Yep. Tradeoffs, tradeoffs. Multivariate optimization. Maybe there's a global optimum. Maybe not. Maybe there are several optima. Maybe there are none. That's engineering in a nutshell.
Tom
Tom
Good question!
Attachments
I am a software guy by trade (please have mercy), and I am dealing with operational reasearch and optimization systems (via IP, MIP or ML approaches).
I suspect there could be an interesting optimization case here. And since it would be at design time, it wouldn’t matter if a model needed to run for hours.
Capacitor number 1:
Suppose the B+ raises to 300V before the tubes warm up.
In that case, you want the 0.47uF 250V coupling cap to be rated for more than 250V.
Use a quality 400V cap there.
I would not consider a paper cap to be quality for this part of the circuit (RC coupling cap).
If the unloaded B+ is 400V or more before the tubes warm up, use a 600V or 630V rated cap.
Cap voltages need to be rated for all turn-on, quiescent, warmed up, large signal swings, and power down voltage conditions.
Capacitors number 2:
It is not just the voltage rating that is important. Electrolytic caps can / sometimes do have leakage currents.
The 3.9k resistor to ground may allow there to be some leakage current going to the headphones, if there is any, it will displace the headphon voice coil, de-centering it from the 'zero' position.
Be sure to use a very low leakage current electrolytic.
A 100V or 400V electrolytic might have less leakage current than a 60V cap.
In any case, the quality here is paramount.
Similar requirements for the parallel capacitor (plastic, teflon, etc.).
Caution:
If the output tube warms up faster than the input tube, you may have near the B+ voltage on the output tube grid, and the output tube's cathode will follow that. That high voltage can potentially destroy the two output caps (well, that just means they have to have higher voltage ratings than the B+ voltage.
And I would not want all that warm-up charge current of those output caps to go through the headphones. Let the amplifier stabilize the output caps by charging through the 3.9k output resistor first, and then connect the headphones.
And when you turn the amplifier off, the output caps will discharge. The discharge current is through the 330 Ohm cathode resistor, ground, the ground end of the headphones in parallel with the 3.9k, and back to the other side of the headphones and 3.9k Resistor.
Well, we 'eliminated' the need for an output transformer, but you can see the additional potential problems we got instead.
Most generally, output transformers do not have any DC current or leakage current out that a capacitor has.
Design is an exercise in tradeoffs.
All the above are just my opinions.
Your Mileage May Vary.
Suppose the B+ raises to 300V before the tubes warm up.
In that case, you want the 0.47uF 250V coupling cap to be rated for more than 250V.
Use a quality 400V cap there.
I would not consider a paper cap to be quality for this part of the circuit (RC coupling cap).
If the unloaded B+ is 400V or more before the tubes warm up, use a 600V or 630V rated cap.
Cap voltages need to be rated for all turn-on, quiescent, warmed up, large signal swings, and power down voltage conditions.
Capacitors number 2:
It is not just the voltage rating that is important. Electrolytic caps can / sometimes do have leakage currents.
The 3.9k resistor to ground may allow there to be some leakage current going to the headphones, if there is any, it will displace the headphon voice coil, de-centering it from the 'zero' position.
Be sure to use a very low leakage current electrolytic.
A 100V or 400V electrolytic might have less leakage current than a 60V cap.
In any case, the quality here is paramount.
Similar requirements for the parallel capacitor (plastic, teflon, etc.).
Caution:
If the output tube warms up faster than the input tube, you may have near the B+ voltage on the output tube grid, and the output tube's cathode will follow that. That high voltage can potentially destroy the two output caps (well, that just means they have to have higher voltage ratings than the B+ voltage.
And I would not want all that warm-up charge current of those output caps to go through the headphones. Let the amplifier stabilize the output caps by charging through the 3.9k output resistor first, and then connect the headphones.
And when you turn the amplifier off, the output caps will discharge. The discharge current is through the 330 Ohm cathode resistor, ground, the ground end of the headphones in parallel with the 3.9k, and back to the other side of the headphones and 3.9k Resistor.
Well, we 'eliminated' the need for an output transformer, but you can see the additional potential problems we got instead.
