Headphone speakers

[Goodrat]

Anyone do much experimenting with headphone speakers?
Using 150 ohm speakers is typical, although that is 75 ohms since they are wired in parallel. I see some 500 ohm Kobitone speakers at Mouser that have close to full range. I will give them a try. Plus the spec indiactes they should be louder than typical 150 ohm speakers. It seems to me that being higher than the 150 ohm series resistor that is present in most machines, should be better than a 75 ohm load.
When I run the numbers, I think I should be fine.
When the volume pot on the phones is set to a comfortable level, the phones are more like 400 ohms anyway with the KB's set I have.

[bbsailor]

Re: Headphone speakers

Try putting a SPDT 2 position toggle switch in the headphones to change the headphone elements impedance from being in series at 300 ohms, or in parallel at 75 ohms. Just go to some guitar sites that feature guitar wiring (Dimarzio or Seymour Duncan) and see how humbucking pickups can be switched from series to parallel. It is the same with two headphone elements.

Another thing. Get an inexpensive sound level meter from Radio Shack to measure the output to see which combinations of speakers actually are the lowdest. Just place the meter through a hole in a piece of plywood, the size of your ear to seal up the hearing cavity. The take measurments with different machines, headphone elements and series/parallel connections to see what produces the loudest sound.

I have seen some 600 ohm headphones with over 110 db sensitivity. Remember, the power levels double at each 3db sound level increment and the voltage levels double at each 6db sound level increment. This will help you keep track of the data and have something scientific to base your work on. It will aslo help you post a good report on this forum about you findings.

To get some real sensitive headphones requires optmizing the signal to noise ratio... less noise and/or more signal or both. The trouble with just amplifying the signal is that you can really blast your ears. That is why adding a boost with some form of hard limiting, such as through a diode limiter, is helpful. While you are at it and building your headphones, plan on putting a switch in the headphones to turn on the limiting diodes. You can even switch between silicon and germanium types (using a center off DPDT) to limit at .6 V or .3V depending on what sounds better. Then what ever outboard amplification you add will not blast your ears. There are some 3V op amps that will do just fine. You could build a prototype with both an input level control as well as a variable gain. Just make sure you put it in a good metal case to keep from picking up the radiated coil signal. Also, use shielded wire between the outboard amp and the metal detector and to the headphones. Real high impedance headphones are suseptable to picking up coil radiation. If this is a problem, then you need to shield each headphone cup and use a metal screen over each speaker.

One quick way to tell if one set of headphones is more sensitive than another is to set one just at the threshhold of hearing then change to the other headphones and see if the level changes. The ear can normally just discern a 3 db level change otherwise you need a sound level meter for finer levels of sound level discrimination. With a sound level meter you must set the reference level for something mid-range like a 60 db level, then change headphones and remeasure to see any difference.

This discussion brings back some memories of acoustic theory. While you are at it, look up the Fletcher-Munsen Curve (not sure of spelling) which depicts the frequencies that the human ear is most sensitive. It would be efficient if you could make the natural resonance of the headphone speakers and the output tone of the MD to work in this most theoritically sensitive area of human hearing. Maybe that is why it practical to use a square wave rather than a pure sine wave for the MD output signal. The fundamental tone can be lower with the harmonics being higher in the more sensitive hearing area.

Let us know about your progress.

bbsailor

[bbsailor]

oops! SPDT should be DPDT

correction!

[Goodrat]

Re: oops! SPDT should be DPDT

Thak you for that response.
Since I am limiting my speaker choices to what I can find in small quantities, I need to buy at digikey. The Korbitone 25S500 is 500 ohms with a sensitivity rating of 117db @ 1mW. Comparing to the Audiosears 2463 in Killer B's headphones with 150 ohms, 73db .
Just to compare what it takes to get 1mW with parallel speakers, the 75 ohm needs .0037A and .28V and the 250 ohm needs .002A and .5V.
Yes, the final test is just to see wich sounds good. I do have a sound level meter, I'll try that.
What my major point is, many manufacturers boast how their phones are 150 ohms, as if that was a magic number because of the typical source impeadance of most MD machines (due to a 150 ohm series resistor). The highest power transfer is when the load is 150 ohms, of coarse, but this is not always the best sound depending on other fac

[Goodrat]

Re: limiter

Oh, and I have done some experimenting with LED's for a good simple limiter. That is covered.

[knondike clad]

Re: limiter

DO you have circuit on putting led limiter in stereo head phone with volume controls?

[Goodrat]

Re: limiter

Try this. Take red LED's and connect accross the volume controls. Cathodes to ground. Is this for music or a metal detector. Stereo or dual volume control mono?

