The Phantom SR45 Build - Snow Creek Shuttle Episode 1-A Bottlehead Saga

[ssssly]

Why...thank you. I tried to make it aesthetically pleasing as well as functional. Unfortunately the Parafeed cap didn't quite clear. It's stuffed tight into the corner between the plate choke and the SR PCB.

Moved the bracket over a couple mm and printed V5 last night. In the ultrasonic now.

And if you think that's cool wait until I get to the Paramour rebuild portion of the shuffle. I'm going to do fully 3D printed and electroformed in place wiring.

[ssssly]

I mocked up the full wave rectifier for the 45 filament and...I don't really like it. Even with the low Vf of the 1N5820s it just seemed to struggle to provide a stable 2.5v. Set to 2.5v it always wanted to sag and hover around 2.1-2.2. Set to 2.7v around 2.1-2.3v and set to 2.8v it would swing from 2.2-2.8v.

After trying to troubleshoot it for a day I was back to thinking just drop a simple LM317 based constant current heater circuit in there.

Then I remembered seeing some ADP linear voltage regulators that operate around the operating points of the 45 filament. And I came across this little bugger.

https://www.digikey.com/en/products/detail/analog-devices-inc/ADP3339AKC-2-5-RL7/11534905

up to 6v in, 2.5v @ 1.5a out, and only needs 1uf esr agnostic stabilizing caps.

Been staring at the specs for some time now wondering why this won't work a treat and not finding any.

So give it to me...why won't this regulator work?

[Paul Birkeland]

Heat through a regulator can be problematic.  Remember that you need to cover the dropout voltage plus some safety margin.

I don't know what "setting for 2.7" means.  You setup the full wave rectifier with a resistor between the rectifiers and the first cap, then adjust the resistor value to get yourself about 2.5V.  I would aim for 2.4V-2.6V.

[ssssly]

I calculated the values for the resistors using my pet AI. Since the resistors were the only variable I referred to them as "set" resistors.i bought a range of resistors from 0.2-0.5ohm. According to the calculations greater than 0-0.2 ohm should result in ~2.7v. I'm now getting ~1.5v loaded with a tube instead of a dummy load.

It says the dropout is only 0.23v at 1.5A for that regulator. Actually why it caught my attention for more than a second. That and the 85/105c ambient/junction temp. Which leads me to believe it could be fine temp wise.

But since I have never seen them used in a heater, I assumed there was something I was missing.

But they are kind of an obscure oop regulator. And the only ones that operate at or near those points. So figured it could just be one of those things where nobody has used one because nobody was looking for one. Since most people would probably just use a transformer designed for what they need.

[Paul Joppa]

Regulated filaments can be tricky to design. It's the one area where PSUD is not as accurate as  you might wish, due largely to the inaccurate diode models. You may need to use a Schottky diode rated 15 amps or more. Then you need to subtract the peak ripple voltage from the average DC voltage to keep the regulator from dropping out. I subtract another 10% to allow for power line fluctuations. And the DCR of the transformer (including the reflected primary DCR) adds to the series resistance. And the high voltage winding rectifier/filter adds to the primary voltage losses.

[ssssly]

Regulated filaments can be tricky to design. It's the one area where PSUD is not as accurate as  you might wish, due largely to the inaccurate diode models. You may need to use a Schottky diode rated 15 amps or more. Then you need to subtract the peak ripple voltage from the average DC voltage to keep the regulator from dropping out. I subtract another 10% to allow for power line fluctuations. And the DCR of the transformer (including the reflected primary DCR) adds to the series resistance. And the high voltage winding rectifier/filter adds to the primary voltage losses.

Certainly seems that way. I have by far the least experience designing anything power supply related. Just not something you come across in mechatronics much unless you are designing large scale industrial, which I don't. And even then it's pretty much just IDing COTS stuff that fits the bill.

I may just slap the PT-2s I have left in these SR-45s and experiment with the PT-7s more when I do my Paramour rebuilds.

I currently have some 6BX7 amps powering the horns in my upstairs system and it's making it near impossible to find a good crossover point. They just sound completely different than the Paramounts powering the woofers. And seem to exaggerated the beamy-ness of the 500b horns.

So I kind of just want to get these built to get the 6Bs out of my upstairs system.

[ssssly]

And there we have it. Amp 1 wired up. On to checks tomorrow.

[ssssly]

Well these certainly weren't the numbers I was expecting.

DRV P (225)
KREG (18)
REG P (355)
KDR (7.8 )
B+ (355) that's actually what I was expecting
CK (397)

A1,2 (67)
A3 (318)
A4 (26.2)

B4,5 (3.2)

What's really boggling my brain at the moment is the heater voltages.

[Paul Birkeland]

Which voltage is wrong?

[ssssly]

The 45 socket voltages were not what I was expecting.

I was expecting A1,2 sub 2vac
A3 ~ 300vac


And the 45 got so hot so quick it kind of alarmed me, given 60x the voltage I was expecting on A1,2.

But if those numbers are good and my expectations were wrong, happy to chalk it up in the win column and see what it sounds like.

[Doc B.]

I'm not following this. I am not looking at the schematic, but on a 45 (or 2A3 or 300B) the filament is A1 and A4. A2 is the grid. To measure AC fil voltage on these tubes you measure AC across from A1 to A4. Note that the entire filament can be operating on AC voltage and also be DC biased with respect to ground. So you could measure 2.5 VAC across the filament pins and also some higher DC voltage with respect to ground on either of those pins.

[ssssly]

I used your funky numbering convention from an old paper P'mour schematic. 

1-2 filament, 3 plate, 4 grid.

Data sheet for the 45, P'mour & mount manuals are 1-4 filament, 2 plate, 3 grid. BeePree is 2 grid, 3 plate.

Bonehead problem solved. I was measuring to ground. Filament is 2.7vac.

However, the data sheet lists max 275v to the Plate and min 31.5 for the grid and I have 318v and 26.2v respectively.

I

[Doc B.]

I'm not helping much, huh. Pin 2 is plate, pin 3 is grid.

[ssssly]

I'm not helping much, huh. Pin 2 is plate, pin 3 is grid.

Always nice to see to chime in though.

This is my first 45 Amp. Do 45 tubes normally run considerably hotter to the touch than 2A3s or 300bs? Like...lick your fingers and remind yourself never to touch that again hot?

[ssssly]

Well, it makes music. Much better sounding music than the 6b amp it replaced. Also surprisingly louder than that amp, which in theory should be in the 10watt range.

And far less hum. 4x less to be precise. And quieter than my Paramounts by a milivolt.

Is that 318v at the plate in the ballpark of what it should be or should I reduce it?

Sounds great. Just don't want it to be a tube eating machine if that's not the intended operating point.