There are so much quality learning to be found on Youtube. For example 'Uncle Doug' and 'Blueglow electronics' has many excellent tutorials, that explains the inner workings of valve amps well for beginners. I have learnt a lot from from these guys.
If you combine these brilliant videos with a little written theory, and study schematics for old amps, you will get a much better understanding - and not just "paint by numbers"
If you try that 807 amp in Post # 30 for example, do not do exactly like the schematic says, it has one error.
It uses a 3rd output transformer winding for Ultra Linear operation.
Unfortunately, it is not center tapped, and there is no voltage to that winding to supply voltage for the screens.
Schematics . . . complete and accurate, versus all the other schematics.
Do not bother with Un-obtain-ium, Obsole-tium, and Highpriced-ium old output output transformers.
Just purchase one of todays Ultra Linear transformers with % taps on the primary.
And, you can wire it as UL, or you can live with lower power and Triode Wire those 807, which will allow you to get rid of the global negative feedback.
Then you can easily adjust which distortion is dominant, the high frequency roll off, and the damping factor (like perhaps a 2 Ohm resistor in series from the amplifier's output tap to resistor, and from there to the speaker.
Or build a simpler single ended amplifier, with Triode Wired Pentodes/Beam Power tubes, or Real Triodes such as the 6CK4.
High frequency rolloff, and damping factor adjustments are real easy.
Have Fun!
It uses a 3rd output transformer winding for Ultra Linear operation.
Unfortunately, it is not center tapped, and there is no voltage to that winding to supply voltage for the screens.
Schematics . . . complete and accurate, versus all the other schematics.
Do not bother with Un-obtain-ium, Obsole-tium, and Highpriced-ium old output output transformers.
Just purchase one of todays Ultra Linear transformers with % taps on the primary.
And, you can wire it as UL, or you can live with lower power and Triode Wire those 807, which will allow you to get rid of the global negative feedback.
Then you can easily adjust which distortion is dominant, the high frequency roll off, and the damping factor (like perhaps a 2 Ohm resistor in series from the amplifier's output tap to resistor, and from there to the speaker.
Or build a simpler single ended amplifier, with Triode Wired Pentodes/Beam Power tubes, or Real Triodes such as the 6CK4.
High frequency rolloff, and damping factor adjustments are real easy.
Have Fun!
Last edited:
I can't answer that, but the Quicksilver audio looks like a relatively good, but simple, push-pull design. It has grid bias and ultra linear output transformer and some feedback. So some good design features for low distortion, but perhaps not the most refined design of this kind (Have a look at the famous Williamson amp from 1949 for comparison)
I am guessing a bit, but I think it would have a moderate (pleasing?) level of distortion - but not "gross level" like a Single ended triode amp (typically) has.
The 8417 is a Beam Power tube.
The KT77 is a Beam Power tube.
. . .
The EL34 is a Pentode tube.
The EL34 has the same specs as a KT77, but do they have the same sound?
Is there anybody that remembers the Quicksilver triode wired beam power tube push pull amplifier?
I think it had No global negative feedback.
Many schematics have errors. The connection from the lower output tube plate to the 8 Ohm tap is the error.
Some schematics have extra wires; some schematics have missing wires.
As to the sound of the amplifier: View attachment 1280056
Perhaps it is the straightforward signal path and the simplicity of the whole circuit.
"You should make things as simple as possible, but no simpler" - Albert Einstein
“Those who do not know History, are bound to make things worse than ever” - Me
The KT77 is a Beam Power tube.
. . .
The EL34 is a Pentode tube.
The EL34 has the same specs as a KT77, but do they have the same sound?
Is there anybody that remembers the Quicksilver triode wired beam power tube push pull amplifier?
I think it had No global negative feedback.
Many schematics have errors. The connection from the lower output tube plate to the 8 Ohm tap is the error.
Some schematics have extra wires; some schematics have missing wires.
As to the sound of the amplifier: View attachment 1280056
Perhaps it is the straightforward signal path and the simplicity of the whole circuit.
"You should make things as simple as possible, but no simpler" - Albert Einstein
“Those who do not know History, are bound to make things worse than ever” - Me
Last edited:
It’s straight from Quicksilver manual, so, I’m not sure. BTW, thanks for posting as I’m a huge fan of your YouTube channel! Your approach is as good as any I’ve seen…. Keep up the good work!!
Interestingly, just about every 1930's radio with a field coil loudspeaker has a couple of exposed metal sockets for an external speaker with B+ on one of the plugs. I always disconnect them as the first thing I do. When I was a lad I was reaching around the back of a set to reconnect the mains input plug and had a nasty shock.
Owners Manuals and Tech Manuals do have errors. Many are never corrected/updated.
