A new hybrid concept: C-Vac - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 30th April 2013, 09:00 AM   #1
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Default A new hybrid concept: C-Vac

Presentation:

I am going to describe a new class of hybrids, a Composite of Vacuum tubes and semiconductors having remarkable properties.
  • Many hybrid schemes end up in an unhappy marriage, combining the flaw of one technology with the inconvenience of the other.
  • On the contrary, C-Vac is designed with a well-defined goal in mind: generate near-ideal substitutes of vacuum tubes using a "model" tube and semiconductor circuitry.
The main application I had in mind when developping the concept was to make OTL amplifiers easily accessible: very few commonly available tubes are capable of driving a normal speaker load directly, and even those that can like the 6C33 work on the edge and need parallelling to operate comfortably.

The OPT in tube amplifiers is one of the main sources of problems: even the best generate non-linearities, they are costly, very specific, and their parasitic parameters severely limit the amount of feedback that can safely be applied to combat the said non-linearities.

Tube sound is one thing, "magnetic" sound is something completely different and rather unpleasant: it generates odd harmonics at a low level because of the sigmoid transfer function.
Benefiting from the tube sound without the unpleasantness of the magnetic sound is certainly worthwhile, and those having listened to OTL amps report a greater transparency, unachievable by other means.

How it works?

C-Vac is essentially a high precision current multiplier, using a low power tube (or other device) as a model. The virtual tube resulting from the multiplication process is the exact copy of the model, just very much larger: instead of mA's, the anode current becomes amperes, but the curves remain absolutely identical, apart from the scale factor.
The scale factor can range from 1 to many thousands without accuracy loss, and the dynamic performance extends to tens of MHz.
Below is an illustration of the faithfullness of the multiplication: the green curves are the reference scaled by the correct factor to enable comparison, and the red ones belong to the substitute based on the same 12AT7 tube.
The curves are almost perfectly overlaid, except at low voltages where divergences appear.
These divergences are caused by the fact that the substitute uses a enhancement device, requiring a minimum voltage to begin to conduct, and there are no auxiliary bias sources or connections.
But this also means that the substitute can be used in any configuration or topology, without restriction.
Anyway, circuits using tubes in the linear domain normally provide a minimum A-K voltage, and if this low voltage issue really is a problem, there are easy workarounds with auxiliary bias connections.

From the red curves, we can see that this virtual tube is in fact equivalent to 303 paralleled 12AT7 triodes: the C-Vac circuit acts as an exoskeleton boosting the "muscles" of its "pilot":

Click the image to open in full size.

The resulting composite device is totally dominated by its tube model, the semiconductor part being almost ideally transparent and neutral.

Of course, purists are bound to say that "it cannot possibly sound like a real tube", but all the others will see it as a sensible and realistic approach, combining the best of both worlds.

I have posted this subject in the Solid State section, because C-Vac is completely based on semiconductors, but I will also describe a real application example, an OTL amplifier in the Tubes section.

To be continued....
Attached Images
File Type: png Cvac.png (49.9 KB, 262 views)
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪

Last edited by Elvee; 30th April 2013 at 12:56 PM. Reason: typo
  Reply With Quote
Old 30th April 2013, 09:28 AM   #2
diyAudio Member
 
Project16's Avatar
 
Join Date: Nov 2012
Strongly on!
Another thread that might be exciting.

Regards!
  Reply With Quote
Old 30th April 2013, 09:37 AM   #3
ilimzn is offline ilimzn  Croatia
diyAudio Member
 
Join Date: Feb 2005
Location: Zagreb
The simplest form of such a device is the CFP, the 'input' device dictates the characteristics. To get proper multiplication one needs to work on the way the 'output' device is interfaced, somewhat more complex than just a resistor in the collector/drain/anode circuit of the reference device
Knowing the author's previous work, I am very much looking forward to this thread!
  Reply With Quote
Old 30th April 2013, 12:52 PM   #4
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Default After the appetizers, here is some real meat

Let us begin to examine some circuits.

Many implementations are possible and have been evaluated, I have selected the ones offering the best performance to complexity ratio.

In C-Vac1, the cathode current of the "pilot" is sensed by R2, and compared to the boosted current by a Rush cascode, compensated by Q3 and Q4.

The circuit is used in a follower and a "naked" tube is also shown for reference.
The boosting factor is set by the ratio of R2 to R1, in this case 303x.
The loads are in the same ratio, yielding 30.3K for R3.

