[intlplby]

exactly how much of an impact does making a PPA class A impact the battery life of the amp when used portably?

 

[amb]

Quote:

Originally Posted by intlplby exactly how much of an impact does making a PPA class A impact the battery life of the amp when used portably?

"Class A" is a fuzzy term unless you qualify it with an expected peak output current. Such current is determined by the maximum voltage swing you require and the headphone impedance that amp is going to drive. Then you set the quiescent current to half the peak output current to make sure the amp remains in class A for those given conditions. While in class A the output transistors will not "turn off", thus completely eliminating crossover distortion. The price you pay is increased heat dissipation and power consumption.

Note that with PPA v1 (with the HA5002 buffers) you cannot vary the quiescent current other than in four discrete steps, determined by the number of buffers you install. In PPA v2 with the discrete diamond buffer you will be able to set the current by adjusting a trimpot.

 

[tangent]

If you're referring to the op-amp biasing, then how much impact it has depends on how you adjust R9. You can go anywhere from very low current up to high current, over a good 10:1 range, around a few mA.

 

[steinchen]

Quote:

Originally Posted by tangent If you're referring to the op-amp biasing, then how much impact it has depends on how you adjust R9. You can go anywhere from very low current up to high current, over a good 10:1 range, around a few mA.

Is there a rule of thumb how much opamp bias is reasonable or does it rather depend on the opamp how much bias is needed to improve distortion ? I've tried just a few settings on my PPA, to my ears it seemed that classA biasing improves sound considerably (with any opamp I've tried so far), the amount of bias (0.5mA - 1.5mA) makes no audible difference (for me). Is there a technical explanation to this ? Does anybody share my experience ?

Has biasing the opamp any impact on the buffers or is it confined to the opamp ?

 

[__redruM]

I've asked this before as well. What are the trade-offs in setting R9? I'm at the 1ma stock setting, and battery life is over 5 hours, But I haven't let the amp run all the way to empty yet. My problem is the Zener Diode for my LED cuts out after about 3 hours, but the amp is still going strong.

My setup is: 12 AAA at 700mah. With a 1 ma bias. My Zener is 11V.

Is there any benifit to raising the bias setting? What about a different bias for the ground channel? Example L and R at 1.5ma and ground at 0.5ma. Has anyone who has really played with this gotten better sound with a value other than 1ma per channel?

I'm thinking about changing the Zener diode to 9V. 11V sounds like a reasonable setting, but with the amp running, the voltage level drops somewhat. So even though the cells may still have 1.0-1.2 volts left, the voltage drop from running the amp cuts the light out early.

 

[Excalibur]

This is weird. I just asked Tangent about this same subject yesterday! I'm glad this thread started so I can get some ideas on the best setting for R9. I would imagine it would be slightly different for different types of opamps. Anyway, if someone has a favorite "procedure" I would be glad to hear about it.

 

[intlplby]

lets say my load was 32ohms so i would need a high bias of like 22mA

what would be the impact on battery life

22mA isnt too high for OPA627s is it?

 

[__redruM]

Quote:

Originally Posted by intlplby lets say my load was 32ohms so i would need a high bias of like 22mA

How'd you get 22mA? What formula are you using? Is this 22mA on each channel or the sum of all three channles. I'd guess that you'd need a 44V power supply to get 22mA bias in each channel. Given that each volt dropped across R10 is 1 mA.

 

[tangent]

Quote:

Is there a rule of thumb how much opamp bias is reasonable

Not that I've ever seen. Everything I've ever heard on the subject is that, in the end, it's subjective. That's one reason why we went with adjustable bias in PPA v1.1, thinking that people would start to experiement and report on their results. It's so easy, yet I have yet to see any results. I guess people are just using my 1mA default and leaving it at that.

Quote:

Has biasing the opamp any impact on the buffers

Nothing significant.

Quote:

My problem is the Zener Diode for my LED cuts out after about 3 hours, but the amp is still going strong.

I can see that the docs weren't clear enough on picking the zener value, then. I've rewritten that section to be clearer.

Quote:

lets say my load was 32ohms so i would need a high bias of like 22mA

You're confused. That would apply if this were the bias for the output buffers. Q1/Q2/R9 adjusts the bias for the op-amp only, and its load is the input of the buffer, which has an input impedance of 3Mohms. That might lead you to believe that you only need microamps of bias to reach class A, but again, subjective tests have shown that more is wanted for best sound.

 

[steinchen]

I have to revise one of my statements. I did some min/max bias testing on my PPA and found the more bias the better. With changing bias from (min) 0.3mA to (max) 1.8mA (like on PIMETA) sound improved significantly. Bass became deeper and tighter, highs smoother, the overall sound warmer. With low to none bias sound was almost harsh.

On PPA v2 I'll probably jumper across the bias trim pots and safe the 5 bucks

Increasing bias for about 0.2mA and listening for a while showed only little to none audible difference, I guess because the step is too small and you get used to the sound. To get objective results one must do double blind a/b testing, therefore I decided to build a little relay based a/b switch. Since I got only one PPA (

) I'm only able to test PIMETA vs PPA atm and do some cable testing. I'm curious how PPA v1.1 vs PPA v2 vs M3 compare ...

 

[tangent]

Quote:

On PPA v2 I'll probably jumper across the bias trim pots and safe the 5 bucks

Then you'll wind up with uneven bias across the channels, unless you hand-match your biasing JFETs.

 

[__redruM]

Quote:

Originally Posted by tangent I can see that the docs weren't clear enough on picking the zener value, then. I've rewritten that section to be clearer.

Thanks, for the chosen example.

What do you think about running the ground channel at a lower bias than Left and Right?

 

[__redruM]

8 hours...

12 aaa 700mah cells, 1ma on each channel, gave me 8 Hours of amp time. I've got the opa627s and started to get a little white noise at the 8 hour mark.

 

[tangent]

Quote:

What do you think about running the ground channel at a lower bias than Left and Right?

The whole point of having a current source on the ground channel is to increase commonality between the channels. Making them uneven on purpose seems silly to me.

Quote:

12 aaa 700mah cells, 1ma on each channel, gave me 8 Hours of amp time. I've got the opa627s and started to get a little white noise at the 8 hour mark.

It sounds like your amp's quiescent current is about 90mA, then. I don't see how +/- 1mA additional is going to make any significant difference.