If you ever wondered where Audio Anonymous is, you're already there.
Some of the resident evil are trying to inhale an spdif joint, a rare mocking bird reverted to sniffing glue fumes.
( Marijuana Anonymous - The Netherlands )
How right you are . Never underestimate how the pick up arm you already have might only be wrong as far as mass is concerned . Even the old SME 3009 ( 6 g effective ) with some mass added might work .
Can anyone assist with a headphone amp ? How many volts required etc ? I can just about get 1 Vrms from a NE5532 into 32R from memory .
I have next to zero PCB space as this was an afterthought . It doesn't need to be the best ever . Better than in most computers wound be fine . I can bias the output slightly into class A . I suspect 4 mA would be about right ( a resistor to -ve about my limit for PCB space if not using a JFET CRD diode ) .
The chip that looks possible is this one below . Does anyone have enough headphones to know the facts and figures ?
http://www.mouser.com/ds/2/294/NJM4556A_E-218655.pdf
Looking carefully at this next chip it might work . I supect I could pull more in class A without problems . That should avoid crossover distortion . Doubtless into 32 R not too bad if so ? It is almost like an op amp . I might splay the SMD pins to make it through hole . If so it might work quite well .
http://www.farnell.com/datasheets/51452.pdf
The space I have is about that of 3 x DIL 8 side by side . I think I will have to use output capacitors as the DC offset is a problem .
Opinions vary greatly about this . Some say a standard op amp will drive anything . Other have > 5 V into 32 R . I doubt any 32 R headphone needs much more than 1 V rms ?
I made some progress . 2 V rms seems possible . 10 nV noise , Unity gain stable . GBWP 8 MHz . +/- 18 V . > 70 mA 150 R . DIL 8 . Dual device .
This is the constant current source I will use to have class A at lower volume .
Current Limiting Diode CRD 4.5mA | Rapid Online
This is the constant current source I will use to have class A at lower volume .
Current Limiting Diode CRD 4.5mA | Rapid Online
Nigel,
I understand the hypothesis about forcing an opamp into class A making things better, but there are a number of issues with it. Looking at you schematic, you are introducing three issues imo.
Firstly, you put a capacitor in the signal path (the 10 uF coupling fb to ground) that otherwise wouldn't be there. Now, I am not affraid of using coupling caps myself, but in the spirit of audiophile purity, this should be a no go.
Secondly, I looked at the specs of the current limiting diodes from your earlier post. As good as they are, they are not completely linear. It will be straightened out by NFB, but why introduce something that has to be straightened out in the first place, unless it is absolutely needed.
BTW, you would seem to need a coupling cap at the output, that's the third point.
I understand the hypothesis about forcing an opamp into class A making things better, but there are a number of issues with it. Looking at you schematic, you are introducing three issues imo.
Firstly, you put a capacitor in the signal path (the 10 uF coupling fb to ground) that otherwise wouldn't be there. Now, I am not affraid of using coupling caps myself, but in the spirit of audiophile purity, this should be a no go.
Secondly, I looked at the specs of the current limiting diodes from your earlier post. As good as they are, they are not completely linear. It will be straightened out by NFB, but why introduce something that has to be straightened out in the first place, unless it is absolutely needed.
BTW, you would seem to need a coupling cap at the output, that's the third point.
I can recommend two good solutions for limited space headphone amps (prehaps going too far in terms of space). These both sound pretty good. First this Audio - Headphone Amplifier - TPA6133A2 - TI.com its programmed to start at full volume so you do not need to program it for volume to use it. I badgered TI to modify the TPA6130 to do this. The AKM is another option: AK4201EU | Product | AKM - Asahi Kasei Microdevices that has sounded good.
If you have a little real estate and can handle really wideband design rules (copy the demo board) the TPA6120 Audio - Headphone Amplifier - TPA6120A2 - TI.com can sound very good. I use a paralleled R/L on the output like a poweramp. It will run from the existing power supplies and drive any remotely normal dynamic headphone. The 100 MHz power band will be the design challenge.
If you have a little real estate and can handle really wideband design rules (copy the demo board) the TPA6120 Audio - Headphone Amplifier - TPA6120A2 - TI.com can sound very good. I use a paralleled R/L on the output like a poweramp. It will run from the existing power supplies and drive any remotely normal dynamic headphone. The 100 MHz power band will be the design challenge.
the TPA6120 is CFA - so you have the opportunity to overcompensate by simply increasing the feedback R - I used 5 kOhm feedback for unity gain
the real challenge is soldering it down to enough Cu to use the potential power
I used a largish (1812) chip inductor rated for >2 A to decouple/isolate the output from destabilizing cable Cload
the real challenge is soldering it down to enough Cu to use the potential power
I used a largish (1812) chip inductor rated for >2 A to decouple/isolate the output from destabilizing cable Cload
Last edited:
It is a shame that the TPA6120A isn't an all in one solution , it has wonderful spec .
I quite like my little idea . Like throwing a clay pot I am just getting it on the wheel in my sketch . It might not be so daft to do it this way . 1 to 2 mV DC offset is not too bad . Of all my options it is the space saver . If Andrew T reads this I wonder if he would see any value in trying to get 4.5 mA of class A , the 1 mW figures suggest is might be ? I am using a J FET . If I was to use a resistor it could be more . The load is taken to be 32 R . There is some debate as to how to calculate headphone output so don't take my figures as correct . AKG say the K1000 will go loud . It can take 11 Vrms . It will be a 8 pin DIL socket if using this option . Often an OP2604 or whatever will do a better job .
I didn't look at TPA6133A as to PSU . Can the charge pump be replaced by a conventional supply ?
If I ditch the feedback cap the output will need to be 470 uF non polar . It might be safer to do that .
I did the obvious and hard biased various op amps into class A . They seldom shine when doing this . The surprise was LM324 ( LM358 tried ? ) . It looks a bit like a comparator inside . The major bad points being noise and crossover distortion ( harder to find than people say ) . Pulling it into class A made it look rather nice . I did get very low distortion at 22 kHz at gain of 5 . It was 30 seconds to install so was for fun .
One thing that scares me is designing something then already it is obsolete . If NJM 4556 goes OBS there are perhaps 10 obvious substitutes . The dreaded NE 5532 would be usable . One thing that sort of works is TLO74 ( too big ) . Despite every text book saying no it can be hooked up as one big op amp . Trying is as it should be made no difference . It's primitive current limiting makes it instantly able to do this . None of the JFET's demands input current , sharing feedback resistors seems OK . I guess that's what you would expect . I thought DC offset would be better with separate feedback and 47R output resistors . Nothing obvious . The layout of quad op amps makes this easy .
Here is a table relating to a healthy op amp . It was a feed-forward zero bias current booster ( 20 R feed-forward BC327/337 ) . The figures are related to where the op amp still supplies the load ( say 0.5 V rms for safety if 32 R ) . At 1 kHz it looks very good . At 10 kHz a bit awful if looking harder . However will 10 kHz > 0.5 V rms ever happen in real music ? The distortion was below - 80dB if 1 kHz at any volume . It is possible to reconfigure this as class AB . If so 2 x 15 R emitter resistors ( 0R if zero bias class B ) , 2 x 1N4148 , 2 x 4K7 , same feed-forward ( 100 R works fine ) . At +/- 14 V . BTW +/- 14.3 V is the regulated power at that point . I have 4 VA to spare .
A speculation . Any op amp of good design into 20 K will be exactly as class A ? The speculation being that all good op amps are sufficiently biased to be in the A off AB ? The output standing current of lets say 2.5 mA being enough ?
I quite like my little idea . Like throwing a clay pot I am just getting it on the wheel in my sketch . It might not be so daft to do it this way . 1 to 2 mV DC offset is not too bad . Of all my options it is the space saver . If Andrew T reads this I wonder if he would see any value in trying to get 4.5 mA of class A , the 1 mW figures suggest is might be ? I am using a J FET . If I was to use a resistor it could be more . The load is taken to be 32 R . There is some debate as to how to calculate headphone output so don't take my figures as correct . AKG say the K1000 will go loud . It can take 11 Vrms . It will be a 8 pin DIL socket if using this option . Often an OP2604 or whatever will do a better job .
