I'm just about ready to start construction of my headphone amp, and have pretty much decided on the Gilmore Dynalo. But I've also discovered the Dynahi, and bridged versions of both. (Thanks Dan Gardner.) One webpage, dedicated to a bridged Dynahi version, measured an output of 19Wrms into an 8-ohm load. My Beyerdynamic DT-770's only need 5Vrms@20ma to achieve full output. I realize that you want some dynamic headroom, but how much is enough?
Will I really notice any sonic improvement with almost 20db of headroom? Are there any "rules-of-thumb" on this?
Thanks all;
Artie
Will I really notice any sonic improvement with almost 20db of headroom? Are there any "rules-of-thumb" on this?
Thanks all;
Artie
for the 250 Ohm Beyers I find 96 dB re 1 mW
5 Vrms will give ~+20 dB for 116 dB max SPL
if you occasionally listen at near live levels this probably isn't enough for the +120 dB peak SPL of some genre of music:
http://headwize.com/articles/hearing_art.htm
20 dB peak-to-average is very rare in popular music recordings given the "Loudness War" so many may never need even 100 dB capability if they listen at safe levels for hrs/day the ave should be kept well below 90 dB SPL
for myself, I like to know the amp can cleanly drive the headphones to a little over 120 dB SPL
it is extremely unlikely that music dynamic peaks (snare, crash cymbal) will overheat a headphone - max ratings are for continuous levels
over excursion/voice coil bottoming on bass notes is a possibility with some cans though
5 Vrms will give ~+20 dB for 116 dB max SPL
if you occasionally listen at near live levels this probably isn't enough for the +120 dB peak SPL of some genre of music:
http://headwize.com/articles/hearing_art.htm
20 dB peak-to-average is very rare in popular music recordings given the "Loudness War" so many may never need even 100 dB capability if they listen at safe levels for hrs/day the ave should be kept well below 90 dB SPL
for myself, I like to know the amp can cleanly drive the headphones to a little over 120 dB SPL
it is extremely unlikely that music dynamic peaks (snare, crash cymbal) will overheat a headphone - max ratings are for continuous levels
over excursion/voice coil bottoming on bass notes is a possibility with some cans though
Just to be clear, the 20db figure was just comparing the 100mW max of the Beyer's to the 19Wrms of the Dynamite. (22db, to be exact.)
I understand that increased headroom will give me more "punch" and dynamics, but I'm wondering how much is enough. If my headphone amp could only provide 200mW's, would that be sufficient? Would 1W be better? ("Better" being a subjective term, I realize.)
I understand that increased headroom will give me more "punch" and dynamics, but I'm wondering how much is enough. If my headphone amp could only provide 200mW's, would that be sufficient? Would 1W be better? ("Better" being a subjective term, I realize.)
you must compare power into equal load impedance or must compare voltage at the load to find the difference in overhead.
19W into 8r0 is very different from 19W into 250r0.
5Vrms into 250r0 is very different from 5Vrms into anything other than 250r0.
1mW into 250r0 is 0.5Vrms
100mW into 250r0 is 5Vrms.
1W into 250r0 is 15.8Vrms.
5Vrms ~=7Vpk
7Vpk into 96dB/mW headphones is ~119db peak.
That is very loud.
19W into 8r0 is 17.4Vrms = 24.7Vpk. That is too loud, ~130dB peak SPL.
and the current available to drive the 8ohm load is simply a waste of resources.
An amplifier designed to deliver 10Vpk into 50r0 will drive your 96dB, 250ohm headphone as loud as you will ever want.
19W into 8r0 is very different from 19W into 250r0.
5Vrms into 250r0 is very different from 5Vrms into anything other than 250r0.
1mW into 250r0 is 0.5Vrms
100mW into 250r0 is 5Vrms.
1W into 250r0 is 15.8Vrms.
5Vrms ~=7Vpk
7Vpk into 96dB/mW headphones is ~119db peak.
That is very loud.
19W into 8r0 is 17.4Vrms = 24.7Vpk. That is too loud, ~130dB peak SPL.
and the current available to drive the 8ohm load is simply a waste of resources.
An amplifier designed to deliver 10Vpk into 50r0 will drive your 96dB, 250ohm headphone as loud as you will ever want.
Thanks Andrew. I probably shouldn't have included a technical spec in a tongue-in-cheek remark. I should have simply remarked that the Dynamite was wretched excess. (Or so it seems.) 
I guess what I'm asking is: if you were going to design a headphone amp for yourself, how much headroom would you provide for?
Thanks again. Sorry if I'm confusing the issue.
I guess what I'm asking is: if you were going to design a headphone amp for yourself, how much headroom would you provide for?
Thanks again. Sorry if I'm confusing the issue.
