[RollsDownWindowsManually]

I have an older headroom Micro DAC and Micro Amp and want to level match the output with my Pixel XL phone and laptop headphone output to see if I can tell the difference with different headphones (Ety ER-4XR, HD800, Beyer DT-250, Koss Portapro, Sony MDR-7506, etc).

Just randomly trying to do so without scientific level matching, I don't think I can tell a difference with the Etymotic, Koss or Sonys, but when I step it up to the HD800 the volume difference makes things noticeable, moreso on the Pixel phone vs the Micro Amp.

Still, I'd like to do a test where I level match the ampflifier & headphone outs.

Could I do this with just a multi-meter measuring the voltage and an A/B swticher like this one https://www.amazon.com/Stereo-Switcher-Splitter-Adapter-Headphones/dp/B01MXVIYVC/ref=sr_1_3?s=electronics&ie=UTF8&qid=1497133678&sr=1-3&keywords=a%2Fb+switch+headphone ? To measure the voltage, could I just cut open an old headphone cable, attach the stripped ends to the multi-meter and match voltages, or would there be some risk of damaging the electronics?

Also, if I wanted to measure the output impedance of my sources, what's the best way to do so? I understand I will need a resistor. Could i just put something like this https://www.amazon.com/a12050500ux0...id=1497134070&sr=1-1&keywords=25+ohm+resistor in line with the positive lead for each stereo channel, play a 60hz sine wave, and compare the voltage I meaure against the no-load voltage to calculate the output impedance, then use the Zo = RL(V-VL)/VL formula.

Thanks, sorry if this has been covered, search didn't yeild the answers. Sorry, I am not much of an electronic tinkerer, the extend of my multimeter experience has just been setting gains on 12v car stereo amplifiers. I'm the type of person that can't tell the difference between 192 kbps LAME encoded MP3's and lossless (I think my hearing tops out at 16 khz) so I'm always interested in seeing if I can save money/hassle/disc space by eliminating audio components that yield no audible benefit to me.

 

[Speedskater]

Does your multi-meter have low voltage AC ranges. The DMM doesn't need to be true RMS or even vary frequency accurate.

 

[RollsDownWindowsManually]

Not sure, it's one of the cheap yellow radio shack style multi-meters that I've had for 10 years or so, away from home for work right now, else I'd check.

 

[VNandor]

For the volume matching part, you don't have to tear down a headphone cable if you happen to have jack to anything (to jack, rca, 6.5mm jack, etc.) cable. You could use a sine wave for the volume matching, just make sure your multimeter is on the correct measuring range and it's measuring AC not DC. The switch isn't necessary for the volume matching but it could come handy for the actual blind test because shorter switching time helps to spot subtle differences.

I would say the easiest way to measure the output impedance without any calculations is to have a variable resistor something that could go to let's say 100 ohms. Measure the output voltage without the load, then connect the variable resistor into the circuit. Change the resistance of the resistor until the output voltage dropped to the half of the original value, then measure the the variable resistor's resistance. It should be the same as the output impedance.

 

[castleofargh]

as suggested above, having several resistor values might be cool sometimes, if only to confirm the results in case there is some odd behavior for the device into a given load or if you messed up something(not that it would EVER happen to me). for my stuff, anything below 100ohm is Russian roulette, nothing I have is precise enough to deal with such small voltage variations. but I guess it's highly related to my cheap gears and even cheaper resistors.
it could be cool to use your computer input to check that your test signal is ok (not clipped or massively distorted under a given load), because that could result in the wrong voltage reading(usually it matters only if you go for max output, but with cheap voltmeter that might be necessary).

to test if the voltmeter can do the job, ideally you'd find a usable reference. some device where a given output has been measured elsewhere. I had a sansa clip and a fiio X1. both have been measured pretty extensively on the web so I had a few known values including max voltage and impedance. it was handy to check how bad I am at playing the little engineer, and to realize that my multimeter really sucked at measuring the audio range. I've settled for a 300hz tone, somehow I always assumed 50/60hz would be the best for a multimeter but it didn't go that way in practice???

