Author Topic: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier  (Read 48728 times)

0 Members and 1 Guest are viewing this topic.

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 2589
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #150 on: January 25, 2016, 06:13:49 pm »
Drawing an  analogy to a J-FET is simply wrong.
A more apt analogy would be a MOS-FET since there is an insulator between the elements like in a vacuum tube.


Sorry Dave Ya got it wrong.

Sorry for the late reply, but comparison to a jfet is pretty much right. 

Both the tube's grid and jfet's gate behave the same way. Small capacitance for negative voltages, a diode for positive voltages (with respect to cathode / source).
 

Offline AF6LJ

  • Supporter
  • ****
  • Posts: 2901
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #151 on: January 26, 2016, 04:06:57 pm »
Drawing an  analogy to a J-FET is simply wrong.
A more apt analogy would be a MOS-FET since there is an insulator between the elements like in a vacuum tube.


Sorry Dave Ya got it wrong.

Sorry for the late reply, but comparison to a jfet is pretty much right. 

Both the tube's grid and jfet's gate behave the same way. Small capacitance for negative voltages, a diode for positive voltages (with respect to cathode / source).
J-FETs have a lower impedance there is physical contact between elements, and a J-FET cannot operate with current flowing on the gate, a vacuum tube can.

I hate to put too sharp of a point on it but;
Drawing that analogy dumbs down the whole discussion. Vacuum tubes are unique as an active device, even though they share some commonalities with solid state field effect devices.
Sue AF6LJ
Test Equipment Addict, And Proud Of It.
 

Offline bktemp

  • Super Contributor
  • ***
  • Posts: 1623
  • Country: de
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #152 on: January 26, 2016, 04:45:04 pm »
Drawing an  analogy to a J-FET is simply wrong.
A more apt analogy would be a MOS-FET since there is an insulator between the elements like in a vacuum tube.


Sorry Dave Ya got it wrong.

Sorry for the late reply, but comparison to a jfet is pretty much right. 

Both the tube's grid and jfet's gate behave the same way. Small capacitance for negative voltages, a diode for positive voltages (with respect to cathode / source).
J-FETs have a lower impedance there is physical contact between elements, and a J-FET cannot operate with current flowing on the gate, a vacuum tube can.

I hate to put too sharp of a point on it but;
Drawing that analogy dumbs down the whole discussion. Vacuum tubes are unique as an active device, even though they share some commonalities with solid state field effect devices.
A J-FET can be operated with a current flowing into the gate. If the junction gets forward biased, current will flow. And this will decrease the drain-source resistance further.
The operating principle may be different, but electrically their behaviour is similar (Id/Ug curve), except J-FETs work on a much lower voltage than tubes and drain and source are interchangeable.
At the end of the tube era some tubes were substituted with a J-FET input stage driving a larger transistor.
 

Online Brumby

  • Supporter
  • ****
  • Posts: 9215
  • Country: au
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #153 on: January 27, 2016, 04:23:26 am »
Drawing that analogy dumbs down the whole discussion.

A simplification it may be - but that comment was still more intelligent than the product.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 2589
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #154 on: January 27, 2016, 02:28:40 pm »
I don't feel offended when talking about the JFET analogy for tubes.  :)

It is an analogy, not a direct replacement having same operation voltage and same high impedances...
 

Offline AF6LJ

  • Supporter
  • ****
  • Posts: 2901
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #155 on: January 27, 2016, 03:02:10 pm »
I am not offended; I feel the analogy dumbs down the subject.
Sue AF6LJ
Test Equipment Addict, And Proud Of It.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 2589
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #156 on: January 27, 2016, 04:59:26 pm »
You seem to have a better analogy, don't you?   :D
 

Offline AF6LJ

  • Supporter
  • ****
  • Posts: 2901
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #157 on: January 27, 2016, 07:08:38 pm »
You seem to have a better analogy, don't you?   :D

This one situation where an analogy serves no good.
Go learn how vacuum tubes actually work.

Just as a small example the J-FET analogy doesn't explain why a tetrode under certain conditions draws negative screen current.

Bad analogies are a trap for young players.
Sue AF6LJ
Test Equipment Addict, And Proud Of It.
 

