Of course I can't comment on each piece of equipment, but isn't it as usual, means you have to know about some sources of problems/errors to avoid these?
Relais manufacturers (category signal relais) usually rate their types for maximal and nominal load conditions and while most people will realize that using contacts above their maximal ratings will most likely create problems. Far less intuitive is that using such relais contacts far below their nominal rating will most likely create problems too.
Using reed relais for ATE might be the ultimate solution but is it really a good idea for audio gear (in any case)?
So, one has to know about a potential problem source, has to know for which data to look for and has to evaluate if solutions from other fields are really suitable for audio applications.
All depends on signal levels. There have been measurements done and published in Linear audio. For Phono level signals reeds make a lot of sense. For line level a good quality sealed relay seems to be the best choice. Shock horror pick the right part for the job which was a solved problem when I was still in short trousers.
10+ years ago I built a number of microphone amps for a friend in a studio. They use Pana sealed relays for switching the signal at microphone level, everything from ribbons to SM58s.
They still work perfectly, every day.
Next...
They still work perfectly, every day.
Next...
As the force on the voice coil does not directly relate to the position of the voice coil under all operating conditions its not even worth trying with you. For every occasion that it's pointed out that the vast majority of speakers out there are designed for voltage drive you ignore that. Pointless to engage
That it depends on signal levels was exactly my point (more precisely one of my points); the relevant literature on these problems is quite old - but still fascinating read IMO - but it seems to be part of human nature to run in cycles, which means to solve problems and to forget about it during the following decades. Especially when analog design is somewhat less popular these days.
As good audio reproduction is a surprisingly complex matter, involving several fields, it is more likely to face problems, that were already solved but are not (anymore) part of the normal education.
The linear audio article is from 2017, so hardly 'old'. But the minimum current requirement for connection on a modern panasonic relay is low enough that its only daft eejits like me who are switching sub mV signals who need to consider reed relays.
@SJ 3rd harmonic gets higher in a reed relay with currently above about 1mA. Still very very very low mind. But if you run your signal through 50 relays as a studio might needs to be considered.
Last edited:
There are lots of sealed standard electromechanical relays with contacts appropriate for audio. Many also seem to work well below their minimum ratings. Those NEC (now Kemet) are an example and ubiquitous. They are used in older AP analyzers, maybe new too but haven’t seen a teardown..
I've been switching uA/mV for years with no (zero) problems at sub 10 ppm distortion (Panasonic AG series is my chosen weapon - YMMV).
If the relay is sealed there is no problem. If its not sealed you may have an issue, but then there's self cleaning contact relays for that kind of stuff and you'd never use them to switch small signal.
I suspect the lower switching 'limit' on small signal relays has more to do with simply spec'ing the test envelope boundaries than actual limitations on the relay. I imagine high speed testing of relays at perhaps thousands of units an hour would pose some difficult technical challenges if you wanted to characterize them for switching say 50uV at 1 or 2 uA. Thermocouple effects for one. But maybe someone has the inside spill on this.
BTW, we thrashed this subject to death a few years ago on this very thread.
On reflection, its a good thing we are all well physically separated from each other. I can imagine a good solid punch-up if we were all in the same place.
Naturally I'd be watching the proceedings from the side-lines with a beer at hand.
If the relay is sealed there is no problem. If its not sealed you may have an issue, but then there's self cleaning contact relays for that kind of stuff and you'd never use them to switch small signal.
I suspect the lower switching 'limit' on small signal relays has more to do with simply spec'ing the test envelope boundaries than actual limitations on the relay. I imagine high speed testing of relays at perhaps thousands of units an hour would pose some difficult technical challenges if you wanted to characterize them for switching say 50uV at 1 or 2 uA. Thermocouple effects for one. But maybe someone has the inside spill on this.
BTW, we thrashed this subject to death a few years ago on this very thread.
On reflection, its a good thing we are all well physically separated from each other. I can imagine a good solid punch-up if we were all in the same place.
Naturally I'd be watching the proceedings from the side-lines with a beer at hand.
I'm pretty sure Faraday understood complex algebra he certainly knew the current and voltage in an inductor were not in phase. Adding the modulus of the impedance of the resistor to that of the inductor is the wrong answer.
I suspect you missed my critique of your friends article. The statement that there is no error signal in a feedback system unless the voltage and current are out of phase is also wrong. You should both go back to Black's original formulation from 1928 or so.
Last edited:
Why pick a fight with an unarmed man?
Your fortress is a fortress of solitude. No-one else can touch it.
The issue started with Scott Wurcer denying that there were problems with dirty contacts not passing signal until a voltage threshold was reached and then they would work.
He then raised the issue of gas dielectric break over which really had nothing to do with the normal connector contact issue of air carried contamination.
After having a few folks confirm that audio contacts do get dirty and cleaning them helps he has gone back to his bit as how do relays work reliably.
That answer of course is use the right relay and for his particular case those would be sealed relays that for some strange reason don't suffer from air exposure.
Then reasonable folks asked why reed relays aren't used more for audio and the answer is simple, the reed needs to be magnetic and that introduces hysteresis in the signal based magnetic field. Yes that is at an extremely low level but since it is easily avoided by other types of relays reed relays are rarely used.
Western Electric developed the most reliable relays for use in the telephone system. These use bifurcated crossbar contacts so that there are four points of high pressure contact for each switch. They then made their own alloy to withstand the enormous number of closures they expected the relays to withstand. The modern version is of course much smaller and often uses gold over a silver base contacts. At the lowest levels the gold, which does not fail due to atmospheric junk, provides the contact surfaces. When the relay is used at higher voltage and current the gold burns off and you have the more rugged silver contacts. They would be effected by air carried junk except that the relay is handling enough voltage to clear such stuff in normal use. The problem is that once the relay has switched a high enough level to burn off the gold it should not be reused to carry the lower level signals.
He then raised the issue of gas dielectric break over which really had nothing to do with the normal connector contact issue of air carried contamination.
After having a few folks confirm that audio contacts do get dirty and cleaning them helps he has gone back to his bit as how do relays work reliably.
That answer of course is use the right relay and for his particular case those would be sealed relays that for some strange reason don't suffer from air exposure.
Then reasonable folks asked why reed relays aren't used more for audio and the answer is simple, the reed needs to be magnetic and that introduces hysteresis in the signal based magnetic field. Yes that is at an extremely low level but since it is easily avoided by other types of relays reed relays are rarely used.
Western Electric developed the most reliable relays for use in the telephone system. These use bifurcated crossbar contacts so that there are four points of high pressure contact for each switch. They then made their own alloy to withstand the enormous number of closures they expected the relays to withstand. The modern version is of course much smaller and often uses gold over a silver base contacts. At the lowest levels the gold, which does not fail due to atmospheric junk, provides the contact surfaces. When the relay is used at higher voltage and current the gold burns off and you have the more rugged silver contacts. They would be effected by air carried junk except that the relay is handling enough voltage to clear such stuff in normal use. The problem is that once the relay has switched a high enough level to burn off the gold it should not be reused to carry the lower level signals.
Last edited:
Joe,
Tutorial on Ohm's Law for AC circuits using complex algebra:
https://www.usna.edu/ECE/ee301/Lect...s and Complex Numbers in AC and Impedance.pdf
Tutorial on Ohm's Law for AC circuits using complex algebra:
https://www.usna.edu/ECE/ee301/Lect...s and Complex Numbers in AC and Impedance.pdf
0.1mV across 100Ohms is 1uA. Waay below specs for the AG series. Given a nice gent gave me a bag of ATE grade reed relays was a no brainer for me for the phono stage.
- Status
- Not open for further replies.