 

[pinnahertz]

Still, I'd like to do a test where I level match the ampflifier & headphone outs.

Could I do this with just a multi-meter measuring the voltage and an A/B swticher like this one https://www.amazon.com/Stereo-Switcher-Splitter-Adapter-Headphones/dp/B01MXVIYVC/ref=sr_1_3?s=electronics&ie=UTF8&qid=1497133678&sr=1-3&keywords=a%2Fb+switch+headphone ? To measure the voltage, could I just cut open an old headphone cable, attach the stripped ends to the multi-meter and match voltages, or would there be some risk of damaging the electronics?

You need to level-match with each headphone as the load, measuring the voltage applied to the headphone under test. You can't just level match the amp/dac outputs unloaded, the headphone load may change the output voltage. If you got a headphone Y adapter, something like a 3.5mm male to 2x 3.5mm females, plug the headphones into one jack, then using a 3.5mm to stereo RCA cable in the other jack, measure the voltage across one of the RCA connectors, and level match that way.

The switch box you linked to will not permit ABX testing, only AB testing, which may get you some information, but your observations will be biased. True ABX switching is not a trivial thing. The right device is fairly expensive, and not easily replicated. The problem is switching between A, B and X, which will be A or B, but unknown to the tester.

Also, if I wanted to measure the output impedance of my sources, what's the best way to do so? I understand I will need a resistor. Could i just put something like this https://www.amazon.com/a12050500ux0...id=1497134070&sr=1-1&keywords=25+ohm+resistor in line with the positive lead for each stereo channel, play a 60hz sine wave, and compare the voltage I meaure against the no-load voltage to calculate the output impedance, then use the Zo = RL(V-VL)/VL formula.

The concept is right, but that huge resistor is not necessary. A 1 watt unit would be just fine;
https://www.amazon.com/Wirewound-Re...167538&sr=1-1&keywords=25+ohm+resistor+1+watt

Some have suggested adjusting the resistor looking for a loss of 1/2 the unloaded voltage. This may work in some cases, but not likely in most because the output impedance of amps can be very low, but they don't have the ability to drive a clean signal into a load that matches their output Z. Your technique will be more accurate, providing you have means of accurate voltage measurement at low levels. You could be measuring around .1 volts, and doing that with precision requires a pretty good meter. The cheap way to do this is with software and your sound card. The free version of TrueRTA will measure RMS volts down to very low levels. Just read and follow the calibration procedure. Absolute accuracy isn't as important as measuring the loaded and unloaded voltage difference.
https://trueaudio.com/rta_abt1.htm

 

[RollsDownWindowsManually]

Could I do an ABX style test with the A/B switcher if I had another person administer the test in an ABX fashion, randomly deciding which source to be X, A & B while I cannot see the electronics, or would that too induce some kind of bias?

Although upon further consideration I'm thinking none of this will be worth the trouble.

Also, and I know this is going off on a tangent, but has anyone every done an AB or ABX style blind test of the Sennheiser HD800 vs HD800s? It seems like it would be possible if the subject were blindfolded, due to the similar if not identical feel of the two headphones on a person's head.

 

[pinnahertz]

Could I do an ABX style test with the A/B switcher if I had another person administer the test in an ABX fashion, randomly deciding which source to be X, A & B while I cannot see the electronics, or would that too induce some kind of bias?

Technically, no, that would introduce bias. X must be unknown to the tester and the testee, and the switching system must not present any clues as to which choice is live.

Although upon further consideration I'm thinking none of this will be worth the trouble.

Also, and I know this is going off on a tangent, but has anyone every done an AB or ABX style blind test of the Sennheiser HD800 vs HD800s? It seems like it would be possible if the subject were blindfolded, due to the similar if not identical feel of the two headphones on a person's head.

