How to - Distortion Measurements with REW

The 50 Hz peak is a cabling issue.
Maybe I use too high impedance on the end of the measuring cable.
 

Attachments

  • Meas_self_1kHz_1.PNG
    Meas_self_1kHz_1.PNG
    309.2 KB · Views: 349
  • Meas_self_1kHz_2.PNG
    Meas_self_1kHz_2.PNG
    301.2 KB · Views: 345
Yes they are 24 bit interfaces. You might have the Focusrite ASIO driver installed already. Go to the REW preferences and see if there is an ASIO driver (instead of JAVA). If so try using the ASIO driver and try selecting 25 bit dither instead of 16 bit dither in the generator.

If not look at this Focusrite page and search for the text "make sure you select the Focusrite ASIO driver in the audio settings for that software"
How to install Focusrite drivers on Windows – Focusrite Audio Engineering
 
Member
Joined 2016
Paid Member
Before switching to a QA401 I had a scarlett 2i4. Wasn't to bad actually but you need to use an external attenuator to get the voltage down to the sweet spot which was 0.62v and I found the the output voltage was limited.

I did find that virtins multi instrument was alot better than REW.

You can download a trial and if you search the net it will tell you how to reset the trial period if you need to.

Multi-Instrument, a powerful multi-function virtual instrument software - Virtins Technology

Enjoy. Screenshot_20210327-064809_Photos.jpg
 
Before switching to a QA401 I had a scarlett 2i4. Wasn't to bad actually but you need to use an external attenuator to get the voltage down to the sweet spot which was 0.62v and I found the the output voltage was limited.

I did find that virtins multi instrument was alot better than REW.

You can download a trial and if you search the net it will tell you how to reset the trial period if you need to.

Multi-Instrument, a powerful multi-function virtual instrument software - Virtins Technology

Enjoy. View attachment 935987

Thanks,
I use a voltage divider.
The 2i4 1st gen has a better output, and the Solo 3rd gen has better input.
I used the Solo for the measurements.
Thanks for advice, I will try it.
 
Member
Joined 2016
Paid Member
Thanks,

I use a voltage divider.

That's actually not the best approach you can loose alot of drive current using a voltage divider. You are better off measuring the input impedance and then using a Pi filter.

If you can. The ultimate wiring setup in my opinion is putting a attenuator on the positive and negative inputs and then take your measurements differentially. This way you get rid of any common mode noise.

Also take care not to create a ground loop with the scarlett

Here is a link to explain it all in more detail.

QA401 - Differential measurements with attenuation - QA401 - QuantAsylum Forum

I also sorted the distortion issue on my amplifier as well it was the those doggy ebay load resistors.

Check out post 2444 here for the final results

diyAB Amp - The "Honey Badger"


diyAB Amp - The "Honey Badger"
 
Last edited:
If I understand it correctly you used an Ebay cement wire wound resistor with poor results? Did you end up with good results from the parallel & series combination of a number of those cement wire wound resistors? Or did you use a resistor different type in the end to get good results?
 
Member
Joined 2016
Paid Member
If I understand it correctly you used an Ebay cement wire wound resistor with poor results? Did you end up with good results from the parallel & series combination of a number of those cement wire wound resistors? Or did you use a resistor different type in the end to get good results?

Yes exactly what you said.
I'm going to use the parallel & series resistor array that I created as a baseline as I search for a solution with higher wattage as that current setup is only good for one or two short tests before taking a break to allow the load resistor to cool. I did order a few more of those resistors so I can make a 5x5 array.

Something to note with evey parallel string you reduce the inductance.
 
Member
Joined 2016
Paid Member
I was thinking of putting four 16 Ohm 100W aluminum case resistors in parallel to get 4 Ohms. I wonder if that is a good configuration for low inductance.
Do you have a way to measure the inductance? If not you might like to purchase a few of these to use as a baseline to complete the other resistor you are considering.

I ended up with 1uH.
But its not just about that. As keantoken said in my thread its about how the magnetic field from the collapsing inductance injections more harmonics into the measured signal.

AU $2.46 5%OFF | 10pcs 10W 5% Cement Resistor Power Resistance 0.1 ~ 10K 0.1R 0.5R 10R 50R 0.22 0.33 0.5 1 2 5 8 10 15 20 25 30 100 1K 2K 3K ohm
10pcs 10W 5% Cement Resistor Power Resistance 0.1 ~ 10K 0.1R 0.5R 10R 50R 0.22 0.33 0.5 1 2 5 8 10 15 20 25 30 100 1K 2K 3K ohm|Resistors| - AliExpress 20210327_172037.jpg Screenshot_20210328-071129_Photos.jpg
 
Last edited:
Last edited:
Member
Joined 2016
Paid Member
Personally I was looking at this as its spec is low inductance.

284-NHS200-8.0F NHS200 8R0 1% ARCOL / Ohmite | Mouser Australia

But I'm now cautious of the metal casing. But have an enquiry in with mouser to get the exact inductance.

I have also been looking at this.
100W 8 Ohm(8R) inductive Resistor Tube Power Audio Amplifier Test Dummy Load 1Pc | eBay

Or this

1000 WATT 8 ohm test DUMMY LOAD non inductive for audio amplifier | eBay
 
Member
Joined 2016
Paid Member
Last edited:
That's actually not the best approach you can loose alot of drive current using a voltage divider. You are better off measuring the input impedance and then using a Pi filter.

If you can. The ultimate wiring setup in my opinion is putting a attenuator on the positive and negative inputs and then take your measurements differentially. This way you get rid of any common mode noise.

Also take care not to create a ground loop with the scarlett

Here is a link to explain it all in more detail.

QA401 - Differential measurements with attenuation - QA401 - QuantAsylum Forum

That's my simmetrical voltage divider = attenuator.

I also sorted the distortion issue on my amplifier as well it was the those doggy ebay load resistors.

Check out post 2444 here for the final results

diyAB Amp - The "Honey Badger"


diyAB Amp - The "Honey Badger"

Thats my attenuator = simetrical voltage devider
 

Attachments

  • IMG_20210322_123232 (2).jpg
    IMG_20210322_123232 (2).jpg
    213.2 KB · Views: 394
Member
Joined 2016
Paid Member
Thats my attenuator = simetrical voltage devider
Good work.
Can you give me some numbers?

Amplifier output voltage Vrms or Vpk-pk

And your target input voltage which you possibly attained through loopback testing to get the lowest distortion levels
Mine was 630mv rms

Also have you measure your input impedance of your scarlett.

Mine was 13,675 ohms
 
Last edited: