Lme49600 headphone amp - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 18th May 2010, 08:50 PM   #11
diyAudio Member
 
theAnonymous1's Avatar
 
Join Date: Feb 2004
Location: Anonymityville
Quote:
Originally Posted by Sebastiaan View Post
Sorry, I didn't look at this sheet yet, I see it now.
No problem, I was just confused as it seemed like that circuit was being intentionally disregarded here for some reason.

I've built it already and it performs very well. I can't imagine ever needing more than one buffer per channel for headphones though.
  Reply With Quote
Old 18th May 2010, 08:53 PM   #12
diyAudio Member
 
Join Date: Jun 2004
Location: the Netherlands
Quote:
Originally Posted by theAnonymous1 View Post
No problem, I was just confused as it seemed like that circuit was being intentionally disregarded here for some reason.
Oh no, I like the circuit from the sheet! A good designed servo with equal legs for both inputs from the servo input. One can never go wrong with this circuit and must sound good. If the OP can live with a gain of 3, I would highly recommend to copy/paste this circuit. If he want to stick with a voltage gain of 20, then some adjustments need to make to the servo circuit.

With kind regards,
Bas
  Reply With Quote
Old 18th May 2010, 09:05 PM   #13
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
Default Toto, I've a feeling we're not in Kansas any more. ...

I'm afraid some of the suggestions are past their "use by date" - when you're wrapping a 50+MHz GBW op amp's feedback loop around a 180 MHz buffer some of yesteryears app note and published headphone amp project topologies need modification

Headphone cable C can surprise you - 1 nF is not impossible even in the manufacturer's standard cable due to flexible cables putting the wires very close together - custom cables of exotic construction can be worse

Load C is a bigger problem as loop intercept frequencies rise - what mostly worked with 4-8 MHz GBW op amps may not with 6-10x faster parts

resistance in series with the output, inside the feedback loop is definitely a bad idea unless you know your load C and add some more parts to define a 2nd order filter function - and you'd want a 50+ MHz GBW oscilloscope to verify that you were successful


much smaller R after the feedback point, in series with the load, helps damp the reaction of cable load C with the complex (rising with frequency ~+ inductance) output impedance of feedback amplifiers; 10-25 Ohms if you don't mind the R in series with the headphone - or use lossy ferrite or parallel R,L like many SS audio power amps

series R at the input of the buffer can be good - the LME data sheet claims it is stable with any C load if terminated in 50 Ohms on the input
you need to rearrange the earlier suggested series R to be effective at much higher frequencies - as a RC "Zobel" to gnd instead

likewise toss out any "audiophile capacitor" nonsense for the ps bypass at the chips - you need ultra low parasitic inductance - use monolithic ceramic chip caps at the op amp, buffer supply pins - to a solid gnd plane - the "order of magnitude scaled parallel bypass caps" rule of thumb also should be looked at with suspicion - at these frequencies you can get resonant peaks between the SRF of the different type/value caps that can make ps worse rather than better

Last edited by jcx; 18th May 2010 at 09:09 PM.
  Reply With Quote
Old 18th May 2010, 09:12 PM   #14
diyAudio Member
 
Join Date: Jun 2004
Location: the Netherlands
Quote:
Originally Posted by jcx View Post
I'm afraid some of the suggestions are past their "use by date" - when you're wrapping a 50+MHz GBW op amp's feedback loop around a 180 MHz buffer some of yesteryears app note and published headphone amp project topologies need modification

Headphone cable C can surprise you - 1 nF is not impossible even in the manufacturer's standard cable due to flexible cables putting the wires very close together - custom cables of exotic construction can be worse

Load C is a bigger problem as loop intercept frequencies rise - what mostly worked with 4-8 MHz GBW op amps may not with 6-10x faster parts

resistance in series with the output, inside the feedback loop is definitely a bad idea unless you know your load C and add some more parts to define a 2nd order filter function - and you'd want a 50+ MHz GBW oscilloscope to verify that you were successful


much smaller R after the feedback point, in series with the load, helps damp the reaction of cable load C with the complex (rising with frequency ~+ inductance) output impedance of feedback amplifiers; 10-25 Ohms if you don't mind the R in series with the headphone - or use lossy ferrite or parallel R,L like many SS audio power amps

series R at the input of the buffer can be good - the LME data sheet claims it is stable with any C load if terminated in 50 Ohms on the input
you need to rearrange the earlier suggested series R to be effective at much higher frequencies - as a RC "Zobel" to gnd instead

likewise toss out any "audiophile capacitor" nonsense for the ps bypass at the chips - you need ultra low parasitic inductance - use monolithic ceramic chip caps at the op amp, buffer supply pins - to a solid gnd plane - the "order of magnitude scaled parallel bypass caps" rule of thumb also should be looked at with suspicion - at these frequencies you can get resonant peaks between the SRF of the different type/value caps that can make ps worse rather than better
Oeps. I am not familiar with those opamps, and I didn't know they have such a high band-with open-loop gain. In that case it would't be a bad idea to compensate the design a little by simply adding a small capacitor from the drivers opamp output to the negative input to reduce the slew-rate and bandwidth to more a more safe zone.

