The GAP in my slow coming knowledge...

Perhaps the biggest gap in my (still very little) electronics knowledge is audio transformers. Could someone help me understand this seemingly simple question that everyone but I seems to know...

How does one determine the kind of transformer necessary? I'm fine with Ohm's law in application, I'm just not sure what figures I'm supposed to be working with with a tranny. More specifically, if I wanted to add a transformer to an unbalanced circuit what are my calculations to consist of? As well as, adding a transformer to an opamp balanced circuit...

Thanks!
-E

> what figures I'm supposed to be working with with a tranny.

Impedance in, impedance out, frequency response, maximum level at lowest frequency of interest.

It's pretty simple if you break it down into catagories.

1. 600:600 Bridging
2. Mic to tube grid: 1:10 ratio which equals 600:60 K impedance
3. Innerstage: (BA6a or Pultec EQP) 1:2 turns ratio
You want to go from Plate to Grid, which would mean about a 20k (plate) to 80k (grid)
4. Output transformer: 4:1 turns , which might mean 10K to 600 ohms.

Then you have some ratios inbetween, ilke the 1:7 stuff on the API's,
and the 75 ohm to 300 ohm like the API output.

You just figure out what the impedance is of the two devices you want to match up and compute from there.
Turns ratio is the square root of the impedance ratio.

Also determine what level signals you will be processing to give you a db rating of the transformer.

There really are not that many types of audio transformers. Most catalogs cover the whole spectrum for one particular series on one page.
And they all tend to boost a 20 to 20k frequency response.

If you leave out the innerstage stuff, which is rare anyway, you then only have three catagories, bridging, mic input and output to 600 ohms.
cj

I love those classic PRR answers that give you just enough to see the light (that is of course assuming you don't already know this stuff), but you have to work a little to fully get there.

Now, I'm off to re-read the transfomer and impedence (the later I'm even more befuddled about) sections in AOE a dozen times and study schematics until I go nuts...or just fall asleep...

Thanks!
-E

Thanks CJ!

When I posted this I actually had both of you in mind!

-E

impedance has too many sylables for beginners, so just substitute ohms.
I have always hated that word, sounds too complicated for what it really represents.

:guinness:

I like the word ohms too
:roll:
but it seems to suit DC circuits and impedance seems to show the complications that might be in an AC circuit.

In the DC world an antennae might seem like a short circuit BUT at RF it just fine.

You may also want to read "RadioDesigners Handbook - Ch.5 - Transformers"

http://www.gyraf.dk/schematics/RadioDesigners%20Handbook%20-%20Ch.5%20-%20Transformers.pdf

Jakob E.

:razz: Hey! 1000 posts!!! :sam:

"Impedance" is just "Ohms" at every frequency.

For the longest time, whenever I would see the word "impedence," I would mentally change it to resistance, but now I think I'm starting to get it...

-E

ac ohm, thats all is
buzz words like impedance are for people to use durring job interviews.

Sorry, it's not always quite that simple, and even saying it three times doesn't make it true :wink:

Ok, can I hijack this thread for a minute :?: :green:

So I have an api output transformer, which has three secondaries rated at 75 ohms each I believe. I know that I can hook them all up and get one 600 ohm output, but my ? has to do with when you don't do that, and when you use each one to split the signal to different sources. I'm not saying I'm going to do this, but I'd like to understand what the other types of equipment will think about getting this lower impedence feed.

Is it fine, or.... Here are my example feed points:

1. A tranformer balanced compressor, like an LA2A

2. A tranformerless balanced input to say a DAW

3. A mic input transformer, say an API 312, or a tube mic pre.

Regards

ju

Go for it!
Remember, all those windings are the same, so you could say you have three primary windings and one sec. winding, two pri's and two sec's, or one pri and three sec's.
A vu meter gets one winding on some of the schematics I have seen, Why do yoyu want to feed another mic pre with a mic pre?

who said I was using the transformer in a mic pre? =)

> impedance has too many sylables for beginners, so just substitute ohms.

In many "nice" situations, impedance is a fancy way to say "resistance that varies, but not enough to hurt us".

