stability issues output triple and high freq extension ?

[hypertiger]

I've seen member thimios test 200khz square waves , the only burn out
was by using "slow" output devices (21193/4's).

Could you please share more details on the test? Was it standalone test for EF3 w/o any kind of slew rate limit? any info on load, power supply?

Thanks in advance.

 

[sumaudioguy]

Did I miss the schematic of Vzaichenko's "21'st century?" Is that schematic somewhere here? Can you link it?

 

[ostripper]

Could you please share more details on the test? Was it standalone test for EF3 w/o any kind of slew rate limit? any info on load, power supply?

Thanks in advance.

Both thimios and Vzaichenko fried OPS devices using unlimited slew CFA
stages running through an EF3 with low Ft power devices.

Vziachenko characterized the failure as "cross-conduction".
-8R load
-50V rails
Sumo - 21'st century schema below , worked every day since early
2015.

OS

 

[vzaichenko]

Exactly - I stopped using 21193/4 as my front-ends are fast and they can kill the OPS if its closing time is not good enough

21'st century control board - Jeff Wilhelm just mentioned he's got some spare new PCBs >HERE<

 

[rhythmsandy]

so we cant use slower devices to a fast VAS stage..

 

[thejohn]

So if one was limited to something similar to 21193/4 or say a TO-3 EF3, how far could one push the IP/VAS design before running into problems? Would all of the VFA Slews be safe?

 

[ostripper]

So if one was limited to something similar to 21193/4 or say a TO-3 EF3, how far could one push the IP/VAS design before running into problems? Would all of the VFA Slews be safe?

Blameless(wolverine) , symasui are 40-80V/uS depending on miller cap
(39-68pf). Might not push a slow OPS too hard.
Edit - spooky leach might be too much (>125V/us).
OS

 

[thejohn]

I do like the Symasui idea. The follow up I guess being if the T03 outputs are mounted to a heatsink, wired back to the pcb would the stability problem be likely to get worse due to parasitics?

 

[hypertiger]

Both thimios and Vzaichenko fried OPS devices using unlimited slew CFA
stages running through an EF3 with low Ft power devices.

I see, quite predictable.

 

[ostripper]

I do like the Symasui idea. The follow up I guess being if the T03 outputs are mounted to a heatsink, wired back to the pcb would the stability problem be likely to get worse due to parasitics?

I'd put basestopper's right on the TO-3's pins. I've ran devices for test out on
wires from the OPS's before.
Would depend on how long and what those wires were near.

OS

 

[wahab]

Provided they are adequatly stabilised triple EFs have a huge advantage over double EF, often an amp will perform much better at about 0$ added cost.

Also the MJL2119x are nothing else than beta binned MJ1502x in a different package...

 

[mlloyd1]

hmmmm .... interesting observation.
tomlinson holman made good use of mj15024/25 in the EF3 used the apt 1 many years ago. the write-up on the design in electronics design from 1980 or so is worth reading if one can find it.

mlloyd1