Project: Simple bass guitar headphone monitor mixer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I would reconsider using the LM386 for a headphone amp, it's a total crap house IMHO.

Why not either go for LME49600/ BUF634 or an opamp stage with follower transistors (for instance BC327/BC337, or BD139/BD140 if you need more clout)?

I've built this now, the power amp section is on a separate board so it is simple to change out if the LM386's aren't up for it. It sounds pretty good though. I've ordered a couple of TDA7052A so I might try these instead. A buffer section is possible, but it has to be a small one since I really don't have any space left.

Found and solved bugs:
- The line in signal was too high compared to the bass, bigger resistors on the line in inputs solved this.
- The bass response was really bad at first, I needed much bigger coupling caps, 10uF on the line in and 22uF on the bass in works great (there is a slight bass loss on the line in, but this is how I want it)
- The pots where noisy, which turned out to be DC on the pots:
1) I had forgot the coupling cap after the FET-buffer (duh)
2) There was still some DC measured after putting in the cap, so I tried to change the polarity of the input caps (+ towards the opamp), this solved it!
(usually you'd use non-polar caps here, but I had to use electros here because of the high values required.)

Unsolved bugs:
- The most annoying thing is the loud pop when powering on/off, and there isn't really any simple solution for that with opamps and the lm386. Or actually I have a simple solution: A DP3T-switch which first switches on the power, then the signal path to the headphones, but it seems impossible to find a small make-before-break slide switch for this (a normal BBM wouldn't work as it would momentarily cut the power when switching from "standby" to "on").
- There is a hiss when used with my earbud headphones. The two other ones I've tried worked great (AKG K272, 55 ohms and Skullcandy of unknown impedance). The earbuds are pretty cheap Sony's, I don't know what impedance, but probably pretty low. My conclusion is that either the earbuds are bad (unlikely since they don't hiss otherwise) or else their impedance is too low for the LM386 to drive (it's the less powerful 0.33W LM386). If the problem is still there when I get my Shure in-ears, I'll have to redisign the power section.

This is the design as built (currently):

bass_monitor_mixer_2.0.png


/Fredde
 
Ok, I tested it with USB power (5V), and got a nasty surprise. USB power, it seems, is very noisy. Putting in an RC filter at the power input mostly solved this. There is still some noise, but it's quiet enough for normal use. I've also only tried it with one computer, some computers might be noisier than others.

However, 5 volts, especielly with the RC filter, which lowers it further, is really pushing it headroom-wise (+/-2.5V is the rated minimum of the OPA2134). The bass signal will clip if you crank the input volume. This is also pretty much a non-issue, since I wouldn't run my headphones that loud anyway. But it might be a thing to consider if you have inefficient headphones or like loud levels. (The 2 x 220uF was what I had handy and would fit into the case, but you could put in more and/or bigger caps for even better filtering.)

Here's a new version with the RC filter in place:

bass_monitor_mixer_2.1.png


I guess you could call this design confirmed and working, so feel free to build one if you like! I haven't yet tried it with the real in-ear monitor headphones (Shure SE215, 20 ohms) that I've ordered, and I still might have to do a major redesign if it doesn't work with these.

/Fredde
 
Just experimented with putting resistors in series with the headphone output to allow for low impedance headphones. 22 ohm resistors almost completely removed the hiss from the earbuds (they are 16 ohms). This also lowers the output volume, which could make the clipping when run on 5 volts more of an issue, so I won't make it permanent until I've tried the Shure headphones.
 
For an opamp followed by BC327/BC337, look at the headphone stage of this:
Mixout-VCpanner

Haven't tried this myself, so I don't know how it sounds, but it is VERY simple.

The opamp with the BD139/BD140 follower transistors can be found here:
Headphone Amplifier

The 100uF caps aren't that great an idea, and if you want to go slightly more elaborate you can use a more proper current source to bias the power transistors.

As for LME49600, this driver actually has a headphone amp mentioned in the app note:
http://www.ti.com/lit/ds/symlink/lme49600.pdf

I built one without the servo, and it sounds fantastic (mind you, you'll burn your headphones if any dc is present, so the servo isn't that bad an idea to include).
The BUF634 is supposedly the exact same thing.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.