tief;
That was back in 2003
I don't recall the models, but you can do a search on their Legacy amp pages
I see. Perhaps you recall the Crown-specific term for their circlotron amp devices and the years of manufacturing. I have found only this:
Crown Amplifier Technical Information
Discontinued Amplifier Products
Or was it maybe a quasi compl. topology like this:
http://www.crownaudio.com/pdf/legacy/pl_three_schematic.pdf
Crest Audio : Schematics
It is possible to implement this topology using many types of output devices. For instance, Atmashpere uses it for their triode OTL amps.
James Bongiorno addition tot he knowledge pool regarding this topology is his patent that demonstrates how to automatically bias the output stage if BJTs are used in it. That being said, GENERALLY the bias for the output stages equals the voltage between the load midpoint and the driver ground. With this in mind it's possible to implement proper bias for any type of output device. For a MOSFET, a Vgs multiplier fed with a constant current would do the trick.
SITs and tubes are a bit more tricky. In theory, you need to compensate for power supply voltage variation, as both act as triodes - anode/drain voltage variations translate to anode/drain current variations at constant grid/gate voltage. If both output halves devices were perfectly equal, the variations would cancel out differentially. In reality, the differences translate into generation fo intermodulation components with the power supply ripple. This is somewhat less critical with tubes than with VFET/SITs just becasue tubes used generally have lower mu and gm. SITs also require the bias voltage to be present before the output stage power supply voltages appear. With tubes thi is mostly automatically true if the bias voltage is generated using semiconductor rectifiers (as indeed happens in 99.9% of cases) so it's present before the tubes warm up enough to conduct. With VFETs, worng power sequencing means dead VFETs, as simple as that.
James Bongiorno addition tot he knowledge pool regarding this topology is his patent that demonstrates how to automatically bias the output stage if BJTs are used in it. That being said, GENERALLY the bias for the output stages equals the voltage between the load midpoint and the driver ground. With this in mind it's possible to implement proper bias for any type of output device. For a MOSFET, a Vgs multiplier fed with a constant current would do the trick.
SITs and tubes are a bit more tricky. In theory, you need to compensate for power supply voltage variation, as both act as triodes - anode/drain voltage variations translate to anode/drain current variations at constant grid/gate voltage. If both output halves devices were perfectly equal, the variations would cancel out differentially. In reality, the differences translate into generation fo intermodulation components with the power supply ripple. This is somewhat less critical with tubes than with VFET/SITs just becasue tubes used generally have lower mu and gm. SITs also require the bias voltage to be present before the output stage power supply voltages appear. With tubes thi is mostly automatically true if the bias voltage is generated using semiconductor rectifiers (as indeed happens in 99.9% of cases) so it's present before the tubes warm up enough to conduct. With VFETs, worng power sequencing means dead VFETs, as simple as that.
http://www.diyaudio.com/forums/solid-state/23888-sumo-power-amp-4.html#post631753
Post 39 and 40 are what I'm interested in. I have a Sumo 9+ and was wondering if anyone replaced the stock op amp with what was suggested in posts 39 or 40? If so what were your results?
Also what other things would you do to update the stock amp? Someone else suggested this:
"There are a pair of low quality 0.1µF coupling caps per channel, replacing with a higher quality cap will really help. I would also consider adding a 22µF bypass cap on top of each of the four 10,000µF main filter caps."
Are these mods sound advice for a stock 9+ amp?
Post 39 and 40 are what I'm interested in. I have a Sumo 9+ and was wondering if anyone replaced the stock op amp with what was suggested in posts 39 or 40? If so what were your results?
Also what other things would you do to update the stock amp? Someone else suggested this:
"There are a pair of low quality 0.1µF coupling caps per channel, replacing with a higher quality cap will really help. I would also consider adding a 22µF bypass cap on top of each of the four 10,000µF main filter caps."
Are these mods sound advice for a stock 9+ amp?
I posed an a set of questions to a tech who rebuilds these amps and asked him some questions. This is how it went:
Question- "The stock fan is pretty quiet, but may be worth while to replace (it's old). It's a 120VAC type with a 390R series resistor."
His Response- The fans can be noisy but a bigger problem is that with time and heat they fail and the amp will run too hot if this is the case. The Nine Plus actually runs quieter than the original Nine. If it is getting noisy it is wearing out.
Q- "The Sumo is a push-pull class A amplifier with sliding bias, it increases its bias as the output increases. James Bongiorno received a patent for this circuit, and of course others have now built similar sliding bias circuits."
