Let's keep a good thing going.

Official Mini³ website

---

[Quoting amb]

Let's keep a good thing going.

---

I have built 2 Mini^3 amps - one long-life and one high-performance. Both sound GREAT and were a lot of fun to build.

Thanks AMB for your work on this project, it is a great service to the DIY community.

I've not used one when listening to the amp, just as a very convenient battery charger - swapping the plug from my laptop to the amp at the end of the day.

you mean like this one?

I will use them with K401s, maybe K601.

I wonder, has anybody tried the Mini3 with low impedance IEMs like Etymotics ER6i? What's expected? they're way low impedance (16Ohms) and high efficiency (108dB), see www.etymotic.com/ephp/er6i-ts.aspx

I usually take the ER6i on trips... the ipod powers them enough, but guess there's room for improvement by using a good amp and taking the line out from the ipod's dock connector?

[Edited by pukka on 12-14-2007 at 03:25 AM.]

But be warned, the better the recording the better the combo sounds, mini3 + etymotic is very revealing. You may have upgrade your mp3s to take full advantage of the combo, 128kbs is not good enough. I'm currently reripping my collection to FLAC + lame mp3 at the -V4 setting which works great with my Archos 504 player.

Kevin

Both G=2, black is high performance, silver is extended runtime.

Extra points if the finish is actually furry.

1) replace R1 with a capacitor (1 uF polypropylene or polyester, or 2 x 0.47 in parallel), or

2) cut the traces on the bottom of the PCB and add an SMT capacitor between the volume control and R1 (probably a Panasonic ECP-U(A) 1 uF)

Given that the input impedence is 100 kOhms, a 1 uF cap should give me a 3db-down frequency of about 2 Hz (if my calcs are correct) so it shouldn't affect base response to any large degree.

Are there any other options?

How do you think a cap in these locations will affect the sound quality?

Thanks,
Bricksie

I think a better idea is to leave the amp as is, but make a special line out cable for your MP3 cable and put the caps inline (similar to what Red Wine Audio does in their iMod line-out dock cable for iPod 5G/5.5G). That cable is needed for their modification on the iPod 5G/5.5G, because it bypasses the output opamp but introduces a DC offset at the line out.

An even better idea would be to use a source that doesn't have significant DC offset.

So if I put a 22 uF cap in series with each line (left and right) of the cable, and put a 4.7 kOhm resistor after each cap going to ground, I should get the same 3 db rolloff (1.5 hz).

Is this how it's done?

EDIT: removed

[Edited by bricksie on 12-17-2007 at 11:40 AM.]

---

[Quoting bricksie]

So if I put a 22 uF cap in series with each line (left and right) of the cable, and put a 4.7 kOhm resistor after each cap going to ground, I should get the same 3 db rolloff (1.5 hz).

Is this how it's done?

---

The Black Gate caps used by the iMod people are rather small for all values, so I could go as high as 47 uF and still be only 6x7 mm in size.

I think I have some experience soldering SO8 chips as another hobby is RC and have "fetted" about 150-200 miniz boards:

Nevertheless, I was a bit anxious after reading all this thread the 8397 seemed to be a chip fairly picky... So I decided to be extra careful when soldering it.

My bad... when I had the 8397 carefully and perfectly soldered, I noticed I had done it to the wrong place (U4).

I desoldered it in a pinch and resoldered it at the U5 position.

I used a method I have done a million times when modding, if done properly is the one that less stresses both the PCB pads and the chip: flow solder on all the legs on one side and another, alternating between sides, and the chip soon floats and sticks to the soldering iron tip. See www.insflug.org/shots/fetremove.asf

I've removed these fets and even SO16 optocoupler this way and never damaged them, is a matter of timing. But maybe mine got damaged

When I did the initial checks, idle current was 100mA. IG-OL DC Offset showed 3VDC, the IG-OR was 1mV and IG-OG 2mV.

I checked for shorts prior to that and found none. I reflowed the 8307 pads, but no joy.

Seems like half 8397 is damaged?

All the remaining checks were fine.

[Edited by pukka on 12-19-2007 at 04:15 AM.]

I think this chip is specially fragile, and probably more in the E-pad version than the regular packaging version. With such a low C/W ratio, I guess it means that both is very able both to let heat go out, and... go in. Maybe somewhat the normal, non e-pad package prevents both geat escaping and heat getting in.

Do you think is really needed on this amp? Would a non-e version heat enough to distort in this scenario?

I already placed an order for another one on your shop, in the meantime I'm going to cook-up a mini DIP8 to SO8 adapter and try some 2132 / 627 /637 AD826 I have around.

[Edited by pukka on 12-19-2007 at 01:18 PM.]

I suggest not trying to opamp-roll in this amp. Repeated soldering/desoldering will cause the PCB pads to delaminate.

I\'ve soldered a DIP socket to small awg cables and soldered that to the pads, so will try the opamps in the socket, not by soldering/desoldering.

Anyway, a problem remains. Idle current seems fine (about 14mA with a 2132) DC offsets are bearable, 9 millivolts, but must have a cold joint somewhere because the left channel crackles and suddenly is loud, suddenly lowers the volume when touching the headphone cable; the right doesn\'t sound, sometimes does when touching the input cable.

I didn\'t have any shorts, checked that prior to powering the amp. And always discharge the caps prior to check.

what I want is to try to sort out where\'s the problem; if I fry an opamp, it won\'t be another 8397

Will re-check the initial checks.

[Edited by pukka on 12-19-2007 at 02:53 PM.]

AFAIK SO-8 double opamp pinout is the same as DIP8, isn't it?

The input and output sockets are solidly soldered.

DC offset very low, 0.6mV IG-OG, 0.0 IG-OL (looks odd!??), 0.9mV IG-OR, using a couble AD627 in a burr brown adapter.

[Edited by pukka on 12-20-2007 at 05:52 AM.]