Audio Transformer reference

[acheld]

I've been nosing around for more detailed info regarding output transformers, found this:

https://www.jensen-transformers.com/wp-content/uploads/2014/09/Audio-Transformers-Chapter.pdf

Yes, it's a dry read . . .

[HotBluePlates]

I've been nosing around for more detailed info regarding output transformers ...

What are you wanting to know about output transformers?
(We tend to ascribe "magic" to them, but fail to see simple facts about how they work)

[dwinstonwood]

Another good resource:
http://www.radioremembered.org/outimp.htm

[acheld]

What are you wanting to know about output transformers?

I know the basics, but understanding "reflected load" has not been intuitive to me -- and that drives me nuts.   The more I read, the more I get it that Ohm's law is describing most of what I need to know.   

[shooter]

just a couple quick-e searches for more fun n knowledge with AC electronics


https://en.wikipedia.org/wiki/Resonant_inductive_coupling
https://en.wikipedia.org/wiki/Kirchhoff's_circuit_laws

[PRR]

..."reflected load" has not been intuitive to me --

Do you understand car or bicycle gears?

[HotBluePlates]

https://en.wikipedia.org/wiki/Resonant_inductive_coupling
...

The "resonant" part of this link makes it a poor choice for folks interested in guitar amps.  That's because our circuits don't "resonate" the way this is describing.

Instead:

I know the basics, but understanding "reflected load" has not been intuitive to me ...   The more I read, the more I get it that Ohm's law is describing most of what I need to know.

PRR is about to get to the core of it, but grasp a simple fact first:  Inductive Coupling (scratch that "resonant " stuff for a bit).

Just accept that there is inductive coupling.  What then happens is there is current we control that passes through a coil of wire, and the magnetic field created by that current induces a current in all other nearby coils.  The induction is such that the volts-per-turn applied to the driven coil induces the same volts-per-turn on all the other coils.

   -  The fundamental property we exploit with transformers is "volts-per-turn is the same on all windings."
   -  We then give each winding a different number of turns, depending on whether we want to step voltage Up, Down, or Keep-Same.

   -  Ohm's Law: just a relationship between Volts, Current, Impedance

   -  The transformer primary has no "intrinsic impedance" because this would just waste power as heat.
   -  We attach a load to a secondary, and that defines a ratio of Volts and Current according to Ohm's Law:
      -  10v Secondary, 5Ω Load ---> 10v / 5Ω = 2A ---> Volts : Current = 10v : 2A
      -  10v Secondary 10kΩ Load ---> 10v / 10kΩ = 1mA ---> Volts : Current = 10v : 1mA
   -  Assume an Ideal Transformer:  Power In (primary) = Power Out (secondary)
   -  The transformer's Turns Ratio changes how many volts we apply to the primary to get our "10v output" on the secondary.

   -  Power = Volts x Current, and
      Current = Volts / Impedance, so
      Power = Volts² / Impedance, so
   -  Since Power In = Power Out, the Impedance "felt" at the Primary is the square of the Volts-change imposed by the ratio of turns between Primary & Secondary.


   -  Why the current induced in the Secondary has the ability to generate "back-pressure" and define the Impedance of the Primary is a bit easier to grasp if you ever played with Ring Magnets on a Post.

[acheld]

Do you understand car or bicycle gears?

Well yes.       But I never heard the term "reflected work" when talking transmissions or pedaling my bikes.  It's just work.

Just accept that there is inductive coupling.

LOL.  I have accepted this; the event occurred in East Lansing after a brutal mid-term exam many years ago.

HBP, nice summary!   Much appreciated.

I think the term "induced load" is more descriptive than "reflected load."   Yeah, I get it, they're the same thing, I see that now.

[Lectroid]

HBP, nice summary!   Much appreciated.

I think the term "induced load" is more descriptive than "reflected load."   Yeah, I get it, they're the same thing, I see that now.

A very good explanation, HBP!  Most texts are too wedded to jargon--not enough explanations in plain language. No source I've read ever laid this out in simple basics terms like you did.  Induced = Reflected, full stop.

The bicycle gears analogy from PRR is helpful too.  We all remember downshifting our expensive French derailleurs, don't we?  Lower crankshaft RPMs = more torque for the same power input? 

Just Curious Dept:
If I haven't stretched this analogy far past the breaking point, one more question...

My understanding is that an OT changes a high voltage at low current into a lower voltage and a proportionally  higher current.   