With kind regards,
Bas
  Reply With Quote
Old 18th May 2010, 09:24 PM   #15
diyAudio Member
 
theAnonymous1's Avatar
 
Join Date: Feb 2004
Location: Anonymityville
Is all of that really necessary though?

I don't have a scope, but the board I built has zero instability problems that I can tell. I've driven multiple different headphones and even small speakers with it and it never misbehaves.

I'm not an expert with this stuff, but I do know that reality is sometimes very different than simulations and assumptions.
  Reply With Quote
Old 18th May 2010, 10:13 PM   #16
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
you do have some added stability margin with a high amplifier gain of 10 - but your loop intercept frequency is ~ 5 MHz - higher than most discrete power amps - including most Mosfet output designs, so you should be prepared for similar problems - many "diy safe" project amps have loop intercept frequency below 500 KHz -as you would in a op amp project with a older 4-8 MHz GBW op amp

feedback circuits can "work", giving "acceptable" audio even while oscillating, possibly at a level that doesn’t heat the chips too much

you really can't be sure of a design unless you’ve looked at it with a 'scope faster than the fastest device in the feedback loop - look for Bob Pease comments on practical stability testing and Linear Technology's High Speed Op Amp design app notes

R in series with the output, inside the feedback loop, reacting with load C is accurately modeled - 100 Ohms * 1 nF ~= 1.6 MHz, you would be adding >70 degrees phase shift to the loop at 5 MHz => over 2x overshoot (armchair guestimate - it can be calculated/simmed to as much accuracy as you know the op amp GBW at those frequencies)

the local feedback C around the input op amp does slow the loop and can tune the damping of the Cload overshoot - but it continues the feedback of the op amp to unity gain at 50 MhZ - requiring good ps bypass performance to beyond that frequency

you may not need to apply every high speed design trick for some instances of these circuits to work - but you are walking near the edge and may not realize when a "simple" change takes you over into high speed instability

Last edited by jcx; 18th May 2010 at 10:15 PM.
  Reply With Quote
Old 18th May 2010, 11:27 PM   #17
diyAudio Member
 
Join Date: Jun 2004
Location: the Netherlands
Dear,

For fun I made some quick simulation. I don't have Spice models from this national devices, but I used the OPA637/Buf634 instead. The Buf634 is in specs exactually equal to the LME49600. The Opa637 comes close to the LME49720.

This simulations are not the real world and this parts are not equal, but you can be sure with such a high open-loop bandwidth that the situation with this headphone amp comes close, or worser. Simulated is first the schematic from national, with 100 ohm load, and 100Ohm/1nF load. The second attachment I compensated the circuit slightly with 47pF as can be seen in the schematic. For my taste still a to big bandwidth to almost MHz! I would even limit it further.

Personally I would change the circuit and add a lowpass filter at the input (but resistor values have to change then in order to make it work).

The original circuit is working on the edge of oscillation I fear.

With kind regards,
Bas
Attached Images
File Type: png headamp2.png (73.1 KB, 787 views)
File Type: png headamp1.png (67.8 KB, 765 views)
  Reply With Quote
Old 18th May 2010, 11:28 PM   #18
diyAudio Member
 
Join Date: Jun 2004
Location: the Netherlands
I would like to add, that working with a bigger voltage gain would improve things drastically, and this is in the benefit of the OP who wanted to work with a voltage gain of 20. However I wonder if that is not way to much gain with modern sources to feed a headphone.

With kind regards,
Bas
  Reply With Quote
Old 19th May 2010, 01:15 AM   #19
diyAudio Member
 
theAnonymous1's Avatar
 
Join Date: Feb 2004
Location: Anonymityville
Hmmm, I'm curious now to know how much, if any, oscillation is taking place with my board.

Anyone here interested in putting in under their scope? You can keep it when your done.

jcx?
  Reply With Quote
Old 19th May 2010, 03:32 AM   #20
T in AZ is offline T in AZ  United States
diyAudio Member
 
Join Date: Jul 2009
Quote:
Originally Posted by Dr. Popo View Post
Does anybody have any suggestions as to how i should solder LME49600 to pc board? I am good at soldering but never have come across ic in power pad package!
How To Solder Surface Mount Devices & Chips With Solder Paste & Hot Air. SMD Soldering! I made a setup similar to this and I use a micro butane torch instead of a hot air pencil, a halogen spot on the under side with a dimmer for preheat.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
What's the difference Between LME49600 and LME49610? bashenan4455 Chip Amps 4 20th September 2011 04:30 AM
WTB: LME49600 preamp-board from ebay Nisbeth Swap Meet 5 6th April 2010 03:46 PM
Compound OpAmp w LME49600 jackinnj Parts 1 25th February 2010 05:36 PM
What's the difference Between LME49600 and LME49610? bashenan4455 Chip Amps 3 14th February 2010 12:22 PM
LME49600/LME49740 Headphome Amp theAnonymous1 Solid State 28 31st January 2010 03:25 AM


New To Site? Need Help?

All times are GMT. The time now is 01:38 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2