Simple Class A Headphone Amp Using THAT1646

[limacon]

Nice!

Where is the third transistor/Vbe multiplier? I see the "bias" port on the sch but no circuit.

You don't really need a servo in this circuit. The output Vos is low - it's the sum of the 1646 output offsets plus 1/2 the Vbe difference of the NPN/PNP output pairs. Typically its under 5 mV.

I would remove C2 and C8.

Thanks Wayne! It’s WIP. I’ve been learning TinaTI and wanted to sim your circuit.

Noted about the servo, thanks! I’ll happily send you a board and source files once completed and tested.

 

[dreamth]

Built this with drv134 back in 2012...Really cool amp at the time.I had some akg 600 ohms headphones back then.I might want do it again as i still have some smd drv134 chips from a scraped sound card.

 

[mlloyd1]

dreamth:
was that a soundblaster card?

 

[limacon]

Would it be worth adding

a). small start up delay on the power rails for just the the THAT1686 parts? A voltage ramp start to prevent compression driver damage? I think I can squeeze that from another dual transistor part and a few passives.

b). input level peak meter. Again from another dual transistor part and a few passives and led.

c). do the transistor banks need to permanently loaded with ?mA in any way to prevent shutdown?

 

[dreamth]

dreamth:
was that a soundblaster card?

I can't exactly remember what it was, but certainly no soundblaster...It was part of a professional two compound video + sound editing tv station rendering card, basically 2 cards of 30/10 cm size .I will look for the whole thing and tell you if i find it.I think it was something canadian...

 

[limacon]

Nice!

Where is the third transistor/Vbe multiplier? It provides thermal feedback. I see the "bias" port on the sch but no circuit.

You don't really need a servo in this circuit. The output Vos is low - it's the sum of the 1646 output offsets plus 1/2 the Vbe difference of the NPN/PNP output pairs. Typically its under 5 mV.

I would remove C2 and C8.

Wayne,

Please find the thread here. 1st draft almost done. Some checking to do after sleep.

https://www.diyaudio.com/community/threads/smd-class-a-headphone-amp-that1646.384945/

 

[mlloyd1]

I can't exactly remember what it was, but certainly no soundblaster...It was part of a professional two compound video + sound editing tv station rendering card, basically 2 cards of 30/10 cm size .I will look for the whole thing and tell you if i find it.I think it was something canadian...

Thanks, I didn't think that it was, based on some of the chips I saw in your picture. I was just curious, so no hurry.

 

[dreamth]

Thanks, I didn't think that it was, based on some of the chips I saw in your picture. I was just curious, so no hurry.

This was just half of the audio board which had some of the video processing on it too...ad811 , ad 828 op275 op-amps, ssm line drivers snd receivers...top of the line 2001 soundcard.the removed components were just some voltage regulators , two drv134 drivers and some ssm line receivers i used in a headphones amp made in 2012 that i gave later to a friend in 2015.

 

[dreamth]

Found a third smaller board of the whole editing thing ...Matrox

 

[mediatechnology]

The 47 uF across the Vbe multiplier provides start-up delay.

How do you propose to heatsink all those SMT output transistors?

 

[limacon]

The 47 uF across the Vbe multiplier provides start-up delay.

How do you propose to heatsink all those SMT output transistors?

Thanks!

Lots of 0.15mm vias to 4 layers and a pool of potting compound… hopefully