Most generally, output transformers do not have any DC current or leakage current out that a capacitor has.
Design is an exercise in tradeoffs.
All the above are just my opinions.
Your Mileage May Vary.
Last edited:
Lambo was really a tractor company. The boss used tractor profits (and Fiat parts-bin) to fund & build a sport-car addiction. (Since 1973 the sportycar side spun-out through Chrysler and etc to VW. Lamborghini Trattori still builds dirt-machines.)
Porsche made a lot of farm tractors.
Porsche also designed an engine for another company for a light commercial truck which did not go to production, which I think they recycled in a front-engine sporty car.
A US(?) deluxe tractor company decided to use the 1930s Chrysler Straight Eight. It would not only pull a plow, and the motor be fixed at any Plymouth dealer, it would run fairly fast on the road to take the family to church in the optional full cab (what we call Crew Cab today).
Ferguson and Standard both needed engines but neither could do it alone.
"...a 2,088 cc inline four cylinder petrol engine ... Originally developed concurrently for passenger car use and for the Ferguson TE20 tractor, it was widely used for Standard passenger cars of the 1950s, most notably the Vanguard. Later it was successfully used in Standard's popular early generation Triumph TR series sports cars."
23.9 bhp (Ferguson TE-A20 tractor)
68 bhp (Standard Vanguard)
https://en.wikipedia.org/wiki/Standard_wet_liner_inline-four_engine
Triodes are very compliant. Surround them with plate and cathode resistors, not too bad-valued, they will usually find a working condition.
I was wondering, would it be possible to simulate this circuit in something like LTSpice?
I remember trying that tool before and it was quite a learning curve, plus I’m not sure if there good models for the 6C19.
I remember trying that tool before and it was quite a learning curve, plus I’m not sure if there good models for the 6C19.
Yes you can. I use it very often. DC working points are pretty accurate. AC sometimes is fine, but sometimes not so much. THD is always on the optimistic side.
LTspice is quite good at telling you what not to do.
Thanks!
That’s what I hoped, a virtual bench that would allow me to tinker while I learn.
Do you know of any good tutorials?
Also, it seems to get repackaged a lot or built with various tweaks on different OSes, is there a version of LTSpice that is tried and true?
SPICE is the original Berkeley public domain program from the 70s.
Others have adapted it with their own variations, including Linear Technology (LT), now part of Analog Devices Inc.
They're all somewhat different from one another, so try some of them out.
https://learn.sparkfun.com/tutorials/getting-started-with-ltspice/all
http://www.simonbramble.co.uk/lt_spice/ltspice_lt_spice.htm
Others have adapted it with their own variations, including Linear Technology (LT), now part of Analog Devices Inc.
They're all somewhat different from one another, so try some of them out.
https://learn.sparkfun.com/tutorials/getting-started-with-ltspice/all
http://www.simonbramble.co.uk/lt_spice/ltspice_lt_spice.htm
Last edited:
Other books to read:
'Designing Power Supplies for Tube Amplifiers' - Merlin Blencowe
'Designing High-Fidelity Tube Preamps' - Merlin Blencowe
Both, IMHO, easier to follow than Jones and as, or more, informative.
Well, I learned enough to make this:
Now, I need to figure out the magic LTSpice directives to make it tell me things...
Because +B1 and +B2 are kinda important in the above schematic, I tried to reverse engineer the power supply of that circuit by looking at its PCB...
Here is what I am seeing (I suck at reading 2-layer boards!):
Compared to similar things I have seen online, the contraption after the choke looks a little different. (It's probably an obvious pattern, but I am so new at this, it's all lost on me.)
On the left side, there should in fact be the 120-0-120 secondary of a power transformer, but I didn't figure out how to do that.
Does that look rational?
Here is what I am seeing (I suck at reading 2-layer boards!):
Compared to similar things I have seen online, the contraption after the choke looks a little different. (It's probably an obvious pattern, but I am so new at this, it's all lost on me.)
On the left side, there should in fact be the 120-0-120 secondary of a power transformer, but I didn't figure out how to do that.
Does that look rational?