Revisiting some "old" ideas from 1970's - IPS, OPS

[jwilhelm]

I finally managed to come up with a layout for one channel of the CFA/VFA combination design. This will still mount flat nicely on a 6" x 12" heatsink. One more channel to go.

 

[vzaichenko]

I finally managed to come up with a layout for one channel of the CFA/VFA combination design. This will still mount flat nicely on a 6" x 12" heatsink. One more channel to go.

Nice! We will probably add 2 small capacitors to CFA option - I will let you know later today.

 

[vzaichenko]

Jeff, please check that C24, C25 are in place - see IPS schematic here:
VERTICAL CFA IPS + NS-OPS

Those caps can be 10pF or 15pF - does not matter. But they need to be there (eliminating probability of light ringing at high swings).

I was just afraid they could be removed in the Diptrace file I've sent you earlier. Schematics in the thread are all ok.

 

[jwilhelm]

Mirror image twins (almost)!

 

[vzaichenko]

Serious stuff
Looking forward to testing

 

[jwilhelm]

Me too, but customs seems to still be admiring my heat sinks.

 

[vzaichenko]

Me too, but customs seems to still be admiring my heat sinks.

Customs likes them

Last year I have received a big compartment from the Italian manufacturer - it came to me very quickly. Went through customs in 2 hours. So, it was interesting for me to see what the specified in the documents - what they actually called it.

"Empty metal box"

 

[jwilhelm]

I've got two CFA versions ready for testing, once I figure out how to set the bias.

 

[vzaichenko]

Jeff, they look fantastic!

For your reference - you can download the bias setting guide here:
Download build instructions

 

[still4given]

Hi Valery,

Will the NS-OPS work with MJL41194/93?

 

[vzaichenko]

Hi Valery,

Will the NS-OPS work with MJL41194/93?

By the way! It will.
Just be careful when testing with a square wave - don't go above, say, 100KHz. They close really slowly, comparing to modern outputs.

 

[still4given]

He, he, I can barely hear 10khz. I'll give it a try with a pair before I load it up. Thanks

 

[damon1983]

He, he, I can barely hear 10khz. I'll give it a try with a pair before I load it up. Thanks

MJL3281+MJL1302 have much better sound than MJL21194+93.

 

[vzaichenko]

There will be no difference in sound. In a proper A/B blind test you will never notice any difference in such a design.

 

[AndrewT]

MJL3281+MJL1302 have much better sound than MJL21194+93.

The design has to suit EITHER device combination.
Rarely can one just swap devices.

 

[vzaichenko]

The design has to suit EITHER device combination.
Rarely can one just swap devices.

This is actually right. In most cases, the amp is designed for particular output devices, including compensation scheme. So, if someone gust swaps the devices, the whole thing may start ringing/oscillating.

For this particular case - MJL21194/93 will be ok, although when using them in such high-speed designs there's always some risk of trouble at high frequencies.

 

[damon1983]

There will be no difference in sound. In a proper A/B blind test you will never notice any difference in such a design.

ok. Maybe to use KT864+KT865?..

 

[jwilhelm]

I've got one channel up and running on a 63V supply. Sounds great so far. The 60hZ hum isn't there with open input either. This one's dead silent!

 

[vzaichenko]

I've got one channel up and running on a 63V supply. Sounds great so far. The 60hZ hum isn't there with open input either. This one's dead silent!

Excellent! Is your input lifted or not (10R or jumper)?

 

[jwilhelm]

The input is grounded. 0 ohm resistor.