Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Quality headset audio amp

Status
Not open for further replies.

Oznog

Active Member
I'm looking at building a newer version of my ultralight aircraft headset amps. It supports music as well as voice. I want quality here.

My previous version all ran off of 12v. It used a very powerful OPA634 unity gain amp chip, which due to its low gain and sigficant output impedance was surrounded with a TL082 to make it put out a regulated voltage. The center point was created by another OPA634/TL082 rather than a cap because I ran calcs and the impedance really does appear to affect the bass response. It would be loaded with 2 headsets in parallel, 32 ohm I think.

So I wonder if this whole thing is even such a good idea. For one, OPA634 is an expensive chip. Second thing that comes to mind is I don't have a lot of protection against overvoltage spikes, makes me think maybe I should switch to a 5v regulated bus which is needed for the logic in the new design anyways. In 5v I can get rail-to-rail amps too.

I guess there's always the LM386 chip. That thing always seemed to be pretty crummy in general to me though. Too low of a gain for an external feedback loop and aren't they kinda noisy? Without feedback compensation the impedance can mess with freq response, and I might have to readjust the volume when adding the load of a second headset which I find far too inconvenient.

Any recommendations on what the best circuit would be? Is there a cheap but powerful and quality headset amp out there that I'm missing?
 
I think you're getting carried away here?.

It's only a headphone amp!, and for use on an ultralight!, an LM386 would be far better quality than required.

I don't know what your simulations said?, but I've never seen any audio amplifier that has to provide frequency compensation based on the output load impedance?.

Don't mess about with simulators, build the thing and use it!.
 
Nigel Goodwin said:
I don't know what your simulations said?, but I've never seen any audio amplifier that has to provide frequency compensation based on the output load impedance?.

Lower freq present a lower impedance load on the amp, so amp output impedance tends to have a rolloff in power as the ratio of amp impedance to speaker impedance gets higher. This cases the bass response to suck.
 
Oznog said:
Nigel Goodwin said:
I don't know what your simulations said?, but I've never seen any audio amplifier that has to provide frequency compensation based on the output load impedance?.

Lower freq present a lower impedance load on the amp, so amp output impedance tends to have a rolloff in power as the ratio of amp impedance to speaker impedance gets higher. This cases the bass response to suck.

NO!.

An 8 ohm speaker has a DC resistance of about 6 ohms, other speaker impedances are a similar percentage - so your 32 ohm headphones would have a DC resistance of about 24 ohms. At audio bass frequencies the impedance will be even higher than at DC, but even so the change is negligible against the very low output impedance of the amplifier (which should be well under 1 ohm, and probably under 0.1 ohms).

Also, rather obviously?, the lower impedance of the speaker will INCREASE the power supplied to the speaker, not DECREASE it!.

So if you've been getting a lack of bass, you're doing something badly wrong somewhere?.
 
Nigel Goodwin said:
At audio bass frequencies the impedance will be even higher than at DC, but even so the change is negligible against the very low output impedance of the amplifier (which should be well under 1 ohm, and probably under 0.1 ohms).

LM386 has an output impedance of around ~10 ohms from what I read, but data on this issue is sparse.

Another thing is, there may be one or two headsets plugged in and I'd rather not have to readjust vol when that happens.
 
The output impedance of an LM386 with its default gain of 20 is much lower than 10 ohms. Its max peak-to-peak output voltage gets limited by its lack of infinite current for low impedance loads but its actual output level doesn't change much with a change of load.

Try it! I'll bet its output level hardly changes between it having no load to when it has your 16 ohm load. Just make certain that its input isn't too high to cause distortion when loaded.
 
Oznog said:
LM386 has an output impedance of around ~10 ohms from what I read, but data on this issue is sparse.

As I've already said, and 'audioguru' has, it will be FAR lower than that!, it's not something very often given in the specs, as it's difficult to measure, and not something you really need to know.

Another thing is, there may be one or two headsets plugged in and I'd rather not have to readjust vol when that happens.

You won't, the very low output impedance of a transistor power amplifier will keep the voltage pretty well constant, certainly well below the threshold of your hearing ability. If you're planning adding LOT's of headphones in this way, you should feed each one via a series resistance to keep the overall load impedance high enough, but for just two it won't be a problem at all.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top