Maybe try a little square of cardboard that will fit around north south east and west of the LDRs. Glue it to the board and pour hot wax in. Just melt a candle in the oven in a jar but put the jar in a larger container full of water so that the water will warm the jar and melt the wax that way. I suppose you could boil water on the stove and put the jar in the boiling water. When you want to remove excess wax from the jar just put the jar in the freezer and it will come off of the jar easily after a few hours.
Uriah
Uriah
Candle? Can I use any wax? I supposed that a special wax was used for these purposes.
Looks easy method. Thank you Uriah.
Jose if you use 2 x 8 pin DIL opamp scockets to mount your NSL32SR2S's into you can insert and change them easily, then it's very easy to solder them with a pointed tip iron into the sockets once your satisfied you have a good quad match set inserted.
In the early days I used the clear plastic case cut in half from Tic Tac mints, as the case was slightly tapered (important) which is a must once the wax has hardened so it can slip off easily. And it fitted perfectly over the 2 x DIL sockets then pour the wax in.
I use Snow Ski base wax as it is harder and has a higher melting point, just in case your in the tropics or somewhere hot.
Cheers George
In the early days I used the clear plastic case cut in half from Tic Tac mints, as the case was slightly tapered (important) which is a must once the wax has hardened so it can slip off easily. And it fitted perfectly over the 2 x DIL sockets then pour the wax in.
I use Snow Ski base wax as it is harder and has a higher melting point, just in case your in the tropics or somewhere hot.
Cheers George
Last edited:
in old tube amp trafos they used bees wax
I think its from what you call the bee comb
why not use araldite/epoxy filler ?
or maybe I misunderstood something
These are good for permanent potting Tinitus, but no good for any, if needed, service work.
With using hard wax if ever someone fed something nasty into the inputs or outputs of the Lightspeed Attenuator and blew the NSL32SR2S's I can easily turn the unit upside down, direct a hot air gun onto the wax and it drips away in seconds to reveal all the NSL's to be replaced.
Cheers George
I bought 32sr3, ten numbers. I wanted to try matching them, but as their range is quite high, could not get 2 good matched pairs.
Then I said to myself, why do I need to match them. If I mismatch them by the same amount more or less, I should get a well balanced(Left and Right) volume control.
Lastly I just used George's 100K pot circuit with the 2 presets and 4 series Rs w LEDs.
I reduced their range by reducing the pot value by fixed Rs. The max R of LDR goes to 26K. Other channel goes to 22k approx.
And I have a volume control that deviates (between 2 channels)to 6mV with 700mV input, that too with no matching of LDRs.
Gajanan Phadte
Then I said to myself, why do I need to match them. If I mismatch them by the same amount more or less, I should get a well balanced(Left and Right) volume control.
Lastly I just used George's 100K pot circuit with the 2 presets and 4 series Rs w LEDs.
I reduced their range by reducing the pot value by fixed Rs. The max R of LDR goes to 26K. Other channel goes to 22k approx.
And I have a volume control that deviates (between 2 channels)to 6mV with 700mV input, that too with no matching of LDRs.
Gajanan Phadte
Yes you can do it that way, but it will have vastly differing i/o impedances between channels at different levels, some claim this can be heard with different systems configurations re: h/f & l/f filtering and loading, I would rather have everything symmetrical between channels, that way there's no doubt in ones mind. Can you imagine the backlash I would get if I did this to the production Lightspeed Attenuator.
Cheers George
Cheers George
A matched volume control as far as having the same volume for each channel yes, but not the same i/o impedances for each channel.
eg: one channel could be 200ohms series and 100ohms shunt for a given volume. Yet the other channel could be 400ohms series and 200ohms shunt, for the same volume, yet the i/o impedances as seen by each channel of the source and poweramp will be different.
Cheers George
Ok. With this sockets is more easy all. May be used how protoboard too. Good idea. The first is get a matched quad.
Thank you again.
Regards and good week.
Jose
A 10position 4way switch would allow 10 different volume settings. All of them would be matched for impedance and for attenuation and for balance.
Every step of every way can be trimmed to any accuracy that the builder deems necessary.
