Go Back   Home > Forums > >

Headphone Systems Everything to do with Headphones

ESP P113 Modifications
ESP P113 Modifications
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
Old 3rd June 2021, 12:56 PM   #11
joensd is offline joensd  Germany
diyAudio Member
 
Join Date: Dec 2001
Location: Berlin, Germany
Of course you do not always have to study every circuit to the fullest before you build it but itīs good to understand the basics at least (and it is an opportunity to learn and maybe even have more fun)
In this case you are dealing with a composite amp and possible instability so either you are able to verify stability or stay very close to the recommendations/parts given. (though it is generally recommended to check stability, expected current draw etc.)

I built the "original" with a NE5534 and the performance was stellar.
Also notice the different biasing with the LED.
Build This High Quality Headphone Amplifier. – HeadWize Memorial

Last edited by joensd; 3rd June 2021 at 01:01 PM.
  Reply With Quote
Old 3rd June 2021, 01:08 PM   #12
BucketInABucket is offline BucketInABucket  Hong Kong
diyAudio Member
 
Join Date: Aug 2015
Location: Audio heaven
I actually did experiment with using LEDs to bias the circuit, I know red LEDs are exceptionally quiet parts.

Indeed, that's a very similar circuit, and they also use that small capacitance cap between opamp input and output like I mentioned I thought about earlier.

I'm new to amp design so I intend to check the basics qt least to make sure everything is functional. I'm definitely having fun doing this too, and I can't wait to actually build this circuit in real life.

Last edited by BucketInABucket; 3rd June 2021 at 01:17 PM.
  Reply With Quote
Old 3rd June 2021, 01:50 PM   #13
joensd is offline joensd  Germany
diyAudio Member
 
Join Date: Dec 2001
Location: Berlin, Germany
Quote:
they also use that small capacitance cap between opamp input and output
Thatīs miller compensation and is one of many compensation techniques used in composite amplifier designs. (dominant pole compensation, lead/lag-compensation you will see most often)

Quote:
can't wait to actually build this circuit in real life
Iīm sure you will like the amp!
  Reply With Quote
Old 3rd June 2021, 02:38 PM   #14
BucketInABucket is offline BucketInABucket  Hong Kong
diyAudio Member
 
Join Date: Aug 2015
Location: Audio heaven
Just realized I forgot to answer your uni question. Only a first year student here, got through the basic maths and physics courses, nothing specialized yet.
Quote:
Originally Posted by joensd View Post
Thatīs miller compensation and is one of many compensation techniques used in composite amplifier designs. (dominant pole compensation, lead/lag-compensation you will see most often)
Ah, makes a lot of sense. Only 22pF was enough to make the amp stable with a 1uF in parallel with the 32R load, quite amazing stuff. (Yes R5 is 2k2, forgot to change it heh).
Click the image to open in full size.
I also tried the LED bias method again, this time with some proper 1.7V red LED models, and I've decided to stick with it. Performs more than well enough for my requirements and I'm always down to use less parts.
Quote:
Originally Posted by joensd View Post
Iīm sure you will like the amp!
I hope so! I've put quite a lot of hours into it now

By the way, what are C3 and C4 for? Still haven't worked it out.

Edit:

Using bias LEDs have of course resulted in much higher Ic so emitters are going back to 10R. Still results in 1.18W/transistor which is more than before, and 2.59W output power.
Click the image to open in full size.

Last edited by BucketInABucket; 3rd June 2021 at 02:57 PM.
  Reply With Quote
Old 3rd June 2021, 03:29 PM   #15
BucketInABucket is offline BucketInABucket  Hong Kong
diyAudio Member
 
Join Date: Aug 2015
Location: Audio heaven
On the other hand taming a square wave with 600R/100n requires two 100p caps in parallel or a single 220p cap. Fair dues.
  Reply With Quote
Old 3rd June 2021, 03:51 PM   #16
joensd is offline joensd  Germany
diyAudio Member
 
Join Date: Dec 2001
Location: Berlin, Germany
Quote:
By the way, what are C3 and C4 for? Still haven't worked it out.
They couple the signal to the base of the transistors and stabilize bias voltages.

Quote:
Ah, makes a lot of sense. Only 22pF was enough to make the amp stable with a 1uF in parallel with the 32R load, quite amazing stuff.
Itīs nice to see those things working right? It is also very educational when you start doing AC-analyses with LTspice and look at the bode plots.
(keywords: open loop gain, closed-loop-gain, tian-probe; see links from knoppers)

But keep in mind that the opamp-model is not close to reality!
So the OPA1652 will behave very differently!
(just like the 0.1uF caps next to your voltage sources donīt do much other than being charged because they are connected to a (unrealistic) ideal voltage source)
A composite amplifier is also not a good place to start rolling opamps.

If I were you Iīd stick with one of the original circuits. (meaning same bias, same opamp; same everything actually and go from there if you like)
I havenīt simulated that circuit but I donīt think youīll ever need a LED instead of 1xdiode (2.1V instead of 0.6V). 30mA bias current through the transistors should be plenty. The Earle Eaton amp is different, it uses 1x LED for both base-emitter-junctions.
If you are doing your PCB, you could put 2 diodes in your schematic instead of one. Bridge one at first and you can "hotrod" your circuit later and install the second. (though Iīd guess that you wonīt need it)

Last edited by joensd; 3rd June 2021 at 03:57 PM.
  Reply With Quote
Old 3rd June 2021, 04:20 PM   #17
kn0ppers is offline kn0ppers  Germany
diyAudio Member
 
Join Date: Mar 2019
I think SPICE can be educational if you know what to expect and why, so you can actually interpret the simulation results. Otherwise you quickly end up chasing ghosts in a haunted castle. Also the models can be quite accurate, depending on what parameters we are talking about. But one must always remember the assumptions made, the limitations of models and simulation in general, the spread of parameters in the real world, layout parasitics etc.

About the simulation: You can't use 1kHz sinewave and a closed loop for loop stability analysis. Also 100nF is very unrealistic. I would stick with stepping the capacitance in the range from 100pF to a maximum of 10nF. Even the 10nF already represent spools full of cable.

Last edited by kn0ppers; 3rd June 2021 at 04:23 PM.
  Reply With Quote
Old 3rd June 2021, 08:30 PM   #18
BucketInABucket is offline BucketInABucket  Hong Kong
diyAudio Member
 
Join Date: Aug 2015
Location: Audio heaven
Quote:
Originally Posted by joensd View Post
They couple the signal to the base of the transistors and stabilize bias voltages.
Makes sense.
Quote:
Originally Posted by joensd View Post
Itīs nice to see those things working right? It is also very educational when you start doing AC-analyses with LTspice and look at the bode plots.
(keywords: open loop gain, closed-loop-gain, tian-probe; see links from knoppers)

But keep in mind that the opamp-model is not close to reality!
So the OPA1652 will behave very differently!
(just like the 0.1uF caps next to your voltage sources donīt do much other than being charged because they are connected to a (unrealistic) ideal voltage source)
A composite amplifier is also not a good place to start rolling opamps.

If I were you Iīd stick with one of the original circuits. (meaning same bias, same opamp; same everything actually and go from there if you like)
I did start from there in LTspice, and then I wanted to see what would happen if I increased the bias current, and then here I am. What a journey
Quote:
Originally Posted by joensd View Post
I havenīt simulated that circuit but I donīt think youīll ever need a LED instead of 1xdiode (2.1V instead of 0.6V). 30mA bias current through the transistors should be plenty. The Earle Eaton amp is different, it uses 1x LED for both base-emitter-junctions.

If you are doing your PCB, you could put 2 diodes in your schematic instead of one. Bridge one at first and you can "hotrod" your circuit later and install the second. (though Iīd guess that you wonīt need it)
I'm deliberately trying to bias the circuit in class A so that's why I'm using the LEDs instead of diodes

Quote:
Originally Posted by kn0ppers View Post
About the simulation: You can't use 1kHz sinewave and a closed loop for loop stability analysis. Also 100nF is very unrealistic. I would stick with stepping the capacitance in the range from 100pF to a maximum of 10nF. Even the 10nF already represent spools full of cable.
You're right really, I also used square waves and varied the capacitance too. I just wanted to see what worst-case was. Not sure how to do analysis without the closed loop though, how would I go about this?

I gotta say, PCB layout is my forte, not circuit design. This is fun though!

Edit: now that I think about it, I already did AC analysis: Click the image to open in full size.
Doesn't really show my modifications were detrimental though cause the original circuit was this:
Click the image to open in full size.

Last edited by BucketInABucket; 3rd June 2021 at 08:52 PM.
  Reply With Quote
Old 3rd June 2021, 09:14 PM   #19
joensd is offline joensd  Germany
diyAudio Member
 
Join Date: Dec 2001
Location: Berlin, Germany
Quote:
I'm deliberately trying to bias the circuit in class A so that's why I'm using the LEDs instead of diodes
I think the composite arrangement of these amps will result in a much lower THD than the bias currents alone would suggest.
Would love to see THD measurements between the standard version and one with higher bias. Donīt think the difference will be worth it.

If you want pure Class-A with good sound for pennys, you could try this one:
(Iīm sure it "breadboards" just fine)
3 Transistor HP Amplifier with low dist
  Reply With Quote
Old 4th June 2021, 11:05 AM   #20
BucketInABucket is offline BucketInABucket  Hong Kong
diyAudio Member
 
Join Date: Aug 2015
Location: Audio heaven
I'm taking your advice and have decided to just go for the original bias levels in the end, that will also let me fit it in the case I originally planned to use rather than a larger one with a lot of wasted space.
  Reply to this post

Reply


ESP P113 ModificationsHide this!Advertise here!
Thread Tools

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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
ESP P113 Compact Version BucketInABucket Headphone Systems 6 4th June 2021 11:09 AM
ESP P113 volume control problem NealJ Headphone Systems 9 21st February 2021 07:00 AM
Rod Elliot P113 headphone amp - balanced version? abza Headphone Systems 7 22nd February 2019 07:06 PM
Balanced HP amp using P113 - suggestions? abza Headphone Systems 2 30th October 2018 08:12 PM
ESP P113 voltage gain of 1, will it work? DanielNielsen Headphone Systems 4 12th February 2013 02:30 PM


New To Site? Need Help?

All times are GMT. The time now is 01:03 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio
Wiki