Questions:
In the bicycle analogy, would the bicycles's crankshaft speed be analogous to voltage, and torque analogous to current? 

Also, In this scenario, is the speaker load analogous to the "hill" faced by the bicyclist ?

[shooter]

The "resonant"

apologies, it was beat'n into me for many years 

Just Curious Dept:

a tube amp is a voltage amplifier, the speaker is a current driven "load" so the OT acts as translator

[HotBluePlates]

Do you understand car or bicycle gears?

Well yes.       But I never heard the term "reflected work" when talking transmissions or pedaling my bikes.  It's just work.

But you know if you pick the wrong gear in the car you lug the engine & stall.  And if you pick the wrong gear on the bike going up that hill that you're not going anywhere.

I think the term "induced load" is more descriptive than "reflected load." 

... there is current we control that passes through a coil of wire, and the magnetic field created by that current induces a current in all other nearby coils.  ...
...
   -  Since Power In = Power Out, the Impedance "felt" at the Primary is the square of the Volts-change imposed by the ratio of turns between Primary & Secondary.


   -  ... the current induced in the Secondary has the ability to generate "back-pressure" and define the Impedance of the Primary ...

There are 2 core fallacies that people seem to apply with tube amps and output transformers, and it wrecks their ability to see how simple the power section really is:

   1.  The output tubes "make power."

   2.  The output tubes "want a certain load."


Output Tubes Pull Current

The output tubes is receive a grid-voltage signal and they respond by attempting to draw more/less plate current.  That's all they do.

The OT Primary Impedance Creates a Voltage-Drop

When a tube pulls plate current, the current passes through the OT primary.  Its primary impedance creates a voltage-drop in accordance with Ohm's Law:  Volts = Current x Impedance.  The voltage-drop momentarily leaves less voltage at the tube plate.

Here, we're only thinking about the half-cycle that goes from "idle current" to "peak plate current" when thinking about the tube's plate current creating a voltage-drop across the OT primary impedance.  Even so, we can understand most of what we want to know about the tube's operation from this half-cycle.

   -  We can acknowledge the tube's plate voltage shoots above its idle value as plate current falls below idle current and to zero-current.  However, Zero-Current x Any-Voltage = Zero Power, so we care about this part of the cycle mainly with regard to transformer winding insulation.

Output Tubes Don't "Make Power"

Power = Volts x Current, and our output tubes only have control over the plate current they attempt to pull through the OT primary impedance.  The resulting voltage-drop above is multiplied by the late current the tube pulled to create the voltage-drop.  And that is our power output; it's partly a byproduct of the OT primary impedance causing a voltage-drop.

   -  If we make the primary impedance smaller, there is a smaller voltage-drop for a given tube plate current.  Power Output goes down.  Power Output is being constrained by tube plate current (the tube's curves show a maximum plate current for a given condition & applied-signal).

   -  If we If we make the primary impedance larger, there is a larger voltage-drop for a given tube plate current.  Power Output goes up, to a limit.

   -  Our output tubes cannot pass plate current with 0v on their plates; typically, they need 50-100v remaining on the plate to pass current.  We have a fixed supply voltage, so the most-voltage available to drop across the OT primary impedance is "Supply Volts - 100v" (or 50v or wherever the tube's "plate current knee" lands for that screen voltage).

   -  Going to an OT primary impedance "higher than the limit" noted above attempts to pull plate voltage down to a level that stops plate current.  Power Output falls because the "Current" part of our equation is being constrained.

This image shows:
   -   "Optimum" loading (Blue)
   -   "Too-small primary impedance" loading (Red): power output is constrained by lessened voltage-swing
   -   "Too-large primary impedance" loading (Brown): power output is constrained by lessened current-swing

[HotBluePlates]

The OT has no "Primary Impedance" only a "Reflected Impedance"

The OT's primary has a DC Resistance very much less than the reflected impedance it was designed to have.  That's to avoid wasting audio-power as heat.

Some folks have looked at a transformer's Primary Inductance, then tried to calculate the impedance implied by that for different frequencies.  That's a mistake...  The primary inductance makes sure that when I say, "just accept that there is inductive coupling," that it actually happens.

Primary inductance makes the transformer behave like a transformer.  It also sets a lower frequency limit below which the transformer stops exhibiting "ideal transformer behavior."  If we need to pass signal-over below that limit, we need to construct the transformer differently (add turns, increase core size, change core material) to boost the primary inductance.

