• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Theoretical question about interstage transformers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Our venerable Kondo San was a professor of electronic engineering and molecular metallurgy, one of his design goals with his Ongaku was the minimal use of iron, not a very popular idea though.

A coupling cap is cheaper, smaller, wider bandwidth, less distorting, less prone to hum pickup etc.

Absolutely agree, the curious is that folks who hate capacitors are fans of interstage transformers.

Faradophobia, I guess... :p:D

Try to understand two types of interstage. one with magnetization current and other without.

Could you please explain how does a transformer work without magnetization current?
 
Our venerable Kondo San was a professor of electronic engineering and molecular metallurgy, one of his design goals with his Ongaku was the minimal use of iron, not a very popular idea though.

I am afraid Juan but Kondo San made for himself things that he never put for sale and are better than his commercial products. In fact he possibly used those personal items for reference! Business and personal reference not always match....

Something like this "funny" looking thing:

YouTube

Having said that I have myself amps that follow his commercial and successful designs like the Neiro.

I think it's a good general representation. Only a fool would pay 400+ dollars for an IT when a great tube based circuit would cost about 50 IMHO.

Also I should mention that I was running open loop.

If I had $100 I would never spend 50 on a basic IT and 50 for the rest of the amp. Instead a proper $400 IT not only will have perfect 10 KHz square waves but also BEYOND 20 KHz perfect square waves which means extended and smooth frequency response (that also facilitates the application of feedback), generous inductance and headroom. A minimum of 200 KHz bandwidth is not just no match for the Hammond but very often even for RC or directly coupled stages!

These two IT's below cost actually about $325. The first is for SE and the second for PP:

http://www.monolithmagnetics.com/si...IT-05 interstage transformer prelim rev01.pdf
http://www.monolithmagnetics.com/si...stage transformer balanced bifilar prelim.pdf

N.B.
The 50pF cap in the load is not necessary anymore when the power tube is connected.

This one costs more because it is more complex to make but achieves the same minimum 200KHz bandwidth without bifilar winding:
http://www.monolithmagnetics.com/si...3 25 interstage transformer prelim rev 01.pdf

It's impossible to get blood out of turnips.
 
waxx said:
But each to his own off course, this is a subjective opinion.
You asked for a technical reason. I gave you a technical reason: low mu in DHTs. I also gave a social reason: nostalgia.

This is an objective fact: a coupling capacitor is much closer to being an ideal component than an interstage transformer. This, of course, says nothing whatsoever about which sound you will prefer; it merely says something about faithful reproduction of a signal.
 
This is an objective fact: a coupling capacitor is much closer to being an ideal component than an interstage transformer. This, of course, says nothing whatsoever about which sound you will prefer; it merely says something about faithful reproduction of a signal.

It depends on the application. And it's just only about DHT's. Have you ever measured how much grid current a common EL34 can draw when approaching full modulation (Class AB1)? When a smallish grid current will appear things will be the other way around with the capacitor being anything but ideal or nearby. So one either uses a transformer or proper DC coupled stage that can handle that. None of the latter solutions are common in mass production. RC coupling in power amps is mostly for mass production because is cheap and simple. Surely not the best. It might be the best option in some cases where grid current is not a concern.
 
With a smallish grid current … a capacitor isn't ideal. But… a transformer or proper DC coupled stage can handle this.

Or, a sufficiently large coupling capacitor.

One of the reasons coupling capacitors — ones far, far from ideal — get “a clean pass” in service as DC blocking AC pass-thru interstage elements (contrary to the majority of hullabaloo regarding being 'there') is because as DF96 once made me realize, interstage transformers pass almost no current nominally. Relatively-negatively biassed input grids conduct very little current — those (again hat-tip to DF96) some grid current is expected.

At very high loading impedance, even a smallish interstage coupling cap has an audio-range attenuation approaching zero at the lowest audible frequencies. But as you note, that situation changes dramatically as soon as the downwind load becomes appreciably greater than near-zero-current.

Yet the solution is so straight forward that I find it amusing that it is so often overlooked: estimate the grid impedance when operating at positive instantaneous input grid conditions, then calculate at 10 Hz what value of interstage coupling capacitor would really be needed to deliver that current without appreciable attenuation.

Voilá. Then the bumped up interstage capacitor is just as applicable serving duty to remove the previous stage's DC anode influence, as the interstage transformer. And per what I wrote in my original comment here, with modern ultra-performance plastic films and evaporated aluminum 'plates', surprisingly high value non-electrolytic capacitors are readily available, and far smaller than a woodchuck in form factor.

