difference? Do I have my setup wrong?
I think I am on the right track but just not there yet.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
First map the gains (and losses) in each known stage. Use dB or you will go nuts. Decide where to use dBV or dBM. I suggest dBV (dB referenced to 1V) its easier.
Also figure out what you want to know-
Amp gain
Frequency response
SNR
Distortion
Then measure what is going into the amp in dbV.
Measure what is coming out in dBV - the difference is the voltage gain of the amp in dB. (Power gain will give you a headache and isn't useful here.)
The QA400 offers several measurement units. If you don't need to know absolute levels (the PK interface can do that for you) use the dBFS which should show close to overload you are. The scaling option with absolute voltages works but you really need an accurate voltage source to calibrate it and when you change the input attenuator on the PK interface the numbers all need to change again. I would leave that until you are getting good readings and understand what you are seeing.
Distortion-
there are several ways of expressing distortion. the HP339 only knows one of them: THD+N in a frequency range, measured average indicating RMS. The QA400 will show THD (the RMS sum of the harmonics only, ignoring the noise) or THD+N including noise (I forgot what band adjustments there are). The noise band is important. If it includes hum frequencies they can dominate a measurement. Second, if noise is included and the source has pure white noise the difference between a 30 KHz band and an 80 KHz band will be 80/30 more noise since the noise is constant and directly proportional to the bandwidth. Noise rarely is white so its not this simple but it explains the idea.
I think understanding the harmonic spectrum is more important, particularly the higher harmonics.
I hope this helps and expect it will confuse some so please keep asking.
Also figure out what you want to know-
Amp gain
Frequency response
SNR
Distortion
Then measure what is going into the amp in dbV.
Measure what is coming out in dBV - the difference is the voltage gain of the amp in dB. (Power gain will give you a headache and isn't useful here.)
The QA400 offers several measurement units. If you don't need to know absolute levels (the PK interface can do that for you) use the dBFS which should show close to overload you are. The scaling option with absolute voltages works but you really need an accurate voltage source to calibrate it and when you change the input attenuator on the PK interface the numbers all need to change again. I would leave that until you are getting good readings and understand what you are seeing.
Distortion-
there are several ways of expressing distortion. the HP339 only knows one of them: THD+N in a frequency range, measured average indicating RMS. The QA400 will show THD (the RMS sum of the harmonics only, ignoring the noise) or THD+N including noise (I forgot what band adjustments there are). The noise band is important. If it includes hum frequencies they can dominate a measurement. Second, if noise is included and the source has pure white noise the difference between a 30 KHz band and an 80 KHz band will be 80/30 more noise since the noise is constant and directly proportional to the bandwidth. Noise rarely is white so its not this simple but it explains the idea.
I think understanding the harmonic spectrum is more important, particularly the higher harmonics.
I hope this helps and expect it will confuse some so please keep asking.
I did some more experimenting today.
Yes they are in good agreement. But I believe my testing method was flawed. After I measured the distortion reported on the QA400 I moved the input cable from the QA400 to the HP339. This setup includes the HP339 Oscillator to the Amplifier to Pete's Interface to the HP339 Analyzer.
I believe the limiting component in my complete setup is Pete's Interface. The noise floor is high (+15 dB) compared to both the HP339 Oscillator and the QA400.
If I measure distortion directly off of the dummy load I measure (-83dB) (0.007%).
I am going to have to take a serious look at this Soundcard Interface. The generator side of the Interface is extremely noisy and not usable. I may have something wrong with my build.