NE5534 Bal/Comp Pins?

[Tenson]

Hi Chaps,

I'm wanting to make a circuit with NE5534 chips (directly swappable for AD797?) and I just wanted to ask what I need to do with the Bal and Comp pins? If not used, do I just leave them floating?

The application is a differential input buffer with a gain of about 1.3x I know it needs a cap on one of the pins to stabilise it at low gain levels, is this just bypassing the feedback arm or is it on one of these additional pins the 5534 has compared to the 5532?

 

[mchambin]

At this low gain, near unity, you need the cap at the compensation pins ( 8 - cap - 5 ).
When offset triming is not required, the adjust pin is left floating ( 1- unconnected )

 

[Tenson]

Thanks. I should mention that even though gain is near unity, it does have 4 1k resistors around it, so does it still need compensation or is that only needed if you have a direct link on the feedback?

 

[mchambin]

Compensation, only has to do with the gain value, whatever resistors used to set the gain..

The NE5534 is unity gain stable with 22pF. With no cap it is stable at gain > 3.

Have a ook here: http://www.diyaudio.com/forums/solid-state/73827-ne5534-compensation.html

 

[Tenson]

I thought that would be the case, thanks for clarification. One last question if I may! You said the NE5534 needs the compensation cap between pins 8 and 5, while in the AD797 datasheet they show it between pin 8 and 6 (output). Does this mean they are not directly swappable?

I could of course make a small jumper to choose if pin 8 is cap coupled to pin 5 or pin 6.

 

[mchambin]

From the AD 797 NE5534 data sheet, I see: The exta pins 1, 8, 5 . are not compatible.
NE5534 cap between 8 and 5
AD797 cap between 8 and 6 but not doing the same ( it is unclear to me what it actually does ) and not the same values.

 

[dchisholm]

. . . You said the NE5534 needs the compensation cap between pins 8 and 5, while in the AD797 datasheet they show it between pin 8 and 6 (output). Does this mean they are not directly swappable? . . .

No, they are not directly swappable (at least not with predictable results) if external compensation is used. That is the general case for most opamps when external compensation, or external offset adjustment, is used. That's because the internal circuit details for how the compensation or offset adjustment works are different for different opamp circuits.

(That's also why you almost never see either the offset or compensation functionality implemented in opamp SPICE models: since there is a wide variation in how the effects are accomplished internally it is difficult to impossible to include accurate models of the effects in the standard Boyle macromodel or its derivatives.)

Dale