[diditmyself]

A class A amplifier is inefficient by nature and single end is the least inefficient of them all, so these kind of amps will generate some heat. Besides being ininefficient, a single end amp can't operate in class AB, so you'd want some margins which will make it even hotter. Since you don't need that much power driving headphones, this is no major problem in a stationary amp. Just use a proper heatsink.

 

[diditmyself]

 

[Spacehead]

thank you for those schematics, diditmyself

I forgot to order LM317, but I am getting 12.5 ohm resistors (0.6W) and extra BD139s
but I don't have 6.8 ohm resistors (at least I don't remember having).

Could I use 7905 and 12.5 ohm resistors to create a constant current source for BD139?

 

[Spacehead]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif The LM317 pulling 300mA+ is getting super hot though, I'm worried it'll go into thermal shutdown. I can't keep my finger on the heatsink for even a part of a second. Thinking of what to do. Even the 100R resistors are getting super hot, it's like a thermos in there, wow! Can't touch the resistors, can't touch LM317, only output transistors are kept at reasonable temps.

How much current you think this whole circuit takes? I am trying to build a small power supply that would be powered from a 6 W DC-DC converter.

 

[FallenAngel]

Well, it's about about 100mA per channel for that BD139 buffer.

All this talk about this thing made me want to build another one from scrap parts. Waiting for PSU now.

 

[Spacehead]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Well, it's about about 100mA per channel for that BD139 buffer. All this talk about this thing made me want to build another one from scrap parts. Waiting for PSU now.

That's pretty cool you are hooked again

But 100 mA with that 5 Watt resistor? High voltages? I am wondering if that 6 W DC-DC converter will be enough... It was cheap so I had to buy it for fun!

 

[diditmyself]

The values in the 1st schematic isn't absolute. They should give a constant current of about 100 mA, but the 6.8R resistor might have to be adjusted. I don't know if 100 mA is sufficient for you. It depends on the impedance of your phones and how loud you want to listen. When using LM317 you can calculate the resistor value and be sure that it matches reality.

 

[Spacehead]

I have now learned through experimentation, that only the ground channel is important if better current driving ability is needed. I am now listening to amp that has 2SD612K in ground channel. RMAA says that noise level is -88 dB and dynamic range also -88 dB.
I am having hard time getting THD low enough. It is now 0.2%. I am using batteries (6x1.5V NiMH).
Dual mono cmoy configuration , with OPA1611 x 3

FallenAngel, did you get that PSU already?



edit: both amplifiers give some ringing on high frequencies. I don't know what is causing that. Maybe I should use only electrolytics.

edit2: ringing fixed with proper decoupling of op amps. THD dropped to 0.039%. But now I notice that it doesn't properly give current to my 24 ohm cans. I need to add some power transistors to left and right channel op amps feedback loop. then it rocks, and with batteries!!

 

[FallenAngel]

2x 15V transformer has landed, making an LM317/LM337 split supply now, should have photos up this weekend.

 

[Spacehead]

Cool

I didn't want to play with AC mains voltages and split the wall wart supply with TLE2426 buffered with NJM4556(bridged). My wall wart gives 17V, so I get about +6.56V and -6.56V with 7805/7915 (made adjustable). Now that I think of it, the rail voltages are pretty low. Bridged NJM4556 is parallel with bridged NJM2068 , so I get about 250-300 mA virtual ground. Parts don't get hot except BD139 and their resistors.
Next wall wart that I have, gives 33 Volts at no-load. That is quite a lot , after regulating I should get 14.5V rails. That would be plenty, but some op amps can't stand that high voltages.
I have also been modding my Sound Blaster X-FI. But something is wrong as I get bad THD value.

 

[Spacehead]

I had to test the 24V 300 mA wall wart. The regulators output 11V. I couldn't hear any difference in sound. I think it is as good as it can be.

One thing I would like to do. I would like to add BD140 so that it would be Push-Pull configuration. How about DC then? And how would the circuit look? Adding another transistor in parallel should decrease distortions and noise.

 

[Spacehead]

Quote:

Originally Posted by diditmyself /img/forum/go_quote.gif

ok I bought LM317LZ and not LM317T

Is it ok to use LM317LZ, it handles only 100 mA max ?!

I also have LM337T. How can I proceed to do a CCS for BD139s?

I also have couple JFETS, 2SK363 and 2SK246 ...

 

[TzeYang]

short answer no. use the LM317T.

 

[Spacehead]

Quote:

Originally Posted by TzeYang /img/forum/go_quote.gif short answer no. use the LM317T.

ok, I calculated the power dissipations and came into conclusion that LM317LZ can't take it.
Too bad. I guess I need to settle in for wirewound resistor.

 

[diditmyself]

You can use your LM337T. Connect IN to negative rail, ADJ to the emitter of BD139 and OUT via a resistor to the emitter of BD139.

BTW, when looking at the schematics I posted: I hope you understand that R9 is just there to illustrate the load of a headphone and not a resistor that should be soldered in the amp.