Perf-board 500 series headphone amp

Advertise with us
shop deals on amazon
Shop deals on ebay
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
G

geoff004

Well-known member
Joined
Apr 6, 2008
Messages
343
Location
Bucks County, PA
I had 1 slot left in the 51x rack and needed a headphone amp.
It's inspired by Gustav's head grinder and built mostly off the Jakob/Gyraf schematic.
I've enjoyed building off perf-board.  It's like point-to-point in that you can build in 3 dimensions and the PCBs for so many projects are readily available at Radio Shack.  Who knew?

Anyhow, sound is great.  It makes listening to music enjoyable.
Here's the breakdown - THAT1200 ICs in to TL074s for distribution amps.  Then in to 3 sets of the design mentioned above.  3 volume knobs.  I haven't used it enough to say how hot it'll get but so far so good.

Pictures:

hp1.jpg


hp2.jpg


hp3.jpg
 
So I've been using this thing a good bit and it sounds superb. It'll get 3 sets if headphones amazingly loud.
But it gets quite hot.
So my question is this: is there a way to not have it get so loud and in turn have it operate at a lower temperature? -possibly limit the power to the transistors or something?

The schematic can be found here:

https://pcbgrinder.com/download/HG_docu.pdf
 
output transistors are working at 20mA idle so up to there in clas A, over there in class AB. You could change output config to get cooler working point, you could change leds 1 to 4 for 1N4448 diodes and R1 to 4 to a much lower value 8.2Ω for example or stay with leds and put a higher value of resistors 1 to 4 to 100Ω for about 10mA idle for example, in any case they be less hot when no signal and will start getting hot with signal rising depending on the load, if you use 600Ω headphones it will be harder to the amp to get hot.

Other than that, voltage swing seems too much for a headphone amp... usually something under 10V is used to feed and a resistor at the output to limit the current at the HP and protect them. Also AC cupled output to protect HP from a DC current that this could be putting into them.. The circuit is a nice assisted opamp but miss couple of things to work properly as HP amp.

JS
 
> operate at a lower temperature?

1) Use bigger heatsinks.

2) Silicon is tough. Who cares how hot it runs? (I had one ran so hot I put warnings on it.)

3) Use lower voltage. As JS is saying, even high-Z phones don't need over +/-12V rails. Low-Z phones can go LOUD with +/-7V.

!! The plan as shown is not protected against shorts. Shorts happen. A burnt-up headphone amp may be acceptable to some, not to others. In this case, it's not the few-cent parts cost, it is downtime and aggravation.

IMHO, about 30 ohms series resistance does little to no harm to headphone sound, but will make this amp short-proof; also limit the damage which can be done to low-Z phones.
 
As rule of thimb some study says that 10ºC more will make half life to silicon... BD are cheap but one burned BD may but one expensive HP isn't cheap... 30Ω will help but not burning BD will also help... Also change a transistor is a PITA specially in a middle of a session. I know 5 years or 10 are both long lifes for a 20cent component but 10 is longer for a small heat sink price.

JS
 
That is a bunch of class A current but probably why it sounds so good.  ;D

Heatsink the transistors and if you must increase the value of 47 ohm emitter resistors (68-100 ohm). maybe use more serious transistors.

I like PRR's suggestion of adding a circa 30 ohm series buildout resistor. It will not be noticed in series with hi-Z cans, and will protect agains low Z cans. I made a nice healthy headphone amp at Peavey (using TO-220 power transistors)  that could drive speakers, but would pop the thermal fuse in the 1A wall wart when driven hard, so I added a small resistance in series with the transformer winding so at high current the PS voltage would sag and not take out the lump.

Keep the rail voltage (IMO)... but maybe a little less class A current might be OK, if the heat sink can't cut it.

JR


 
geoff004 said:
- THAT1200 ICs in to TL074s for distribution amps. 
Click to expand...

For this distribution part of the circuit, is it as simple as just running the incoming signal to a THAT1200 which in turn feeds a TL074 before going into the headphone amp?  And that is repeated three times?

And why use TL074 and not TL072?

Cheers

Rob
 
If you look up the data sheet for the tl074 there is a schematic for making a distribution amp.  Basically you feed one of the channels of it with your signal, which in turn feed the other 3 channels.  The THAT1200 is simply for turning balanced signal into unbalanced.

My layout worked pretty well overall. The unbalancing/distribution part is a little cramped but works.

While I'm posting I'll add that this is probably the single most useful thing I've built for myself in a long time.  Heat has not been any issue.  I've used it on a number of longer sessions and it's been great. The only major thing I've changed since my first post is that I changed out all my wires for shielded ones. I was having minor noise issues with the unshielded wire - nothing noticeable when there was music playing, but it was a distraction between takes.  Total cost probably didn't exceed $100.  Hopefully that answers it.
 
