100w amplifier with a single pair of 2SC5200 and 2SA1943

Partiendo de una potencia de +-45 V, me gustaría armar un amplificador de 100 w con un solo par de 2SC5200 y 2SA 1943. ¿Me pueden recomendar un esquema fiable? Adjunto algunos esquemas que he encontrado de los cuales no tengo opiniones. Serían con solo un par de transistores de salida. Gracias.
 

Attachments

  • 200w-transistor-amplifier-circuit-diagram.png
    200w-transistor-amplifier-circuit-diagram.png
    49.9 KB · Views: 759
  • 420020301_780307550785823_8968404361649064907_n.jpg
    420020301_780307550785823_8968404361649064907_n.jpg
    78.4 KB · Views: 641
  • 1600992434718822-1.jpg
    1600992434718822-1.jpg
    50.5 KB · Views: 595
"Starting from a power of +-45 V, I would like to put together a 100 w amplifier with a single pair of 2SC5200 and 2SA 1943. Can you recommend a reliable scheme? I attach some schemes that I have found of which I have no opinions. They would be with only a couple of output transistors. Thank you."

Attachments​


1. A single pair of C5200/A1943 is asking for failure with a ±50V supply.
2. Tip41/42 are too slow for drivers.
3. Separate 100 Ohm resistors to turn off the outputs is too slow and asking for shoot-through current failure. They should be cross-coupled.
4. While VI protection can cause issues, you should at least have current limit protection.
5. A current mirror on the LTP is better for turn-on/off thumps as well as general performance.
6. Especially with a current mirror, the LTP needs to be degenerated to reduce slew distortion.
7. A Darlington VAS improves THD. See Douglas Self's web site.
8. The feedback decoupling cap should be voltage clamped.
...
 
"Starting from a power of +-45 V, I would like to put together a 100 w amplifier with a single pair of 2SC5200 and 2SA 1943. Can you recommend a reliable scheme? I attach some schemes that I have found of which I have no opinions. They would be with only a couple of output transistors. Thank you."

Attachments​


1. A single pair of C5200/A1943 is asking for failure with a ±50V supply.
2. Tip41/42 are too slow for drivers.
3. Separate 100 Ohm resistors to turn off the outputs is too slow and asking for shoot-through current failure. They should be cross-coupled.
4. While VI protection can cause issues, you should at least have current limit protection.
5. A current mirror on the LTP is better for turn-on/off thumps as well as general performance.
6. Especially with a current mirror, the LTP needs to be degenerated to reduce slew distortion.
7. A Darlington VAS improves THD. See Douglas Self's web site.
8. The feedback decoupling cap should be voltage clamped.
...
Would this design fit the characteristics you mentioned?

Thank you
 

Attachments

Would this design fit the characteristics you mentioned?
And likely few hundred more versions.
Very common topology.
People would change few things here and there.
But for the most part would work.
Or a final design would be similar.

Has VI limiters
which many would leave out.
Depends on application.
Few strange values here and there.

Sure others will post more eventually.
 
  • Like
Reactions: rafaelct
The design you are proposing is a EF3 output stage. Is this an existing design with tested PCBs? If not, do you have experience biasing and stabilizing a EF3 output stage? It can be tricky even for experienced designers.

Have you considered the Wolverine? It's a EF3 that have been tested with 2SC5200/2SA1943. In its EF3 3-pair configuration with 49V rails, it's rated for 100W into 8 ohms. This seems to be what you are looking for and is a proven design with excellent support and high performance.

If you really want something you can design yourself, I'd suggest starting with an EF2.
 
In my design each module already has an assigned space. Power supply, preamplifier and power stage. Since my stage idea was with a single pair of 2SC5200-2SA1943, I installed a radiator with the two final transistors. I no longer have the possibility of placing another radiator, so the best thing would be to put two more transistors on that same radiator. But I already have the power supply so I must take advantage of it. That is why I was asking you specifically about that type of amplifier. Thank you
 

Attachments

  • WhatsApp Image 2024-07-04 at 17.44.10 (2).jpg
    WhatsApp Image 2024-07-04 at 17.44.10 (2).jpg
    56.3 KB · Views: 104
  • WhatsApp Image 2024-07-04 at 17.44.10.jpg
    WhatsApp Image 2024-07-04 at 17.44.10.jpg
    153 KB · Views: 98
I don't see a path for a ±45V supply and a single pair of 2SC5200-2SA1943 without a risky SOA. I suspect your best option if you want to retain the more expensive items (chassis and PSU), is a design with a pair of lateral mosfets. These should be fine with a ±45V supply.

Mooly's CFP lateral mosfet singleton might fit your needs. It's runs from one pair of lateral MOSFETs on a ±45V supply and is good for around 100W into 8R. It's also a proven design with many existing PCB variants: My MOSFET amplifier designed for music
 
  • Like
Reactions: rafaelct