Loudspeaker Relays

[Bob Cordell]

I would like to learn what experiences people here have had with relays in the output path of a power amplifier, and what recommendations anyone might have for a very good-performing PC-mountable loudspeaker relay.

When I built my listening comparison AB switch box, I spent quite a bit of time making distortion measurements on relays as candidate for use in that application, and found substantial differences among different relays. Most of those relays were large, high-current relays that were not PC-mountable. Some were automotive relays. PC mountability and size were not an issue in that application.

Cheers,
Bob

[mat02ah]

Originally posted by Bob Cordell
Most of those relays were large, high-current relays that were not PC-mountable. Some were automotive relays. PC mountability and size were not an issue in that application.

The problem with all these relays is that they are designed to switch a load. Typically relays for speaker output are switched without a load (dry). This is a more like a low-power, signal application.

So a good speaker relay is a high-power signal relay, which is not very common, to say the least ...

For me, the best (but expensive) compromise is a Matsushita (Nais, Panasonic, SDS ...) S Relay. Minimum switchable current is 100uA, max. is 4A (which makes it unsuitable for high power amps). I am using two contacts in parallel.

Thomas

[richie00boy]

It's to do with the relay construction, Doug Self wrote an excellent article in Electronics World about 8-10 years ago.

[fotios]

Originally posted by Bob Cordell
I would like to learn what experiences people here have had with relays in the output path of a power amplifier, and what recommendations anyone might have for a very good-performing PC-mountable loudspeaker relay.

When I built my listening comparison AB switch box, I spent quite a bit of time making distortion measurements on relays as candidate for use in that application, and found substantial differences among different relays. Most of those relays were large, high-current relays that were not PC-mountable. Some were automotive relays. PC mountability and size were not an issue in that application.

Cheers,
Bob

This is from my long experience with P.A. power amplifiers such Peavey.
Peavey in all models of the old good series CS used a type of relay named AROMAT. After an extensive research i discovered that those relays was direct similar with the NAIS (Matsushita) "JTN 1aS-PA-f-xxV. Concretelly these relays are of type 1 Form A which means SPNO with a switch capability of 30A. If its contact material which is Silver alloy it is not a problem for you, then i am in position to confirm that these relays are very strong and never presented any problem in their operation. The price it is about 2,23Euros per piece.

Fotios

[fotios]

Originally posted by mat02ah


The problem with all these relays is that they are designed to switch a load. Typically relays for speaker output are switched without a load (dry). This is a more like a low-power, signal application.

So a good speaker relay is a high-power signal relay, which is not very common, to say the least ...

For me, the best (but expensive) compromise is a Matsushita (Nais, Panasonic, SDS ...) S Relay. Minimum switchable current is 100uA, max. is 4A (which makes it unsuitable for high power amps). I am using two contacts in parallel.

Thomas

Hi Thomas

For the record only, i had a bad experience with the paralleled contacts, because the precision of activation mechanism to close the contacts. Some times, the one was closed faster from the other contact and thus all the load passed through one contact. I have seen in practice this as such: the one contact was damaged
instead the second was very clean. The synchronisation of moving the contacts in such type arrangements it is very critical. For this reason exactly i prefer the single contact relays.

Fotios

[Bob Cordell]

Originally posted by fotios


This is from my long experience with P.A. power amplifiers such Peavey.
Peavey in all models of the old good series CS used a type of relay named AROMAT. After an extensive research i discovered that those relays was direct similar with the NAIS (Matsushita) "JTN 1aS-PA-f-xxV. Concretelly these relays are of type 1 Form A which means SPNO with a switch capability of 30A. If its contact material which is Silver alloy it is not a problem for you, then i am in position to confirm that these relays are very strong and never presented any problem in their operation. The price it is about 2,23Euros per piece.

Fotios

Thanks for the tip. I'll try to get hold of some of these and measure them for distortion.

Cheers,
Bob

[syn08]

Bob,

I've tested myself quite a bunch of relays. Strictly for PCB mounting, this is the best I was able to identify and measure:

http://search.digikey.com/scripts/Dk...?name=PB945-ND

I though prefer the auto models of 40A or more, socketable. As long as they have silver contacts, almost all models measure good and consistent.

[EchoWars]

For the record only, i had a bad experience with the paralleled contacts, because the precision of activation mechanism to close the contacts. Some times, the one was closed faster from the other contact and thus all the load passed through one contact. I have seen in practice this as such: the one contact was damaged
instead the second was very clean. The synchronisation of moving the contacts in such type arrangements it is very critical.

Parallel contacts are used everywhere in consumer audio...why do we care if one relay closes a few mS before the other one? It is not an issue...

[mat02ah]

Originally posted by fotios

For the record only, i had a bad experience with the paralleled contacts, because the precision of activation mechanism to close the contacts. Some times, the one was closed faster from the other contact and thus all the load passed through one contact. I have seen in practice this as such: the one contact was damaged
instead the second was very clean. The synchronisation of moving the contacts in such type arrangements it is very critical. For this reason exactly i prefer the single contact relays.

Good point, but this only applies if you switch the load. In this case paralleling contacts won't help for the reason you described. So if you want to use a relay to switch from speaker 'A' to 'B', make sure that a single contact is rated for the load.

But typically speaker relays are used for muting after turning on the amp and will be switched 'dry'. In this case paralleling helps to reduce the contact resistance.

You are right again if you want to use the relay for protection (e.g. DC protection). In this case the spec of one contact counts. But for DC protection relays are close to useless anyway, because you won't find relays for DC >40V and high-current apps. All you can do is to use a change-over contact, short circuit the speaker to ground. The relay will be defective after that.

Thomas

[fotios]

Originally posted by EchoWars
Parallel contacts are used everywhere in consumer audio...why do we care if one relay closes a few mS before the other one? It is not an issue...

Under normal conditions of load of course it is not an issue because the existence only of the little iddle current throwing to the load. These conditions are: 1) During powering up the amplifier there is not signal in its input (i.e. the input it is grounded via the vol. pot. in its minimum place) so there is not a high current throwing to load. 2) During operation at high power levels, there is not any triggering of relay from the protection circuit (i.e. false detection between input and output due to large V.I. phase shifts caused from the reactive components of the load) 3) There is not any heavy shorting of output in the case of signal presented in input (and thus current flow to load). 4) There is not frequently presented failures of the output due to shorted power transistors which activates the DC protection circuit and then if the main fuses are not blown (a usual phenomenon in many amplifiers) then all the current pass through the contacts of relay during its deactivation which is the worst case for damaging the contacts (when the coil armature try to open the contacts under loading condition, the power it is by 30% bigger than of the case of closing the contacts).
Finally, why so many verbosity about relays with single or double contacts? The common thought says that if the price it is the same between a relay with single pole single touch (SPST) with a current rating of 30A why we must use a double pole single touch (DPST) with a current rating of 15A per contact?
One more remark it is that, the same relay when formed as 1A (SPST or SP N.O.) it has a current rating of 30A, and when formed as 1C (SPDT or SP C.O.) it has 20A.
To not misunderstand me (and thus to we not start a tiring debate without sense as usual happens in this forum) my thought is formed under my occupation with big power amplifiers (from +/-60Vcc up to +/-100Vcc) for 25 years and i can to confirm you that this is an issue in such devices. I have replaced enough relays with sticked contacts after a short circuit of their outputs; this related by 90% with relays of two or more contacts.

Regards

Fotios