Elevated heaters - Where to take DC ?

[kagliostro]

Where is the best place to take DC for the Helevated Heater PS ?

Is better to take the voltage for the divider at the end of the filter chain or at the begin ?

Of course taking DC after the last filter will give a smoothed DC but will this be a way to compromise in some way the noise at preamp level ?

K

[tubeswell]

I usually put a voltage divider for the DC reference between HT and ground at the reservoir cap - it doubles as the bleeder path


(Using a humdinger - instead of the heater winding CT - for the reference would be superior in terms of bucking heater noise, but its more parts to fiddle with. YMMV)

[sluckey]

+1

[archaos]

+1 here.

[jjasilli]

Question:  in a cathode biased power amp, why not take the dc reference from the top of the cathode resistor?  Is it preferable to build a separate circuit?

[kagliostro]

Thanks Tubeswell - Steve - Archaos

@ Jjasilli

actually my friend, in his pencil schematic, has draw the elevated heater connected to the cathodes of the final tubes, it was my fault to draw

the schematic with a partition resistor, after I recognize my fault I've think that a partition resistor will give a higher DC reference voltage that

I think can be a better choice also if not sure 100% about.

May be a higher DC voltage elevation will be useful in some projects as to have a lower voltage difference between cathode and heaters

In accordance with your council about the DC voltage source this would be my layout



Grazie

Franco

p.s.: do you think an electrolytic rated for 100v will be enough as to smooth the elevated DC ?

[HotBluePlates]

p.s.: do you think an electrolytic rated for 100v will be enough as to smooth the elevated DC ?

The elevated voltage is already smoothed by the power supply. And after it is divided-down, any ripple present is reduced by the voltage divider, along with the desired d.c. reference. The 47uF 100v cap is working against 195kΩ to reduce ripple to the reference further. For 120Hz ripple, the 47uF cap looks like ~28Ω; 5v of ripple at the 1st power supply filter cap will be reduced by the 195kΩ resistance and the 47uF cap to 627mV (~8:1 reduction) with less starting ripple being even smaller.

Further, because this is applied to the heater center-tap, both ends bounce up and down with any ripple left.

"But wouldn't the cap possibly see the B+ voltage at turn-on? Shouldn't it be rated for more than 100v?"

For the Tweed Bluesmeister, the steady state is B+ of 414v and a reference voltage of 41-42v. The divider passes ~1.9-2mA of current, and drops ~372v across the 195kΩ resistance. The worst-case happens at switch-on, when the B+ rises to 325v*1.414 = 460v (due to no current draw) and the 47uF reference divider cap looks like a short-circuit due to no charge/no voltage. With a full 460v across the 195kΩ resistor, it will pass ~2.36mA and the 47uF voltage will rise from zero up to the desired 41-42v. So the 100v rating of the cap is ample, and the resistance will dissipate about 1w (hence the 2w total rating of a pair of 390kΩ 1w resistors in parallel).

[sluckey]

K, this has nothing to do with your topic. But I just noticed the 1µF/50V cap at the left end of your board. I believe the positive end should be pointing down.

[kagliostro]

Grazie for the explanation HotBluePlates

"But wouldn't the cap possibly see the B+ voltage at turn-on? Shouldn't it be rated for more than 100v?"

That is exactly what I've think

---

Grazie Steve

I turn it the wrong way, I go to correct it

Franco

[kagliostro]

The partition resistor used as source for the helevated heaters reference voltage can be considered as a substitute of the bleed resistor ?



Thanks

Franco

[John]

Yes, your 2 voltage divider resistors will make a 206k bleeder for the B+.  For what it's worth, I would have them a little further down your power rail. At least past your first filter cap. Not sure if it really matters though. Sluckey can tell you. 


And Merry Christmas, since I'm just now getting in the spirit. 

[kagliostro]

Thanks for the answer and the wishes


---


The partition resistor is in that position because .... it was asked where to place it 


Franco

[sluckey]

The boys at ax84.com are fond of that circuit. I've used that same circuit in two amps. (See November and Stuffed Vox)

[tubeswell]

Watch the B+.  Heater elevation works best somewhere between about 20V and about 60V (remembering h-k volts). (27% of 400 and something is over 100 and something).

[kagliostro]

The PT is 230v-0-230v AC, so I think I'll have ~300v DC

using 150k + 56k on the voltage divider the voltage will be near 80v, may be effectively a bit high

with 180k + 47k it will be near 60v

Thanks

Franco