elevated filaments and relay supply

[CopperWings]

Hello everyone! I've never realy worked with relays before so I'm wondering if it's possible to have the filament winding virtualy center-tapped and referenced to ~50V by means of a voltage divider on the end of the HT supply and rectify it to power one or two two-coil latching relays, provided that the filament winding can handle the total load. The main reason for insisting on elevating to some voltage is the existance of a dc coupled cathode follower that I feel would live a happier life (especialy if it's a 12ax7). Would I just have to block the DC component by means of caps going into a bridge rectifier or what? Any thoughts?

[Gaz]

Hello everyone! I've never realy worked with relays before so I'm wondering if it's possible to have the filament winding virtualy center-tapped and referenced to ~50V by means of a voltage divider on the end of the HT supply and rectify it to power one or two two-coil latching relays, provided that the filament winding can handle the total load. The main reason for insisting on elevating to some voltage is the existance of a dc coupled cathode follower that I feel would live a happier life (especialy if it's a 12ax7). Would I just have to block the DC component by means of caps going into a bridge rectifier or what? Any thoughts?

Unfortunately you can't have your cake and eat it too. 50 volts is 50 volts, and that's what'll be on the heater winding. It's okay because running relays off the filament string can be noisy anyway. It's much easier to use a little 5-6.3v transformer, and just use that to power the relays. Xicon makes some, which are available on Mouser for about $5.00.

[R.G.]

Even better: you can get small signal DPDT relays with dual coil latching for under $3.00. Just did some of that. These were TE-Axicom, but there are items in there for $3-4 from Fujitsu and Omron.

[defaced]

Unfortunately you can't have your cake and eat it too. 50 volts is 50 volts, and that's what'll be on the heater winding. It's okay because running relays off the filament string can be noisy anyway. It's much easier to use a little 5-6.3v transformer, and just use that to power the relays. Xicon makes some, which are available on Mouser for about $5.00.

The heaters are floating at 50v, it doesn't mean that the relay coils will have 50v across them, the rectified power to relay coils will just float with with it. So yes, you can have elevated heaters and run a relay supply off of it provided the relay supply is NOT referenced to ground, it will have to float. Now whether that will induce noise is another topic best suited for experimentation. The only limitation to this I can think of would be any rating on maximum coil to switch contact voltage.

[loudthud]

Seems to me the attached circuit would work. You may need to fine tune the capacitor values. The diodes across the 1000uF caps protect against reverse voltage at startup. They get charged by the positive bias supply.

[Gaz]

The heaters are floating at 50v, it doesn't mean that the relay coils will have 50v across them, the rectified power to relay coils will just float with with it. So yes, you can have elevated heaters and run a relay supply off of it provided the relay supply is NOT referenced to ground, it will have to float. Now whether that will induce noise is another topic best suited for experimentation. The only limitation to this I can think of would be any rating on maximum coil to switch contact voltage.

Yeah, I didn't think about floating the relay supply (make sure to isolate your footswitch jack, of course). I just looked at a Omron 5V relay datasheet and the max contact rating was 60vdc, so you're good there.

I've never seen the schem LT posted - cool!

[Chuck H]

I have my reservations about that. Is floating the supply at 56.3 volts with a contact rating of 60V really a good idea? For that matter... What about the fact that a user is likely to be holding on to his/her guitar strings (grounded to the amp chassis) at the same time as handling the footswitch with that voltage floating on the sleeve!!!

[Steve Conner]

Floating the footswitch is asking for trouble. Say you use a jack plug with a metal body to hook it up, the body will be at 56V.

An alternative to Loudthud's idea (which indeed ought to work fine) is to float the coil supply and use optocouplers to trigger the relays. It depends on the relative costs of optocouplers and bigass caps.

[Zozobra]

Its probably better to build a small dedicated power supply using a small transformer for the switching circuit. I've built a few things using bridge rectified heaters to supply relays and they always had an annoying buzz to them. When I eventually got an o-scope I traced the noise to the heaters and I blame the BR for doing it!

[R.G.]

I've built a few things using bridge rectified heaters to supply relays and they always had an annoying buzz to them. When I eventually got an o-scope I traced the noise to the heaters and I blame the BR for doing it!

You were probably right. General-purpose rectifiers have a turn-off that is both slow and abrupt. They keep conducting for a time after they're reverse biased as the charge carriers are swept out of the junction region, then turn off with a *wham*. That wham incites the parasitic inductances and capacitances connected to the diode to ring; the ringing is generally at RF, so you get a -squark!- of RF every time a rectifier turns off, 120 (or 100) times a second. This is carried and radiated throughout the circuit and then accidentally detected and made into spikes at 2x power line frequency. Sounds like a "buzz".

Ordinary rectifiers need snubbed, or replaced with rectifiers which have a soft shutoff.

I very much agree - a second power supply to run relays is easy and cheap. Not only that, it can be tucked somewhere in the box far away from the input circuits.

As a comment about elevated heaters in general, unless your tubes are compromised, good layout and wiring can achieve low hum without elevated heaters.

[Chuck H]

unless your tubes are compromised, good layout and wiring can achieve low hum without elevated heaters.

The OP indicated that the elevated heaters, in his case, were to safe margin a cathode follower circuit. Cathode followers had been a problem with some modern preamp tubes. I don't know if it still is or if the tube manufacturers caught on and changed the designs.

I did notice some hum (a really tiny amount) crept into one of my amps when I raised one of the preamp cathode resistors from the typical 1.5k to 6.8k (unbypassed). So I added an elevation circuit. That seemed to remove it. But I'll admit that I never scoped or measured anything so I can't offer anything but perception.

[CopperWings]

Thanks you guys! yeah my main concern is current production tubes. I don't feel comfortable floating the relay supply so I just might use non-elevated heaters and see how a 12ax7 or maybe ecc81 fair. But since rectifier switching noise and ringing are an issue I'll probably go for a small extra transformer to power the relays, this way I can elevate the heaters if there are issues with tubes or hum...

[Zozobra]

When I was looking into this a while ago I found some pretty small 5/9V encapsulated transformers that should mount straight to perf and power a hand full of relays for only a few pounds. I've since moved away from channel switching amps though.

[loudthud]

A simple experiment I did on an amp with an elevated heater supply that was derived from a voltage divider. Ground the elevated supply with a clip lead, then power the amp up. Carefully disconnect the ground lead and listen for the change in hum levels.