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

[Paul Birkeland]

What value did you use for your cathode resistor? 

PJ originally drew up the SR-45 for use with the available 3K iron from Magnequest, but with what you have I would be aiming for a little less quiescent current.

[ssssly]

1.6k

[ssssly]

And if you wouldn't mind, what is the math for calculating the value?

While I have largely considered my audio hobby as electric Legos, I would like to begin actually wrapping my head around the EE side of these circuit topologies.

[Paul Birkeland]

That seems pretty reasonable to me.  To figure out what cathode resistor value you want, you would start by marking the desired operating point on the plate curves, then you could have the bias voltage (67V) you desire and the operating current (about 40mA).  V=I*R, so you have 67=0.040*R, so you end up with 1675 ohms as the ideal calculated resistor.  The 1.6K and the hum pot arrangement gets pretty close.

[Paul Joppa]

I'm away from my notes at present, so I haven't commented yet. Specifically, I don't remember the operating point I chose. But the voltages seem a bit too high. You could be over-heating the 45.

It remains unclear what the voltages are - that "26v at the grid" might be 26mV(?) - or else the grid resistor is not connected(?).

[Paul Birkeland]

He's running about 250V plate-filament voltage and 40mA of current, so more or less at the limit for #45 operation.  I would presume since the apparent bias voltage is correct that the "26" is indeed mV.  With the PT3/7/10 doubler winding, regulating the B+ a bit higher keeps the top C4S a bit cooler and plays better with 4-5K output iron.

A little extra cathode resistance might not be the worst idea to bump the plate dissipation down a little bit.

[ssssly]

It is indeed 26mV. Fat fingered it. But that is also a dead cold measurement. The tube was getting so hot I wasn't leaving it running long between measurements.

I ended up swapping the PT-2s into it.

The output iron is 4.9k.

The resistors in the PCB were changed out to get the B+ to ~360v per PBs previous recommendation. Dual 1.8ks if I remember correctly.

Also changed the Ufs for Cree diodes. So .7v vs .2v drop per.

The 45 tube gets really hot, really quick. Way hotter than any 2A3 or 300B I've ever touched. Have to wait ~5 min before I can even think about touching it. And I'll yank a 2A3 or 300B immediately after turning off an amp. I have a welder's left hand and it burned me the first time.

Happy to take any additional measurements. I built it with plenty of room to poke around.

Which really makes the question, what was the intended operating point PJ chose?

Looking at tube charts, for anything but an EML 45, I'd guess 270v plate to ground. But I amiditly have little idea what I'm talking about here.

[Paul Birkeland]

The modifications I specified for raising the B+ will not work with the PT-2.  This isn't the kind of circuit where you can change one think without a cascade of other changes being required, so I would put the other power transformers back if at all possible, and this is especially true if you left the voltage doubler in place and used that with the PT-2.


Yes, tubes get too hot to touch, that's normal. 

[ssssly]

Nope. No voltage doubler.

Back home looking at the schematic and the original operating points are 320v plate, 13.5 v bias for the first half of the 6CM7, 200v plate, 8v bias for the second. And 300v for the 45 plate.

Looking at the tube curves im guessing my issue with the PT-2 running the 6cm7 at those voltages is the risk of running out of current with the 45 running beyond max.

And I'm guessing the reason for running the 45 at 300+v is to extend the bass and high treble response at a slight hit to linearity and tube life while maxing out amplification.

I however don't need bass or watts out of these amps. They will live as dedicated horn compression drivers amps. They'll never see below 500hz. And the compression drivers they will live hooked up to break up in the 12-15k region.

If my previous guesses are correct, would it then make more sense to decrease the plate voltage of the 45s to the 260v range, where the curves seem more linear, to keep or increase the voltage on the 6cm7 which seems more linear approaching 450v?

On these amps in particular id much rather rob Peter of bass and watts to pay Paul in linearity and midrange. And id greatly prefer to stick with the PT-2 at this point. Swapping out that transformer will be a PITA I'd rather avoid.

[Paul Birkeland]

The plate voltage is low with the PT-2 to provide compliance for the shunt regulator and the extra voltage isn't really helpful when running a 3K output transformer.  If you're using a 4-5K transformer, then you could take advantage of a little more B+ and run the #45 a lot closer to the typical book operating point into the typical load impedance.

The plate voltage of your amps is the voltage appearing at the plate minus the average of the voltage appearing at pins 1 and 4.  You are currently running well below 300V.

I would highly recommend putting the PT-7 transformers and the doublers back in place and leaving things as they are, with the caveat that you could probably go to 1.8K as a cathode bias resistor on the output stage if you want to.

If you want to go back to the PT-2, then the lower regulated B+ and 3K iron makes a lot more sense.

Again there is nothing alarming about how hot the #45s are getting.  They are rated for 10W of plate dissipation and you were right at 10W of plate dissipation under the conditions you initially posted.

[Paul Joppa]

From the data sheet, 250v plate to cathode at 40mA would be -48v bias and thus a 1200 ohm cathode resistor 298v plate to ground.

320v plate to cathode at 40mA would be -67v bias and thus a 1675 ohm cathode resistor; 387v plate to ground. And 12.8 watts plate dissipation, which is too much.

[Paul Birkeland]

From the data sheet, 250v plate to cathode at 40mA would be -48v bias and thus a 1200 ohm cathode resistor 298v plate to ground.

320v plate to cathode at 40mA would be -67v bias and thus a 1675 ohm cathode resistor; 387v plate to ground. And 12.8 watts plate dissipation, which is too much.

A1/4 (67)
A2 (318)

This is 251V P-K and about 41mA of plate current, so 10.3W. 

With the PT3/7/10, nudging B+ up to more like 340V allows one to move a bit closer to the next operating point in the datasheet of 275V/36mA with 56V of bias, which calls for a bit less than 1.6K of cathode bias resistor.  To be honest I just borrowed the operating point from the #45 instructions in the Stereomour I manual   Perhaps there actually is a bit of positive grid voltage at the #45 causing it to draw a bit more current than it should be?

[ssssly]

Well, let's see.

I've got some 1.2k 12w resistors laying around. They are mills non inductive wirewound, not sand. But bypassed by a 100uf cap I wouldn't think the value should change much until the vhf range.

Gonna throw one in and get some better measurements with things warmed up more than when taking the first couple sets.

[Paul Joppa]

Don't reduce the cathode resistance below its current 1600 ohm value until we understand the voltages right.

251v plate to cathode a 40mA is indeed a good operating point for a 3000 ohm load, but it is NOT consistent with the plate curves of a 45 at -67v bias. Not even close.

[2wo]

If you have some 100-200 ohm resistors, 1/2w or preferably higher you can add one in series to the cathode resistor and recheck your operating point till you get what you want then replace with a single HP resistor.

The added resistor is just for short term testing...John