I will post an old plan of mine which has lost two web-hosts so far.
 

Attachments

  • Tori-Amp-NPN.gif
    Tori-Amp-NPN.gif
    5.2 KB · Views: 183
geoff004 said:
If you look up the data sheet for the tl074 there is a schematic for making a distribution amp.  Basically you feed one of the channels of it with your signal, which in turn feed the other 3 channels.  The THAT1200 is simply for turning balanced signal into unbalanced.

My layout worked pretty well overall. The unbalancing/distribution part is a little cramped but works.

While I'm posting I'll add that this is probably the single most useful thing I've built for myself in a long time.  Heat has not been any issue.  I've used it on a number of longer sessions and it's been great. The only major thing I've changed since my first post is that I changed out all my wires for shielded ones. I was having minor noise issues with the unshielded wire - nothing noticeable when there was music playing, but it was a distraction between takes.  Total cost probably didn't exceed $100.  Hopefully that answers it.
Click to expand...

And no issues running NE5534 at 16V?  I presume not since you have not mentioned it, but thought I'd ask.

I see Gustav even powers his circuit with + / - 18V which seems like it would drastically shorten the life of a chip rated for 15V max.
 
No issues.
I think actually max voltage might be as high as +/-22v for the 5534 - maybe a little less for the other ICs  Either way it's not a problem.
 
geoff004 said:
No issues.
I think actually max voltage might be as high as +/-22v for the 5534 - maybe a little less for the other ICs  Either way it's not a problem.
Click to expand...

My mistake. Farnell had + / - 15V as the max supply voltage for NE5534 on their website, but I see that it is indeed 22V looking at the datasheet
 
Hi Geoff,

So I've got a channel of this running (THAT1200 > TL074 > 1st channel of Gyraf headphone amp) and sound comes out, but it is very distorted.  I presume it's something to do with mis-matched impedances but after spending last night and today reading a lot of different stuff online, I am none the wiser about how to fix this...

I anticipated that I might need to do some tweaking as my headphones are 75ohm which I think is lower than the Gyraf amp likes, but I was wondering whether or not you changed any other component values?  And have you tried headphones as low as 75ohm?

I included a 430ohm build out resistor between TL074 and input of headphone amp, and I'll also try a build out resistor on the output of the headphone amp.  But if you made any other changes to component values or made any additions, would you mind sharing them?

Cheers

Rob
 
Rob-
I only changed that one resistor value I mentioned above.  But it really had no impact on the sound.
As far as headphones - I've used a good mix of different headphones and haven't had any problems.
Are you overdriving the input?
My DAC is a Metric Halo LIO-8.  I set the output to 'monitor', which is something between +4 and -10.  I can't remember how it was with a +4 input; -10 works fine.
I really don't have any other thoughts.  My entire unit is 3 stock schematics with no adjustments or alterations (other than 1 component).
Perhaps you should start without the added resistors, then go from there.
 
Thanks.  I double checked everything and I'd made a stupid error - no 1M feedback resistor on the TL074s.  I added those and the distortion is gone.

This is WAY too loud for my headphones.  Ear damage would probably occur before 12:00 on the volume pot so I think some experimenting with build out resistors is in order.  But in principle it works!  Now to build the other two channels.
 
> no 1M feedback resistor on the TL074s

???

I don't see 1meg on any of the plans in this thread.

Which plan did you use?
 
PRR said:
> no 1M feedback resistor on the TL074s

???

I don't see 1meg on any of the plans in this thread.

Which plan did you use?
Click to expand...

I used this : http://www.ti.com/lit/ds/symlink/tl074.pdf

Figure 25.

John Roberts did mention lowering resistance in the feedback network in a different thread, however I built this stock just to see how it worked on the basis that I could experiment later.
 

Similar threads

rock soderstrom
DIY project: A handy power supply PCB for small tube amps like preamps, buffers or microphones.
2
Replies
33
Views
6K
JohnRoberts
JohnRoberts
Studio Mollan
  • Sticky
StudioMollan: TIRDI-500, PROTO-500.
Replies
19
Views
8K
znjamznjam
Z
amplexus
McCurdy AU300 clone - development thread
2 3
Replies
52
Views
9K
My3gger
M
SparkleBear
KM54c SDC Tube Mic Build / Gefell mv691 Mod
2
Replies
25
Views
10K
SparkleBear
SparkleBear
rob_gould
(Another) 500 series headphone amp
Replies
1
Views
1K
shabtek
shabtek

Latest posts

Back
Top