Elevated Heaters From A PT With No CT, WTF ?

[Ben Waylin]

Can anyone draw this out for me ?

Thanks in advance.

 

[SpiderWars]

It depends on the amp so you’ll need to state that. I think you connect the two 100ohm filament artificial center tap resistors to a newly created power supply node instead of ground. Or if it is a cathode biased amp you could connect to the power tube cathodes.

 

[Ben Waylin]

It depends on the amp so you’ll need to state that. I think you connect the two 100ohm filament artificial center tap resistors to a newly created power supply node

I thought it was to connect them to an existing node off the existing power supply, prior to one of the filter caps connection to ground .. ?

What you said makes perfect sense to me.. that's kind of how I worked it out in my head.

But I didn't think you constructed a new node.. just used an existing node to replace the artificial center tap ground connection.

 

[glpg80]

You have to be very careful what you’re doing here not to short out your power supply through an unintentional ground loop.

At this level I need a schematic of whatever secondary you’re working with and how the filtering is implemented to make sure you don’t smoke the resistors or burn up your PT. What you don’t want to do is complete circuit from each side of the secondaries of each zero crossing through one diode which halve wave rectifies the signal before crossing the zero axis and doing it again on the other side.

Heaters can get away with no center tap because the circuits aren’t referencing ground elsewhere and are essentially floating at AC voltage just pulling a ton of amps. The fake center tap provides an adjustment for ground referenced to the center crossing of the ac wave for the whole load at the same time - but remember everywhere else is floating at AC voltage and nothing references ground.

With no center tap on the secondary but loads that do reference ground, you have to be careful otherwise you’ll create an unintentional completed circuit through the ground and filter caps back through the diodes and burn up your resistors or your PT or blow a fuse - whichever is weakest.

Now if your goal is to float a heater that doesn’t have a center tap but the secondary side of the B+ is full wave rectified and does have a center tap referenced to ground, then yes you’d just float the fake ground center tap made by a resistor divider with the correct ratio B+ resistor divider referenced to ground that gets your desired DC offset amount.

 

[Ben Waylin]

You have to be very careful what you’re doing here not to short out your power supply through an unintentional ground loop.

At this level I need a schematic of whatever secondary you’re working with and how the filtering is implemented to make sure you don’t smoke the resistors or burn up your PT. What you don’t want to do is complete circuit from each side of the secondaries of each zero crossing through one diode which halve wave rectifies the signal before crossing the zero axis and doing it again on the other side.

Heaters can get away with no center tap because the circuits aren’t referencing ground elsewhere and are essentially floating at AC voltage just pulling a ton of amps. The fake center tap provides an adjustment for ground referenced to the center crossing of the ac wave for the whole load at the same time - but remember everywhere else is floating at AC voltage and nothing references ground.

With no center tap but loads that do reference ground, you have to be careful otherwise you’ll create an unintentional completed circuit through the ground and filter caps back through the diodes and burn up your resistors or your PT or blow a fuse - whichever is weakest.

Here's the PT I'm using:

 

[glpg80]

So the secondary of the main B+ AC secondaries and the heater taps both aren’t center tapped, is the circuit off of the B+ secondaries full wave rectified?

 

[Ben Waylin]

So the secondary of the main line and the heater both aren’t center tapped, is the circuit off of the B+ secondaries full wave rectified?

Yes, I'm going to use a 3A, 1,000V bridge rectifier.

 

[glpg80]

Exactly.

Yes, I'm going to use a 3A, 1,000V bridge rectifier.

Three 30µF 450V filter caps.

A center tapped full wave rectifier obviously requires a center tap which you don’t have. You can use a bridge rectifier but you cannot connect the - side of the bridge rectifier to the center tap of the secondary (where the node would then be strapped to ground) since you don’t have one. Therefore it should be attached to electrical ground. The + side you’d send to your caps as normal. The top and bottom get the normal ~ feeds.

Rectify and filter what you need. The resistor divider to pull off a reference DC voltage needs to go to the same ground potential as the caps and the - side of the bridge rectifier. That resistor divider then can be used to float the center of your fake ground heater tap to elevate your heaters.

I’d recommend using precision accurate resistors to elevate the heaters as you want both sides of the sine wave elevated the same amount and you don’t want thermal runaway since you’re not regulating this at all in any manner. You shouldn’t be pulling any current at all but if a preamp tube shorts to ground then you just shorted your B+ through your heater windings and those resistors will take a severe beating before a fuse pops. So don’t use wimpy resistors for this so that a fire doesn’t start. If it were me I’d probably try to find precision wirewound resistors or at least flame retardant ones.

I believe this will work but you’d need to research it further as normally I like at least one side of the secondaries of the PT to have a center tap where I can reference it to ground. This guarantees each Half of a sine wave is crossing a true zero crossing right off the secondary.