The schematic in question connects the bottom output tube's plate to the output transformer primary, . . . and to the 8 Ohm tap which feeds the parallel RC feedback network.
. . . Nobody is going to be shocked.
The secondary common is returned to ground.
So the B+ circuit will pump lots of current to the center tap, then through the 1/2 DCR of the primary, and then through the large wire gauge of the secondary; notice the secondary common terminal is grounded.
That secondary wire is not going to open, the primary wire is a much much finer gauge, and so is the secondary wire gauge of the B+ transformer.
I am sure that any technician will be able to fix the amplifier if it is broken for any reason.
But a diy newbie could perhaps build an amplifier by folloing the schematic. Ouch!
The schematic in question connects the bottom output tube's plate to the output transformer primary, . . . and to the 8 Ohm tap which feeds the parallel RC feedback network.
. . . Nobody is going to be shocked.
The secondary common is returned to ground.
So the B+ circuit will pump lots of current to the center tap, then through the 1/2 DCR of the primary, and then through the large wire gauge of the secondary; notice the secondary common terminal is grounded.
That secondary wire is not going to open, the primary wire is a much much finer gauge, and so is the secondary wire gauge of the B+ transformer.
I am sure that any technician will be able to fix the amplifier if it is broken for any reason.
But a diy newbie could perhaps build an amplifier by folloing the schematic. Ouch!
Just an FYI, I posted the quicksilver schematic. Unfortunately, I’m not skilled to see potential problems. Perhaps, I shouldn’t post schematics; however, the point of this thread was to gain knowledge and by some of the back and fourth, I thought it was appropriate to post the schematic, of an amp that I enjoyed.
caryking,
1. I am glad you posted the Quicksilver schematic. It is a good example tube circuit. (***) see below
The schematic error of the extra wire can Not cause a shock.
Shocks do Not come from Shorted B+.
Instead, shocks come when you Touch working B+ (Working as in when B+ is Not shorted)
2. (***) That Quicksilver amplifier is simple. But also is a great discussion tool:
A.
It has the two most common types of compensation for global negative feedback (RC global negative feedback compensation, and the RC dominant pole compensation);
Ultra Linear mode (Which "sits" between Beam Power mode operation (0%), and Triode Wired Beam Power mode operation (100%).
Has one of the most often used output tube bias modes (Fixed Bias)
Uses one of the most popular phase splitters, the split load Concertina.
Uses the most popular biasing of the Concertina, by DC coupling the input stage plate to the Concertina grid.
The input tube cathode's two resistors in series, with global negative feedback to the junction is very often used.
B.
Simple. Effective. I bet it sounds good.
Remove the one wire that is in error, and you have a working amplifier.
Now, to give a comparison to the Quicksilver circuit . . .
Go and find an amplifier circuit that does Not have any of the circuit topology items that are between the letter A. and the letter B above.
Hint: there are hundreds that meet that criteria.
Probability results in a Mean thing about Karl Friedrich Gauss, an Average man at the Peak of his career, who was at the Center of the Gaussian Curve.
1. I am glad you posted the Quicksilver schematic. It is a good example tube circuit. (***) see below
The schematic error of the extra wire can Not cause a shock.
Shocks do Not come from Shorted B+.
Instead, shocks come when you Touch working B+ (Working as in when B+ is Not shorted)
2. (***) That Quicksilver amplifier is simple. But also is a great discussion tool:
A.
It has the two most common types of compensation for global negative feedback (RC global negative feedback compensation, and the RC dominant pole compensation);
Ultra Linear mode (Which "sits" between Beam Power mode operation (0%), and Triode Wired Beam Power mode operation (100%).
Has one of the most often used output tube bias modes (Fixed Bias)
Uses one of the most popular phase splitters, the split load Concertina.
Uses the most popular biasing of the Concertina, by DC coupling the input stage plate to the Concertina grid.
The input tube cathode's two resistors in series, with global negative feedback to the junction is very often used.
B.
Simple. Effective. I bet it sounds good.
Remove the one wire that is in error, and you have a working amplifier.
Now, to give a comparison to the Quicksilver circuit . . .
Go and find an amplifier circuit that does Not have any of the circuit topology items that are between the letter A. and the letter B above.
Hint: there are hundreds that meet that criteria.
Probability results in a Mean thing about Karl Friedrich Gauss, an Average man at the Peak of his career, who was at the Center of the Gaussian Curve.
Last edited:
I have heard of manufacturers putting out schematics with intentional errors but that one goes a bit far.
caryking,
grovergardner is right; but do not loose hope . . .
