Cheap source/step'd-attenuator switches:

[uvodee]

Originally posted by Alex S


BUT the basic principle is just a simple voltage divider - I hope you know what that is.

If you don't understand what I'm saying, don't proceed.

Please take note of the fact, that since '92 I do not take any remarks like this from any one!!!!!
It is fun for me to e.g. make a balanced line stage, SOZ, ZenLite, GC etc... allthough I am not an EE and I learned how to solder by reading books from the library and not in school.
I do not need to take any c... from any person participating on this forum!

I feel offended!

Jean-Pierre

leve de Koning, Vive le Roi,
Leve Belgie, Vive la Belgique!

[Alex S]

I'm sorry I've offended you. I have not read postings sufficiently far back to know of your background.

Please accept my sincere apologies.

[uvodee]

I accept it.


Jean-Pierre

[uvodee]

complete ineptitude is one thing, but trying something out that I have not seen here nor anywhere else ( I hope someone else has done it before anyway!) was...... going for the idea of using the cheap pots from all electronic.
Again I must admit I may have sinned against something Holy or I may have had some beginners luck but ..... here is what I tried.


I had bought a packet of 610 resistors (61 different kinds) for about $12 and was thinking of what I could do with all these ....
i decided to take use ALL the 2K2's (+ one 2K7) by soldering them between the lugs of this particular pot. on the center lug is the IN wire attached and the last lug in series(with the R's soldered) has the outwire going to the IN pin of the LM 3875 ....

Can someone tell me here why this works perfectly ?
the sound does not increase in, let's say normal increments, but my guess is that I can work on it by using a spreadsheet from decdun or matjans or Peter....

Is this damaging to any part of the schematics/source or speakers?

What is the reason why this can be made so cheap and is not used anywhere else? (cost was $1.75 for the pot and 22 cents for all the 11 Resistors making it less than $2.00 per channel)


J-P.

[Morse]

Hi guys;

Interesting topic. Here are a few ideas and experiences - maybe there's potential for some unorthodox cheapskate use?

1. For source selection, I've had good luck with conventional DPDT toggle switches. The switching is fast enough that "click" or "bounce" has not been a problem - at least here. I was lucky enough to have a few solid silver contact jobs on hand, and they don't seem to introduce any sonic degradation.

2. For use as a stepped attenuator, I can see a couple of possibilities. One would be to use a cheap, easily available "break before make" type rotary switches and add bounce caps (say tiny ceramic discs) between each set of contacts. I've done that for a long time on power switches (usually use .01uF 3kV rated ceramic discs there) so I don't see a problem with it in this app. Does anyone else? I've not tried it yet, so I can't comment that it does or doesn't work in this app.

3. For a really oddball solution to the stepped attenuator problem, why not consider a series of DPDT's, each wired as a bypass to a 2 channel voltage divider. You could have one for 1dB cut (fine adjustment), one for 3dB cut, one for 6dB cut, one for 10dB cut, one for 20dB cut, another for 30dB cut, and a last "mute" function that would knock things down by either using the chip's onboard "mute" (if applicable) or maybe 50dB of cut, failing that. It wouldn't have the intuitive feel of a rotary switch, but I don't see why it wouldn't work, and you could use cheap, easily available DPDT's. Since the switching speed is fast on a toggle, I doubt "click" or "bounce" would be a problem, but if it was, "bounce caps" as above suggest themselves as a solution.

So, what do you guys think of these approaches? Any potential, or have I finally gone off the deep end of pennypinching?

All the best,
Morse

[GregGC]

Originally posted by Morse
Hi guys;

Interesting topic. Here are a few ideas and experiences - maybe there's potential for some unorthodox cheapskate use?

1. For source selection, I've had good luck with conventional DPDT toggle switches. The switching is fast enough that "click" or "bounce" has not been a problem - at least here. I was lucky enough to have a few solid silver contact jobs on hand, and they don't seem to introduce any sonic degradation.

2. For use as a stepped attenuator, I can see a couple of possibilities. One would be to use a cheap, easily available "break before make" type rotary switches and add bounce caps (say tiny ceramic discs) between each set of contacts. I've done that for a long time on power switches (usually use .01uF 3kV rated ceramic discs there) so I don't see a problem with it in this app. Does anyone else? I've not tried it yet, so I can't comment that it does or doesn't work in this app.

3. For a really oddball solution to the stepped attenuator problem, why not consider a series of DPDT's, each wired as a bypass to a 2 channel voltage divider. You could have one for 1dB cut (fine adjustment), one for 3dB cut, one for 6dB cut, one for 10dB cut, one for 20dB cut, another for 30dB cut, and a last "mute" function that would knock things down by either using the chip's onboard "mute" (if applicable) or maybe 50dB of cut, failing that. It wouldn't have the intuitive feel of a rotary switch, but I don't see why it wouldn't work, and you could use cheap, easily available DPDT's. Since the switching speed is fast on a toggle, I doubt "click" or "bounce" would be a problem, but if it was, "bounce caps" as above suggest themselves as a solution.