R- This is just a description of the biasing circuit which is technically not really correct as it is a true Class-a design. The point is that it is Class-A, it draws more current than a normal amp from the wall, runs warmer and it sounds good.
Q- "The opamp is an LM353(selected), and pretty long-in-the-tooth by todays standards. An OPA2604 is a drop-in replacement, and much better sounding. Others could be used, but they would have to be selected for high-voltage operation. "
R- The opamp can be replaced, although I get excellent results with the stock opamp. I'm listening to a rebuilt Nine right now and it sound quite good with the LM353.
Q- "There are a pair of low quality 0.1µF coupling caps per channel, replacing with a higher quality cap will really help. I would also consider adding a 22µF bypass cap on top of each of the four 10,000µF main filter caps. "
R- See the Sumo Nine board pictures on my site in the Gallery. There are around 10 ceramic disc capacitors per channel that inhibit the sound in there own way as there is a solid improvement in purity of the sound when they are replaced. A better approach if you are going through the trouble to add 22uf bypass capacitors to the power supply filter caps, since the power supply PCB has to be removed to get at the connections, is to reolace the main filters with the better sounding/high temperature rated caps available today. Sonically better than any bypass.
Q-"Does what he suggest sound like good advice for the 9 amp? Would you have any comments on mods to this amp yourself?"
R- You can see pictures of what I do on mysite GASaudio.net. Listed by level of improvement I'd replace the disc caps, next the speaker terminals are crap and always worn out and making poor connection, input connectors (some versions are worse than others), the main power supply filters, then replacing the resistors with 1/2 watt low noise metal film. Hopefully I have answered your questions. Please feel free to ask if any others come up. Mike
So does what he said sound easy enough to do and is it better then what was suggested before in my last post? Do his mods make good sense?
Question- "The stock fan is pretty quiet, but may be worth while to replace (it's old). It's a 120VAC type with a 390R series resistor."
His Response- The fans can be noisy but a bigger problem is that with time and heat they fail and the amp will run too hot if this is the case. The Nine Plus actually runs quieter than the original Nine. If it is getting noisy it is wearing out.
Q- "The Sumo is a push-pull class A amplifier with sliding bias, it increases its bias as the output increases. James Bongiorno received a patent for this circuit, and of course others have now built similar sliding bias circuits."
R- This is just a description of the biasing circuit which is technically not really correct as it is a true Class-a design. The point is that it is Class-A, it draws more current than a normal amp from the wall, runs warmer and it sounds good.
Q- "The opamp is an LM353(selected), and pretty long-in-the-tooth by todays standards. An OPA2604 is a drop-in replacement, and much better sounding. Others could be used, but they would have to be selected for high-voltage operation. "
R- The opamp can be replaced, although I get excellent results with the stock opamp. I'm listening to a rebuilt Nine right now and it sound quite good with the LM353.
Q- "There are a pair of low quality 0.1µF coupling caps per channel, replacing with a higher quality cap will really help. I would also consider adding a 22µF bypass cap on top of each of the four 10,000µF main filter caps. "
R- See the Sumo Nine board pictures on my site in the Gallery. There are around 10 ceramic disc capacitors per channel that inhibit the sound in there own way as there is a solid improvement in purity of the sound when they are replaced. A better approach if you are going through the trouble to add 22uf bypass capacitors to the power supply filter caps, since the power supply PCB has to be removed to get at the connections, is to reolace the main filters with the better sounding/high temperature rated caps available today. Sonically better than any bypass.
Q-"Does what he suggest sound like good advice for the 9 amp? Would you have any comments on mods to this amp yourself?"
R- You can see pictures of what I do on mysite GASaudio.net. Listed by level of improvement I'd replace the disc caps, next the speaker terminals are crap and always worn out and making poor connection, input connectors (some versions are worse than others), the main power supply filters, then replacing the resistors with 1/2 watt low noise metal film. Hopefully I have answered your questions. Please feel free to ask if any others come up. Mike
So does what he said sound easy enough to do and is it better then what was suggested before in my last post? Do his mods make good sense?
hi, i just bought this amp today from a dealer and it sounds pretty good. it's satisfactory, in fact. it's tad veiled though which i hope the op-amp swap will improve.
what kind of pcb does it use, btw? looks like antiquate fiberglass. do anyone know the technical name for the material?
what kind of pcb does it use, btw? looks like antiquate fiberglass. do anyone know the technical name for the material?
