API 2520 informations

I did some research but i could not find anything about it...
what is the open loop impedance of this guy?
and also what is the maximum voltage you could apply to it without friying ?

thaks in advance.

Best,
Mattia.

Here you go:



I wouldn't recommend going beyond +-18V though. In most API applications, the supplies are limited to +-16V.

What do you mean by open loop impedance? That's not a meaningful term...

Samuel

output impedance?

thank you samuel for the data sheet!!!

Best,
Mattia.

It ain't my datasheet. :wink:

Can't find a trustworthy schematic right now to give detailed comments on output impedance; but IIRC the opamp uses a single emitter follower which will make the o/l output impedance quite high at low frequencies, probably around 1k. At some kHz it will likely fall to about 100 Ohm or a bit lower, then rise again in the MHz region.

But why do you care about its o/l output impedance?

Samuel

There is some discussion about accuracy of published schematics but the open loop source impedance looks like emitter follower with a few ohms in series. Should be adequately low Z.

JR

The emitter follower is driven by a high impedance collector (100k-ish), that's why the o/l output impedance likely ends up around ~1k (100k/hFE) at low frequencies and not just "very low". At higher frequencies the collector impedance is reduced due to local feedback of the compensation capacitor and the inherent follower output impedance becomes dominant.

Samuel

[quote author="Samuel Groner"]But why do you care about its o/l output impedance?
[/quote]

well I would like to hook it up to different trafos in order to "taste" the different flavours and impedance matching is my primary concern.


or should it be not...?

Mattia.

Sorry to mingle myself in this, but as Samuel & John actually already said or implicated, whatever the o/l output impedance is, it gets reduced by a massive factor (essentially the open loop transfer of the opamp * ), since without much doubt you'll be using some kind of negative feedback. So while the above stated 200,000 ('=' 106dB) isn't available for all frequencies, it'll be clear that the resulting 'small-signal output impedance' is pretty small.

Unless there are series resistors added, following the closed loop structure... unless they're very small these will usually doninate/determine the driving source impedance that's seen by the TX.

Regards,

Peter

*: possibly reduced by an amount equal to the closed loop gain, too lazy/hot here to think about that now, but in either case you'll have at least a whopping 50...60dB reduction of that o/l value.

What Peter said.

You don't need to match impedances; for flat frequency response you simply want to load (i.e. a resistor across the secondary winding) the transformer with the rated impedance (typically 600 Ohm for line transformers)--but there is no absolute need to do this. Typically you want to drive a transformer from a very low source impedance to keep distortion low, and an opamp provides just this.

Samuel

thank you guys,
quick and easy explanations,
really appreciate it!!!

Mattia.

> impedance matching is my primary concern.
or should it be not...?

"Matching" is for 1933, when every tube was expensive.

By 1966, transistors were cheap, gain was cheap. Then it works best to UN-match. Low-impedance outputs. High-impedance inputs. Multiple inputs on one output without re-matching.

The spec you need is "Minimum Load Impedance: 75 ohms".

This means you can freely connect 600R transformers. You can connect 150R transformers. You can even connect two 150R or eight 600R.

With nominal +/-15V supply the output can swing 11V peak. At 600R that is +20dBm which is plenty for running all over the house. If you use 150:600 iron you can do +26dBm, which is +8VU with 18dB headroom, suitable for long telco lines.

Transformers are source sensitive. You get one response with the near-Zero output on the module, another if you add 560R passive resistor in series with the transformer. Objective performance is generally worse; sound may have more flavor.

> what is the maximum voltage

The max +/-20V will increase the above levels 2.7dB. Really, if "x"dB is not enough, "x"+2.7dB won't cause great joy either.

And if THAT is not enough, you should stop fooling with pocket size modules and use a Power Amplifier.

> the o/l output impedance likely ends up around ~1k (100k/hFE) at low frequencies and not just "very low".

Yes, but in practical use it will be closed-loop. GBW is 50MHz, forward gain at 1KHz is 50,000. Closed-loop gain is likely under 100. Then Z(out) is reduced at least 500 times. It is surely just a few ohms. Say 2 ohms.

Minimum load is 75R. 2 ohms is "very small" compared to 75R. From no-load to full-load, gain sags to 0.974. 2.5% change. On a precision absolute-voltage measurement system, that's more than the width of the meter pointer. In real-world audio, 0.2dB is essentially equal to zero, and 2 ohms vanishes in cable resistance and general tolerances (75R 5% is 3.8R slop).

If Mattia decided that a certain iron needed exactly 543R source, he would want a 541 ohm resistor. Except I bet nobody could tell the difference 520R or 570R.