[mgalusha]

A TAP-X owner recently asked me if I would like to have a listen to his unit and and be willing to take some basic measurements with the dScope. I agreed and those are presented below. Unfortunately the listening portion was a little more costly. I missed it so much after shipping it back I had to contact John Chapman and order the parts to build my own. 

The unit I measured was balanced, so all measurements were made in balanced mode with the TAP-X set to "54", which is unity gain. I suspect most measurements would actually improve at lower gains but I have not tested to verify this hypothesis.



This is an FFT plot, 2V RMS input. While (auto)transfomers are not active devices, there is some distortion. The good news is that even the highest distortion product is 2nd harmonic and is -108dB below the fundamental. All the really nasty high order stuff is even lower and over 120dB down from the fundamental.




This is the THD+N vs frequency at +4dBu. Since the level is fairly high, the plot is above the noise floor and essentially shows just the distortion.




This is the intermodulation measurement, 0dBu input level at 19 and 20kHz. The 1kHz difference product is about -105dB down and equates to .0005% IMD.




Frequency response. Note the full scale of the plot is +/- .1dB, so the TAP-X has less than .01dB deviation from 10hZ to 10kHz and at worst is up ~.02dB at 30kHz.

mike

[TRADERXFAN]

Wow, I thought you had essentially the same thing in your manual Slagleformer you had built... what were the  differences here?

-Tony

[mgalusha]

Wow, I thought you had essentially the same thing in your manual Slagleformer you had built... what were the  differences here?

They are the same autoformers but Mr. Chapmans implementation provides greater flexibility, finer gain steps, multiple inputs and a balance control. I built the little one to use at work, it sits between a Lindemann USB DAC and some Mackie powered monitors. I also built my TAP-X as a balanced version since I have both a balanced source and amps.

I didn't measure the Slagleformer/manual switch version but I imagine it won't be quite as good, the almost unavoidable rats nest of wires connecting the switches to the autofomers must introduce some crosstalk and other anomalies. John's implementation is super clean with high precision relays right at the transformer terminals.

[jtwrace]

Pretty awesome!

[mgalusha]

I should add, the THD+N plot is likely the residual of the generator. It's specified as -105dB and H2 in the graph is about -110dB from the fundamental.

[poseidonsvoice]

I should add, the THD+N plot is likely the residual of the generator. It's specified as -105dB and H2 in the graph is about -110dB from the fundamental.

And you stated in your 1st post that all measurements were made at Unity Gain. So in all honesty, the worst case scenario is already incredible!

Nice 

Best,
Anand.

[John Chapman]

Hello!

Thanks Mike for posting this info. Everyone knows that measurements can't predict sonics but I am a firm believer that they should be at least decent. If something does not measure well it's misbehaving in a sense and therefore will react differently in different systems. Sometimes this works out (two wrongs making a right?) but it does make it hard to get a handle on things when in another system the 'wrongs' add instead of cabncelling each other out.... . I feel better knowing that stuff should behave itself well. Mike's measurements match closely with the measurements I get here when doing testing and QC checks prior to shipping.

BTW - I can't take any credit for these nice measurements! The credit has to go to Dave Slagle who makes the autoformers. I'm just wrapping them in a package - he makes the magic bits inside.

Thansk!

John