They go really smooth :
http://www.produktinfo.conrad.com/d...-da-01-de-DRAHTPOTI_ABW2_2W_1_KOHM_10_LIN.pdf
http://www.produktinfo.conrad.com/d...-da-01-de-DRAHTPOTI_ABW2_2W_1_KOHM_10_LIN.pdf
On my things on my to do list is to get 10 of each type of relay possibly suitable for critical audio use and measuring the contact distortion and coil switching noise.
Do you have any candidate models to include in the tests?
The killer problem with relays is Wetting current - Wikipedia, the free encyclopedia - I went round and round in circles at one stage trying to find a 'solution' ...
I like the Panasonic relays. The DS2E types or TQ for high level. Low level I think are TSX or SX I have to check tomorrow. Unfortunately these later ones are NLA but the only designed and specked for uV levels and affordable. I think Teledyne had some but about 100$ each and in a prior career I replaced more than one. Panasonics under 10 failures in 20+ years and many thousands. Boring in a good way.
I already stated above:
http://www.diyaudio.com/forums/anal...ch-preamplifier-part-ii-4550.html#post3733668
That I realised that LDRs probably wouldn't be quiet enough.
My stated interest in them subsequent to this post is for a line level only volume control only - hope that clarifies things.
Even then, I'm not convinced they are the best way to go but since Richard seems to be enthusiastic I thought I would take another look at them.
mike
Last edited:
Here are a few candidates:
http://www.teledynerelays.com/pdf/electromechanical/732.pdf
http://www.rhopointcomponents.com/media/blfa_files/2900.pdf (2920)
They are both expensive ($25+ Ea). The mercury wetted relays are supposed to be essentially ideal but they have magnetic elements. The TO5 can relays are wonderful but very premium. They can be used to 1 GHz and higher.
Here is some other interesting reading. http://www.digikey.com/Web Export/S...witchingLowMicrovoltsSignals.pdf?redirected=1
You can buy precision programmable resistors with relays that won't compromise the precision of a .01% resistor. But they cost what a Constellation preamp costs.
Here is another way to go for programmable loading- use a programmable resistor, like this DS1267B Dual Digital Potentiometer - Overview with a transformer to match it to the target range. In the usable range of the transformer it can be very accurate. Would also work with the LDR's. Still not cheap.
And if you want the other extreme use a carbon pile resistor and a servo motor to squeeze it to the desired resistance. No sliding contacts. You may be able to mod the carbon "receiver" from an old phone to do this.
http://www.teledynerelays.com/pdf/electromechanical/732.pdf
http://www.rhopointcomponents.com/media/blfa_files/2900.pdf (2920)
They are both expensive ($25+ Ea). The mercury wetted relays are supposed to be essentially ideal but they have magnetic elements. The TO5 can relays are wonderful but very premium. They can be used to 1 GHz and higher.
Here is some other interesting reading. http://www.digikey.com/Web Export/S...witchingLowMicrovoltsSignals.pdf?redirected=1
You can buy precision programmable resistors with relays that won't compromise the precision of a .01% resistor. But they cost what a Constellation preamp costs.
Here is another way to go for programmable loading- use a programmable resistor, like this DS1267B Dual Digital Potentiometer - Overview with a transformer to match it to the target range. In the usable range of the transformer it can be very accurate. Would also work with the LDR's. Still not cheap.
And if you want the other extreme use a carbon pile resistor and a servo motor to squeeze it to the desired resistance. No sliding contacts. You may be able to mod the carbon "receiver" from an old phone to do this.
Last edited:
I've used sealed reed relays at femto-amp currents with no problem, years of sevice and 100's of thousands of operations.
Ultrasonic is an understatement, to have intermodulation you need to excite the resonance. I remain skeptical that a cart playing an LP generates much energy at 7MHz. Also my experience is that those ultra high Q factors are physically unreal due to unmodeled radiational, series resistance, and skin losses.
Indeed. The famous COTO relays at the front end of almost every semiconductor test system.
Billions of them.
yes, i also find those resonances grotesque high. Nevertheless i know people that claim that they can hear tiny changes in loading. I could imagine that a very high priced phono stage may not be sellable without that feature if that makes sense or not.
I also know certain people that listen to different loading just for the fun of it. Somehow they get enjoyment from it. Not me. For me that is hard work.
I also know certain people that listen to different loading just for the fun of it. Somehow they get enjoyment from it. Not me. For me that is hard work.
Hi Richard,
You comments above really got me thinking
I'm wondering did you actually hear this new Benchmark DAC ?
Personally to date, I just havn't liked the "sound" of any trimmers & pots have heard, not sure what it is, but suspect it's a mix of the nature of resistive element & the nature of the contact.
Perhaps the guys at Benchmark found some really good pots & trimmers but better figures alone does not convince me that it will sound good.
The only Contactless VCs I'm aware of are LDR & DVC
My experience of DVC is with Sabre 32 bit DAC - this is the first VC I ever heard that does not seem to have an audible character. I know the noise floor comes closer to the signal the quieter you go but with my reasonably low gain power amp I really have not noticed any problem.
Previously when I was participating in the "lightspeed" LDR thread some of us were trying every single noise reduction measure we could to get the best possible result and even at line levels every change was clearly audible. ( which suggests to me that the noise levels were not quite as low as was desirable ) Never-the-less, at the time some of us thought our efforts had resulted in the best sounding VC of any other analogue options we had tried including TVC and a large selection of traditional pots.
However, subsequent to that on the two occasions I compared LDR with DVC for me the DVC was the clear winner.
I would be very interested to hear if anyone here has heard any kind of a analogue VC that sounds better than a well implemented 32 bit digital VC.
I understand that it may be possible in theory for an analogue VC to be better but I cannot imagine it actually being achieved without spending a lot of money on it.
Of course you guys that have analogue sources in your systems have no choice you have to choose some kind of analogue VC.
My system is only digital so I can choose - I'd love to hear practical suggestions for an analogue VC that sounds better than a DVC in a 32 bit, very low noise digital system.
mike
Last edited:
Do you really try to misread everything? The cite was on open relays. Sealed relays fail at high currents as they cannot dissipate the ionization remnants.
The folks at Western Electric and Bell Labs who worked out reliable switching for the telephone networks did have a clue. They go for .0001% failure rate and get it or better.
Do let me know how your small sealed reed relays handle a lightning strike.
Did you sit on something sharp, or what?
The topic is switching low level signals in the absence of any standing currents. i.e signal switching. NOT telephone like level switching at the end of 1Km of lightning conductor.
Lightning will certainly zap a small reed relay, but if you are designing your MC front end to resist a direct lightning hit .... good luck with that!
Please disclose all facets of the problem or issue and folks might understand. BTW the 150uA through an ordinary hookup wire causing oxidation was a joke wasn't it?
- Status
- Not open for further replies.