It is not possible to ABX headphones. The switching time between the two is far too long (auditory memory is very, very short), the assistant will know which is which and be a source of bias, and the likelihood of two pairs of headphones of even slightly different models feeling identical on the head is actually fairly small.

 

[castleofargh]

a possible problem beyond not being double blind is that a basic switch tends to make some sort of noise that will be audible for the listener, so "randomizing" might not go without cues.

 

[pinnahertz]

a possible problem beyond not being double blind is that a basic switche tends to make some sort of noise that will be audible for the listener, so "randomizing" might not go without cues.

That's what I was getting at when I said "and the switching system must not present any clues as to which choice is live. "

 

[gregorio]

Could I do an ABX style test with the A/B switcher if I had another person administer the test in an ABX fashion, randomly deciding which source to be X, A & B while I cannot see the electronics, or would that too induce some kind of bias?

That depends on the confidence level required. Personally, I have often done it as you've described, simply because of time constraints and the cost of the right ABX equipment. I have always tried to minimise bias though, I typically used to get my secretary to switch and she had no idea what she was connecting/switching between, no idea what I was testing or testing for. Technically this would effectively be a double blind or semi-double blind test, rather than an ABX but I was testing for my own curiosity/benefit, not to publish scientific papers.

G

 

[RollsDownWindowsManually]

Can anyone recommend a decent multi-meter for measuring low voltages such as headphone outputs? My old one only has a 200v & 500v AC setting, which I don't think would work very well. It takes an umcommon 12v battery anyway, which is depleted.

Also, could I just use the 75 ohm Etymotic ER-4P to S conversion cable as a resistor for my output impedance test? I forgot that it might work, and it's basically just a 75 ohm resistor AFAIK.

 

[pinnahertz]

Can anyone recommend a decent multi-meter for measuring low voltages such as headphone outputs? My old one only has a 200v & 500v AC setting, which I don't think would work very well. It takes an umcommon 12v battery anyway, which is depleted.

Almost anything can measure low AC voltages these days, but remember a DVM is only accurate over a relatively limited range of frequencies, like 50Hz to 400Hz, typically. They'll measure outside of that but the reading will be in error. Multi-meters are not audio devices. Another solution is to use software and your sound card, and just calibrate with the DVM.

This one looks kind of interesting:
https://www.parts-express.com/true-...etooth-40-android-and-ios-compatible--390-737

Also, could I just use the 75 ohm Etymotic ER-4P to S conversion cable as a resistor for my output impedance test? I forgot that it might work, and it's basically just a 75 ohm resistor AFAIK.

Real resistors are very cheap, just get an assortment of values to play with.

 

[castleofargh]

Can anyone recommend a decent multi-meter for measuring low voltages such as headphone outputs? My old one only has a 200v & 500v AC setting, which I don't think would work very well. It takes an umcommon 12v battery anyway, which is depleted.

Also, could I just use the 75 ohm Etymotic ER-4P to S conversion cable as a resistor for my output impedance test? I forgot that it might work, and it's basically just a 75 ohm resistor AFAIK.

in time you will probably like to have several values depending on the situation or to validate your other measurement, but if you have nothing else then sure the ety adaptater cable will do. if you start playing with the pins, you can even put both channels in series and double the impedance, as this cable uses one resistor in each channel. ^_^
I don't remember if it's exactly 75ohm and don't have one anymore but usually when Etymotic provides a value it's correct.

as for a multimeter, I only know the one I have and wouldn't recommend it. I had to go through way too much troubles just to make sure I had one usable reference I could rely on. it's very fine for impedance and rather high voltages at low frequencies, but minimum reading is crap for most audio applications and I often end up like @pinnahertz said, using the input of my ADC(scarlett 2i2) to measure low voltage variations with a test tone.

 

[Speedskater]

You can make a frequency correction chart for a budget DMM.
All it requires is a CD player and a test tone disc. Most CD player are more than accurate enough. At least from 30 Hz to 15 kHz.
Burn a WAV not a MP3.