Last edited:
Some CROWN amps has the same...
..........Interesting. Please let me know some models - thank you.........
tief;
That was back in 2003. I don't recall the models, but you can do a search on their Legacy amp pages.
................I see. Perhaps you recall the Crown-specific term for their circlotron amp devices and the years of manufacturing..............
A technican by a Crown service center tell me, that Crown never used the circlotron topologySorry tief but I'm so away from that issue now, that I had cleared my RAM for other projects
Last edited:
Not without the risk of the occur of unwanted oscillation effects. I guess, several steps of re-design by compensation are necessary for optimizing the roll-off in the upper frequency aera respective the square-wave behaviour. The LME49860 (also 2xAD797) are much more better than the model "LF421CN".
For Dual-AD797 OP-AMP order ebay item No 220677386848
For additional redesign ask Mr. Scott Wurcer (diyaudiomember) - go to
http://www.diyaudio.com/forums/members/scott-wurcer.html
Also a discrete OP-AMP solution could be very interesting.
This would be much more easy for me, because I can investigate the behaviour for stability on the simulation aera by CAD software (not really possible by integrated OP-Amps).
Here a thread about this:
http://www.diyaudio.com/forums/soli...screte-clone-comparison-seven-variations.html
In the attachement the schema from the death links by post #2 and post #21
Last edited:
Interesting simple design of A class Hybrid Circlotron (PPP) by Audiodesign : http://www.diyaudio.com/forums/tubes-valves/178536-hybrid-circlotron-amplifier-only-3-components-signal-path.html
"does it have to be a fet-input opamp? "
Not at all, but remember that the input capacitor and resistance values are designed for a very high input impedance, and whatever you use must be able to work with the high voltage involved (the LF 353 was selected). You are in effect redesigning the amplifier and compensation and bias will also need to be examined.
I think changing out the ceramic caps will do more for the sound quality than changing the opamps. I chose the OPA 2604 on the basis that it has high input impedance, can run on the higher voltages without having to be selected, and I like the sound. JB suggested that he liked the AD 823 better, but that you need to select it for the higher than the spec sheet voltage used in the Sumo.
Not at all, but remember that the input capacitor and resistance values are designed for a very high input impedance, and whatever you use must be able to work with the high voltage involved (the LF 353 was selected). You are in effect redesigning the amplifier and compensation and bias will also need to be examined.
I think changing out the ceramic caps will do more for the sound quality than changing the opamps. I chose the OPA 2604 on the basis that it has high input impedance, can run on the higher voltages without having to be selected, and I like the sound. JB suggested that he liked the AD 823 better, but that you need to select it for the higher than the spec sheet voltage used in the Sumo.
Hi Ultimatex86,
The diagram in post #64 is conceptual, but should basically work. For an actual build, I would cascode the front end JFETs so they see less than half their max Vds rating, or just replace them with small-signal MOSFETs having a higher Vds rating.
It's also critical that the output of the bias current source drop slightly with rising temperature to keep the output MOSFET bias stable. And I would include some form of output protection, if only to add rail fuses to each floating power supply.
Of course, I would also suggest that you to get input from someone who has actually built this type of circuit. I always seem to learn something important from building (and listening) that didn't surface on paper or in simulation.
I have good result with 2SK1058, but I would like try with IRF, one is enough for 100W 4ohm
http://www.diyaudio.com/forums/solid-state/261515-vssa-circlotron-7.html#post4441226
http://www.diyaudio.com/forums/solid-state/261515-vssa-circlotron-7.html#post4441226
Last edited:
I'm not sure how well behaved these circlotrons are, but more conventional op-amp input designs cannot use an OPA2604 as a drop in for an LF353/412. They work fine on a light bulb limiter or variac at low power, but as soon as you drive it to clipping the op amp blows out. And about half the time the turnon transient will take it out. Then it puts DC to the speaker of course. They just don't like being driven beyond the rails or with excessive differential voltage - even for an instant - and even diode clamps don't help. I finally had to give up on the 2604 in those applications. A 353 will take repeated input overvoltage as long as there is a few k ohms to limit the current. Even the full brunt of a plus or minus 140 volt rail thru a 47k feedback resistor. Try that with a 2604 and you'll be putting in another, and another and another.....
What do you think about using LME39830 with mosfet ?
Maybe a 10 or 11 with bipolar transistor ...
Maybe a 10 or 11 with bipolar transistor ...
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.