Offline Pentium100

  • Regular Contributor
  • *
  • Posts: 245
  • Country: lt
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #158 on: January 27, 2016, 09:23:57 pm »
JFET is similar to a pentode (operating with the screen and suppressor grids connected to fixed voltages), but I do not think there is a solid state device that has the IV curves of a triode.
 

Offline Groucho2005

  • Regular Contributor
  • *
  • Posts: 84
  • Country: es
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #159 on: January 27, 2016, 09:34:01 pm »
but I do not think there is a solid state device that has the IV curves of a triode.
Yes there is, Static Induction Transistor (SIT), some info here: http://www.firstwatt.com/sitintro.html
 

Offline helius

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #160 on: January 27, 2016, 09:43:53 pm »
Yes there is, Static Induction Transistor (SIT), some info here: http://www.firstwatt.com/sitintro.html
I was reading with interest until he said "First, it allows a single gain stage...without a feedback loop or degeneration." As is widely known, a triode amplifier doesn't work "without feedback": the feedback happens internally.
 

Online blueskull

  • Supporter
  • ****
  • Posts: 12284
  • Country: cn
  • Power Electronics Guy
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #161 on: January 28, 2016, 02:50:46 am »
but I do not think there is a solid state device that has the IV curves of a triode.
Yes there is, Static Induction Transistor (SIT), some info here: http://www.firstwatt.com/sitintro.html

Their shit outputs 10W from 200W input at 4.5% THD at full load and 0.18% THD at no load??? :wtf:
 

Offline AF6LJ

  • Supporter
  • ****
  • Posts: 2901
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #162 on: January 28, 2016, 04:14:44 am »
JFET is similar to a pentode (operating with the screen and suppressor grids connected to fixed voltages), but I do not think there is a solid state device that has the IV curves of a triode.

I haven't seen one...


In this day and age vacuum tubes are for the most part at the end of their technological evolution. The last place you will find vacuum tube tech used in are devices like ion implanters, Ion discharge machining and similar industrial applications, where the work is being done inside a vacuum with an electron emitter and grids, anodes and the like to accelerate ions.

Vacuum tubes are still fun to play with. :)
Sue AF6LJ
Test Equipment Addict, And Proud Of It.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 2095
  • Country: gb
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #163 on: January 28, 2016, 07:42:20 am »
Their shit outputs 10W from 200W input at 4.5% THD at full load and 0.18% THD at no load??? :wtf:

Where did you get the 200W input from?  I see 18dB gain and 10W output.

The distortions figures are quite impressive for a single transistor power amp with no feedback, try getting anywhere near that with a single bipolar or MOSFET.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 2589
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #164 on: January 28, 2016, 11:35:15 am »
AF6LJ: Yolo, there are also mosfet tetrodes. Usually HF stuff. Also behaving well like a tetrode (pentode) valve, having the g1 input capacitance ridiculously low.
 

Offline AF6LJ

  • Supporter
  • ****
  • Posts: 2901
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #165 on: January 28, 2016, 01:31:18 pm »
AF6LJ: Yolo, there are also mosfet tetrodes. Usually HF stuff. Also behaving well like a tetrode (pentode) valve, having the g1 input capacitance ridiculously low.

I know about dual gate MOS-FETS They don't behave fquite like a tetrode but are a beast all their own.
Sue AF6LJ
Test Equipment Addict, And Proud Of It.
 

Offline Groucho2005

  • Regular Contributor
  • *
  • Posts: 84
  • Country: es
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #166 on: January 28, 2016, 03:47:27 pm »
Their shit outputs 10W from 200W input at 4.5% THD at full load and 0.18% THD at no load??? :wtf:

Where did you get the 200W input from?
Go to the product page and look at the SIT1/SIT2 amplifier pages. Yes, this is an extremely inefficient design and there are barely any speakers it can drive properly.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 2095
  • Country: gb
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #167 on: January 28, 2016, 03:51:08 pm »
Their shit outputs 10W from 200W input at 4.5% THD at full load and 0.18% THD at no load??? :wtf:

Where did you get the 200W input from?
Go to the product page and look at the SIT1/SIT2 amplifier pages. Yes, this is an extremely inefficient design and there are barely any speakers it can drive properly.

