advertisement


Chinese stepped attenuators on eBay...

M

mikesnowdon

resU deretsigeR
Can anyone recommend a decent one? I remember some guys had good reports on a particular one using smt resistors. Are there any others worth considering?
 
Thats the one. Make sure you get one thats log + stereo, and the right k for what your planning. I use one in my cmoy build, it replaced with Alps blue and there was a noticible improvement in the sound.

And to top it all, they are cheap enough that you can just give it a go

:)

Sam
 
I've just bought a 20k one for my new Starfish build. Once I get home and finish the build I'll have to report!
 
The one I used on my ott cmoy headphone amp sounded a lot better after ~100 hrs... But then that was EVERYTHING burning in...
Sam
 
I have a couple of the gigaworks ones and have to agree they are a noticeable improvement over the blue alps (in my M^3 headphone amp).

The only gripe is the limited number of steps, on some disks I find myself continually toggling between settings. To the point that I wonder if it's sensible to put 2 in series for finer adjustment.
 
Reduce the overall gain, so that the attenuator has a smaller effect over its full sweep.
 
Guys,
Not sure about the step sizes on the gigaworks, but had thought about this:
Put a resistor in series with the incoming signal which is equivatent to half the step size (say the step is 3dB in the main part of the pot travel, then use a resistor equivalent to 1.5dB) with a switch across this resistor so it can be left in circuit to provide 1/2 stop attenuation or shorted to remove its effect. [1.5 dB on a 50k pot would be 9.5K]
Best,
AP
 
I have to agree that the steps are a tad too course for my listening environment. I am having to twiddle back and forth across one step, if my wife was out I wouldn't have the problem.

Soundwise it was fine in both my Starfish and B4 pre-amps. Certainly didn't screw the sound up in either case, no harsh highs.

Just about to replace it with a JVE, RN60D resistors, 64 steps so we shall see.
 
Hmm, recently ordered a gigaworks attenuator, so its probbably going to be a bit coarse.
Hopefully I can just float the pot up on a resistor as I always listen at one volume.
I did so a 23 step on fleabay going for $25, otherwise it looks like big money, or a pair of Gigaworks.
 
Bemused said:
Just thinking what a neat trick, but your right it wont work with a log taper.
Click to expand...

Guys,
I can assure you it will work. The pot becomes moved down. See below

A: R in = 0, B: R in = 9900 (= -1.57dB)
Wiper, Gain A dB, Wiper, Gain B dB , Δ Gain
1, -93.98, 1, -95.55, 1.57
1.4, -90.85, 1.4, -92.42, 1.57
2.1, -87.71, 2.1, -89.28, 1.57
3.0, -84.58, 3.0, -86.15, 1.57
4.2, -81.45, 4.2, -83.02, 1.57
6.1, -78.32, 6.1, -79.89, 1.57
8.7, -75.18, 8.7, -76.75, 1.57
12.5, -72.05, 12.5, -73.62, 1.57
17.9, -68.92, 17.9, -70.49, 1.57
25.7, -65.79, 25.7, -67.35, 1.57
36.8, -62.65, 36.8, -64.22, 1.57
52.8, -59.52, 52.8, -61.09, 1.57
75.8, -56.39, 75.8, -57.96, 1.57
109, -53.25, 109, -54.82, 1.57
156, -50.12, 156, -51.69, 1.57
224, -46.99, 224, -48.56, 1.57
321, -43.86, 321, -45.43, 1.57
460, -40.72, 460, -42.29, 1.57
660, -37.59, 660, -39.16, 1.57
946, -34.46, 946, -36.03, 1.57
1357, -31.33, 1357, -32.90, 1.57
1947, -28.19, 1947, -29.76, 1.57
2792, -25.06, 2792, -26.63, 1.57
4004, -21.93, 4004, -23.50, 1.57
5743, -18.80, 5743, -20.36, 1.57
8238, -15.66, 8238, -17.23, 1.57
11815, -12.53, 11815, -14.10, 1.57
16946, -9.40, 16946, -10.97, 1.57
24306, -6.27, 24306, -7.83, 1.57
34861, -3.13, 34861, -4.70, 1.57
50000, 0.00, 50000, -1.57, 1.57

This is adapted from a calculator I devised to try and even out the mis-match between channels on a Noble pot I bought.
The columns 1 and 2 show the wiper positon (ohms) and dB equivalent, for a spacing of equal dB increments with R in front of pot = 0 ohm
The columns 3 and 4 show the wiper positon (ohms) and dB equivalent, for a spacing of equal dB increments with R in front of pot = 9900 ohm.
Column 5 gives the difference between the two gain values, which is constant.
[Sorry, but tabulated results don't show very well]

Best,
AP
 
You must log in or register to reply here.


advertisement


Back
Top