If you need assistance, please send an email to forum at 4hv dot org. To ensure your email is not marked as spam, please include the phrase "4hv help" in the subject line. You can also find assistance via IRC, at irc.shadowworld.net, room #hvcomm.
Support 4hv.org!
Donate:
4hv.org is hosted on a dedicated server. Unfortunately, this server costs and we rely on the help of site members to keep 4hv.org running. Please consider donating. We will place your name on the thanks list and you'll be helping to keep 4hv.org alive and free for everyone. Members whose names appear in red bold have donated recently. Green bold denotes those who have recently donated to keep the server carbon neutral.
Special Thanks To:
Aaron Holmes
Aaron Wheeler
Adam Horden
Alan Scrimgeour
Andre
Andrew Haynes
Anonymous000
asabase
Austin Weil
barney
Barry
Bert Hickman
Bill Kukowski
Blitzorn
Brandon Paradelas
Bruce Bowling
BubeeMike
Byong Park
Cesiumsponge
Chris F.
Chris Hooper
Corey Worthington
Derek Woodroffe
Dalus
Dan Strother
Daniel Davis
Daniel Uhrenholt
datasheetarchive
Dave Billington
Dave Marshall
David F.
Dennis Rogers
drelectrix
Dr. John Gudenas
Dr. Spark
E.TexasTesla
eastvoltresearch
Eirik Taylor
Erik Dyakov
Erlend^SE
Finn Hammer
Firebug24k
GalliumMan
Gary Peterson
George Slade
GhostNull
Gordon Mcknight
Graham Armitage
Grant
GreySoul
Henry H
IamSmooth
In memory of Leo Powning
Jacob Cash
James Howells
James Pawson
Jeff Greenfield
Jeff Thomas
Jesse Frost
Jim Mitchell
jlr134
Joe Mastroianni
John Forcina
John Oberg
John Willcutt
Jon Newcomb
klugesmith
Leslie Wright
Lutz Hoffman
Mads Barnkob
Martin King
Mats Karlsson
Matt Gibson
Matthew Guidry
mbd
Michael D'Angelo
Mikkel
mileswaldron
mister_rf
Neil Foster
Nick de Smith
Nick Soroka
nicklenorp
Nik
Norman Stanley
Patrick Coleman
Paul Brodie
Paul Jordan
Paul Montgomery
Ped
Peter Krogen
Peter Terren
PhilGood
Richard Feldman
Robert Bush
Royce Bailey
Scott Fusare
Scott Newman
smiffy
Stella
Steven Busic
Steve Conner
Steve Jones
Steve Ward
Sulaiman
Thomas Coyle
Thomas A. Wallace
Thomas W
Timo
Torch
Ulf Jonsson
vasil
Vaxian
vladi mazzilli
wastehl
Weston
William Kim
William N.
William Stehl
Wesley Venis
The aforementioned have contributed financially to the continuing triumph of 4hv.org. They are deserving of my most heartfelt thanks.
Hello all. I've been wanting to audio modulate one of my SSTCs for a while now, and I've really been wanting to try a Class A amplifier topology. I know it can be done via the enable pins by feeding a VCO with the audio signal and using the VCO output to FM modulate the interuption rate, and I may use that method in the future once I learn more about VCOs and PWM, but for now I would like to take a serious stab at Class A operation.
The SSTC I'll be using is a "standard" antenna feedback to 74HC14 to UCC 22/21 pair driving a mains fed halfbridge via a variac. I plan to do audio modulation with the coil running at less than 500VA input at all times, and most commonly probably far lower around 100VA for long run times. I know the coil and driver can handle this.
I haven't worked out how much or what cooling I'll need for the linear biased transistor, but I do have a transistor in mind: FDL100N50 from fairchild. It is a 500V 100A Nchannel MOSFET with 0.055ohm RdsON when saturated that can handle 2500W Pd at maximum. I know that I'll have to play around and find the optimal quiescent bias point for the gate as well as the optimal amplitude for the audio signal which will result in the lowest level of distortion. Anyway, ignore cooling method for now and let's talk about circuit design and expected parameters.
I have never actually built an audio amplifier before, unless you count a 2N222A transistor roughly biased somewhere in the linear region to make a buzzer louder, a couple of years ago before I really started studying EE. So, I would greatly appreciate it if someone would review my basic circuit design. Note; I left off most of the components of the TC side of things for simplification, they're all there in actuality and the coil is performing beautifully.
I apologise if it is a bit hard to read, I had to use my cell phone camera as my main camera is on loan to a family member. To the right of the linear amp is a bypass switch which turns on and off the audio modulation.
If I have everything bassically correct and this should work, how would I determine the power dissipation of Q3 (the linear biased mosfet)? The On resistance would be continuously varying with the audio signal, so I assume the power dissipation of the transistor would also be continuously varying, but how would I guestimate an average value, so that I can choose adequate cooling?
