audio output transformer specs meaning?

[jocamo]

Greetings,

I've got an audio output transformer here that's listed as 1000ohms - 8ohms.

Does this imply it will ONLY accept inputs of 1000ohms, or does it mean it will accept inputs "up to" 1000ohms.
Likewise, on the other end, will this apply to the 8ohms side.

I've got what I think is a 600ohm source output, and a 16ohm destination speaker.


I'm trying to narrow down the proper transformer for the project.

To any and all information, thank you.

 

[duke37]

The 1000Ω and the 8Ω is the impedances that the transformer will match. If connected to an 8Ω speaker, the input to the transformer will appear like 1kΩ.

If you connect to a 16Ω load, the input will appear like 2kΩ which is three times the optimum but may be all right. Remember that the 16Ω speaker will have a load impedance very different (higher) to this at some frequencies.

 

[jocamo]

So, given that my source output impedance is about 600ohm, i need to find a transformer that is more in line with 600:16? I.e, match the transformer to the both the source output and the destination.

I thought the idea of the transformer was to nullify the speakers affecting the draw on the source--to force it to the destination? But you're saying, that because the of the "8", putting a 16 on the end goes right back to the idea of the speaker influencing the draw from the amplifier? Still learning a lot here.

 

[kellys_eye]

What precisely IS the 600Ω source? It sounds like a low level audio source rather than something with the 'power' to drive a speaker (even via a transformer).

Similarly, what's the amplifier system? Is it valve based or is it simply a system designed to drive speakers via a 100V line?

 

[jocamo]

Sorry,
It's an aircraft intercom.

I'm researching The figures but I'm told the output from the phono jack is 600 ohms; standard aviation headsets are 300ohms each wired in parallel...so that would be 150, correct?

Another project recommended that transformer but I'm getting a slight "pop" or crackle when loud transmission are received and made.

I don't have much more than that.

 

[kellys_eye]

You could put the headphone is SERIES to match the 600 ohm requirement.

But what is the 'power' of the 600Ω signal source - possibly noted in dBm.

Most 600Ω signal sources require some form of amplification in order to drive loads.

 

[jocamo]

well, the headphones are only 16ohm ear buds. I've not thought about how they are wired since I've never cut back the sheathing. Would standard ear buds be parallel or series?

I don't know the power of the source--only that standard headsets are in parallel to give 150ohms.


so, i tried out the #273-1380 1000:8ohm transformer using some standard earbuds and plugged it into a regular aircraft intercom.

The audio works very well. However, when speaking, the side-tone of my own voice has a slight sizzle to it when transmitting. I didn't notice it at first, so it may have been there the entire time. Maybe not. Is it possible the signal of transmission (using a standard headset's mic in this experiment) is amplified differently and gives the little ear buds trouble when sent back through as side-tone?

I did a radio check with another person and was told the volume and clarity were 5x5.

 

[Audioguru]

You could put the headphone is SERIES to match the 600 ohm requirement.

Have you tried it? Headphones have a common ground. If you feed the "live" wires with the signal then the earphones in series will be out-of-phase and sound weird.

I think the "sizzle" of your sidetone is because it is played with high fidelity. The radio system cuts all high audio frequencies when it is receiving.

 

[jocamo]

Have you tried it? Headphones have a common ground. If you feed the "live" wires with the signal then the earphones in series will be out-of-phase and sound weird.

I think the "sizzle" of your sidetone is because it is played with high fidelity. The radio system cuts all high audio frequencies when it is receiving.

Hmm, is there a solution to that, without sounding foolish here? The whole reason for the project was I didn't want ANR headphones, and the only passive kind are huge and bulky for what I'm doing. Honestly gaming headphones are lightweight and far more comfortable--as are the ear buds I had--wear them for hours without fatigue. Would a passive crossover work to cut out the sizzle? "Could" that be put in line. It's been years since I've worked with anything like those for my truck.

 

[jocamo]

I did learn the output impedance is 600ohms.

Normal aviation headsets are 300 (150 parallel).

Curious why 600 when headsets would be much lower. Ensure stronger output?

The audio sounds great. Just the distortion in the sidetone when speaking. May have damaged the earbuds.

