5VDC from center-tapped 6.3VAC?

[keithv]

I'm trying to create a regulated 5VDC supply using the 6.3VAC heater supply.  However, my 6.3VAC secondary has a center tap.  Using a 4700uF cap and 1N4007 full wave rectifier, I get 7.5VDC, and 5VDC out of the regulator, but only on its own virtual ground.  If I connect the negative end of the rectifier to chassis ground, the negative half of the rectifier explodes.  I can only get 3.5v from half wave rectification. 

I did some looking around on the web and this was all I could find about the subject http://www.valvewizard.co.uk/heater.html  It suggests that the negative rectified DC voltage be kept separate from the center tap (chassis) ground.  I want to use the 5VDC to operate a relay via a footswitch.  Am I supposed to keep the ground of the relay circuit isolated from chassis ground?  Or should I remove the center tap, and then make the chassis ground reference after the rectifier?

[keithv]

Figured it out by experimenting.  The answer was to remove the center tap and reference the negative rectified DC voltage to ground.  Works now and no more exploding diodes!

[sluckey]

I get 7.5VDC, and 5VDC out of the regulator, but only on its own virtual ground.

That's right. There is no reason to connect any part of the 5V supply to chassis. Just let the supply float. Use isolated jacks for any footswitch connection also.

It suggests that the negative rectified DC voltage be kept separate from the center tap (chassis) ground.  I want to use the 5VDC to operate a relay via a footswitch.  Am I supposed to keep the ground of the relay circuit isolated from chassis ground?

That's right. There is no reason to connect any part of the 5V supply to chassis. Just let the supply float. Use isolated jacks for any footswitch connection also.

I purposely repeated myself. 

Figured it out by experimenting.  The answer was to remove the center tap and reference the negative rectified DC voltage to ground.  Works now and no more exploding diodes!

That makes the 5 volt supply happy, but now your filament circuit has no reference to ground and filament hum will probably increase. The better solution is to float the 5 volt supply and keep the center tap tied to ground to reduce filament hum.
.

[Willabe]

Here's a link for more information on hooking up relay power supplies; 

http://el34world.com/Forum/index.php?topic=14348.0

[HotBluePlates]

...  Using a ... 1N4007 full wave rectifier ...  If I connect the negative end of the rectifier to chassis ground, the negative half of the rectifier explodes.  ...

Terminology Police:
If you used a 4-pin rectifier package, with +, - and a.c. input ("~") terminals, that is a full-wave bridge, or simply "bridge," rectifier. A "full-wave rectifier" uses only diodes from the a.c. input to the + output, and a transformer winding with a center-tap ground reference.

Both rectifier arrangements are "full-wave" but one requires a center-tapped transformer winding to develop any output, while the other has diodes which explode if you connect that center-tap. 

[keithv]

...  Using a ... 1N4007 full wave rectifier ...  If I connect the negative end of the rectifier to chassis ground, the negative half of the rectifier explodes.  ...

Terminology Police:
If you used a 4-pin rectifier package, with +, - and a.c. input ("~") terminals, that is a full-wave bridge, or simply "bridge," rectifier. A "full-wave rectifier" uses only diodes from the a.c. input to the + output, and a transformer winding with a center-tap ground reference.

Both rectifier arrangements are "full-wave" but one requires a center-tapped transformer winding to develop any output, while the other has diodes which explode if you connect that center-tap. 

LOL

Anyhow...  It hums like a sonofabitch now.  There's really no way to do this without keeping the 5V regulated supply ground isolated?  What about taking an approach like with the Marshall 1959 circuit, where they used a full-wave bridge rectifier and a center tapped PT for the B+ voltage?  The center tap of the PT is referenced on the AC side through a pair of .22uF caps and by a pair of 100uF caps on the DC side.  Could something like this be applied to the 6.3VAC secondary?

What about something like a Cockcroft-Walton rectifier to get the voltage up high enough to be regulated and still reference the center tap to chassis ground?

[keithv]

I'm gonna try a single stage Cockcroft-Walton and see if I can get 6 or 7 VDC out of it with the center tap to ground.  wish me luck.

[DummyLoad]

why do you need regulation for a relay?

[keithv]

I suppose I don't, but if I can get enough voltage out of the 6.3VAC supply to regulate it, I might as well, no?  My problem was mostly that I couldn't even get 5v unregulated without full wave bridge rectification.  And I would prefer it if I could use chassis ground for the entire circuit and not have to float an isolated ground just for the relay.

Update:  So far, the single stage Cockcroft-Walton rectifier works and produces 7.7VDC with the center tap to chassis ground.  Time to see how noisy it is.

[Willabe]

And I would prefer it if I could use chassis ground for the entire circuit and not have to float an isolated ground just for the relay.

Why?

[PRR]

Regulating relay-power is like using distilled water on your lawn.

[keithv]

And I would prefer it if I could use chassis ground for the entire circuit and not have to float an isolated ground just for the relay.

Why?

What if the body of the plug of the cable I'm using for the relay footswitch touches one of the dozens of things that are connected to chassis ground?  If it were possible, wouldn't you prefer not to have an isolated ground to worry about?

[HotBluePlates]

... wouldn't you prefer not to have an isolated ground to worry about?

I'd prefer to have things working without exploding. 

[sluckey]

Isolated ground is not something I worry about.

[PRR]

When carrying a load bigger than a volt meter, those caps need to be much larger than the 0.1uFd shown.

Apologies if you already knew that, and the diagram is "concept" not "production-ready".

For 5V relays I would arm myself with 1,000uFd caps and see where it came out.

[keithv]

When carrying a load bigger than a volt meter, those caps need to be much larger than the 0.1uFd shown.

Apologies if you already knew that, and the diagram is "concept" not "production-ready".

For 5V relays I would arm myself with 1,000uFd caps and see where it came out.

Thanks for the tip.  No, this is definitely not "production ready".  I haven't tried it under any significant loads yet.  I'll keep tinkering and let you guys know of any progress.

[Willabe]

What if the body of the plug of the cable I'm using for the relay footswitch touches one of the dozens of things that are connected to chassis ground?

What dozen things? The only thing that I can think of that you might touch the plugs body to the amps chassis? And how are you going to be able to do that anyway?
 
If you plug it in with the amp off, then turn on the amp, no chance that can happen, even if you touch the chassis.

You can get a 1/4" plug with a plastic shell body. 

[TheKT88KilledJFK]

PLEASE let me know if it is bad form to bump this, but I had the same issue as op so I obviously had to search this forum for the experience of others here. 

Just converting the rectifier section to full wave instead of bridge would take care of his issue or is this wrong?  It seems this has not been mentioned and I am curious as to why since most of you have reasons for everything you do or don't do.

Edit: tried it on the bench.  I see what the author meant.  Full wave only gives 3.5v which makes total sense.  I was confused by the "half wave" terminology.     

[kagliostro]

Mesa do it this way

on Heartbreaker


on Tremoverb


Franco

[92Volts]

If the negative output of the bridge rectifier is grounded (instead of grounding the center tap), shouldn't that provide a ground reference for the AC line as well?

Maybe increased hum in this configuration is from increased current/noise on the ground line. This doesn't seem like it would let the AC heater line "float" any more than grounding the center tap

[TheKT88KilledJFK]

Ground will no longer be the center reference point for the AC wave of the heater regardless.  The bridge rectifier pushes the potential of the transformer up so that it's AC waveform is never going negative.  This is a problem because when you ground pin 9 of the preamp tubes it's seeing twice the amplitude of ripple than you would see otherwise.  It may only be positive thus not more power, but the voltage difference is greater so the hum propagates easier. 

However, I have never bench tested this or encountered this. 

[sluckey]

This is a problem because when you ground pin 9 of the preamp tubes it's seeing twice the amplitude of ripple than you would see otherwise.

Why would you ground pin 9?

[TheKT88KilledJFK]

This is a problem because when you ground pin 9 of the preamp tubes it's seeing twice the amplitude of ripple than you would see otherwise.

Why would you ground pin 9?

I'm assuming OP left it in this configuration.  Not saying you have to, or should all the time.  But it's also a problem relative the grid & cathode, is it not?  I assume that's how the noise is getting amplified. 

[sluckey]

The OP never mentioned preamp tubes, let along grounding pin 9.

[TheKT88KilledJFK]

The OP never mentioned preamp tubes, let along grounding pin 9.

You're right but it's a configuration commonly used by hobbyists on this forum so I figured I'd list a case in which a problem would present itself. 

Pushing the heater tap up with a grounded bridge rectifier affects all the tubes on that circuit which is why you wouldn't want to ground the bridge rectifier.  You'd want to keep it isolated and ground the Ct of the heater winding instead.  This keeps more options open for heater attachment arrangements and schemes.

Since it's easier to float the bridge rectifier and you have to isolate the foot switch jack carrying the relay's signals anyways, the circuit pretty much has to be isolated from ground.

 

[sluckey]

I know how to properly get 5VDC from a 6.3VAC filament winding with center tap. My question and comment was dealing specifically with your comment about grounding pin 9 of the preamp tubes. That went out of style with the '50s tweed amps.

[TheKT88KilledJFK]

I know how to properly get 5VDC from a 6.3VAC filament winding with center tap. My question and comment was dealing specifically with your comment about grounding pin 9 of the preamp tubes. That went out of style with the '50s tweed amps.

Why's that?  You've posted before that it's a better configuration. 

[sluckey]

I know how to properly get 5VDC from a 6.3VAC filament winding with center tap. My question and comment was dealing specifically with your comment about grounding pin 9 of the preamp tubes. That went out of style with the '50s tweed amps.

Why's that?  You've posted before that it's a better configuration.

In the '50s it was common to ground one side of the filament winding and one side of the tube filament to chassis. This was to simplify filament wiring. You only needed one wire to string from one socket to the next. Later on it was discovered that a center tapped filament winding (or artificial center tap) would decrease filament hum in comparison to the "one wire" filament string. By the '60s most of the quality manufacturers had made the switch.

Please provide a link where I have said that grounding pin 9 is a better configuration. I don't recall having ever said that when talking about 6.3v filaments.

[TheKT88KilledJFK]

I know how to properly get 5VDC from a 6.3VAC filament winding with center tap. My question and comment was dealing specifically with your comment about grounding pin 9 of the preamp tubes. That went out of style with the '50s tweed amps.

Why's that?  You've posted before that it's a better configuration.

In the '50s it was common to ground one side of the filament winding and one side of the tube filament to chassis. This was to simplify filament wiring. You only needed one wire to string from one socket to the next. Later on it was discovered that a center tapped filament winding (or artificial center tap) would decrease filament hum in comparison to the "one wire" filament string. By the '60s most of the quality manufacturers had made the switch.

Please provide a link where I have said that grounding pin 9 is a better configuration. I don't recall having ever said that when talking about 6.3v filaments.

You haven't said anything about 6.3 but I've been using that configuration with success.  Sometimes people use reduced heater voltages for noise suppression and to affect the linearity of the tube's response although some think it hurts the tube.   

[sluckey]

Please share a schematic of some of your examples. Maybe we can all learn some new tricks.