I've got a technique for building prototype output stages for experimentation and testing with solid-state amps that seems to work quite well, so I thought I'd share it - nothing very earth shattering, but seems to work quite well.
Using stripboard and a chunk of aluminium as a heatsink/thermal mass. I tend to put output devices, drivers, any associated caps/resistors on the board, sometimes the whole bias generator, sometimes just a thermal compensation transistor. The connecting wires are solid-core suitable for direct breadboard insertion.
(note the thermal pads are just kapton tape with a little thermal paste on top!)
This is the latest, an EF output stage using MJE15030/1, NJW3281/1302's.
I've done this also for CFP stage, lateral MOSFET stage, and a quasi-comp IGBT stage.
Some of them included Zobel and inductor, but I've a separate board for that now.
This makes it easy to experiment with amplifier topologies on a breadboard, firstly without the OS, then tie in an OS once that's satisfactory. Less chance of making mistakes (although adjustable current limited supply is useful for that).
I'm thinking of making some little Vbe multipliers that plug into a breadboard too, and generally modularizing to make it easier to play with basic topologies.
For a finished design I would always layout a PCB, this isn't a standard of construction I'd like in a finished amp(!)
Using stripboard and a chunk of aluminium as a heatsink/thermal mass. I tend to put output devices, drivers, any associated caps/resistors on the board, sometimes the whole bias generator, sometimes just a thermal compensation transistor. The connecting wires are solid-core suitable for direct breadboard insertion.
(note the thermal pads are just kapton tape with a little thermal paste on top!)
This is the latest, an EF output stage using MJE15030/1, NJW3281/1302's.
I've done this also for CFP stage, lateral MOSFET stage, and a quasi-comp IGBT stage.
Some of them included Zobel and inductor, but I've a separate board for that now.
This makes it easy to experiment with amplifier topologies on a breadboard, firstly without the OS, then tie in an OS once that's satisfactory. Less chance of making mistakes (although adjustable current limited supply is useful for that).
I'm thinking of making some little Vbe multipliers that plug into a breadboard too, and generally modularizing to make it easier to play with basic topologies.
For a finished design I would always layout a PCB, this isn't a standard of construction I'd like in a finished amp(!)
Last edited:
Very wise. One of my own early designs used strip board.
I accidentally shorted the output and it vapourised the ground line on the strip board before the fuse blew.
With the likes of JLCPCB etc PCB's are very cheap now.
Just watch out for cheap PCB's but then they add on extra to the postage fee.
Well naturally, but this is an output section, the stray capacitance is totally dwarfed by the device capacitance by several orders of magnitude I think.
I remember when I started out to build and test Amplifiers, I lived in Asia that time, in Thailand, and I took a 3mm thick wooden Board and drew the schematic in bigger size on to the Board with a Magic Pen.. this was back in the early 70's.
After that I took Brass Nails and "hammered" them into the board at the place of connection or where ever a Components Wire would end.. this just for none heating components.. After that connecting the Nails with wires and soldering the components to the nails.. it worked quite well. also was easy to change components.
After that I took Brass Nails and "hammered" them into the board at the place of connection or where ever a Components Wire would end.. this just for none heating components.. After that connecting the Nails with wires and soldering the components to the nails.. it worked quite well. also was easy to change components.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.