[bbsailor]

Re: oops! SPDT should be DPDT

Download the PDF file for that 25SP500 speaker. Note the graph that shows the response sensitivity for various frequencies.

Below 150 Hz it falls off pretty fast. From about 150 Hz to about 3 KHz it is pretty flat at about 93 to 94 db. At about 3 KHz it rises to about 103db. It seems to have the classic double resonant hump between 5 and 10 KHz.

If you put a 5 to 10K ohm isolation resistor in series with the speaker and a variable signal generator and measure the voltage with a high Z VOM or VTVM at the speaker, you will see the resonant points as you sweep between 5 and 10 KHz. That is not a bad thing to do with all your candidate speaker elements. Keep good notes. The final test is what sounds good and works. Then with your notes you will begin to narrow down the candidate headphone elements that have potential.

Check to see if this speaker really puts out 117 db as claimed with 1mw of power. The variation of +/- 4 db is a pretty wide variance. It is most likely -4db. Other headphones put out 130+ db at full rated power. Look at the specifications for this speaker element very carefully to make your comparisons.
Go to the headphone web site I gave you before to see how headphone sensitivity is rated.

This speaker may well work very well in your setup. Experiment, however know the facts!

Remember the most efficient speaker elements have the largest and/or most powerful magnets. A lower impedance very efficient speaker may work better than a lower efficiency high impedance speaker. Just clip various speakers to the same sound source and measure at the same distance for a quick check of this. Most of these experiments can be done before you assemble the final headphones.

The reason why headphone vendors boast about 150 ohm headphones is because number 150 is a larger number than 32 (typical headphone impedance). This is more driven by marketing hype than any true expectation the the general public will appreciate the subtle nature of impedance matching and efficiency.

To get the lowdest output, get the most efficient speaker driver. Then, go inside your metal detector and reduce the series resistor by half by soldering a paraller value equal to what is already there.

bbsailor

[bbsailor]

Headphone article reference

Try this site.

They have a table of various commercial headphones, indicating the impedance and sensitivity. This is very educational.

http://headwize2.powerpill.org/tech/dbohn1_tech.htm

bbsailor

[Goodrat]

What I want to hear

What it comes down to is, should I buy a set of these speakers? By the way they are at Mouser, not Digikey, for $5. I've ordered 10 pots, and will order just a set or two of speakers for now. But I'm sure after I build, a few friends will want be to build more.

I don't intend to open or modify my X5. (still afraid to open it due to its cost) So hopefully the series resistor in the machine can stay.

It would be interesting if the wider freq response of this speaker will be beneficial to the 1KHz square wave, which is about 8v p-p by the way. There may be plenty of power to spare. Killer B headphones will blast your ears full up! And again, they are at 75 ohms in that state, and at normal volume more like 450 ohms. So, for my 300 ohm setup, I may be fine.
I intend to put two 10K audio pots in parallel across the input, then connect from wiper to the speakers to ground. This makes wiring easier.

By the way, I've tried speakers from the older type telephones. Not bad at all. They are probably 600 ohm. Pleny loud.

[bbsailor]

Re: What I want to hear

The rule of thumb for matching pot values to impedance is 10 to one. If an impedance source is 1K then 10 K pots should not provide too much of a load and this is often what is done in commercial designs.

If you find that you loose some high frequencies when you drop the volume, try putting a .01, .05 to .1 cap from the pot center to the hot side. Try different values to suite your hearing.

Also, you may want to try the single transistor Depthmaster circuit on Patents portion the Geotech web site. This will eliminate any voltage losses in attempting to match what ever value series resistor is in you MD. It only uses a single 1.5 V battery and it does a very good job of amplifying low level signals. The only down side is that it comes in a plastic box which picks up the coil radiated signal.

To ensure that your headphones don't pick up the signal, use flexible stranded center conductor shielded wire. Run the wire near the MD coil to ensure that you are not picking up the coil signal. Do this before you finalize your design.

bbsailor

[Goodrat]

Re: cap

Ah yes, the .01 on the volume control trick. I recall that from my guitar amp building days. I'll keep that in mind.

[Goodrat]

Or this

Get a short two headphone adapter. One output to your headphones, the other plug in a stereo plug with an led from each stereo pin to ground. Also try diodes or two led's in oposite polarity on each left/right pin.

[Klindike Clad]

Re: limiter

I will try this on a pair of white headphone.
do I use one led or two per side or is there a circuit I could look at? thanks I will be using them on a metal detector XL pro(6000)

[Goodrat]

Mono example, one LED