Another realization of Bob Cordell's THD Analyzer

I've completed the set of drawings. Made them with sPlan 7. If you want, I can provide the original file as well.
 

Attachments

Pots cause a lot of distortion. You're better off using relays. Even the less expensive ones are better than pots. If you want a lot of resolution you can binary weight them and get 256 frequencies out of eight relays. Log weighting is bit more complicated.
 
Pots cause a lot of distortion. You're better off using relays. Even the less expensive ones are better than pots. If you want a lot of resolution you can binary weight them and get 256 frequencies out of eight relays. Log weighting is bit more complicated.

Pot distortion is greatly reduced when there is no loading, dc or ac, on the wiper.

In regard to 8 relays giving 256 different resistances, are you referring to some kind of R2R arrangement, like used in some DACs?

Cheers,
Bob
 
Bob & Giuliano, thank you for the kind words 🙂

About the potmeters, I wasn't aware of the exsistance of those sections. But what worries me more is the tolerance between the sections. And that can cause distortion...
 
Last edited:
Pots cause a lot of distortion. You're better off using relays. Even the less expensive ones are better than pots. If you want a lot of resolution you can binary weight them and get 256 frequencies out of eight relays. Log weighting is bit more complicated.

I'm thinking about all of the quality test gear that I've used with pots in the main signal path.
 
In regard to 8 relays giving 256 different resistances, are you referring to some kind of R2R
arrangement, like used in some DACs?

For a minimal resistor count (8 resistors for 1 part in 255 resolution), they are switched in various
combinations with decoding logic, either all in series, or all in parallel.

For a series circuit, you need 8 resistors with relative Ohms values: 1, 2, 4, 8, 16, 32, 64, 128

For the parallel circuit values with 8 resistors, calculate in relative Mhos values: 1, 2, 4, 8, 16, 32, 64, 128

Both the series or parallel methods give you every step from 1 to 255 with the right combinations.
 
Last edited:
Pot distortion is greatly reduced when there is no loading, dc or ac, on the wiper.

In regard to 8 relays giving 256 different resistances, are you referring to some kind of R2R arrangement, like used in some DACs?

Cheers,
Bob

Hi Bob,

Yes it's an R2R arrangement.

This link is for a log taper done with relays.

Logarithmic Attenuator Calculator

The author has a calculator for binary weighted in dB attenuator somewhere on his site.
Click the links and see if you can find it.

I worked out the math for this. I'll look for it later and post it.
The attenuator sections are in decibels and the sections are switched in a cascade fashion. The impedance is maintained with either a constant input or constant output resistance. The constant input resistance form is the one to use for tuning a state variable oscillator.
 
Hi Bob,

Yes it's an R2R arrangement.

This link is for a log taper done with relays.

Logarithmic Attenuator Calculator

The author has a calculator for binary weighted in dB attenuator somewhere on his site.
Click the links and see if you can find it.

I worked out the math for this. I'll look for it later and post it.
The attenuator sections are in decibels and the sections are switched in a cascade fashion. The impedance is maintained with either a constant input or constant output resistance. The constant input resistance form is the one to use for tuning a state variable oscillator.

Rayma and David, thank you!

Cheers,
Bob
 
If anyone is interested in the math for this attenuator. I can post a tutorial for it.
The same circuit can be set up for any taper or graduation. It will take some time to put it together so let me know.

David, thank you for the offer. I am interested in the tutorial.
I am not in any rush and like to learn. In for a penny, in for a pound.
Then we almost went metric and we have kilograms.

When it is for my little girl, she gets grahams as in crackers 🙂

Cheers,

Sync