So, what do you guys think of these approaches? Any potential, or have I finally gone off the deep end of pennypinching?

All the best,
Morse

My source selector uses that kind of switches. #3 suggestion is extremely user-unfriendly. #2, you want to use make before brake. that way you won't hear silence or max volume between the steps (depending if you use ser or par. attenuator). The caps are only useful if you have much higher voltages than the audio signal (prevents arching). If you use the caps you'd really deteriorate the sound.

/Greg

[Morse]

Hi Greg GC;

Absolute agreement that the unorthodox arrangement in (3) is rather user-unfriendly; though if I'm making it for myself that isn't such an issue.

A second thought there would be to use a 2P6T make before break rotary in conjunction with a pair of DPDT. The 2P6T could be wired in 3dB increments, and the DPDT's could have a -15dB cut, each. That would give potentially -45dB of cut, so that with careful selection of gain it would be serviceable.

Absolutely it's better to use the "right" parts (in this case a 24 to 36 position make before break dual pole rotary), but those of us in the cheap seats sometimes have to....ummm...."be creative."

On bounce caps, yes they cut down on arcing in a switch, but they also eliminate "popping" or "clicking" - my understanding is that causation of "popping" on source selectors is similar - just that the transient voltage is of a smaller order (though it's before the preamp stage natuerlich, so it's more sensitive to small voltage transients). Still, I've not tried it yet, so who knows? As far as sonic degradation is concerned, I suspect that could be kept a nonissue with the proper selection of cap value. I was thinking that a nice 10-100pF disc cap might be just the ticket. 100pF would give about 80kr between each set of contacts at 20kcps, so if the resistors in the voltage dividers were chosen wisely it might not be an issue at all....

Good luck on all your projects and all the best!
Morse

[fdegrove]

Hi,

Morse,

It so happens I fully agree with on the silver contact DPDT switches. I even often tie both poles together for beter sonics...

Heck I even switch on my 120W OTL amps with such a tiny switch,
not a single problem since over 10 years of abuse.

Regarding point #2. I doubt the caps will attenuate the full signal switch over, how could they?

Regarding point #3. Not really practical except perhaps for some extra pre-attenuator, a next gear switch, if you like.

If I recall correctly this kind of arrangement was once described by Ben Duncan in Hi-Fi News.

Cheers,

[GregGC]

Originally posted by Morse
Hi Greg GC;

Absolute agreement that the unorthodox arrangement in (3) is rather user-unfriendly; though if I'm making it for myself that isn't such an issue.

A second thought there would be to use a 2P6T make before break rotary in conjunction with a pair of DPDT. The 2P6T could be wired in 3dB increments, and the DPDT's could have a -15dB cut, each. That would give potentially -45dB of cut, so that with careful selection of gain it would be serviceable.

Absolutely it's better to use the "right" parts (in this case a 24 to 36 position make before break dual pole rotary), but those of us in the cheap seats sometimes have to....ummm...."be creative."

On bounce caps, yes they cut down on arcing in a switch, but they also eliminate "popping" or "clicking" - my understanding is that causation of "popping" on source selectors is similar - just that the transient voltage is of a smaller order (though it's before the preamp stage natuerlich, so it's more sensitive to small voltage transients). Still, I've not tried it yet, so who knows? As far as sonic degradation is concerned, I suspect that could be kept a nonissue with the proper selection of cap value. I was thinking that a nice 10-100pF disc cap might be just the ticket. 100pF would give about 80kr between each set of contacts at 20kcps, so if the resistors in the voltage dividers were chosen wisely it might not be an issue at all....

Good luck on all your projects and all the best!
Morse

Morse,

I bought a M-B-B 12 psition rotary switch for $5 from Active. It's not the highest quality, but works just fine. I don't think you'll eliminate the poping with caps, but you never know. If it works for you why not. I'd spend some time to find an inexpensive switch. Heck I think a stere pot will give you better result than B-B-M switch with caps.

If you haven't seen my GC (sounds really good to me), here it is:

http://www.diyaudio.com/forums/showt...939#post272939

[Morse]

Hi Frank;

First off, that's some really GREAT looking layout and wiring work. My compliments! My prototype "DormAmp" (TDA2030A 16w stereo amp) is pretty crude inside in comparison!

Hmmm, thanks for the tip on "Active"; would you have a link to them handy? Locally I've not been able to source a 12 position 2 pole rotary - just 12 position 1 pole, 6 position 2 pole, or 4 position 3 pole switches, with only the last type being MBB. I'd like to try a low cost stepped attenuator some time (everything I do now is low $$....) just for the experience.

'Far as "regular" bounce caps are concerned (YOU know this, but there are others reading this thread....), they can be a real lifesaver (or at least loudspeaker saver!). With the DormAmp, in it's earliest version, the "pop" was really a killer when you turned it off. One pair of bounce caps later, it's dead quiet on shutdown....

Pics of the DormAmp will follow as soon as I get a chance to borrow a friend's digital camera - but that's got to wait 'til my car's running again. Just hope the MK II version of it looks half as good as your GC!

All the best,
Morse