I am not suggesting this as a serious alternative, simply that it would be possible to use almost matched LDRs and use trimming to reduce the errors in matching.
Every step of every way can be trimmed to any accuracy that the builder deems necessary.
I am not suggesting this as a serious alternative, simply that it would be possible to use almost matched LDRs and use trimming to reduce the errors in matching.
I´m matching measury the impedance in five position of the potentiometer. I think must be enough.
Matching could be done fairly easily with electronic needs.
The hard part is figuring out how to compare the resistances of the LDRs without compromising the sound quality.
The fast and dirty way would be to superimpose a dc-signal and put a cap on the output. Think phantom power...
Tap the dc-signal and run a servo to get full matching all the time.
It's not pretty and the cap might do some nasty stuff to the sound but it should have a fair chance of working?
The hard part is figuring out how to compare the resistances of the LDRs without compromising the sound quality.
The fast and dirty way would be to superimpose a dc-signal and put a cap on the output. Think phantom power...
Tap the dc-signal and run a servo to get full matching all the time.
It's not pretty and the cap might do some nasty stuff to the sound but it should have a fair chance of working?
Andrew
I have been thinking how to do balanced the right way but without microprocessors and what you are talking about is what I have been trying to noodle. For instance Electroswitch makes a 4 deck 12 position switch. Could use two of them. One for each channel L and R.
Seems a no brainer and it really is a simple concept but I am trying to figure how to make this useable for DIYers. I end up having to trim 8 LDRs for balanced at 12 steps so no doubt I need 96 trimmers. I dont mind that at all actually. It might take me a few hours to tune all the LDRs so that they all match perfectly at the 12 steps. Then use one more switch which switches one more series LDR (so each part of the signal that is in series with the signal will have 2 LDRs, shunt just one LDR) so this extra LDR would switch on and off between 40R and maybe 500R so that I would have a mid-step of volume giving really 24 steps of volume rather than just 12. Easy to implement. I just am worried that DIYers would not put up with 96 trimmers, plus even chinese trimmers at QTY96 are not cheap.
Another option to make it less expensive and less involved is those coded rotary switches that work in hexidecimal. I have to work out the numbers in a spreadsheet but you could get away with 16 steps (add the 40/500R switch and we have 32 steps) I am just not sure we could get anything close to log volume if I did it that way, but it would seriously cut down on the number of trimmers necessary. Plus the way I am thinking would only require one deck. A lot of LM334 and a trimmer for each one. I think a total of 4 trimmers for each LDR. See the style switch I am talking about http://www.cole-switches.com/pdfs/1800Code.pdf
Plus then you could use the same concept for single ended operation.
I have been thinking how to do balanced the right way but without microprocessors and what you are talking about is what I have been trying to noodle. For instance Electroswitch makes a 4 deck 12 position switch. Could use two of them. One for each channel L and R.
Seems a no brainer and it really is a simple concept but I am trying to figure how to make this useable for DIYers. I end up having to trim 8 LDRs for balanced at 12 steps so no doubt I need 96 trimmers. I dont mind that at all actually. It might take me a few hours to tune all the LDRs so that they all match perfectly at the 12 steps. Then use one more switch which switches one more series LDR (so each part of the signal that is in series with the signal will have 2 LDRs, shunt just one LDR) so this extra LDR would switch on and off between 40R and maybe 500R so that I would have a mid-step of volume giving really 24 steps of volume rather than just 12. Easy to implement. I just am worried that DIYers would not put up with 96 trimmers, plus even chinese trimmers at QTY96 are not cheap.
Another option to make it less expensive and less involved is those coded rotary switches that work in hexidecimal. I have to work out the numbers in a spreadsheet but you could get away with 16 steps (add the 40/500R switch and we have 32 steps) I am just not sure we could get anything close to log volume if I did it that way, but it would seriously cut down on the number of trimmers necessary. Plus the way I am thinking would only require one deck. A lot of LM334 and a trimmer for each one. I think a total of 4 trimmers for each LDR. See the style switch I am talking about http://www.cole-switches.com/pdfs/1800Code.pdf
Plus then you could use the same concept for single ended operation.
Last edited: