building a bridged t attenuator pot

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kdawg

Well-known member
Joined
Jun 9, 2004
Messages
129
Location
California
I decided to try and hack apart an Alpha dual linear 500 ohm pot to see if I could transform it into a 3-deck pot. It sure seems like it would work:

The original dual pot:
alpha_pot1.jpg


Back open on the dual pot:
alpha_pot2.jpg


Taking spring side of another Alpha pot and epoxy it to the dual pot:
alpha_pot3.jpg


What the final thing will look like:
alpha_pot4.jpg


Now my question is - can I use a 5k audio taper pot as the 3rd deck to make this work as a ~500 ohm T attenuator or do I need a different taper (Z taper)?

Thanks
-kdawg
 
That's a brain-twister.

If my dyslexia hasn't tripped me, I think a linear makes a better match, but mid-rotation attenuation is only 6dB.

You are 95% of the way there. At this point it may be quickest (or less brain-pain) to go ahead and tack it together.

If you load the output with 500 ohms, an ohm meter on the input should read 500 across full rotation. It won't, but maybe it will be close to constant. Same should happen if you load the input and measure the output, but if you actually wire a T-pad then it will be symmetrical both ways.

The attenuation curve can be checked with a 1VAC or 1.5VDC source through 500 ohms at the input, and 500 ohm load, measure the output across the rotation. Since the T-pad is mainly for equal source and load impedances, that's how you should measure, and the minimum attenuation will be 6dB due to matching.

In many real cases, the source Z is low and the load Z is high. You can measure that way and see how it curves.
 
Unfortunately I don't have any 5K pots around, 10K is the lowest I've got at home. But my head is spinning too regarding the orientation of this thing.

Seeing as this is going on an 1176-like project that I'm working on, I was trying to match their attenuation if possible. Silk screen says -24 db at half-rotation, but is that realistic? Here's my math (which could be off):

kd_atten2.jpg


db : R1 : R2
-6 : 502 : 498
-12 : 168 : 1.49k
-18 : 72 : 3.47k
-24 : 34 : 7.42k
-30 : 16 : 15.3k
-36 : 8 : 31.0k
-48 : 2 : 125k

So how does that come into play with my setup:
kd_atten1.jpg


I can use either leg of the pot of course if I need the inverse resistance.
 
what you're doing is very clever. i tried something similar, though i never would have come up with adding the third deck that way. bear in mind, however, that i think (could be wrong) that if you tack on a third deck that way, whatever you use will be CCW instead of CW. so if you put on a 1K Z (which is the clarostat name for log) on the end, it will act as a REVERSE log. a solution would be to take the alpha apart a bit more, and put the 1K Z deck in the middle, and the 500 ohm linear on the outside. you can swap the outside terminals of the linear deck to get CW action.

ed
 
also, the first schematic you posted is a "bridged" t attenuator, with two fixed resistor values and two variable pot decks. what you're trying to make is a non-bridged t-attenuator, with three variable pot decks. though i'm sure you have it, as a reference, check the mc77 manual pdf schematic:

http://www.purpleaudio.com/pdflib/PurpleMC77v34.pdf

note that there is an error in the posted schem (which i've asked andrew to correct) which is that the 5k Z deck should be 1k Z. he tried 5k first, and never updated the schematic after settling on 1k.

ed
 
Back
Top