Opamps for servos in tube amps - Christmas gift from LT, 12-12-12!

[Wavebourn]

Hi!

I got e-mail from LT today, looks interesting.

Since the chip draws low current, and slew rate is not very impressive, I would hesitate to spoil tube signal path by them, but it is a great candidate to be used in servos, when vacuum tubes are directly coupled, like in my Pyramids. The opamp has MOSFET inputs, and can run from up to 140V!

LTC6090 - 140V CMOS Rail-to-Rail Output, Picoamp Input Current Op Amp - Linear Technology

I ordered samples, it will be a great Christmas gift!

 

[TheGimp]

Interesting, but care must be exercised for the thermal pad.

 

[gingertube]

Quote:
"I got e-mail from LT today, looks interesting".

Ditto,
I posted the same link over at Ampage. You could use it for output tube drive (if its not against your religion) not just bias servo (10mA source/sink capability). Interest there was also in the simple 100W MOSFET Amp shown on p 19 of the datasheet. Only available in surface mount which is a bit of a bummer.
Cheers,
Ian

 

[Wavebourn]

My religion does not allow to use class B and 11V/uS opamps to drive tubes.

 

[ruben alstad]

i allways wondered about servos in tube amps,
but i had the bias in mind, a servo controlling a pot for the bias (fixed bias).
has anyone tried this?

 

[DF96]

Yes, people have tried this. If done the wrong way it acts like a CCS, fixing the average current instead of the quiescent current.

 

[Wavebourn]

I did that in class AB amp, with S/H and comparator: when signal level was above some level it switched off and kept voltage unchanged. When below, it opened the path closing servo loop.

 

[ruben alstad]

Yes, people have tried this. If done the wrong way it acts like a CCS, fixing the average current instead of the quiescent current.

yeah of course, it would have to be a auto bias button or something alike, when there were no signal on the input you could press it and thus it would re bias itself. that's simpler in my opinion.

 

[DF96]

Wavebourn's method is better than a button. A button would require a very long-term hold (hours), or at least separate arrangements every time you switch on. At most his method need only remember the voltage for the length of a CD.

 

[Wavebourn]

Wavebourn's method is better than a button. A button would require a very long-term hold (hours), or at least separate arrangements every time you switch on. At most his method need only remember the voltage for the length of a CD.

Less, actually: it had shorter time constant on soft volume than on high volume. No need for silence, about 5W for 100W amp is a good threshold to switch from 5 seconds to 300 seconds approximately. On 5W power idle and average currents are almost the same.

 

[DF96]

My thought was that if someone had CDs with continuous compressed pop with dynamic range of about 1dB then the bias circuit would not be able to do anything until the CD finished (74 mins max?). Real music of course has much more variation so plenty of opportunity to adjust bias.

 

[Wavebourn]

My thought was that if someone had CDs with continuous compressed pop with dynamic range of about 1dB then the bias circuit would not be able to do anything until the CD finished (74 mins max?). Real music of course has much more variation so plenty of opportunity to adjust bias.

I don't think for such music precise bias matters. Class B will be fine, no sounds decay.

 

[DF96]

For some music Class C might be sufficient!

 

[palmas]

For some music Class C might be sufficient!

I know I'll be using it to have enough voltage from my function generator to drive tubes for testing...

 

[geraldfryjr]

I got that same email !!!

I think of lots of great stuff for that opamp since National Semiconductors version has been no longer available for many many years !!!!

Cheers !!

jer

 

[geraldfryjr]

I got a New Email Today!!!

I see LT is now making a dual version of the LTC6090 it is called the LTC6091!!!!

http://cds.linear.com/docs/en/datasheet/6091f.pdf

http://www.linear.com/solutions/469..._source=transactional&utm_campaign=LT_Insider

jer

 

[gingertube]

Typically bias servos for Class AB clamp the current monitor signal at 2 x the idle current before averaging.
Menno/Guido do this a bit differently. They use more gain and a much narrower clamp window about the desired operating point.
Like this:

and Rob did his own variation here:
http://rmsacoustics.nl/rmsacoustics/tubeamp/tuba_bias-circuits.html

Cheers,
Ian