Sound Pressure Level at your ears is what's important
as I showed you should use the headphone sensitivity specification with the impedance&power calculations to get the SPL output by the headphones
unlike loudspeakers, headphones vary widely across models/manufacturers on impedance and sensitivity - so you have to have a target headphone spec to say if the amp's I,V output capability is adequate
read the Headwize article for example music SPL - and keep below the OSHA time*SPL exposure limits to avoid rapid hearing loss (euro SPL exposure limits are lower than OSHA's)
the 130 dB SPL symphonic music peak SPL likely is on the stage, not in the audience - professional musicians should wear hearing protection
once you can reach some reasonable music listening max SPL with a given headphone and amplifier combination there is little reason to think even more never used dynamic range is beneficial - except in the case of low feedback designs where the "unused" headroom means the amplifier's outputs are actually using a smaller fraction of the bias range and would be expected to distort less than a similar stage at "just enough" bias
as I showed you should use the headphone sensitivity specification with the impedance&power calculations to get the SPL output by the headphones
unlike loudspeakers, headphones vary widely across models/manufacturers on impedance and sensitivity - so you have to have a target headphone spec to say if the amp's I,V output capability is adequate
read the Headwize article for example music SPL - and keep below the OSHA time*SPL exposure limits to avoid rapid hearing loss (euro SPL exposure limits are lower than OSHA's)
the 130 dB SPL symphonic music peak SPL likely is on the stage, not in the audience - professional musicians should wear hearing protection
once you can reach some reasonable music listening max SPL with a given headphone and amplifier combination there is little reason to think even more never used dynamic range is beneficial - except in the case of low feedback designs where the "unused" headroom means the amplifier's outputs are actually using a smaller fraction of the bias range and would be expected to distort less than a similar stage at "just enough" bias
The 100 mW rating of your DT-770 is the long term maximum that the headphone can withstand without suffering mechanical failure.
As noted, it has a 96 dB for 1mW rating. When we raise sound pressure by 3 dB we double the required power, hence...
99 dB will be produced by 2 mW
102 dB will be produced by 4 mW
105 dB will be produced by 8 mW
108 dB will be produced by 16 mW
111 dB will be produced by 32mW
114 dB will be produced by 64 mW
117 dB will be produced by 128 mW
120 dB dB will be produced by 256 mW
If you have *normal* hearing then by the time you reach 114 dB it will subjectively far too loud for continuous listening. Although transient peaks above this will be OK, a doubling of the sound pressure to a continuous 117 dB (i.e. 128 mW) will actually damage the headphone (and most likely your hearing too).
Overall, I would suggest that for the DT-770 an amp that can source a continuous 64 mW with head-room to peaks of 256 mW will be very suitable for your purpose.
As noted, it has a 96 dB for 1mW rating. When we raise sound pressure by 3 dB we double the required power, hence...
99 dB will be produced by 2 mW
102 dB will be produced by 4 mW
105 dB will be produced by 8 mW
108 dB will be produced by 16 mW
111 dB will be produced by 32mW
114 dB will be produced by 64 mW
117 dB will be produced by 128 mW
120 dB dB will be produced by 256 mW
If you have *normal* hearing then by the time you reach 114 dB it will subjectively far too loud for continuous listening. Although transient peaks above this will be OK, a doubling of the sound pressure to a continuous 117 dB (i.e. 128 mW) will actually damage the headphone (and most likely your hearing too).
Overall, I would suggest that for the DT-770 an amp that can source a continuous 64 mW with head-room to peaks of 256 mW will be very suitable for your purpose.
Gilmore suggests higher Vsupply - and bias changes to make the "Pure Class A Dynamic Headphone Amplifier" more suited to higher Z headphones - the nominal design is for 32 Ohm cans
http://headwize.com/projects/showfile.php?file=gilmore3_prj.htm
or you could try adding the 2nd output stage to make the balanced version of the amp outlined in the article - not the usual "dual mono" balanced - it does require buffered inputs since you end up driving the inverting feedback resistor impedances
http://headwize.com/projects/showfile.php?file=gilmore3_prj.htm
or you could try adding the 2nd output stage to make the balanced version of the amp outlined in the article - not the usual "dual mono" balanced - it does require buffered inputs since you end up driving the inverting feedback resistor impedances
Yup . . . I've just about got that webpage memorized.
I plan on running 20v rails for my Beyer's. I can't wait to hear how this all sounds. My current "reference" system is the all-discrete headphone amp thats built in to a Proton 1100 pre-amp. Its a pretty good amp, but I'm fairly sure its class AB. I'm curious to hear pure class A.
I plan on running 20v rails for my Beyer's. I can't wait to hear how this all sounds. My current "reference" system is the all-discrete headphone amp thats built in to a Proton 1100 pre-amp. Its a pretty good amp, but I'm fairly sure its class AB. I'm curious to hear pure class A.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.