DIY: O2 the Edifier HA11
Jan 8, 2012 at 10:17 PM Thread Starter Post #1 of 5

fubar3

100+ Head-Fier
Joined
Jul 16, 2011
Posts
249
Likes
12
The Edifier is an economy amp with nice aluminum case.  There are photos and comments on this thread http://www.head-fi.org/t/412257/edifier-hall-ha11.
 
It is a shame that nice case hides a poor quality circuit which is prone to oscillation and latch-up. Consider the following schematic:-

I think there is positive feedback through the 33 ohm resistors which increases as the load impedance rises. This seems to cause an overshoot for percusive sounds such as cymbals.  Once you hear it, you can't ignore the effect.  Then there is the RF oscillation which the design failed to remedy with the miscellaneous HF bypass caps found in the circuit.
 
Fortunately, the HA11 can be repaired with an infusion of O2 technology.  I could use an arbitrary circuit but the O2 local feedback design (from NwAvGuy) should be most stable with the hacking I need to do. I used parts from the Objective 2 BOM but got power from the existing HA11 9v regulators. I would love to use an O2 circuit board but it would not fit in the HA11 so a big effort was needed to hack the fix.
 
The parts are soldered point-to-point on a generic hobby board with interconnects for the front panel volume control and headphone jack.  It looks ugly but sounds great with no visible oscillation on the scope and no output offsets.  There probably is some cross-talk and ground looping but that is only a guess.
 
Here is the top view, the board is installed upside-down for routing wires and parts clearance

 

 
Jan 12, 2012 at 12:42 AM Post #3 of 5


Quote:
U1 feedback point is wrong and non-inverting input is not referenced to ground :) Excluding these serious flaws, this is essentially a A47 amp.

The original circuit found in the HA11 is not shown completely in the schematic.  I am trying to show that it's design includes negative and positive feedback. Latch-up occurs because of the design flaw.
 
On the other hand, the O2 topology works perfectly.
 
 
 
Jan 13, 2012 at 1:07 AM Post #5 of 5


Quote:
Again, are you sure U1 feedback is implemented this way? It is connected wrong, should be along with the output.
 
Have a look at this: http://www.intersil.com/data/an/an1111.pdf


Ok, my sketch should show a resistor to ground in the feedback path. There are other components but I lost my paper sketch of the circuit trace.
 
The intersil diagram has the same flaw as the HA11.  It has positive feedback that varies with load impedance.  No load results in latchup.
 
 

Users who are viewing this thread

Back
Top