For a complete accuracy, the 100Ω sense resistor is also included.
In general, the circuits where the tube has to be virtualized will already use some degeneration, and thus the sense resistor can be made part of it.
If there is no degeneration, the sense resistor can be made smaller, it is a tradeoff between the sensing accuracy and the deviation from a non-degenerated device.
100Ω for a low power triode is relatively insignificant and gives excellent accuracy with discrete transistors.
If the transistors are monolithic matched pairs, even greater accuracy can be achieved, allowing for lower values.

We can see that the two outputs are almost perfectly superposed.
The differences in the THD's are also insignificant, and the spectral signatures are similarly near-identical.

For this simple circuit, the frequency response is limited to only ~10MHz, but we will see later that this can be easily extended if required.

Note that a real circuit should also preferably include a gate stopper and a protection zener, but since their effect is negligible, they do not appear in the sim.
Attached Images
File Type: png Cvac2.png (169.0 KB, 318 views)
File Type: png Cvac3.png (83.0 KB, 365 views)
File Type: png Cvac4.png (89.7 KB, 327 views)
Attached Files
File Type: asc Cvac1.asc (2.7 KB, 18 views)
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪
  Reply With Quote
Old 1st May 2013, 04:19 PM   #5
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Let us go a bit deeper: the simple C-Vac1 circuit is accurate, has a good low-frequency behavior, but it also has a weakness: its dynamic performances.

The small signal frequency response is perfectible, but that is not where the problem really lies.
The main weakness is the large signal frequency response: it is caused by the large value of R4 and the Cgd capacitance of the MOS. Under large signal conditions, the current provided by R4 becomes insufficient to charge Cgd.
The 470K value just allows for a full swing at 20KHz, which in theory is sufficient for audio, but is a bit short to say the least.

Why not decrease R4 then? A tenfold reduction would mechanically push the full power bandwidth to 200KHz for example.
The reduction of R4 (and R3 which is linked to it) has a number of inconvenients: it increases the quiescent power, and it makes the virtual tube less perfect.
The resistors in themselves do not actually degrade the accuracy, because their current is returned via R1, and is thus taken into account in the tube simulation, but they put a lower limit to the cut-off current of the tube.
With 47K, the current could not go below ~13mA, and the additional quiescent dissipation at 150V would amount to ~1W.

None of that is catastrophic, and has to be put in perspective: the composite tube now has a maximum current of several amperes, and less than 15mA leakage is not going to degrade anything, and the 300V voltage used for the test is not realistic in an OTL context: I have kept the 300V value to allow for easy comparisons in a tube environment, but in the example, it leads to rather obscene power levels, almost 1kW peak in this case, much more with a speaker load.

Under more realistic conditions, the problems would be less severe, but anyway tackling them by brute force only is not the best solution.

Adding a buffer stage will improve the small signal behavior and ~halve the current requirements, because R3 and R4 can remain of a relatively high value.
C-Vac2 shows an example: the small signal bandwidth is increased to ~50MHz, and it now accepts 100KHz at full power for a marginally increased dissipation and cut-off current. Compensation becomes necessary in this case.
More elaborate improvement schemes are possible, but they involve high voltage small signal transistors, an annoyance that has so far been avoided.
Attached Images
File Type: png Cvac4.png (80.0 KB, 191 views)
File Type: png Cvac5.png (60.0 KB, 140 views)
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪
  Reply With Quote
Old 1st May 2013, 05:17 PM   #6
JMFahey is offline JMFahey  Argentina
diyAudio Member
 
JMFahey's Avatar
 
Join Date: Mar 2009
Location: Buenos Aires - Argentina
Sorry for being frontal, but what is this crazy schematic?

1) R5 is 100 ohms and has 150V across it, it dissipates (idle) 225 Watts

Only possible in "Simulation World"; in any real bench you would be opening all room windows by now to clear the smoke.

2) you did not include or even mention any biasing for the tubes, only show an AC signal, which in the absence of any indication to the contrary, is to be presumed referenced to ground.

Care to elaborate?
Thanks
Attached Images
File Type: gif Cvac2.gif (136.0 KB, 178 views)
  Reply With Quote
Old 2nd May 2013, 12:52 AM   #7
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
There are many other things that make a tube amplifier what it is that may not be adequately captured by this replacement circuit but I love the concept - it's clever!