But maybe I can find a better transformer for what normally is a 600:150 ratio.

I had a terrible time just finding that RadioShack model as they no longer carry it. A local shop had one.

 

[Audioguru]

An output impedance of 600 ohms is low enough for driving the capacitance of a long cable or driving the input of an amplifier. A headphones output sometimes is a high power low impedance for driving a speaker with a 220 ohm resistor in series to reduce the loudness to the headphones. An amplifier made for powering headphones usually has an output impedance that is very low. If you drive a 600 ohms load from a 600 ohms source then half the output level is thrown away. Audio transformers are rarely used today.

The sizzle in your sidetone could be a peak in high frequencies of your microphone or your headphones that is not heard when receiving because the receiver cuts away all high frequencies.

 

[Audioguru]

Voices produce frequencies from about 100Hz to about 14kHz. Much of the understanding of speech is in the high frequency consonants that are cut when a receiver has no output above 3kHz so you need to spell many words and use words for letters.

The sizzle in your sidetone can be reduced with a lowpass filter that cuts high frequencies.
A crossover network is used in a speaker to prevent low frequencies from damaging the tweeter and prevent the woofer from shrieking its high frequency resonances.

 

[kellys_eye]

Voices produce frequencies from about 100Hz to about 14kHz. Much of the understanding of speech is in the high frequency consonants that are cut when a receiver has no output above 3kHz so you need to spell many words and use words for letters.

The sizzle in your sidetone can be reduced with a lowpass filter that cuts high frequencies.
A crossover network is used in a speaker to prevent low frequencies from damaging the tweeter and prevent the woofer from shrieking its high frequency resonances.

??
If you check out the speech spectrum for a 'normal adult' the 'hf' end falls off rapidly over around 2kHz and is down by -50dB over 4khz.
This is why the standard audio bandwidth of 300-3000Hz (more commonly 3400Hz) is used for many audio and radio frequency systems and considered adequate to get the content of the messages over without loss.
Standard AM broadcast system use a slightly wider bandwidth (4.5kHz) to improve the music content reproduction.
If you can't 'hear' a message over the standard audio (limited) spectrum then you need to see your doctor!

 

[Audioguru]

Speech that has frequencies cutoff above 3kHz is horrible. Consonants such as with the letters s, c, t, p, sc, ch and many more are commonly spoken and reach 14kHz. You want to say the letter C, not have to say Charlie. Try to say the letter F and you need to say Foxtrot when the high frequencies are missing. Did you say "Sailing" or "Failing"? They sound the same when the high frequencies are missing. Articulation and understanding speech requires frequencies higher than 3kHz. We do not speak with only vowels.

Polycom make high tech speakerphones. Some of them use two telephone lines to double the high frequency cutoff and they sound much better. Here is their article about The Effect of Bandwidth on the Intelligibility of Speech:

 

[kellys_eye]

I see where you're coming from regarding intelligibility but most speech has context in that failure to understand the overall meaning (when talking over a restricted bandwidth path) is rare - however the phonetic alphabet was created to cover those incidents where clarification was necessary. I've never had to 'spell' a whole sentence!

The article you reference (interesting as it was) only advocates increasing the bandwidth to 7kHz to make a notable increase in intelligibility therefore any increase over the standard (300-3400) will see improvement.... whether you need to use ALL the 7kHz is disputable. Moving to 6kHz, 5kHz or 4kHz will improvements too. What's 'enough'? 3400Hz seems to have served us well for decades.

But since Polycom did the research and Polycom make the product I find there is space for 'self-interest' in the results.

 

[Audioguru]

Speech is understood by what you are used to hearing. I find restricted high frequencies and accents cause poor understanding . Some people in Canada have a strong French accent that I have difficulty understanding. British accents are unintelligible such as them saying Wotah instead of saying Water. A truck is called a lorry? British people Toke, not talk.

 

[kellys_eye]

Accents... yes. I see the problem there but no amount of bandwidth will improve the intelligibility of some accents.

I'm an Englishman living in Scotland - I know all about problems with accents!

I find greater issue with people who use microphones by pressing them right up against their face and 'slurring' the words spoken into them.