Registered Member #30
Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706
For a modulator circuit, see Nelson Pass's Zen Amp, or EVR's audio modulated Class-E coil if you can persuade him to show you the schematic The Zen amp uses a feedback resistor from drain to gate, to stabilise the DC conditions. EVR's modulator uses a high voltage amplifier driving the MOSFET as an emitter follower.
A Class-A amp is about 25% efficient, so you might budget for Q3 to dissipate three-quarters of the input power. The optimal bias point is (approximately) with Q3 dropping half of the supply voltage.
The DC bus bypass capacitors present a capacitive load to the modulator, and the size of them may need careful selection.
Registered Member #15
Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068
yeah, contact me offlist if you need any info / schematic for class-e or the plasmasonic coils i've done in the past. Plasmasonic is just a class-d audio modulator.
Registered Member #1232
Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881
If you want a high-level amplitude modulator that is significantly more efficient than a Class-A amplifier but not as complicated as a Class-D switching amplifier, the Class-H modulator is worth considering:
It is a linear modulator that uses a split rail supply and two power transistors. The increased efficiency is achieved because at the quiescent operating point one transistor is fully saturated and the other is in cutoff. With ideal transistors there is theoretically no dissipation at the quiescent operating point. It's kind of like a unipolar version of a Class B amplifier.
Registered Member #146
Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055
It is a linear modulator that uses a split rail supply and two power transistors. The increased efficiency is achieved because at the quiescent operating point one transistor is fully saturated and the other is in cutoff. With ideal transistors there is theoretically no dissipation at the quiescent operating point. It's kind of like a unipolar version of a Class B amplifier.
So you are basically dropping half the voltage you would otherwise, and your "zero audio" voltage is just what you get with the bottom switch saturated? Sounds like a good thing indeed.
Making a SSTC modulator with class A is going to be HARD, unless you plan on water-cooling the power transistor(s). Your devices 2500W Pdiss rating is assuming you can keep the tab at 25*C, and furthermore, 2500W sounds completely wrong for a fet with those ratings (250W would be more realistic, though some manufactures claim up to 500W for a leaded package). 2500W is about what the big IGBT bricks are rated to dissipate. Anyway, its pretty much impractical to get the max rated dissipation, its almost always cheaper/easier to use multiple transistors with less efficient cooling. Dissipating even 500W requires a rather huge heatsink (probably 6"x12" x 2" with a blower), and you might still burn out the chip if your thermal resistance is too much.
It is possible the datasheet is wrong and they mean less than 2.5kW, but I don't plan on pushing it hard anyway. If I can get it working with even as little as 100VA input, I'll be happy. If I had experience with using valves (tubes) I'd just grab as big of a triode as I could afford and do it that way, but I don't even own a filament transformer, haha.
Well, the main reason that I'm looking at Class-A supply rail modulation is the fact that I have no experience with PWM for anything other than motor speed control and I've had issues trying to modulate at the UCC chips. I'm not even sure if it is possible to accurately reproduce polyphonic life-like audio via varying the PRF of interuption rate on a SSTC - either via the UCC chips' enable pins or by pulsing the DC supply.
I noticed that the 120Hz hum from the DC supply ripple at high current levels was very "nicely" reproduced by the streamers and it gave me the idea to modulate the supply rail as a means of audio modulation instead of trying to work with the UCC chips. Trying to use the enable pins has given me trouble before; as the chips assume an ON state whether the enable pins are held high or left to float several early experiments showed that a 5V squarewave created by a simple astable 555 applied to the enable pins failed to modulate the UCC chips' output as both states of the squarewave resulted in an ON state of the UCC chips. I'm still not sure why that happened, but I assume I would need a square wave generator which not just supplies 5V pulses at a desired frequency, but also shorts the output to ground during output LOWs - too bad I don't know how to build that, haha.
Any alternatives or advice is always welcome and appreciated. If I can get some free time I'd like to try a mock up later today to see if I can even make a simple class A mosfet audio amp to drive a loudspeaker. If that works I'll give my smalled SSTC a shot on low power and see what happens.
Registered Member #30
Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706
Here's how modulation via the enable pins is supposed to work:
The usual setup is a full bridge gate drive using one UCC37321 and one UCC37322. The signal inputs are connected together, and the enable pins also.
When the enable pins are high, the outputs of the two chips assume opposite states. The GDT primary gets driven with +12V or -12V depending if the signal input is 1 or 0. (and assuming a 12V supply obviously :) )
When the enable pins are brought low, the two chips both drive their outputs low. Both ends of the GDT primary are forced to the same voltage, which causes 0V to be applied to all of the gates, turning all the power devices off.
Registered Member #1232
Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881
Hi Steve,
So you are basically dropping half the voltage you would otherwise, and your "zero audio" voltage is just what you get with the bottom switch saturated? Sounds like a good thing indeed.
Yes, this is the best ready-drawn schematic I could find online to help with an explanation:
When there is no audio present (the quiescent operating point) transistor Q2 is fully saturated and Q1 is fully off. This feeds exactly half the maximum supply voltage to the B+ rail of the RF amplifier. For negative excursions of the audio signal Q2 drops out of saturation and conducts progressively less current acting like a series-pass regulator and controlling the output voltage in the range 0 to 50V for this example. For positive excursions of the audio signal Q1 comes out of cutoff and conducts progressively more current. This forward biases D1 and controls the output voltage in the range 50V to 100V in this example. Now Q1 acts as a series-pass regulator and Q2 is fully saturated.