That's simply because they are Class A designs, which are inherently inefficient.  Pure class A amps with reasonable power outputs are room heaters, irrespective of the type of power output device.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 9962
  • Country: 00
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #168 on: January 28, 2016, 03:53:47 pm »
Go to the product page and look at the SIT1/SIT2 amplifier pages. Yes, this is an extremely inefficient design and there are barely any speakers it can drive properly.

It's a class A amplifier, so... that's just the way they work.

They're good for cold winter days. With a small enough heatsink you can even cook eggs.  :)

 

Offline Groucho2005

  • Regular Contributor
  • *
  • Posts: 84
  • Country: es
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #169 on: January 28, 2016, 05:02:03 pm »
Their shit outputs 10W from 200W input at 4.5% THD at full load and 0.18% THD at no load??? :wtf:

Where did you get the 200W input from?
Go to the product page and look at the SIT1/SIT2 amplifier pages. Yes, this is an extremely inefficient design and there are barely any speakers it can drive properly.

That's simply because they are Class A designs, which are inherently inefficient.  Pure class A amps with reasonable power outputs are room heaters, irrespective of the type of power output device.
You have to differentiate a bit. Theoretically, a Class-A amplifier can have up to 50% efficiency. A Class A amplifier with complementary bipolar or MOSFET output devices comes very close to that number.
However, a single ended class a triode amp, which is very popular in the audiophile/audiophool community, will barely reach 10% efficiency which seems pretty insane but there are people who wouldn't have it any other way.

 

Offline rrinker

  • Super Contributor
  • ***
  • Posts: 1885
  • Country: us
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #170 on: January 28, 2016, 08:17:33 pm »
 The Class A design was used as a selling point for the receiver I had in college. Well, at full volume (this sucker goes to 11!), it would barely get warm. But at normal listening levels - yes, I could fry eggs right in my dorm room. All was fine until I developed a habit of going to sleep with the volume really low for background noise - fried the outputs twice on it.

 

Online Brumby

  • Supporter
  • ****
  • Posts: 9215
  • Country: au
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #171 on: January 29, 2016, 02:01:13 am »
That's the one thing about Class A amplifiers I always remembered ....  If you want them to run as cool as possible, you had to run them at full volume.
 

Offline retrolefty

  • Super Contributor
  • ***
  • Posts: 1604
  • Country: us
  • measurement changes behavior
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #172 on: January 29, 2016, 02:35:01 am »
That's the one thing about Class A amplifiers I always remembered ....  If you want them to run as cool as possible, you had to run them at full volume.

 That sounds like a myth to me? Can you link to an analysis to support that.
 

Online Brumby

  • Supporter
  • ****
  • Posts: 9215
  • Country: au
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #173 on: January 29, 2016, 04:32:04 am »
Hmm...

My text books are buried somewhere if not long gone - and Google hasn't been too helpful.  I've found some references to power dissipation in Class A amplifiers, but nothing that gives me what I was looking for - which would be a chart of power dissipation vs output power.

It is my understanding that pure class A amplifiers draw maximum power from their supply at no signal (and therefore no output).  While not being particularly critical of my reading over the years, unless I've read it wrong, as power increased to the load, there is a corresponding drop in the power dissipated by the amplifier.  Therefore the amplifier runs cooler at maximum output.

I'm open to correction if I've got that wrong.

Will still keep looking for something to back it up - but I can't spend any longer on this right now.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 2095
  • Country: gb
Re: EEVblog #837 - Reverse Engineering A Valve Headphone Amplifier
« Reply #174 on: January 29, 2016, 10:20:34 am »

You have to differentiate a bit. Theoretically, a Class-A amplifier can have up to 50% efficiency. A Class A amplifier with complementary bipolar or MOSFET output devices comes very close to that number.

Not true for an equivalent design to the one shown in the SIT datasheet.  The efficiency of a simple class A design is no better than 35% at full load, and much worse at low outputs or under quiescent conditions.  Efficiency can be improved by using active loads rather than a simple resistor or transformer coupling, but even then you are unlikely to reach 50% efficiency.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf