If you do need a damping factor of >80 re: 4 ohms (which is what <0.05 ohms equates to), you have to build your buffer / amp akin to a speaker amp - in particular, with Zobel network and output inductor (typically, 20 turns of 1 mm wire on ~10 mm dia. for ~1.5 µH, in parallel to 4.7-10 ohms or so).
I think you're way overestimating your requirements though. Even the arguably most finicky earphones known to man, UE's super.fi 10 Pro with their impedance response ranging from about 6.5 to a little over 60 ohms, would be happy with about 0.3 ohms if you're shooting for a 0.3 dB FR deviation target (which is about the absolute threshold of audibility - in practice, <1 dB tends to be fine).
Impedance calculator
1/8" headphone jacks alone can be in the 10s of mOhms of contact resistance.
HD600s are a lot less critical than those. For the same 0.3 dB requirement, they need no less than 22 ohms. I've run HD580s (about 95% the same headphone) on anything from near-0-Ohm portable player outputs over 22 and 33 ohm soundcard outputs, 47 ohm BTech BT928 (modified) output, 75 ohm onboard sound output to 220 ohm receiver headphone jack, and found them to be fairly uncritical and consistent in general. The difference between ~0 and 75 ohms is audible, yes (calculated deviation is 0.8 dB and change), but not a big deal at all - in fact I found the bit of added warmth slightly preferable even. By contrast, I think 47 ohms is too much for HD590s (<22 ohms preferred, impedance 100/195 ohms min/max), and 33 ohms too much for 50 ohm HD598s (<10 ohms preferred, impedance 60/280 ohms min/max).
Expected FR deviation for HD600s on 47 ohms is about 0.55 dB, just slightly audible but certainly not anything major. 10 ohms would drop this to a negligible 0.13 dB.
Why do you think headphones like these are high impedance to begin with? It simply makes them less critical in all kinds of respects, from output impedance and noise level and return conductor resistance to current demands and output stage crossover distortion - the source merely needs to be able to deliver enough output voltage.
I think you're way overestimating your requirements though. Even the arguably most finicky earphones known to man, UE's super.fi 10 Pro with their impedance response ranging from about 6.5 to a little over 60 ohms, would be happy with about 0.3 ohms if you're shooting for a 0.3 dB FR deviation target (which is about the absolute threshold of audibility - in practice, <1 dB tends to be fine).
Impedance calculator
1/8" headphone jacks alone can be in the 10s of mOhms of contact resistance.
HD600s are a lot less critical than those. For the same 0.3 dB requirement, they need no less than 22 ohms. I've run HD580s (about 95% the same headphone) on anything from near-0-Ohm portable player outputs over 22 and 33 ohm soundcard outputs, 47 ohm BTech BT928 (modified) output, 75 ohm onboard sound output to 220 ohm receiver headphone jack, and found them to be fairly uncritical and consistent in general. The difference between ~0 and 75 ohms is audible, yes (calculated deviation is 0.8 dB and change), but not a big deal at all - in fact I found the bit of added warmth slightly preferable even. By contrast, I think 47 ohms is too much for HD590s (<22 ohms preferred, impedance 100/195 ohms min/max), and 33 ohms too much for 50 ohm HD598s (<10 ohms preferred, impedance 60/280 ohms min/max).
Expected FR deviation for HD600s on 47 ohms is about 0.55 dB, just slightly audible but certainly not anything major. 10 ohms would drop this to a negligible 0.13 dB.
Why do you think headphones like these are high impedance to begin with? It simply makes them less critical in all kinds of respects, from output impedance and noise level and return conductor resistance to current demands and output stage crossover distortion - the source merely needs to be able to deliver enough output voltage.
Last edited:
Just for interest. The pdf in post #1 has a clickable article under 'references' on page 5 (TI page number 28). That leads to an article 'the 0 ohm headphone amplifier'.
http://www.diyaudio.com/forums/head...erence-amplifiers-headphone-applications.html
https://cdn.shopify.com/s/files/1/0321/7609/files/Headphone_Amplifier_Performance_-_Part_2.pdf?1361
http://www.diyaudio.com/forums/head...erence-amplifiers-headphone-applications.html
https://cdn.shopify.com/s/files/1/0321/7609/files/Headphone_Amplifier_Performance_-_Part_2.pdf?1361
good stuff.