I didn't look at TPA6133A as to PSU . Can the charge pump be replaced by a conventional supply ?
If I ditch the feedback cap the output will need to be 470 uF non polar . It might be safer to do that .
I did the obvious and hard biased various op amps into class A . They seldom shine when doing this . The surprise was LM324 ( LM358 tried ? ) . It looks a bit like a comparator inside . The major bad points being noise and crossover distortion ( harder to find than people say ) . Pulling it into class A made it look rather nice . I did get very low distortion at 22 kHz at gain of 5 . It was 30 seconds to install so was for fun .
One thing that scares me is designing something then already it is obsolete . If NJM 4556 goes OBS there are perhaps 10 obvious substitutes . The dreaded NE 5532 would be usable . One thing that sort of works is TLO74 ( too big ) . Despite every text book saying no it can be hooked up as one big op amp . Trying is as it should be made no difference . It's primitive current limiting makes it instantly able to do this . None of the JFET's demands input current , sharing feedback resistors seems OK . I guess that's what you would expect . I thought DC offset would be better with separate feedback and 47R output resistors . Nothing obvious . The layout of quad op amps makes this easy .
Here is a table relating to a healthy op amp . It was a feed-forward zero bias current booster ( 20 R feed-forward BC327/337 ) . The figures are related to where the op amp still supplies the load ( say 0.5 V rms for safety if 32 R ) . At 1 kHz it looks very good . At 10 kHz a bit awful if looking harder . However will 10 kHz > 0.5 V rms ever happen in real music ? The distortion was below - 80dB if 1 kHz at any volume . It is possible to reconfigure this as class AB . If so 2 x 15 R emitter resistors ( 0R if zero bias class B ) , 2 x 1N4148 , 2 x 4K7 , same feed-forward ( 100 R works fine ) . At +/- 14 V . BTW +/- 14.3 V is the regulated power at that point . I have 4 VA to spare .
A speculation . Any op amp of good design into 20 K will be exactly as class A ? The speculation being that all good op amps are sufficiently biased to be in the A off AB ? The output standing current of lets say 2.5 mA being enough ?
@vacufile
While I agree in principle with your notes, that "spirit of audiophile purity" has let me down so many times that I dislike even hearing it.
"Audiophile purity" is just as full of land mines as popular perceptions, if not more. Parts from various manufacturers are often given a cult status for no obvious reason, simply because somebody made something using that part which sounds really great. They fail to note that its great sound does not come from that part, but from a well balanced whole. Or, lately, from a great sculptured aluminium case with a hefty price tag.
Once again, I note my own H/K Citation 24 power amp. Relatively expensive in the absolute, but way down below many a "High End" device with similar specs, it is still the most neutral amp I have ever heard, at least in my system, and my speakers are an exceptionally easy load to drive. And it is a capacitor coupled design, and worse still, its only capacitor is a simple electrolytic of just a bit better quality than most right at the input.
By comparison, I have heard a slew of DC designs which just didn't cut it. Despite their astronomical price tags and very revered names. By way of example, the Accuphase models I have heard, admittedly the lower portfolio models, failed to impress me in any significant way - solid products, very well built, but hardly exciting.
What did surprise me was a humble little Toshiba/Aurex model SB-A45, a cheap amp by any account, delivering modest 40W/8 Ohms. While it didn't exactly blow me away, it did impress me because it did blow away a surprising number of its competitors, costing its price several times over, on sound quality. It simpley boxes way above its class, if you can live with its natural limitations (e.g. its phono section is below the rest in sound quality). It's an AC design.
Therefore, I wouldn't put much stock in "audiophie purity". I have been ambushed by designs which sounded as they had no right to sound, and have been let down by designs which should have sounded way better than they did, far too many times.
While I agree in principle with your notes, that "spirit of audiophile purity" has let me down so many times that I dislike even hearing it.
"Audiophile purity" is just as full of land mines as popular perceptions, if not more. Parts from various manufacturers are often given a cult status for no obvious reason, simply because somebody made something using that part which sounds really great. They fail to note that its great sound does not come from that part, but from a well balanced whole. Or, lately, from a great sculptured aluminium case with a hefty price tag.
Once again, I note my own H/K Citation 24 power amp. Relatively expensive in the absolute, but way down below many a "High End" device with similar specs, it is still the most neutral amp I have ever heard, at least in my system, and my speakers are an exceptionally easy load to drive. And it is a capacitor coupled design, and worse still, its only capacitor is a simple electrolytic of just a bit better quality than most right at the input.
By comparison, I have heard a slew of DC designs which just didn't cut it. Despite their astronomical price tags and very revered names. By way of example, the Accuphase models I have heard, admittedly the lower portfolio models, failed to impress me in any significant way - solid products, very well built, but hardly exciting.
What did surprise me was a humble little Toshiba/Aurex model SB-A45, a cheap amp by any account, delivering modest 40W/8 Ohms. While it didn't exactly blow me away, it did impress me because it did blow away a surprising number of its competitors, costing its price several times over, on sound quality. It simpley boxes way above its class, if you can live with its natural limitations (e.g. its phono section is below the rest in sound quality). It's an AC design.
Therefore, I wouldn't put much stock in "audiophie purity". I have been ambushed by designs which sounded as they had no right to sound, and have been let down by designs which should have sounded way better than they did, far too many times.
The NAD 3020 was the biggest pile of junk ever . However it sounded pretty dam good and was reliable . Nothing about the circuit , build nor components said " this will work ". When NAD had their own factory Proton made a better NAD . Problem is it wasn't . Thin , harsh and pale sounding by comparison . All they had done was fit Audiophile on a budget type components . I never completely understand how they got it so wrong . It is not rare .
A little Onkyo amp used a Sanyo thick film chip output stage . Nothing about the amp said Audiophile . Many amps I strongly dislike have those Sanyo chips ( Sony ) . The Onkyo was excellent . It had it's own sound which is the Onkyo house sound . Interesting they can do that when using a Sanyo chip . A Sony with the same being the usual dreadful Sony amp . Sony TA1010 , 88 , 5650 being exceptions . I said to someone why is it Sony make OK cheap amps and dreadful expensive amps ? The reason seemed to be the young engineers did the cheaper stuff . They they went into TV leaving their not so gifted boss to do the big stuff .
We had a Sony amp which could give > 100 watts 8 R ( 170 W ) . It was awful . When tested into 2 ohms on bursts it gave 2 watts ( drum beat duration ) . NAD 3020 on the same test 92 watts . The pile of junk won .
A little Onkyo amp used a Sanyo thick film chip output stage . Nothing about the amp said Audiophile . Many amps I strongly dislike have those Sanyo chips ( Sony ) . The Onkyo was excellent . It had it's own sound which is the Onkyo house sound . Interesting they can do that when using a Sanyo chip . A Sony with the same being the usual dreadful Sony amp . Sony TA1010 , 88 , 5650 being exceptions . I said to someone why is it Sony make OK cheap amps and dreadful expensive amps ? The reason seemed to be the young engineers did the cheaper stuff . They they went into TV leaving their not so gifted boss to do the big stuff .
We had a Sony amp which could give > 100 watts 8 R ( 170 W ) . It was awful . When tested into 2 ohms on bursts it gave 2 watts ( drum beat duration ) . NAD 3020 on the same test 92 watts . The pile of junk won .
Well, it's all about the full package, the whole shamozzle ... I always look at the gestalt, the complete shebang - this either works or doesn't, in subjective terms - it's almost like a light switch going on or off, as to whether it's OK or not.
If the first round of using, listening shows a positive result then I hit it with some stress tests: up the volume, more and more; feed it difficult recordings - old, poorly transcribed shellac material; super compressed, in your face pop ... work out where the limits, for that setup, are ...
If the first round of using, listening shows a positive result then I hit it with some stress tests: up the volume, more and more; feed it difficult recordings - old, poorly transcribed shellac material; super compressed, in your face pop ... work out where the limits, for that setup, are ...
Unfortunately, dormant. Health issues, specifically my extreme hardship in walkning, blocked its completion. All the lab tests and various doctors have consumed enormous chunks of my time.
In the end, it seems to have paid off. My neurologist says I have problems with nerve endings in my legs caused by diabetes 1 (the kid brother). He has dished out some rather simple therapy and by George, it's working, and then some. So far, my recovery rate is better than 100 V/uS, third day of therapy and I feel way better.
The trouble with diabetes is that it can develop and do damage even if it itself is not an issue; my blood sugar seems almost quartz locked at 7, which is a bit above the expected notm of 6...6.5, but at 6, I have no idea which planet I'm on, by organism is in perfect state at 6.8...7.2, it's been that way for the last 50 years or so. And so it is today. I might not have even known if at one point it did'nt go up to 24.7, meaning I was clinically dead. I got that in check in just a week or so, and with proper medication it hasn't moved since then. Not being a sweet tooth, the things I had to give up were things I hardly ever used anyway.
In effect, my birth certificate date is catching up with me.
I feel somehow accomplished to say I am "old school". Which means I can get back to the Centurion. It is, to me, a very important project, and in several ways. I simply have to see to the end. For once, I paid much attention to rhyme and reason, rather than use materials and damn the cost. Not stingy, but I honestly tried to be rational about it. And of course, I would love to hear it come on song, even if the development process takes more time and effort.
And it has yielded some interesting offshoots. My power indicator board, for example. It's a standalone board, with its own voltage regulation, which can be used almost universally, with proper adjustment. It has connections for 4 LEDs. The first is a simple power on LED, which also tells you if the + supply is functional. The second LED tells you there is some signal input to the channel it's monitoring, as low as 100 mV, i.e. whether the channel is generally operationsl. The third tells you you are approaching what you have adjusted for to be your nominal output in terms of voltage, and the fourth tells you that you are soon going into the clipping range, so watch out. It's based on an National Semi comparator, which is cheap and widely available, with some passive components and just 1 medium power transistor, which serves as the local voltage regulator. Good for output stage voltages of 18...63 V, which I think covers like 98% of all power amp stages out there.
, cheers,Hi Dvv . I am boarder line with blood sugar . I am told some people eat oats to help control things .
Benefits of Oatmeal – Blood Sugar Control & Diabetes Prevention - Sharecare
I the past oats were used to heal wounds . I have seen this in real life , you would not believe how quickly it works if a bath of oats is used daily . They also help do the same inside us . They are very cheap and to me nice to eat . One problem with artificial sweeteners is they still encourage the body to think it eat sugar . Insulin boosters also are doubtful for the so called type 1 diabetes .
Thanks everyone who helped over headphone amp choice . I will test the cowards version today . To hand I have MC33078 , TL072 , NE55342 , OPA2604 , LM358 . The latter included to test the validity of the class A concept . None will have the 2 V rms 32 R I hope for . NJM 4556 ( AD ) is unity gain stable so should suit a gain of 6 very well . I will include a 1 R series resistance as it might assist ( A la Naim NAP 250 1/4 ohm ) . 32 R 30 mH is what a headphone might be . If 1 to 2 mV DC offset I will chance it being OK . The 10 K I show is the pot . That will have to be it's + input resistance path . I might fit 100 K to have a fall back . Even that challenges PCB space .
I am ill health semi retired . What it means is I design and do repairs . No full time building of stuff . In many ways I never was as happy as I am now . I always try to use full size components as it helps me and it helps availability . Usually capacitors are the bulky items more than DIL or 0.25 watt resistors .