The Class-H modulator arrangement works quite well for audio modulation because audio signals tend to have a high peak-to-average ratio. Meaning that they linger around the quiescent point (0 volts) most of the time and only have occasional big peaks. This makes a linear modulator that is optimised for efficiency around the quiescent point much more efficient overall than a Class A modulator that is at its worst efficiency in this range!
Of course the state-of-the-art high-level modulator in modern digital AM transmitters is the polyphase Class-D modulator. This is really just a whole bunch of PWM modulated synchronous buck converters with their inputs and outputs connected in parallel and their switching instants staggered in phase. This paralleling distributes conduction and switching losses across many switches and chokes, and the phase-shifting makes filtering input and output current ripple much easier and minimises PWM carrier radiation from the antenna.
For an audio-modulated SSTC the Class-H modulator has the advantage that it is more tolerant of the varying load impedance reflected back by a modulated CW arc. The Class-D modulator involves an LC output filter which must be designed for the correct damping ratio into a given resistive load. This design is always going to be a compromise when driving an ampitude modulated CW SSTC, as the TC arc reflects back a varying impedance to the inverter, and this in turn presents a varying impedance to the LC output filter of the Class D modulator. The result is intermodulation distortion in the high-frequencies of the audio as the growing and shrinking arc modulates the damping of the Class-D modulator's output filter!
This problem is unique to audio modulated TCs as far as I know, and doesn't affect AM radio transmitters as much because the antenna always presents a well controlled and fixed impedance to the output of the transmitter regardless of the carrier level throughout a modulation cycle. This means that the load presented by the RF power amp back to the modulator is also more-or-less constant throughout the modulation cycle. So the design of the Class-D modulator output filter is much less of a compromise in this case.
Here's how modulation via the enable pins is supposed to work:
The usual setup is a full bridge gate drive using one UCC37321 and one UCC37322. The signal inputs are connected together, and the enable pins also.
When the enable pins are high, the outputs of the two chips assume opposite states. The GDT primary gets driven with +12V or -12V depending if the signal input is 1 or 0. (and assuming a 12V supply obviously :) )
When the enable pins are brought low, the two chips both drive their outputs low. Both ends of the GDT primary are forced to the same voltage, which causes 0V to be applied to all of the gates, turning all the power devices off.
Righto, I got that. In my tests I found that half the time the circuit would not oscillate at all, or would run CW and not modulate. I didn't have this issue in the very first driver board I made from Steve Ward's SSTC-5 though, its modulation worked perfectly. The only difference between that first board and all the others later on is that I started building modularly; the bridge, 555 modulator, and low voltage power supply are all on a seperate boards from the driver. The very first driver was an all on one board build. I hated working on such a cramped board though so it was the first thing I changed. Could this have caused a problem? The 555 modulator is pretty low frequency, so I don't think lead inductance is an issue.
After I began removing the 555 modulation circuits from the SSTCs which either would not oscillate or would not modulate I noticed something strange with my UCC37321 UCC37322 chips; if I have both enable pins connected together but do not connect the trace/wire connecting them to anything, both chips assume an OFF state and the driver does not oscillate. If I do not connect the two enable pins together, they both float at full 12V supply voltage and the driver oscillates. At the time I shrugged my shoulders and swore off using the enable pins.
I gave the issue some more thought today and think I may have a solution to failed modulation via the enable pins, let me know if this makes sense please. If I were to form a voltage divider with the 555 output at the input (of the divider) and the enable pins at the output, with apropriate resistances between input and output and between output and ground; a LOW state of the 555 should fail to bring the enable pins up FROM ground, instead of relying on the 555 to bring the enable pins TO ground. What voltage does the enable pin need to see in order to keep the UCC chips in an ON state?
UPDATE: 8/7/12 I built my little audio class-a amp and determined that I would need something crazy like TEC + watercooling to even approach 100W output. I did get my audio amp working though and found that the amp would not work as I drew it in my posted schematic - it only worked when the load was between Source and Ground, and not with the load between Drain and Vdd. So that's it for that then, heh. Guess I'd better research various enable pin modulation schemes that are in use now or go with something like Class D (E, F, G, or H) if I want to switch the supply rails.
This site is powered by e107, which is released under the GNU GPL License. All work on this site, except where otherwise noted, is licensed under a Creative Commons Attribution-ShareAlike 2.5 License. By submitting any information to this site, you agree that anything submitted will be so licensed. Please read our Disclaimer and Policies page for information on your rights and responsibilities regarding this site.
Class A audio modulation on a SSTC
The Zen amp uses a feedback resistor from drain to gate, to stabilise the DC conditions. EVR's modulator uses a high voltage amplifier driving the MOSFET as an emitter follower.