Just saying. You can get there from here with capacitors. Just takes some calculating and adjusted component values.

GoatGuy
 
I am afraid Juan but Kondo San made for himself things that he never put for sale and are better than his commercial products. In fact he possibly used those personal items for reference! Business and personal reference not always match....

Sure, maybe in his intimacy he used SS amplifiers, it would be another possibility...

Having said that I have myself amps that follow his commercial and successful designs like the Neiro.

Neiro shows a classic design from Kondo San, 5687 cathode follower DC coupled to output valves.

It depends on the application. And it's just only about DHT's. Have you ever measured how much grid current a common EL34 can draw when approaching full modulation (Class AB1)? When a smallish grid current will appear things will be the other way around with the capacitor being anything but ideal or nearby. So one either uses a transformer or proper DC coupled stage that can handle that. None of the latter solutions are common in mass production. RC coupling in power amps is mostly for mass production because is cheap and simple. Surely not the best. It might be the best option in some cases where grid current is not a concern.

Pierpaolo, we are talking about reasonable designs, even being myself a dumb designer I am aware of blocking distortion.

In 300B based amplifiers, an interstage transformer maybe would help with third harmonic distortion cancellation, the only reason which comes to mind.
 
Or, a sufficiently large coupling capacitor.

Unfortunately sufficiently large capacitor often means less ideal. Have a look at manufacturers datasheet and you will see that 1 uF, often less, is more or less a threshold for higher losses. Capacitors have hysteresis themselves....are not ideal at all in the exact meaning of the word!
 
Last edited:
Pierpaolo, we are talking about reasonable designs...

I am too. I have bought a pair of used LL-1635 from a user of this forum for my latest amp at less than half price. If you buy 2x high quality coupling caps they will cost more! So if one doesn't want to spend much can look at the used market...just an example. Cheap capacitors have non-negligible losses.

Apart from this, cost is subjective and there is no technical argument that will change that!

Not to mention that I can take the supply for the driver from the 250V regulated supply of the power stage. With 205V supply (190V anode voltage and the rest for bias and transformer drop) I can get the full drive which is 138V grid-to-grid. See what you need with RC coupling to get that drive with the same very low distortion....added cost for the power supply at least!!!
 
Last edited:
Cheap capacitors have non-negligible losses.

Capacitor hysteresis is way less than transformer hysteresis, but I am not promoting change a transformer by a capacitor, an RC coupled stage DC coupled to output valve is my preferred solution, simpler, cheaper and with lower distortion.

Not to mention that I can take the supply for the driver from the 250V regulated supply of the power stage. With 205V supply (190V anode voltage and the rest for bias and transformer drop) I can get the full drive which is 138V grid-to-grid. See what you need with RC coupling to get that drive with the same very low distortion....added cost for the power supply at least!!!

That is true, but in my designs, power supplies take more than 60% of available space and this is not a problem to me.
 
Last edited:
Nothing against the DC Kondo-like solution but you need one more active device and its supply. Nothing comes for free.

That is relative.

These two IT's below cost actually about $325.

A good interstage transformer does not come for free either...

You rely too much on catalog data. The really good capacitors are quite expensive!

I use cheap MKP Siemens/Epcos with great success, low losses, low inductance...
 
Last edited:
A good interstage transformer does not come for free either...

I am not the one complaining about costs.

I use cheap MKP Siemens/Epcos with great success, low losses, low inductance...
Try to get around grid current then, as suggested earlier, by bumping-up the capacitor and will see how good they are.
I can only believe you if we are talking about small values but not for larger values which means added complexity and added cost to other areas. Nothing comes for free.
 
Last edited:
Try to get around grid current then, as suggested earlier, by bumping-up the capacitor and will see how good they are.

Which part of "I am aware of blocking distortion" you did not understand?

Capacitor hysteresis is way less than transformer hysteresis, but I am not promoting change a transformer by a capacitor, an RC coupled stage DC coupled to output valve is my preferred solution, simpler, cheaper and with lower distortion.

Fushta! :p:D
 
And so that is. You just can't tell that capacitors have negligible losses. It's true only under certain conditions. Here we start again?
Regarding DC coupling is NOT better as well. It's an alternative solution. This is my point having both to compare one to the other.....
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.