The Impedance implied at any given frequency by the primary inductance is very much higher than the Impedance resulting from the designed-load being attached to the secondary and reflecting an impedance back to the primary.  That results in the reflected impedance dominating the transformer's behavior.

   -  A Super Reverb OT has a 4kΩ reflected primary impedance at 200Hz.
   -  One Super Reverb OT replacement has a Primary Inductance of 12.5H
   -  The Primary Inductance has an impedance of 15.7kΩ at 200Hz.
   -  The reflected primary impedance is in-parallel with the impedance due to primary inductance, so the tubes "see" the reflected impedance.

[PRR]

Say you can pedal 100 pounds(force) at 100rpm.

If you gear 1:1 (BIG back sprocket) you can get 100 pounds thrust and climb a very steep hill (1 in 2 grade?) at 100rpm at the wheel.

If you gear 1:3 (18t back sprocket) you can get 33 pounds thrust which may be fine on the level, and spin 300rpm at the wheel.

We could say these represent mechanical "Ohms". 100lf @ 100rpm is 1 Ohm. 33lf @ 300rpm is 0.11 Ohms. Same legs, same power, but different ratios of force and speed.

[glass54]

Hi Guys
At the risk of being shouted down, I believe some of you have missed the point!

The "resonant".......apologies, it was beat'n into me for many years.......... 

etc, etc.

Nowhere in the bicycle conversation did I see a critical component ie Coffee and Cake mentioned (essential for morning tea and afternoon tea )
I cannot see how you can expand intellectual energy and physical energy without feeding the activity with coffee and associated compliments.

PRR is right  Having just returned from Europe after pedaling the Danube, a little of the Rhine and all of the Mosel Rivers for some 1650 km, on our touring bikes, the gear ratios, coffee and cake were essential components to a successful trip.
I hope I didn't interfere with the preferred resonances of these posting and I do admit that I do enjoy technical articles even if I only absorb 30% of the info. After all we're all here to learn and better our hobbies.
(By the way NO ebikes, can't transport ebikes from Aussie to Europe by air)
Kind Regards
Mirek

[dwinstonwood]

I intend to to re-read this thread until the lightbulb turns on.

One part of this OT conundrum (in my case, anyway), is the matter of choosing different loads and voltages for different tube types.

In the first two columns, this 6L6GC datasheet shows two "Effective Loads" for the same Plate and Screen voltages 360 and 270). The 6,600 Ohm load makes 26.5 watts, and the 3,800 Ohm load makes 18 Watts.

Clearly, we have control over the Output Power by messing around with voltages and loads. 450V and a 5,600 Ohm load would supply the most power. But, maybe one doesn't want maximum power. Maybe they do want an 18 Watt AB1 6L6 amp.

Is this a better route to take than using attenuators? The sound (tone) of the amp would certainly be affected by this, yes?

[acheld]

If you gear 1:1 (BIG back sprocket) you can get 100 pounds thrust and climb a very steep hill (1 in 2 grade?) at 100rpm at the wheel.

On the lighter side, only Sepp Kuss and Jonas Vingegaard can ride with sockets that large on a mountain.

To further the analogy,  the secondary impedance in the bike gearing would be the power applied to the ground by the wheels. 

At a 1:2 gradient, there would be insufficient friction between wheels and ground to maintain forward momentum (of the bike+rider) -- and the wheels would spin -- reducing the secondary impedance as well as "reflected" impedance in the primary gear.

In a pro bike race, it is not common to see gradients higher than about 15% or so; I think the max I've seen is a 25% grade.   Even with their incredible power to weight ratios, it's very tough for riders to climb at those gradients.   And they do lose grip!

[PRR]

I intend to to re-read this thread until the lightbulb turns on. this 6L6GC datasheet shows two "Effective Loads" for the same Plate and Screen voltages 360 and 270). The 6,600 Ohm load makes 26.5 watts, and the 3,800 Ohm load makes 18 Watts.

Confuse-a-cat.

While the numbers were copied-over to the 6L6GC sheet, this goes back to the (almost) first 6L6 data. There is another set of numbers for AB2 (grid-current) operation. The 6k6 load is optimum for one, the 3k8 load is optimum for the other.

With the higher Vg2 rating on later 6L6 types, AB1 operation can reach the AB2 condition. The Bassman 5F6a "is" your 3k8 condition with Vg2 raised from 270v to near 400V, and everything rounded-up like it only had to last 30 days.

