END200 who knows this design ?

[WolfgangL]

i got two amp modules where all TO3 outputs are mounted on a heatsink WITHOUT any isolator. The case of TO3 is collector, so all collectors shorted to ground. Will this work ?

[WolfgangL]

please excuse the bad quality. The OP ist TL072, drivers are MJ150030/31, outputs are MJ15003/04

[Mariodas]

Well, if schematic is correct, this is unusual amplifier.
Do you have more clear schematic image ?

[ilimzn]

This is a rarely used topology, but it is in fact grounding the collectors and swinging the whole (floating) power supply as the output signal. This does complicate things regarding the power supply of the driver, which you can see in the schematic. Still, it would help if the drawing was clearer (without the circled components).

[lineup]

Originally posted by WolfgangL
i got two amp modules where all TO3 outputs are mounted on a heatsink WITHOUT any isolator. The case of TO3 is collector, so all collectors shorted to ground. Will this work ?

TO3 case is = Collector
TO3 have no Pin for Collector
TO3 have B and E pin, only. Only TWO pins


So if output is Common Collector
and you have 4 TO3 power transistors at heatsink
all their Collectors are Connected.


This makes you need no Insulator Bricks ( isolator )
but you have to be careful
and make sure HEATSINK is isolated from GROUND
or any other junction in the amplifier!

Because
if heatsink, common collector output,
in contact with GROUND or other point
= SHORTCUT

[WolfgangL]

i dont have the scheme on paper only as file, so i tried with MS paint to edit. Hope it will help

[WolfgangL]

sorry the scheme

[ilimzn]

Originally posted by lineup
...if output is Common Collector
and you have 4 TO3 power transistors at heatsink
all their Collectors are Connected.... make sure HEATSINK is isolated from GROUND
or any other junction in the amplifier!... Because
if heatsink, common collector output,
in contact with GROUND or other point
= SHORTCUT

Ehhhh.... lineup, my friend, look at the schematic before posting Otput is indeed common collector, and in this topology, it MUST be tied to ground.

If you think about it, collectors are also coupled in a regular topology, it's just that they are coupled through x volts of a power supply, while emitters are coupled through zero volts (simplified). In this design, emitters are coupled through x volts and collectors through zero voltas. Actually, you could theoretically couple all of themeither way as long as you observe that there is the necessary x volts in the complete circuit through transistors, load and power supplies, and of course that the load sees no AC. For instance, in a circlotron amp (like Sumo), each emitter is coupled to each oposing collector through x/2 volts of the total power supply. It's really all the same thing, just the places of the components change around the complete output circuit. It is even possible to make a bridged amplifier in the same manner, see 'grounded bridge' topology.

This particular amp runs the output transistors as common C (i.e. followers) but the positions of the power supply and the load are reversed. This means that:
1) The output is the center point of the power supply (which in this design is derived by AC coupling through the filter caps, note single winding transformer with no center tap).
2) The common collectors are the ground reference point, so NO insulation here, it MUST be grounded. If not, it won't work, perhaps destructively.
3) Like on the more common topologies, the output voltage is developed on the emitters, but in order to compensate for the grounding of the collectors, the power supply has been moved around so that the load sees no DC. In effect, the power supply is 'floating' WRT ground, and the output transistors 'move' it up and down WRT ground by the amopunt of output voltage - hence developing the output, which is why it is not coupled to the ground node at all.
4) This presents a problem for the input stage since it still has to be powered in a more or less conventional manner. because of this, the emitters of the outputs have diodes that are used to suypply power to the driver stages, as the 'moving' of the supply requires additional rectification to get ground referenced power. normally, this would be supplied by an extra winding on the power transformer, but the designers decided to cut a corner or two here.

The whole design is intriguing, and as far as i can surmise, the reason for it was to provide very good reliability, without sacrificing too much to complexity. Think about it: you can mount output transistors without mica, potentially reducing thermal resistance from output transistors to heatsink to a minimum. Also, the AC coupling prevents speakers from blowing if the output fails.

[Mariodas]

Are there any sonic advantages with this design ?
What is power output of this amplifier ?

[lineup]

Originally posted by ilimzn

Ehhhh.... lineup, my friend, look at the schematic before posting Otput is indeed common collector, and in this topology, it MUST be tied to ground.

I was fully aware of the unusual circuitry here. With Grounded Collectors.
It is such a strange creature, I wont even bother to understand the benefits
... hard enough even with a perfect schematic, I would think.


My comment was more of a general advice.
When we can use no isolators for TO3,
is in 49/50 cases,
when common collector point used as output positive terminal.

And when and if we do, be warned!,
we have to insulate heatsink instead
.. to avoid risk for SHORTCUT Disaster


lineup