Headphone Amp Redux

Here you go...

You did it the right way. TL071 has an output drive of only a couple of mA

Years ago I designed a chip... a hearing aid... with a "sliding" class-A output stage... page 5 of this...

So it only bumps up the power as needed. I don't know how this would work with modern headphones with DC in them...

The OP opines that he needs 62.5mA peak, sort of nasty for class-A if portable :-( ...Jim Thompson

Perhaps it's because I lack the experience in analog design to immediately recognize problems like the one I mentioned above. In that case, is it so wrong to use the simulator as an educational tool to help elucidate errors? I worked this design out on paper first as was suggested and it basically worked out of the gate except for this issue. I don't have 40 years of circuit design experience to bring to bear on my projects, so I guess if I have to tweak things in the simulator to get it to work and can learn from the process I don't mind the criticism. Maybe someday I'll have the analog design experience to design "from the heart" as you put it.

I have seen the LM317 used as a current source load for the output stage in several designs - also designs where it is a constant current source load for a class A tube output stage. I might like to build something like that one day. Here's a paper I found comparing the merits of voltage regulator current source topologies:

If you wanted you could make a headphone amplifier with the current source an LM317/LM337. And you could make the driver a voltage regulator IC. You could also make the gain stage a voltage regulator IC. And you could make the voltage regulator a voltage regulator IC. For LM317/337s in TO-92 packages plus external components I bet the total cost of the project wouldn't be more than around $2.00!

Use the simulator to verify.

There was a reason I asked you to verbally (on paper) explain the need for each component... to make you think thru WHY you used it.

You didn't bite, other than platitudinously, vague "buffer", etc.

That doesn't hack it... WHY did you think you needed a buffer, etc ??

In case you haven't guessed, I used to teach... week-long seminars in analog circuit design. I still will, upon request, with sufficient audience to make it worth my while. ...Jim Thompson

I can try to commit a more thorough deconstruction of my thought process to text, but it will probably take me some time to write it up! It may have to wait until after the weekend as I am attempting to complete an entry for the 555 timer design contest:

You may be happy to know my entry is in the "minimalist" category, and I don't believe it has any extraneous parts.

I'd go for something like that - is there realistically still a large enough audience of prospective analog designers in the US to make it happen, though?

Have you seen the lectures on YouTube about analog IC design? From India, some of the lectures are probably 20 years old...

Those are great! :-)

A few more:

BAB - Branch And Burn BABBLE - Backspace And Burn Bad Last Entry BAD - Backup And Destroy

I can't remember what BAC was, or where I heard them.

Ed

I have no idea if there's enough market. In a way, lack of analog skills is beneficial to my consulting business ;-)

I'll check them out. ...Jim Thompson

EIEIO - Enforce In-order Execution of I/O (from the POWER PC, I believe)

Especially important in talking to disk farms. ;)

Cheers

Phil Hobbs

Um, that would be DASD farms. ;-)

Where's my Winchester?

Cheers

Phil Hobbs

Walmart.

-- You can't fix stupid. You can't even put a band-aid on it, because it's Teflon coated.

England, 51° 3'46.58"N 1°18'59.41"W

Winchester Mystery House: 37.31882°N, 121.95009°W ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Remember: Once you go over the hill, you pick up speed

You lost it again? ...must be enough to drive you to drink.

Bitrex,

Is it worth replying to your posts at all. You were offered help (by several people) in designing a headphone amp to your specs but have done nothing.

I pointed out the value of simplicity. You seemed to like that idea but did nothing.

I asked you to remove part of your circuit and seee/post what happens - you have done nothing.

Is it worth replying to your posts at all?

I'm sorry I didn't have a chance to test out your suggestion in a timely fashion; I had a chance to try it this morning. Of course the DC servo can't simply be disconnected from the circuit as it stands as proper biasing depends on it - it runs away to la la land. Rearranging things so that the input stage is biased by a simple voltage divider, I find the simulation says that the output sits at about 300mV. With the servo, the output sits at just a couple of mV above ground. 300mV against a 64 ohm load implies about 4.5 mA of DC current through the headphones. I haven't been able to find exact figures of what the maximum DC current through a pair of 64 ohm headphones is safe, as it probably varies with headphone size and type. I'm pretty sure less has to be better.

If you feel I have been unresponsive to the assistance offered by the NG then I'm sorry, and if you don't wish to continue replying then of course that's your choice. There have been a lot of great general suggestions about design methodology; your point about the value of simplicity has inspired me to attempt a very simple amplifier next to compare performance with. There have also been some suggestions that have consisted of essentially "Here's my design, use this!" Unfortunately, that is not really the kind of help I'm looking for; there are many good textbooks and websites about amplifier design that have ready-made circuits that I could easily build if I were so inclined. In any event, I do greatly appreciate the time that you and others have taken to correspond with me.