It is important for a couple of reasons. The cases for open-back headphones are
Therefore, these headphones are recommended for mixing and mastering (aka monitoring).
- Improved, more natural stereo image (compared to non-existent in closed back headphones)
- Flatter frequency response
- Better, more realistic reproduction of bass
The production-related case for closed-back headphones, and probably the reason why a school would require them, is that open backs "bleed", ie. everyone within 20 meters can hear what you're listening to on the headphones. That is unacceptable when recording (laying down tracks, aka tracking), because any mic nearby would pickup the playback from an open-back set. A good set of closed-back headphones, however, eliminates this completely (imagine 20 students in a classroom with open-backs ...and headaches for all!). I hope that answers that for you
If you bothered to look at the FR of the Sony's I mentioned, you'd realize that they are far flatter than the HD600's. Your idea that open back headphones have a flatter FR is wrong.
No one, not even Sennheiser, as far as I know, recommend HD600's for mastering or recording. They are HiFi headphones, not studio ones. Closed back headphones isolate the listener from the environment to a far greater extent than open back, so they are better when you really need to hear the detail in the recording.
Output series resistance is traditionally used to prevent the chip from overheating due to a short circuit on the output. Since the BUF634 has built in short circuit protection, then output series resistance is not required for that chip.
Why do you assume I haven't bothered? Here, bother for yourself
The HD600 is obviously slightly better in the LF range and considerably better in the HF range.
Last edited:
Why do you assume I haven't bothered? Here, bother for yourself
The HD600 is obviously slightly better in the LF range and considerably better in the HF range.
The HD600 has slightly deeper bass extension but it is not flatter. It has a 5dB boost in the 100-200 Hz region, whereas the MDR7506 is flat.
Ok, I'm done with this thread. You're too stubborn to listen to people who are more experienced than you. You're on your own. Have a nice day.
I'd say that '7506 measurement might have had a bit of an issue with bad seal by the looks of it (that hump in the low mids tends to be an indication). It should normally be looking fairly similar to the V6, its "consumer" cousin.
BTW, you don't actually want headphones that are totally flat in the bass. A bit of extra boost is required to compensate for the lack of body-borne sound. The Harman target response curve, for example, includes a bass boost of about 3 dB at 100 Hz and up to 5 dB at 40 Hz and below.
So you could argue that HD600s are a bit too hot in the low mids, ending up in the "mid mids" being somewhat reduced. People tend to like this though, as that kind of range tends to contain lots of reverb and reducing it sort of unclutters the sound (without affecting tonal balance). So you can see why the classical music guys in particular would be into that, and why they'd be used in monitoring there.
Closed cans may be better out in the field, but they do tend to exhibit internal reflections that subjectively narrow the soundstage and color the sound even when you can barely see them on (smoothed) measurement graphs. People don't put up with the limits of typical open cans in the bass for nothing.
Hint: The Headroom graph builder can also display impedance vs. frequency.
Regular 250 ohm Beyer DT880s are not too fussy either, ranging between about 240 and still less than 300 ohms (so that's like a factor 1.25 tops, or <2 dB). MDR-V6/7506 still show fairly benign variation (<1.5 or ~3 dB), but their lesser minimum impedance of ~80 ohms means that "high impedance" to them is about a factor of 3 lower than for those Beyers. AKG K7xx are in the same ballpark, their lower than average sensitivity (~100 dB/V at 62 ohms) make the output work harder though. HD600s are at like 310..560 ohms, which is approaching a factor of 2, but impedance still is decently high - hence in practice it usually works out OK. HD590s have little more relatively variation but at only about 1/3 the impedance, so they are a lot more likely to run into audible trouble, and HD598 almost halve impedance again with a variation of more than a factor of 4.5 on top of that, so there it's getting pretty fussy.
Orthodynamic headphones tend to have a ruler-flat impedance response as well, but their sensitivity tends to be somewhat to considerably below average, so they are critical in other ways (current / power needs).