Things that I like about tubes - the fact that they are often low power means that the system is put together differently than a SS amp. The speakers have to be more sensitive and this alone makes many tube systems sound different from solid state. Tube amps are often linear enough not to require feedback and this changes things too - not leas their response to speaker back emf. Do you plan to try your circuit in an equivalent system, low power with sensitive speakers, no feedback and a choke loaded power supply ???
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 2nd May 2013, 02:56 AM   #8
diyAudio Member
 
Join Date: Mar 2013
Location: Indianapolis
Quote:
Originally Posted by Bigun View Post
Tube amps are often linear enough not to require feedback and this changes things too...
Woah, hold on there. Are you sure? Most tube amps I've heard without feedback were far from linear...
  Reply With Quote
Old 2nd May 2013, 08:03 AM   #9
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
Quote:
Originally Posted by JMFahey View Post
Sorry for being frontal, but what is this crazy schematic?

1) R5 is 100 ohms and has 150V across it, it dissipates (idle) 225 Watts

Only possible in "Simulation World"; in any real bench you would be opening all room windows by now to clear the smoke.

2) you did not include or even mention any biasing for the tubes, only show an AC signal, which in the absence of any indication to the contrary, is to be presumed referenced to ground.

Care to elaborate?
Quote:
Originally Posted by Elvee View Post
..... and the 300V voltage used for the test is not realistic in an OTL context: I have kept the 300V value to allow for easy comparisons in a tube environment, but in the example, it leads to rather obscene power levels, almost 1kW peak in this case, much more with a speaker load..
Quote:
Originally Posted by Bigun View Post
Tube amps are often linear enough not to require feedback and this changes things too - not leas their response to speaker back emf. Do you plan to try your circuit in an equivalent system, low power with sensitive speakers, no feedback and a choke loaded power supply ???
No, at the moment I have built a test amplifier, an OTL in class B of ~70W.
The benefit of this circuit is precisely to get rid of the current limitations of normal tubes.
One could think oF a SE class A amplifier of the kind you describe, but with muscle enough to drive a standard loudspeaker directly and comfortably. That may be my next project.

Quote:
Originally Posted by iaxxaxxai View Post
Woah, hold on there. Are you sure? Most tube amps I've heard without feedback were far from linear...
The amplifier I have built has almost no feedback, because the open loop gain of the two ECC81 I have used (1 voltage amplifier, 1 phase-splitter and 2 in PP) is barely sufficient to get the full power output with 0.5Vrms in, yet the THD at 70W is under 0.5%. That would be difficult to achieve with semiconductors.
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪

Last edited by Elvee; 2nd May 2013 at 08:06 AM.
  Reply With Quote
Old 4th May 2013, 04:29 PM   #10
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
The circuits described so far are completely "orthodox", and create a simple and faithful copy of the tube they emulate (except for the magnified currents).

It is possible to be more "creative", and to take advantage of the possibilities offered by the additional circuit.

The possibilities are infinite, but I will provide two examples that are probably the most logical and useful in this case.

It is often necessary to create a grid bias, using a cathode resistor or a similar arrangement. This can of course be duplicated in the case of C-Vac, but there is an advantageous alternative: instead of applying the bias to the complete composite, it can be inserted in the cathode of the low power tube alone.
The functional result is identical, but the power wasted is much lower.
I have used this trick in the SolidGlass example.

Another interesting possibility is to use separate supplies for the tube and the high current section.
This is a logical step in the context of OTL amplifiers, because tubes do not like to be starved of voltage, but high supply voltages generate uncomfortably high levels of power with standard speaker impedance's.
The best way to reconcile these conflicting requirements is to split the two supplies.
I didn't use this option in SolidGlass, and even with a 16Ω load the dissipation is quite daunting....
Attached Images
File Type: png Cvacmod1.png (11.2 KB, 186 views)
File Type: png Cvacmod2.png (17.4 KB, 175 views)
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
230 VAC to 2*140 VAC transformer - 20VA gentlevoice Tubes / Valves 14 15th January 2013 01:22 PM
100 VAC - 240 VAC DC pwr. sup. MLStrand56 Power Supplies 10 12th January 2013 07:49 AM
SMPS 230 VAC to 115 VAC 800W TeeJay Power Supplies 88 22nd September 2011 07:30 PM
Sola Transformer 95-130 VAC In 118 VAC Out 60 VA 60 HZ Scott Novak Swap Meet 1 18th April 2009 10:24 AM
Convert 230 VAC amp to 120 VAC?? chuck55 Solid State 8 7th February 2006 01:22 PM


New To Site? Need Help?

All times are GMT. The time now is 11:39 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2