or Connect
Head-Fi.org › Forums › Summit-Fi (High-End Audio) › High-end Audio Forum › Speaker amps for headphones
New Posts  All Forums:Forum Nav:

Speaker amps for headphones - Page 116

post #1726 of 3116
Quote:
Originally Posted by robrob View Post

I did find another error in the formulas and fixed it. I also added the other 2 resistor network.



 



The latest is available here as always: http://robrobinette.com/images/Audio/Headphone_Resistor_Network_Calculator.xls


 



Ignore my previous post.
Most of the time I was using my work computer, which uses Excel.

Last night I used my home computer, which uses an Excel knock-off. It couldn't handle some of the formulas and I got wierd results! Bah! POS!

Just did a re-check with REAL Excel, works quite nicely!

Nice work, BTW!

Hopefully the fans will use it!
Edited by Chris J - 12/4/13 at 9:39am
post #1727 of 3116

Thanks Chris, I actually use LibreOffice on Windows and "Save as" .xls. It probably wouldn't hurt to offer the native .ods file.

post #1728 of 3116

Armaegis, could you take a look at the current version of the network calculator spreadsheet (beta8) and the JavaScript calculator page? I think I've incorporated all your suggestions and corrections.

 

Thanks in advance.

 

Everyone, I'm working on the watt rating calculations now. What should I use for the voltage output from the amp? I'm thinking I should do it like the Amp Output Impedance where I use a standard value but allow the user to change it if they know what it is for their amp.

 

What I've learned so far by playing around with different values:

 

With my 32 ohm HE-500 cans with my tube speaker amp that wants an 8 ohm speaker load the best solution is a 6 ohm Resistor2 and a 2 ohm Resistor3. It gives me an almost perfect Effective Speaker Load of 7.9 and a nice low Effective Headphone Output Impedance of 1.5 with -12.4dB of Attenuation. I was really surprised to see that changing the headphone impedance from 32 ohms to 600 ohms changed the speaker load by only 0.1 ohm and Attenuation by 0.3dB so a 6 & 2 ohm network will be optimal for most headphones and amps that need attenuation for hiss reduction or more volume knob movement.

 

I ordered the suggested resistors from Mouser and I'll see how my HE-500, K240 Studio and modified T50RP's sound with a direct connection, the current configuration of just a 10 ohm load resistor and the suggested resistor network.


Edited by robrob - 12/4/13 at 2:38pm
post #1729 of 3116
Rob rob excellent work.

Al D
post #1730 of 3116
Quote:
Originally Posted by robrob View Post
. I was really surprised to see that changing the headphone impedance from 32 ohms to 600 ohms changed the speaker load by only 0.1 ohm and Attenuation by 0.3dB so a 6 & 2 ohm network will be optimal for most headphones and amps that need attenuation for hiss reduction or more volume knob movement.

 

So is this why Mike's HD600s sound great with a resistor box initially designed for the low-impedance LCD-2s?

post #1731 of 3116

One more thing to point out when ordering resistors for this purpose is to make sure you get "wirewound" type. Here's a great simple article explaining why: http://www.digikey.com/Web%20Export/Supplier%20Content/Riedon_696/PDF/riedon-an-wirewound-vs-film.pdf?redirected=1

 

EDIT: Here's another good article on resistor noise, where again wirewound is what you want in this application

http://www.aikenamps.com/ResistorNoise.htm


Edited by brunk - 12/4/13 at 3:11pm
post #1732 of 3116
Resistor noise wouldn't be of any concern in this application. And if one doesn't mind doing a little hand selection, carbon composition resistors would also be quite suitable.

se
Edited by Steve Eddy - 12/4/13 at 3:31pm
post #1733 of 3116

I went with Vishay/Dale wirewound non-conductive resistors for the resistor network. The exact parts from Mouser were:

 

Mouser part# 71-RS0056R000FB12 Vishay part# RS0056R000FB12 5watts 6ohms 1%

 

Mouser part# 71-NS2B-2 Vishay part# NS02B2R000FB12 3watts 2ohms 1%

 

They were about $2 each with $5 shipping. I combined it with an order for a Hammond choke for my tube guitar amp.

post #1734 of 3116
Quote:
Originally Posted by robrob View Post
 

Armaegis, could you take a look at the current version of the network calculator spreadsheet (beta8) and the JavaScript calculator page? I think I've incorporated all your suggestions and corrections.

 

Thanks in advance.

 

Everyone, I'm working on the watt rating calculations now. What should I use for the voltage output from the amp? I'm thinking I should do it like the Amp Output Impedance where I use a standard value but allow the user to change it if they know what it is for their amp.

 

What I've learned so far by playing around with different values:

 

With my 32 ohm HE-500 cans with my tube speaker amp that wants an 8 ohm speaker load the best solution is a 6 ohm Resistor2 and a 2 ohm Resistor3. It gives me an almost perfect Effective Speaker Load of 7.9 and a nice low Effective Headphone Output Impedance of 1.5 with -12.4dB of Attenuation. I was really surprised to see that changing the headphone impedance from 32 ohms to 600 ohms changed the speaker load by only 0.1 ohm and Attenuation by 0.3dB so a 6 & 2 ohm network will be optimal for most headphones and amps that need attenuation for hiss reduction or more volume knob movement.

 

I ordered the suggested resistors from Mouser and I'll see how my HE-500, K240 Studio and modified T50RP's sound with a direct connection, the current configuration of just a 10 ohm load resistor and the suggested resistor network.

 

Believe it or not, it don't matter much.

You can assume 25 Volts for a starting point, which works out to an 80 Watt Amp into 8 Ohms.

But that's when the amp runs flat out.

 

For Watt ratings, R3 can be 1 Watt, R2 can be 5 Watts.

Keep in mind that average listening level is going to be approx. 1 or 2 Watts from the power amp.

This will work out to between 3 to 5 Volts average music signal outputted by the power amp.

Probably even less.

post #1735 of 3116
Quote:
Originally Posted by Steve Eddy View Post

Resistor noise wouldn't be of any concern in this application. And if one doesn't mind doing a little hand selection, carbon composition resistors would also be quite suitable.

se

 

Yes, fellas, resistor noise is irrelevant in this application.

It will be utterly swamped by the noise from your power amp.

post #1736 of 3116

Thanks, Chris. I'll probably just stick with what's on the spreadsheet right now then. It recommends a 5 watter for R2 and a 3 watter for R3.

post #1737 of 3116
Quote:
Originally Posted by robrob View Post
 

Thanks, Chris. I'll probably just stick with what's on the spreadsheet right now then. It recommends a 5 watter for R2 and a 3 watter for R3.

 

Sorry,  :o 

I forgot to mention, I think that your power value recommendations are more than close enough for this application!

post #1738 of 3116

Yes, thanks Rob for trailblazing this spreadsheet!

 

So, I'm about to start testing my 50-Ohm wirewound resistors, as previously discussed in this thread, with the hope of sufficiently eliminating MG3 hiss, with LESS attenuation than that suffered with my TBI-designed, dual-resistor network.

 

 

I'm surprised to see that amplifier attenuation is calculated as -28.6 dB.   

 

Compare this to the results shown for my current TBI dual-resistor network:

 

 

In the interest of eliminating hiss, I'm currently suffering -15.8 dB attenuation with the dual-resistor TBI design, but the 50-Ohm resistors are going to give me even more attenuation?  -28.6 dB?  

 

I wanted to go the other way - less attenuation.

 

Anybody want to buy $15.00 worth of resistors?  I think I'm done with these, before even getting out the soldering iron.

 

Never mind...  there must be a "bug"  in the spreadsheet: Changing the value of R3 should not affect the results of calculations for the "Attenuation Only" section, where only one resistor is used (R2).

 

When I apply the workaround of specifying a value of "0" for R3, the "Attenuation Only" section looks like this (vs. that seen above when R3 was not set to 0):

 

 

That's much better!  And now, to test it with my ears, expecting to hear -6 dB attenuation, instead of -15.8 dB...

 

:o

 

Mike

post #1739 of 3116

Mike, you have an old version of the spreadsheet. The current one that's linked shows -6dB for the "Attenuation Only" and "Preferred Headphone Interface" (with R3=0). The current version is beta7 shown in the upper right.

post #1740 of 3116
Quote:
Originally Posted by robrob View Post
 

Armaegis, could you take a look at the current version of the network calculator spreadsheet (beta8) and the JavaScript calculator page? I think I've incorporated all your suggestions and corrections.

 

Thanks in advance.

 

It's so hard to make sense of your formulas with all those nested logic statements... I don't have time to pick through it all, but this one sticks out:

 

The "Attenuation Only" scenario: formula for amplifier attenuation is incorrect, it should ignore R3 entirely. I think that's why Mike was getting funny numbers; it looks like it's calculating stuff the same way as "preferred" scenario.

 

I can make a function/macro in libreoffice to clean up the equations... but I don't think it translates into excel.


Edited by Armaegis - 12/4/13 at 5:52pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: High-end Audio Forum
Head-Fi.org › Forums › Summit-Fi (High-End Audio) › High-end Audio Forum › Speaker amps for headphones