Balanced armature IEMs can be quite the opposite in that you'll hear the grass grow with them, which is good for saving battery power in hearing aids (which is what their drivers were originally developed for) but also makes every bit of noise painfully obvious. We are talking ~130 dB / V or more. This is compounded by the rollercoaster type impedance response typical in models with crossovers (see aforementioned extreme example) which demands a low-impedance output, so no cheating with series resistors here. Unfortunately very low noise is hard to achieve with circuitry for very low supply voltages as you tend to find in portable audio gear.
Impedance dips into the single-digit ohm range may be another way of annoying an output stage. The specifics depend on circuitry, but typical rail-to-rail outputs in low-voltage gear tend to be current outputs, and open-loop gain will decrease as load impedance is reduced, driving up distortion even if the output stage didn't have to work harder on top of that, which it does.
You also have to take into account what kind of circuitry you tend to be dealing with. In portable players you'll probably be voltage-limited, and headphones with less than about 100-103 dB/V may be a big no-no (and sometimes even a fair bit higher than that - one particularly wimpy phone output even failed to drive a Sennheider HD25-1, hardly insensitive at 120 dB/V). Output impedance only tends to be in the low single-digit ohms though, maybe 47. By contrast, among hi-fi gear you may have to deal with primitive resistive taps from the speaker power amp stage, which means ample voltage swing but output impedance in the hundreds of ohms. And then there's your average onboard sound headphone output with 0.8-1.2 Vrms tops at 75 ohms.
So really, there's a lot of factors to consider here.
BTW, you don't actually want headphones that are totally flat in the bass. A bit of extra boost is required to compensate for the lack of body-borne sound. The Harman target response curve, for example, includes a bass boost of about 3 dB at 100 Hz and up to 5 dB at 40 Hz and below.
So you could argue that HD600s are a bit too hot in the low mids, ending up in the "mid mids" being somewhat reduced. People tend to like this though, as that kind of range tends to contain lots of reverb and reducing it sort of unclutters the sound (without affecting tonal balance). So you can see why the classical music guys in particular would be into that, and why they'd be used in monitoring there.
Closed cans may be better out in the field, but they do tend to exhibit internal reflections that subjectively narrow the soundstage and color the sound even when you can barely see them on (smoothed) measurement graphs. People don't put up with the limits of typical open cans in the bass for nothing.
I'd call headphones "critical" when they are, say, quite sensitive to output impedance, a result of highly variable and/or low impedance. Take aforementioned super.fi 10 Pros, for example - that's a variation of almost a factor of 10 in impedance. Present those with an approximate current source (a very high-impedance output compared to the load), and frequency response will be warped up to the same degree, which is almost 20 dB. Major, major coloration. By contrast, your average dynamic driver in-ear has an impedance response that's almost ruler-flat, so nothing much would happen there.
Hint: The Headroom graph builder can also display impedance vs. frequency.
Regular 250 ohm Beyer DT880s are not too fussy either, ranging between about 240 and still less than 300 ohms (so that's like a factor 1.25 tops, or <2 dB). MDR-V6/7506 still show fairly benign variation (<1.5 or ~3 dB), but their lesser minimum impedance of ~80 ohms means that "high impedance" to them is about a factor of 3 lower than for those Beyers. AKG K7xx are in the same ballpark, their lower than average sensitivity (~100 dB/V at 62 ohms) make the output work harder though. HD600s are at like 310..560 ohms, which is approaching a factor of 2, but impedance still is decently high - hence in practice it usually works out OK. HD590s have little more relatively variation but at only about 1/3 the impedance, so they are a lot more likely to run into audible trouble, and HD598 almost halve impedance again with a variation of more than a factor of 4.5 on top of that, so there it's getting pretty fussy.
Orthodynamic headphones tend to have a ruler-flat impedance response as well, but their sensitivity tends to be somewhat to considerably below average, so they are critical in other ways (current / power needs).
Balanced armature IEMs can be quite the opposite in that you'll hear the grass grow with them, which is good for saving battery power in hearing aids (which is what their drivers were originally developed for) but also makes every bit of noise painfully obvious. We are talking ~130 dB / V or more. This is compounded by the rollercoaster type impedance response typical in models with crossovers (see aforementioned extreme example) which demands a low-impedance output, so no cheating with series resistors here. Unfortunately very low noise is hard to achieve with circuitry for very low supply voltages as you tend to find in portable audio gear.