For example: a CD recorder input is 22K impedance. With an AC ohms-meter, it really does look like 22K over most of the audio band. Below 10Hz it rises. The DC resistance is many megohms, infinite if the capacitors were perfect. Above 20KHz the AC-ohms falls, maybe to 2K at 400KHz, due to stray and actual capacitors on the input. Further, if we raise the AC voltage from the nominal 2V to say 100V, we see a lower and non-linear "ohms" as the semiconductors break down. But for most practical purposes, we can pretend that the input "is the same as" a 22K resistor.

But what does that mean? You MUST understand impedance matching and mis-matching or none of this makes any sense. I think many people never really "get" matching. But that is a subject for another thread.

> "Impedance" is just "Ohms" at every frequency.

For most practical purposes, the variation with frequency ("reacatance") is the big deal. But remember that there are some strange things in the world, and diodes are strange and common. Also even a "good resistor" will show a change in resistance when you put 1,000V on it. So impedance can cover many odd things.

> Then there's Q and Reactance...

Simple reactance is when "ohms" varies directly as frequency. A perfect coil's "ohms" rises as frequency rises. A perfect cap drops with freq.

There are no perfect coils or caps. One common flaw is a resistance: all coils are made with wire that has resistance, and this series resistance screws-up the reactance (it does not go to zero "ohms" at zero frequency). A high-Q coil has very low series resistance. The actual value of Q will depend on what frequency you are interested in, because the reactive part varies with frequency. The "resistance" part also usually varies with frequency: around 1MHz the current flows only in the skin of the conductor so the "resistance" is higher than the DC measurement. Also all iron-core coils have an invisible shunt resistance, because the iron acts as a 1-turn winding and the electrons run in little circles, dragging on the iron atoms.

But I think many people here would do well to overlook frequency effects, and just understand matching and mis-matching.

> three secondaries rated at 75 ohms each I believe. I know that I can hook them all up and get one 600 ohm output

Three 75Ω windings, in series, adds up to 675Ω, not 600Ω.

This may be a case where the amp does not care 675Ω or 600Ω. And there was a fad for 75Ω impedances once upon a time (look at Jensen 990).

If you use the outputs separately, each one gives 1/3rd the voltage as you expect on the 675Ω connection. So you may not get an "ample" output. However you can load each one down to 67Ω, and the amp sees the same load as the 600Ω connection.

I think many people never really "get" matching. But that is a subject for another thread.

Click to expand...

Please make it a subject for this thread! I am listening intently.

cheers,
kent

I second that motion.

Boys and Girls gather around with your pens and paper...
I would love to be graced with an "impedance matching" primer. :idea:

There's a lot that's been written on the subject of impedance matching, and much of it is available to read online, so there's really no point in giving a lengthy explanation here. For the average audio engineer, it reduces to a few fairly simple rules that will keep you out of trouble most of the time.

For maximum signal voltage, load impedances should be much higher (ten times or more) than source impedances. This is the condition you're concerned about at least 90% of the time when moving audio around your studio. For instance, if your microphone has a source impedance of 150 ohms, your mic preamp should have an input impedance of 1.5K or greater.

For maximum power transfer, load impedance should equal source impedance. We're generally not concerned about power transfer except when driving transducers such as speakers, cutterheads, and so on. And even then, it's typical that the power amp is designed to have an output impedance lower than the load impedance anyway, for better damping.

An audio transformer should have ONE of its windings (primary or secondary) terminated in its rated impedance for flattest response. For instance, consider a 1:10 mic input transformer, which could also be called a 150:15K transformer. (The impedance ratio is the square of the turns ratio). For best response, either the primary must be terminated with 150 ohms, or the secondary with 15K. The source impedance of the mic itself provides the 150 ohm termination for the primary, so the secondary should be left unterminated, or bridged by an impedance much higher than 15K.

Passive filters tend to be fairly critical as regards source and load impedances since these impedances usually form part of the filter itself in actual operation.

These last two items are why some older gear seems to violate the first rule I gave; in other words, this is why you need to terminate the output of your old transformer-coupled tube line amp with 600 ohms even though it reduces the output compared to running it into, say, 10K.

Here's some good reading for you:
http://www.richardhess.com/be/aes-80.htm
http://www.svconline.com/mag/avinstall_transmission_lines_why/
(Note that in the second article, whoever typed it up subsituted "V" for the ohms symbol for some unknown reason).