Transformers and impedance question

[Marco Pancaldi]

Hi.

if I have a 200 ohm nominal out impedance passive mixer, connected to a step-up transformer with 1:4 turns ratio (12 dB gain), what is the minimum load impedance for the transformer output?

I'm not sure of my calculations, as follows: 1:4 turns will be 1:16 impedance ratio, so 200 X 16 = 3.2k. If I use a "matching" impedance load (3.2k input Z), I have 6 dB loss, and the total gain (+12dB step-up transformer AND -6dB transformer loading) will be 6 dB total.

To minimize voltage loss I will need a "bridging" (1:10) of minimum 32k.

Is it correct or I'm wrong?

many thanks
m.p.

 

[JohnRoberts]

Sounds right... some purists object to using dB to describe transformer turns ratios as the input power and output power are the same, but dB are too convenient for overall analysis to not use.

JR

 

[emrr]

Is the output of the passive mixer fixed at 200 ohms, or does it require a 200 ohm load to work correctly? If it's fixed, and truly presents a 200 source to the step up, you will probably be fine with the bridging load on the end. If it's not fixed at 200 then you may need a 3.2K load on the end. Are you sure you need to minimize voltage loss? Sometimes you don't. Lots of transformer gear sounds better matched.

 

[CJ]

"To minimize voltage loss I will need a "bridging" (1:10) of minimum 32k"


Everything is fine up to this point in the post.

Then I do not understand.

Why the xtra xfmr on the output of the 1:4?

The 32 K part is right, but I do not understand why you need another piece of iron?

 

[PRR]

> I have a 200 ohm nominal out impedance passive mixer, connected to a step-up transformer with 1:4 turns ratio

In this situation, in most cases, you want "unloaded" secondary, over 30K. The direct (no transformer) input of a transistor or tube amplifier. If it is a high-gain amplifier, try a 25K-50K pot and take turned-down signal from the wiper (or get a lower gain amplifier).

In some cases there is a ~~20KHz bump which is reduced by loading. Of course loading also reduces midrange. The trick is an R-C network to leave the midrange unloaded but damp the highs. Secondaries over 10K, if intended for very-flat operation, usually have such a network in the specifications. However a 150:2400, even a 600:10K, would not usually show any strong bump and not use an R-C unless you lived and died by 0.3dB errors.

Note that if you can get large levels into a passive mixer, and it accepts unbalanced output, a good transistor amplifier can take signal from 200 ohms with acceptably low noise and boost it any reasonable amount. For example: eight +4dBm sources, several faders on full, gives near 0.1V out of the mixer. If a clean transistor has 1uV hiss, you get 100dB from nominal level to hiss, plus whatever headroom above "+4dBm sources", easily 110dB peak to hiss ratio. Bringing 0.1V back to +4dBm needs only a gain of 12.3 or 22dB, not a big problem, hi-fi line amp.

Normally a 200 ohm passive mixer is worked into a Mike Preamp. Sources and faders are held lower so output is more like 20mV-50mV at the "mike" jack. Then the transformer ratio and loading are in the Mike Amp design, the mixer-out is similar to a microphone, it works without any calculations.

 

[PRR]

> > I will need a "bridging" (1:10)...."

> I do not understand.

I think Marco means "1:10 bridging impedance ratio". Like you have a 200R mike, you'd normally load it in 2K or higher. That "2K" may be marked "for 100 to 300 ohm source" but its actual 2K load means less than about 1dB of voltage loss. Or your CD player has <1K output, your hifi preamp has >10K input. A "bridging" load.

Indeed 200R and 1:4 tranny is 3.2K nominal and 32K to "bridge" that 3.2K for negligible voltage droop.