Impedance dips into the single-digit ohm range may be another way of annoying an output stage. The specifics depend on circuitry, but typical rail-to-rail outputs in low-voltage gear tend to be current outputs, and open-loop gain will decrease as load impedance is reduced, driving up distortion even if the output stage didn't have to work harder on top of that, which it does.
You also have to take into account what kind of circuitry you tend to be dealing with. In portable players you'll probably be voltage-limited, and headphones with less than about 100-103 dB/V may be a big no-no (and sometimes even a fair bit higher than that - one particularly wimpy phone output even failed to drive a Sennheider HD25-1, hardly insensitive at 120 dB/V). Output impedance only tends to be in the low single-digit ohms though, maybe 47. By contrast, among hi-fi gear you may have to deal with primitive resistive taps from the speaker power amp stage, which means ample voltage swing but output impedance in the hundreds of ohms. And then there's your average onboard sound headphone output with 0.8-1.2 Vrms tops at 75 ohms.
So really, there's a lot of factors to consider here.
Thank you for such a knowledgeable post! With all seriousness, this thread contains some priceless resources and information–the above post included–that are invaluable for learning about headphones.
Let me add a few I've read yesterday:
Headphone & Amp Impedance Questions? Find the answers here! - good post, contains so many links inside, technical information, etc.
Golden Ears - Impedance Matching and Bridging
The bottom line for me is, I am not planning on using the headphones for daily use with my phone or at work - for that, I have simple IEMs which are good enough for me for serving that purpose.
What I'm looking for is an affordable headphone monitoring setup that can translate well and make up for the lower frequencies that my monitors lack the ability to produce. I have done a lot of research on the topic, comparing data as well as reported experiences of producers, and weighed the factors in. Consequently, I am keeping my choice of the Sennheiser HD600 for their overall clarity across the audio band, great sensitivity, superior soundstage, tight bass control and overall uncolored sound.
To make the absolute best out of those headphones without spending major dollars I am going to build a 0hm amp (with the generous help of TheAnonymous1) and incorporate it into the power supply project I was already working on.
I hope others will benefit from this thread as I did. A big thank you to everyone who participated!
Let me add a few I've read yesterday:
Headphone & Amp Impedance Questions? Find the answers here! - good post, contains so many links inside, technical information, etc.
Golden Ears - Impedance Matching and Bridging
The bottom line for me is, I am not planning on using the headphones for daily use with my phone or at work - for that, I have simple IEMs which are good enough for me for serving that purpose.
What I'm looking for is an affordable headphone monitoring setup that can translate well and make up for the lower frequencies that my monitors lack the ability to produce. I have done a lot of research on the topic, comparing data as well as reported experiences of producers, and weighed the factors in. Consequently, I am keeping my choice of the Sennheiser HD600 for their overall clarity across the audio band, great sensitivity, superior soundstage, tight bass control and overall uncolored sound.
To make the absolute best out of those headphones without spending major dollars I am going to build a 0hm amp (with the generous help of TheAnonymous1) and incorporate it into the power supply project I was already working on.
I hope others will benefit from this thread as I did. A big thank you to everyone who participated!
You could go balanced for reasonable cost. Don't buy a balanced cable, buy one of these and cut the 1/4" plug off http://www.amazon.com/Replacement-Cable-SENNHEISER-Headphones-HD650/dp/B0028PGXRE
There are 4 wires going all the way to that plug where the two negatives are combined into a single ground connection.
Take a look at The Wire BAL-BAL for an example of a reasonably cheap amp design with very high quality for a balanced out headphone amp (it's a unity gain balanced buffer unless you change the resistors). You can directly replace the LME49600 that is now end of life with BUF634 ICs.
There are 4 wires going all the way to that plug where the two negatives are combined into a single ground connection.
Take a look at The Wire BAL-BAL for an example of a reasonably cheap amp design with very high quality for a balanced out headphone amp (it's a unity gain balanced buffer unless you change the resistors). You can directly replace the LME49600 that is now end of life with BUF634 ICs.
Polarity of 4 wire balanced HD600 headphone cable to 4 pin XLR?
Hello,
Thank you for the reference.
I bought one of these replacement cables.
