Edifier Hall HA11

I got the HA11 from DinoDirect in Hong Kong via air mail for less than $40. I can hear only a slight difference between it and the Fubar3/Mk2 DAC/AMP received recently. The only problem with HA11 was a putt-putt sound caused by the the 50hz, 9vac wall-wart. Of course, I replaced it with a 60hz,12vac spare from my junk box.

The HA11 seems robust with parallel 5532 drivers and oversized linear regulators. It could serve as a base for a DIY project.  Say provide sockets for the opamps.  Or add a small circuit board with buffers for low-impedance phones.

Quote:

Originally Posted by fubar3 

I got the HA11 from DinoDirect in Hong Kong via air mail for less than $40.

Yes, it was cheap and I was not surprised when it has issues. However, I was content to have the metal case which I could not build for lack of a machine shop. The electronics can be fixed since it is an old-school 2-layer board, not SMD. Here are the issues:-

- The power supply is inadequate.

- The opamps are unstable

- The circuit board board input lines/lands and outputs are too close together.

- The opamp topology presents a latch-up issue.

First up.. power supply.  The power input is labelled for 9vdc. There is no common-ground so the 9vac is half-wave rectified to get about 12vdc for the regulator inputs.  But the filter capacitors are only 470uf and should be 2000uf by my reckoning. I discarded the come with wall-wart since the HA11 only issued a putt-putt sound.  A 12vac input gives 16vdc peak with 1.5v ripple.. not great but the regulators work. I added 1000uf caps in parallel with the existing 470s to reduce the ripple to 0.5v but that is optional.

There are 3*5532 opamps. Even the power supply has one connected to the common pins of the fixed requlators.  I traced the circuit which appears to be for ripple reduction. It is a fine example of overengineering.  I chopped out that 5532 and grounded the requlators. Again, that is an option.

Next up... the opamp RF generators. Since the input signal land from the RCA jacks passed underneath the 5532 on it's way to the front panel input select switch, then to the volume control and back to the 5532, I cut the land and bypassed the input switch with a shielded cable.  That did not fix it. Next, I tackled the feedback network which had a 50k/20k divider.  These were atypical values which I replaced with 10k/1k. The bottom side of the board has 2 SMD caps as CA1 and CA2. Since I don't know the correct value, I removed them.  Now the RF generator is dead. (But the 10k/1k should be tweaked for lower gain so the volumen control can be turned up.)

Finally, there is the latch-up issue.  One half of each 5532 is the amp for a channel, the other side is unity-gain buffer.  This buffer has 100 percent negative feedback. Then each opamp output pin is connected to the jack with a 50ohm resistor such that the phone sees about 25ohms. All is well in theory. However, if the phones are not plugged in when power is applied, the buffer drive has nowhere to go except back to itself via the 50 ohm resistors.  so we have 100% negative feedback and 99.99 positive feedback in a race condition. Sometimes the buffer side wins such that it becomes a latch dragging down the 9vdc to 1vdc. Plugging the phones spikes the circuit which becomes free to run.

The final mod is to ground each pin of the phone jack with a 470ohm resistor. No more latching.

I tested the HA11 with the following:-

- Area 31 -- Chesky

- The Stanley Clarke Band

- Metallica with the Phillharmonic

The HA11 input was driven from the Fubar DAC output such that I could compare the HA11 to the Fubar's internal amp. I can't discern any difference. Perhaps I need better ears and better phones.