I talk to three very different groups of Audio people . High end Audio , Pro Audio , DIY Audio . If religion it might be Jewish , Christian , Muslim . All from the same route yet wildly different . I am gradually learning where I am when in each group . Tonight I will drink with Pro Audio friends . You could say all the same deity , different visions of correctness . E.g . I wanted to borrow my Pro Audio friends active crossover . It is fourth order so didn't bother . His mini DSP can be asked to do 1,2,3,4 order ( he didn't know until he looked it up ) . This brings into doubt a lot of what John says . He maintains that Pro Audio speakers will eat hi fi speakers for breakfast if run at hi fi levels . I would say yes if the order of crossover could be changed . To John 4th order means the speakers survive . He owns Meridian active speaks at home . He says all PA gear these days will eat the hi fi versions . He means the best examples when saying that . C-Audio of old is his recommendation for a dual life amplifier bargain . He says anyone who disagrees has a powerful imagination . Pro Audio = Islam ? John say Meyer has a computer program to design the exact set up of a vertical array . This might mean one unit slightly out of line to get it perfect . The room data will be entered to do this . Same rig at a different venue will be set up differently . The adjustment not unlike a satellite dish .
Benefits of Oatmeal – Blood Sugar Control & Diabetes Prevention - Sharecare
I the past oats were used to heal wounds . I have seen this in real life , you would not believe how quickly it works if a bath of oats is used daily . They also help do the same inside us . They are very cheap and to me nice to eat . One problem with artificial sweeteners is they still encourage the body to think it eat sugar . Insulin boosters also are doubtful for the so called type 1 diabetes .
Thanks everyone who helped over headphone amp choice . I will test the cowards version today . To hand I have MC33078 , TL072 , NE55342 , OPA2604 , LM358 . The latter included to test the validity of the class A concept . None will have the 2 V rms 32 R I hope for . NJM 4556 ( AD ) is unity gain stable so should suit a gain of 6 very well . I will include a 1 R series resistance as it might assist ( A la Naim NAP 250 1/4 ohm ) . 32 R 30 mH is what a headphone might be . If 1 to 2 mV DC offset I will chance it being OK . The 10 K I show is the pot . That will have to be it's + input resistance path . I might fit 100 K to have a fall back . Even that challenges PCB space .
I am ill health semi retired . What it means is I design and do repairs . No full time building of stuff . In many ways I never was as happy as I am now . I always try to use full size components as it helps me and it helps availability . Usually capacitors are the bulky items more than DIL or 0.25 watt resistors .
I talk to three very different groups of Audio people . High end Audio , Pro Audio , DIY Audio . If religion it might be Jewish , Christian , Muslim . All from the same route yet wildly different . I am gradually learning where I am when in each group . Tonight I will drink with Pro Audio friends . You could say all the same deity , different visions of correctness . E.g . I wanted to borrow my Pro Audio friends active crossover . It is fourth order so didn't bother . His mini DSP can be asked to do 1,2,3,4 order ( he didn't know until he looked it up ) . This brings into doubt a lot of what John says . He maintains that Pro Audio speakers will eat hi fi speakers for breakfast if run at hi fi levels . I would say yes if the order of crossover could be changed . To John 4th order means the speakers survive . He owns Meridian active speaks at home . He says all PA gear these days will eat the hi fi versions . He means the best examples when saying that . C-Audio of old is his recommendation for a dual life amplifier bargain . He says anyone who disagrees has a powerful imagination . Pro Audio = Islam ? John say Meyer has a computer program to design the exact set up of a vertical array . This might mean one unit slightly out of line to get it perfect . The room data will be entered to do this . Same rig at a different venue will be set up differently . The adjustment not unlike a satellite dish .
Nah, you're talking DJ's.....IME they know their music, but they all have cloth ears, and technically wouldn't know shi_ from clay.Except, Pro Audio when it's wrong, is very, very wrong - excrutiatingly diabolical to listen to - a better word than "listen" is "survive" ... yes, indeed, I would happy to travel down that very bumpy road in a car with rock hard springs, and broken shock absorbers ...
Modern big event audio really is pro-audio nowadays.....the system hang and placement are predicted long before the show, the systems are good, and so are the FOH operators.
Pub/beer garden/restaurant audio can be amateur, very amateur, especially those on stage mixes.
Dan.
- Status
- Not open for further replies.