990 Question about output impedance

[Ethan]

I've been reading about the 990 recently.
In the specs it gives an input impedance, but not the output?
Determined by feedback?

How would you figure out what would be the output impedance that the output transformer sees at it's primary in the "microphone preamplifier" schematic in page four of John Hardy's App sheet:
http://www.johnhardyco.com/pdf/990.pdf

 

[CJ]

I'm sorry, but you have to be a bonefied member to ask a question here. :razz:

 

[Consul]

Bonefied? Is that like when you have a bon- Er, nevermind. :green:

Now that an entire two posts in Ethan's thread are wasted, let's see if he can get an actual answer. :wink:

 

[ciminosound]

It says it can drive loads as low as 75 ohms. For some reason I think the AES paper about it said 10 ohms or less than 10ohms or something. I don't seem to have it handy, will look.

 

[PRR]

> it gives an input impedance, but not the output?

What are you asking: load impedance, or source impedance?

The minimum load impedance for the original 990 is in the 75 ohm range. Someone posted a spec of 275mA peak current, and at 20V peak that's about 75 ohms. That's also two 150 ohm loads, a not-too-unlikely loading in 1970s studios.

The output source impedance, for any good voltage feedback amplifier, will be essentially "zero" compared to the load. The 990 is "good" (great) across the audio band but all good things end: in radio frequencies its output rises just as line or winding capacitive impedance falls. Jensen sells a magic "isolator" to decouple the load above the audio band, and you may need something like that with bifilar-wound transformers. You can use a 20-40 ohm resistor if you can stand a 20-40 ohm output impedance. The "isolator" bypasses the resistor with a coil so it is ~0.1 ohm across the audio band rising to 40 ohms for radio waves; if you use cheap iron this will lower THD.

 

[Ethan]

Thanks PRR.
I guess I was asking if there is an inherent (source) output impedance of the 990.

>The output source impedance, for any good voltage feedback amplifier, will be essentially "zero" compared to the load.

I gots more reading to do.

 

[Gus]

In PRRs post is something important

The output Z is only very low when the opamp is not at its current drive source or sink limits.

Things get messy with the feedback when the current need to correct the loop is not advailbe.

275ma is good drive look at the source and sink output drive curents of some opamps used in audio circuits.

 

[PRR]

> The output Z is only very low when the opamp is not at its current drive source or sink limits.

And if it IS at its limits, we are leaving the world of Hi-Fi and off to fuzz-land.

> if there is an inherent (source) output impedance of the 990.

WithOUT feedback, Zout of a 990 at the top of the audio band will be maybe(*) 100-500 ohms. But we would never pass audio without feedback: it would be much too loud and dull.

With a reasonable amount of NFB to set a good audio gain, Zout will be lower by the amount of NFB. For normal uses, 100 to 10,000 times lower. 5 ohms to under 0.1 ohms. For most loads, this is so very nearly equal to "zero" that we may say it IS zero. (We get a little close with high gains and very heavy 75 ohm loads.)

A lesser amp like 741 also has effective open-loop Zout near 100 ohms. At closed-loop gain of unity, we have 50:1 NFB at the top of the audio band, so it acts like 2 ohms. Minimum load on 741 is 2K, so Zout "is zero" for practical purposes. It degrades as we go to higher closed-loop gains: at gain of 50 Zout becomes 100 ohms but also the -3dB point is 20KHz and THD in the top octave is many percent. We don't see a lot of 741 in "good" audio, especially not at high gain.

(*)I think I made an error estimating open-loop Zout of 990. Does not matter. I've only considered the naked amp, but load enters into it too. Hi-Z devices like pentodes and OTAs, when working with large amounts of NFB, will have super-low Zout. A 6L6 transformed to an 8 ohm load has Zout around 80 ohms; put 20dB NFB around the amp and Zout becomes 0.8 ohms.