I have a pair of Electra-Print 7000:300 balanced output transformers. There are a couple of different output tubes on the bench.
Is there anything close to a standard plug / socket for connecting the 4 wire balanced cord to the chassis.
I am thinking a 4 pin XLR.
Is there a standard polarity?
When you look at the two pin plugs on both the Red – Right, Black – Left there is a polarity or index on each of the headphone connectors, one pin is larger than the other on each plug. If there is a standard, what wires connect to each of the 4 pins of the XLR?
Thanks DT
Hello,
Thank you for the reference.
I bought one of these replacement cables.
I have a pair of Electra-Print 7000:300 balanced output transformers. There are a couple of different output tubes on the bench.
Is there anything close to a standard plug / socket for connecting the 4 wire balanced cord to the chassis.
I am thinking a 4 pin XLR.
Is there a standard polarity?
When you look at the two pin plugs on both the Red – Right, Black – Left there is a polarity or index on each of the headphone connectors, one pin is larger than the other on each plug. If there is a standard, what wires connect to each of the 4 pins of the XLR?
Thanks DT
I found this today.
Any disagreement or comments?
DT
4-pin XLR Used singly for balanced console/desktop amps
Pin 1: Left positive (+)
Pin 2: Left negative (-)
Pin 3: Right positive (+)
Pin 4: Right negative (-)
Shell: ?
The pins are numbered on the bodies of the plug and socket.
There is no official pinout standard, but this is the normal arrangement for the cables shipped with AKG K1000 and HiFiMAN orthodynamic headphones.
I used a 4-pin XLR for my headphone amps, but then I found out many power amps use a pair of 3-pin XLRs. The choice is up to you.
As for the 4-pin pinout, that's what Owen and I have both used as well. Shell should be connected to the cable shield, if the cable you're using has one (ground on the board). Not really something critical.
As for the 4-pin pinout, that's what Owen and I have both used as well. Shell should be connected to the cable shield, if the cable you're using has one (ground on the board). Not really something critical.
Hello Tyler,
Thanks for the reply.
I am going with the 4 pin version, at least this time.
I bought two of the Sennheiser replacement cords and confirm that there are four wires running the length of the cord. The negative wires left and right both attach to the sleeve of the TRS plug.
I got bold and cut off the plug from one of the Sennheiser replacement cords and stripped off the outer insulator material. Inside both of the outer insulation jackets there are two stranded wires with enamel or similar insulation. There is not any separate shielding.
I was unable to get my Fluke DMM to contact bare metal or make any continuity verification with the stranded wire.
I have attached a photo of my thumb and stranded wire.
DT
Thanks for the reply.
I am going with the 4 pin version, at least this time.
I bought two of the Sennheiser replacement cords and confirm that there are four wires running the length of the cord. The negative wires left and right both attach to the sleeve of the TRS plug.
I got bold and cut off the plug from one of the Sennheiser replacement cords and stripped off the outer insulator material. Inside both of the outer insulation jackets there are two stranded wires with enamel or similar insulation. There is not any separate shielding.
I was unable to get my Fluke DMM to contact bare metal or make any continuity verification with the stranded wire.
I have attached a photo of my thumb and stranded wire.
DT
Attachments
the solder wicked nicely into the strands of litz wire
Hello,
I am back from holiday.
I am pleasantly surprised.
I looked up litz wire, some of it is reported as “solderable”. I turned up the temperature on the soldering pencil and applied some 63/37 leaded solder to the litz wire. The solder did not flow immediately however after lingering briefly with the heat the solder wicked nicely into the strands of the litz wire.
No mechanical scraping or chemical enamel removers were required.
I will attach a photo or two of the 4-pin XLR plug being attached to the “balanced” Sennheiser replacement cord.
DT
Hello,
I am back from holiday.
I am pleasantly surprised.
I looked up litz wire, some of it is reported as “solderable”. I turned up the temperature on the soldering pencil and applied some 63/37 leaded solder to the litz wire. The solder did not flow immediately however after lingering briefly with the heat the solder wicked nicely into the strands of the litz wire.
No mechanical scraping or chemical enamel removers were required.
I will attach a photo or two of the 4-pin XLR plug being attached to the “balanced” Sennheiser replacement cord.
DT
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.