I want to design a JFET buffer for a headphone of an impedance of 32ohms. A regular source follower amplifier can barely give me a 0.15 gain and 1.2V peak to peak (undistorted). The headphone is supposed to support 50mW so it can support up to 2.4V peak to peak but I can not really foresee what sort of voltage my buffer should be able to provide.
Or is this a completely wrong approach and I should pick a different buffer stage?
Or is this a completely wrong approach and I should pick a different buffer stage?
Last edited:
I have a couple of Koss headphones and both state absolute max voltage as 5V
I don't know if that is 5Vpk or 5Vac.
2.5Vac into 32ohms is NOT 50mW, it is 195mW
I have a +0dB headphone Buffer (actually 1.1Times = +0.8dB) and it goes way too loud into 40ohms headphones (HD465) from CDP/DVD and loud enough from an FM tuner.
I don't know if that is 5Vpk or 5Vac.
2.5Vac into 32ohms is NOT 50mW, it is 195mW
I have a +0dB headphone Buffer (actually 1.1Times = +0.8dB) and it goes way too loud into 40ohms headphones (HD465) from CDP/DVD and loud enough from an FM tuner.
Last edited:
.
I think the key fact you need is that headphone amps commonly have a gain of about 5 to 7, 10 is a lot.
Don't neglect the fact that your output transistors or op amps must be able to source [provide] perhaps 70 milliamps, more is better. Note that this is milliamps, not milliwatts. It's possible to drive headphones with an output that can source something like 20 milliamps (an average op amp), but this is strictly low rent.
The player you're using (CD player, computer, whatever) has an output voltage of something like 0.75 volts (this varies). A gain of 5 would take this up to something like 3.75 volts max. But of course you don't want to run at maximum volume, which in this case works out fine because the volume control near its center would give something like the 2.4 volts you're interested in.
.
I think the key fact you need is that headphone amps commonly have a gain of about 5 to 7, 10 is a lot.
Don't neglect the fact that your output transistors or op amps must be able to source [provide] perhaps 70 milliamps, more is better. Note that this is milliamps, not milliwatts. It's possible to drive headphones with an output that can source something like 20 milliamps (an average op amp), but this is strictly low rent.
The player you're using (CD player, computer, whatever) has an output voltage of something like 0.75 volts (this varies). A gain of 5 would take this up to something like 3.75 volts max. But of course you don't want to run at maximum volume, which in this case works out fine because the volume control near its center would give something like the 2.4 volts you're interested in.
.
Last edited:
.
It's really kinda picky, but I'm too weak to resist saying this:
<< I want to design a JFET buffer for a headphone of an impedance of 32ohms. A regular source follower amplifier can barely give me a 0.15 gain >>
Buffers, by definition, have unity gain. The whole purpose of a buffer is to have high input impedance, so as not to load the audio source, but low output impedance to reduce noise in following circuits. Commonly a buffer is followed by a gain stage(s).
There, I did say it, and I'm glad.
.
It's really kinda picky, but I'm too weak to resist saying this:
<< I want to design a JFET buffer for a headphone of an impedance of 32ohms. A regular source follower amplifier can barely give me a 0.15 gain >>
Buffers, by definition, have unity gain. The whole purpose of a buffer is to have high input impedance, so as not to load the audio source, but low output impedance to reduce noise in following circuits. Commonly a buffer is followed by a gain stage(s).
There, I did say it, and I'm glad.
.
I measured this very thing when I made my headphone amp and it was,
How much drive is needed... well with music on the Sony cans around 0.8 volts peak/peak (transients) is more than enough for me. Average is around 0.2 to 0.4 volts.
How much output can it deliver... around 4.5 volts peak/peak using the Sonys as a load before clipping using the 6.8 volt supply.
taken from,
http://www.diyaudio.com/forums/head...le-ended-class-headphone-amp.html#post2075959
How much drive is needed... well with music on the Sony cans around 0.8 volts peak/peak (transients) is more than enough for me. Average is around 0.2 to 0.4 volts.
How much output can it deliver... around 4.5 volts peak/peak using the Sonys as a load before clipping using the 6.8 volt supply.
taken from,
http://www.diyaudio.com/forums/head...le-ended-class-headphone-amp.html#post2075959
>2.5Vac into 32ohms is NOT 50mW, it is 195mW
I meant 2.4V peak to peak, so +-1.2V
>Buffers, by definition, have unity gain
I have found that it is difficult to maintain a unity gain for a load as low as 32 ohms. The transistor parameters would need to be very suitable. The new buffer I have designed is a BJT emitter follower with 0.8 gain, which is pretty good.
>How much drive is needed... well with music on the Sony cans around 0.8 volts peak/peak (transients) is more than enough for me. Average is around 0.2 to 0.4 volts.
This was what I was looking for. I actually made an experiment and found that 1Volt across a 32ohm headphone was comfortably audible, i wouldn't call it "loud" though
I meant 2.4V peak to peak, so +-1.2V
>Buffers, by definition, have unity gain
I have found that it is difficult to maintain a unity gain for a load as low as 32 ohms. The transistor parameters would need to be very suitable. The new buffer I have designed is a BJT emitter follower with 0.8 gain, which is pretty good.
>How much drive is needed... well with music on the Sony cans around 0.8 volts peak/peak (transients) is more than enough for me. Average is around 0.2 to 0.4 volts.
This was what I was looking for. I actually made an experiment and found that 1Volt across a 32ohm headphone was comfortably audible, i wouldn't call it "loud" though
you need the sensitivity/efficiency number and some knowledge of SPL Preventing Hearing Damage When Listening With Headphones | HeadWize
the typical spec is dB SPL referred to some drive level normally 1 mWrms, but now some are shifting to 1 Vrms drive V as the reference level
you have to look up the specs, sometimes question when the reference level isn't clearly stated
with 32 Ohm headphones, IEM in particular, some are deafeningly loud with 1 Vrms - hearing damage within minutes
others just loud to vey loud, still most 32 Ohm headphones whose numbers I am familiar with will be damaging you hearing at 1 Vrms over hours per day
you should read the headwize article twice - the 1st time for hearing protection from long term SPL
and again to appreciate that you may want substantial dynamic headroom - be able to reach "deafening" peak SPL for fractions of seconds to seconds without clipping
today even 12 dB peak-to-ave may be "dynamic" with so much "Loudness War" compression used in popular music but well recorded Jazz, Classical can have over +20 dB peak SPL above the average level of the whole track
the typical spec is dB SPL referred to some drive level normally 1 mWrms, but now some are shifting to 1 Vrms drive V as the reference level
you have to look up the specs, sometimes question when the reference level isn't clearly stated
with 32 Ohm headphones, IEM in particular, some are deafeningly loud with 1 Vrms - hearing damage within minutes
others just loud to vey loud, still most 32 Ohm headphones whose numbers I am familiar with will be damaging you hearing at 1 Vrms over hours per day
you should read the headwize article twice - the 1st time for hearing protection from long term SPL
and again to appreciate that you may want substantial dynamic headroom - be able to reach "deafening" peak SPL for fractions of seconds to seconds without clipping
today even 12 dB peak-to-ave may be "dynamic" with so much "Loudness War" compression used in popular music but well recorded Jazz, Classical can have over +20 dB peak SPL above the average level of the whole track
Last edited:
- Status
- Not open for further replies.