[pdf64]

…
Clearly, we have control over the Output Power by messing around with voltages and loads. 450V and a 5,600 Ohm load would supply the most power. But, maybe one doesn't want maximum power. Maybe they do want an 18 Watt AB1 6L6 amp.

Is this a better route to take than using attenuators? The sound (tone) of the amp would certainly be affected by this, yes?

Marshall SV20 (EL34 pair) and JTM20 (5881 pair), both cathode bias, around 300V HT with a 6k OT, seem to be well received. 
https://www.marshallforum.com/threads/cloning-an-sv20.131774/page-2#post-2364273

[bmccowan]

If you gear 1:1 (BIG back sprocket) you can get 100 pounds thrust and climb a very steep hill (1 in 2 grade?) at 100rpm at the wheel.

On the lighter side, only Sepp Kuss and Jonas Vingegaard can ride with sockets that large on a mountain.

Jumping on my saddle here. PRR refers to a big back sprocket. So I think you mean that only Kuss and Vingegaard can ride with a gear that large. The larger the back sprocket is, the easier it is to climb. So if you are a wimpy rider like me, you want a small front chainring - 24 or less and a large rear sprocket 28 or more.
The whole audio transformer thing I know nothing about

[acheld]

big back sprocket.

My bad.  Missed the "back" sprocket.   The rest of the analogy stands.  I am in awe of the pro cyclists . . .

[bmccowan]

Me too. I raced when I was younger (much younger) and was routinely dropped off the back. For years I watched every stage of the Tour de France, and all I could of the Giro and Vuelta. Then the Lance scandal hit me like a brick.
But I am back as a fan. I have several nice steel road bikes that I built up all Campy from framesets - A De'Rosa, a Ron Cooper, a Gianni Motta, etc. Sadly none of my family's younger generation is interested in them - same with my amps.

[shooter]

I was a distance rider at 14, did 20 miles out and 20 back virtually all summer long so I could make out with this drop-dead beautiful blond from Indiana.  When summer ended n I was heartbroken, Mom came to the recue, signed me up for a 100 mile charity ride, I was the youngest rider, took 8.5 hours on my supped-up Swinn(sp) Varsity.  The next summer the neighbor guy got his 12th DWI at age 21, got  a life-time suspension on his drivers license, traded me his 64 slant six 3 speed for my bike, I never looked back 

[dwinstonwood]

PRR, so you're saying that the Bassman 5F6a is prone to enough grid current to run AB2? I'm still trying to unravel what all this stuff means... AB1, AB2, etc.

shooter, I had an orange Varsity. I rode that thing all over the place (but, never competitively). Then came the greatest Christmas present of my entire life, a '75 Yamaha YZ. Pedaling made no sense after that.

[pdf64]

… the Bassman 5F6a is prone to enough grid current to run AB2? …

No, AB2 requires a power amp to drive the output valve grids (because when the signal sends them positive with respect to their cathode, they’re drawing current and voltage x current = power). So it’s a pain to implement, no point bothering unless attempting to achieve the max possible class B2 anode efficiency (audio power output / anode dissipation limit), eg maybe about 100W for a pair of 6L6 GC. Or eliminate bias shift / blocking distortion at the output valve grids.
If there’s RC coupling to the output valve control grids, the possibility of AB2 / B2 operation is pretty much excluded.

[bmccowan]

pdf - what the heck does this have to do with bicycles?

[shooter]

it’s a pain

no point bothering

operation is pretty much excluded

Is that easier to understand, he doesn't like them 

[dwinstonwood]

it’s a pain

no point bothering

operation is pretty much excluded

Is that easier to understand, he doesn't like them 

But, those statements could also apply pedaling, when you can just hang on and ride...

[PRR]

PRR, so you're saying that the Bassman 5F6a is prone to enough grid current to run AB2?

I thought we were talking datasheet. Not some specific amp.

By using 5881 (a 6L6 with 400V rating on G2), Leo got the AB2 power in AB1 mode.

[dwinstonwood]

Thanks PRR. I think I got it. The equivalent power output of the early 6L6's running AB2 by using 5881's at higher voltages in AB1 mode. Right?

[HotBluePlates]

Thanks PRR. I think I got it. The equivalent power output of the early 6L6's running AB2 by using 5881's at higher voltages in AB1 mode. Right?

Yes.