Remove the global negative feedback, and be sure to remove both compensation networks:
The parallel RC from the 8 Ohm tap . . . And the serial RC network (Dominant pole to ground).
Then you can disconnect the screens from the UL taps, and connect the screens through a 100 Ohm resistors to the Plates.
That is Triode Wired mode.
Hope Springs Eternal: . . . Triode Wired mode will . . .
Get back a decent damping factor, and decent output impedance
Get back a decent bandwidth
Get back a decent bass performance
Reduce some of the excess gain that came when you removed the global negative feedback.
The tradeoff: Lower power output . . . but try it, you will like it.
Easy to try, and easy to put back to original.
Have Fun!
Anybody that has never seen a Manufacturer's schematic that has an un-intentional error, has not looked at as many schematics as I have.
grovergardner is right; but do not loose hope . . .
Remove the global negative feedback, and be sure to remove both compensation networks:
The parallel RC from the 8 Ohm tap . . . And the serial RC network (Dominant pole to ground).
Then you can disconnect the screens from the UL taps, and connect the screens through a 100 Ohm resistors to the Plates.
That is Triode Wired mode.
Hope Springs Eternal: . . . Triode Wired mode will . . .
Get back a decent damping factor, and decent output impedance
Get back a decent bandwidth
Get back a decent bass performance
Reduce some of the excess gain that came when you removed the global negative feedback.
The tradeoff: Lower power output . . . but try it, you will like it.
Easy to try, and easy to put back to original.
Have Fun!
Anybody that has never seen a Manufacturer's schematic that has an un-intentional error, has not looked at as many schematics as I have.
Except that Self rejects the notion of thermal distortion. I think he's likely mistaken, not least because the skepticism seems largely based on incredulity that thermal time constants in the order of 0.1s to seconds could somehow 'break' amplifier performance. But also a failure to follow up. (He devoted a lengthy segment only to conclude that a single-ended VAS is still likely better than an LTP/Hitachi VAS, so why not do the homework on thermal effects?)
Here's a testable hypothesis:
Take 'blameless' amplifier 'A'.
Despite THD being <0.001%, there's a purported critical flaw, whereby thermal drift causes Vbe variations in the millivolt range, multiplied by open-loop gain, leading to volt-scale output variations, which are then "fixed" by global NFB.
However, slow DC offset drift (not really DC, just sub-sonic frequencies) then offsets the highly-tuned LTP, preventing proper cancellation of even harmonics. Apply a high dynamic peak to the input, a la Peufeu et al, and see what happens to the balance of the long-tail pair in the following seconds.
Self really seems to like LTPs — they do seem nice in that they cancel a lot of distortion, and can be tweaked and tuned. But the crux seems to be that the better it is, the worse it is!:
Ironically, if the input LTP is relatively crap, and only cancels H-even by 6dB, then any 'phasing' effects from drift are likely to be small. Unless the LTP momentarily traverses a much better operating point, leading to, say 20dB of "bounce" in H2 levels.
But if the LTP is highly tuned with super-strong cancellation, leaving only a string of H3, 5, 7, 9, etc, harmonics, and unmeasurable H2, 4, 6, 8...
then the potential H2 "bounce" from temporary detuning likely to be more severe. So we get a Catch 22: we can only allow such artifacts if we brute-force the design with ultra-low distortion. But even 0.001% THD is questionable, if it turns out that we're not really dealing with THD but something else like IMD or amplitude modulation (or like I suggested, tonality modulation).
For the 'test', an LTP could be deliberately modulated with, for instance, 1-10Hz subsonic sine waves on a DC servo to find out if the above concerns are actually based in reality.
In my own cursory research, cubic exponentiation of distortion levels could be a bigger bane than the absolute level. Quite often, a test signal is raised in 6dB steps (say, from 100mV input to 200mV), and there is a "cliff" where some kind of level-dependent distortion causes a whopping 18dB jump. Therefore: gritty, grainy sound with unmasked artifacts.
I think damping factor is a likely big one. The damping factor wars seem to be alive and well, with competing factions insisting that theirs is the one and only "correct" way to do things.
A couple things to note:
Full range speakers are much more sensitive to IMD and other distortion issues related to passing full-spectrum audio through a single driver, and current drive is known to reduce a lot of those distortions. With voltage drive, the trend is to avoid the issues (usually by kicking the can by increasing the number of speaker ways) rather than figuring out how the problem works in the first place.
The cynic in me thinks the "correctness" of voltage drive is also related to market forces and a desire to sell woofers + mids + mid-tweeters + super-tweeters x active amplification for each damn channel. Face it: if voltage drive forces you to buy a minimum of 3 ways because FR sounds like crap (without complicated doodling with power resistors or DIY amplifier design), then that's what they'll sell.