 

[Ben Waylin]

A center tapped full wave rectifier obviously requires a center tap which you don’t have. You can use a bridge rectifier but you cannot connect the - side of the bridge rectifier to the center tap of the secondary (where the node would then be strapped to ground) since you don’t have one. Therefore it should be attached to electrical ground. The + side you’d send to your caps as normal. The top and bottom get the normal ~ feeds.

Rectify and filter what you need. The resistor divider to pull off a reference DC voltage needs to go to the same ground potential as the caps and the - side of the bridge rectifier. That resistor divider then can be used to float the center of your fake ground heater tap to elevate your heaters.

I’d recommend using precision accurate resistors to elevate the heaters as you want both sides of the sine wave elevated the same amount and you don’t want thermal runaway since you’re not regulating this at all in any manner. You shouldn’t be pulling any current at all but if a preamp tube shorts to ground then you just shorted your B+ through your heater windings and those resistors will take a severe beating before a fuse pops. So don’t use wimpy resistors for this so that a fire doesn’t start. If it were me I’d probably try to find precision wirewound resistors or at least flame retardant ones.

I believe this will work but you’d need to research it further as normally I like at least one side of the secondaries of the PT to have a center tap where I can reference it to ground. This guarantees each Half of a sine wave is crossing a true zero crossing right off the secondary.

Sorry for the crude drawings, but I'm trying to work this out..

Would this be correct ?

 

[glpg80]

Close - you need another resistor divider referenced to ground to drop the 470VDC down to like 80VDC. The center of that divider would then go to the 100 ohms as you have it drawn.

The reason you can just use a resistor divider to drop the voltage is due to the fact no current is being drawn at the fake center of the heater windings.

However under a preamp tube failure this is no longer potentially true therefore include a fuse both on the primary side of the PT and the secondary side. Size your fuses correctly. Use flame proof resistors for this including the fake center tap.

When floating voltages across taps like this especially with no center taps be sure to fuse things correctly. You don’t want a fire if a tube shorts.

 

[Ben Waylin]

Close - you need another resistor divider referenced to ground to drop the 470VDC down to like 80VDC. The center of that divider would then go to the 100 ohms as you have it drawn.

The reason you can just use a resistor divider to drop the voltage is due to the fact no current is being drawn at the fake center of the heater windings. However under a preamp tube failure this is no longer potentially true therefore include a fuse both on the primary side of the PT and the secondary side. Size your fuses correctly. Use flame proof resistors for this including the fake center tap.

When floating voltages across taps like this especially with no center taps be sure to fuse things correctly. You don’t want a fire if a tube shorts.

Ok, but I'll have nowhere near 470VDC there because my PT only has a 200VAC tap and a 180VAC tap for the B+

I was going off this from Robinette's site..

edit Ok, so reading Merlin's site I think I understand now, you do have to create another node off the B+ for this circuit - not use an existing one.

 

[Ben Waylin]

Ok, so the whole circuit drawn out for elevated heaters for a PT with no CT for the 6.3V tap would be this then ?

Then, I just need to solve for the resistor in the "new node" to have between 30V~60V elevation ?

The cap in the new node could be 30µF 450V and that would be sufficient ?

 

[SpiderWars]

You need another resistor between point B and the new node. To form a voltage divider. I would probably just try 220k and then 100k to ground. In my Marshall I used 330k/82k but I had way more voltage.

Also, just my preference but I would tap off of the screen node which looks like point A in your schematic.

 

[glpg80]

I deleted my message, I’ll re write it to make it easier to understand. Apologies on my end.

You will have to make a voltage divider off of the B+ referenced to ground. The center of the voltage divider then feeds your fake ground center tap on the heater supply.

I’ll run the math later in the morning and give you values then explain it a bit easier. But basically it’s this formula:

(Desired Voltage) = (B+) * r2/(r1 + r2)

 

[Ben Waylin]

You need another resistor between point B and the new node. To form a voltage divider. I would probably just try 220k and then 100k to ground. In my Marshall I used 330k/82k but I had way more voltage.

Also, just my preference but I would tap off of the screen node which looks like point A in your schematic.

Ok, I wondered about that..

So I still need the cap and resistor in parallel in the "new node" ?

Or are you saying that resistor gets moved to the + rail of the filter circuit ?

 

[glpg80]

The “No load” node in my image gets connected to the screen node “A” in your image. The bottom of r2 gets connected to ground.

It’s just a voltage divider.

 

[Ben Waylin]

The “No load” node in my image gets connected to the screen node “A” in your image. The bottom of r2 gets connected to ground.

It’s just a voltage divider.

?

 

[glpg80]

?

No

The 100 ohm resistors make a fake center tap for the heater supply. The voltage divider is in addition to those resistors to get the correct voltage TO the center tap

 

[Ben Waylin]

No

The 100 ohm resistors make a fake center tap for the heater supply. The voltage divider is in addition to those resistors to get the correct voltage TO the center tap

So then, the schematic in post #15 is correct ?

 

[glpg80]

No that isn’t right either. One sec I’ll draw something up