tangentsoft.net/audio

Look for the CMOY article.

And basically the older thread has more information for newbies so you have to PM scotty to get the older compilation.

Pretty bad move to baleet the old thread IMO.

what opamp?

Recheck soldering and reflow it, make sure you do not short V+ to V-

This post has been edited by LittleGhost: Nov 11 2009, 03:02 PM

technically, if you have enough skills or knowledge to build an amplifier, you wouldn't be asking which amplifier to build. Your imagination coexists with the amount of your knowledge.

The mini3 is easy to build. Only the opamps are SMD and the rest are through hole.

You can use the same chip and design your own layout if you do not want to buy the board, but I suggest having at least proper equipments as the AD8397 itself is a cranky chip.

Make sure it's realllly stable.

This post has been edited by LittleGhost: Jan 1 2010, 10:00 PM

don't interpret this the wrong way. I only questioned your understanding because you wished to use your own layout and yet you asked about which amplifier to build. The first version of Mini3 even with superior layout was unstable. It was only stable starting on V2 that is why it's a frustrating chip to work with. I want to really make sure you know what you are doing without jumping into the wrong design.

Many people jump ship without going through the basics (CMOY) and end up wasting time and feeling frustrated because the higher end amps did not work out.

This is why there are very little people in this hobby. If you tend to go too fast without knowing what to do, you give up too easily. Now this hobby is quiet and lonely enough, I wouldnt want to have more people leaving this hobby


The AD8397 is unlike any other OPAMP chip. I don't mean to sound condescending but there is a lot of other engineering knowledge you must posses in order to handle it without using the original board. The Mini3 schematic you see on the site will have a higher chance of not working with other board layouts because of the different compensation loop length stability involved. If you already have a grasp of how to deal with loop gain stability, using phase lead compensation, lag lead compensation to cope with the inferior board layout, then I suppose you will do fine regardless. Plus, this chip is not something you can simulate using a program, use a cookie cutter layout, and expect it to work. You must also have the tools IRL to build and stabilize it real time. Heck with the limited tools I have in hand, I would not even touch this chip with my own layout.

So you distill it down to things you can do, you either:

1) Buy the original board from Amb and save the headache of stabilizing it yourself. (or you can ask alvin if he has some boards left, since he's from SG, the price will be cheaper)

2) Use the older Mini3 schematic and see if you can get away for not using the loop bandwidth capacitor.

3) Try a similar amplifier topology but use different opamps. You can always give the PIMETA a try and the PIMETA is easier to build in a stable fashion than the Mini3 is even if you build it on a perfboard.

This post has been edited by LittleGhost: Jan 2 2010, 10:16 AM

^hang two speaker drivers across your ears to emulate the k1000.

PROBLEM SETTLED.

i usually get them from farnell. Though their price is far from being ideal. I don't think you can get them elsewhere over here in malaysia.

Still, they make good standard cases for a variety of board sizes.

i won't get them personally for loose builds, but I'd really get them for perfection purposes

just so to know I've ordered a Tektronik TDS2002B and a gw-Instek 7MHz DDS signal generator. lead time is about 3-4 weeks :\


Going to go crazy when it arrives.

^farnell sells a handful of "cheap" items comparing to the ones you can get locally.

You just need to know what to get. I think there is no one in malaysia that can beat their price when it comes to components you actually use for performance.

dem it, must finish my folded cascode asap :|

Don't mind doing it lol, but my timetable is sorta insane this trimester and might actually take some time to finish it. I thought you've finished building one?

Which reminds me, chchyong89 is also planning a F5 He's my university mate and we might just build one together for his speakers.


I've seen your designs. Your topology reflects similarity with people like douglas self, and DX destroyer, and that guy from sound.westhost. I'm curious why you did not go with more exotic topologies.


I find that eventually all of us will need a scope because simulations are too perfect to be true (though they are accurate most of the time)

Just got my TDS2002B and SFG-2107

Tested a pcbed amplifier that I built and designed without prior scope testing (only done simulations)

Turns out it's stable

This means I get to realize the discrete design I've been working on soon enough

Obligatory show off picture.








have yet to build a test jig (for variable load) for the OSK. The waveforms are taken without load. Amplifier tested is my dBM amplifier using the originally selected OPAMPS (AD825 x2, OPA627 x1)




1KHz Sine Wave 20Vpp (prior clipping)


1KHz Square Wave 20Vpp (prior clipping)


100KHz Square Wave 20Vpp (prior clipping)


Shows very nice results Squarewaves a bit noisy but I suspect it's due to 24VDC wall wart and the lower PSRR of the output stage due to the choice of using resistors for CCS as an alternative to a warmer sound and cheaper budget.

Will get a frequency response graph (dominated by lead compensation) and real time FFT measurements when I build the required test jig.

This post has been edited by LittleGhost: Feb 7 2010, 07:03 PM

SOON!

Maybe we'll meet up after my midterms

desperate times calls for desperate measures

IRF510 is better. IRF520 has shittier specs.

how to bend physics?

cannot leh

EDIT: Okay, conscience kicking in, jokes aside, the only things you can safely socket are the OPAMP socket (since for cmoy you hardly use anything fast), and the input caps (because the lead length means jackshit here), plus all the other things that is not related to the audio circuit (LEDs etc)

---

Added on March 31, 2010, 8:47 pmi kinda hate your new avatar.

i thought you were some new guy or something now that i've grown attached to your old one.



This post has been edited by LittleGhost: Mar 31 2010, 08:47 PM

:\ just use a transformer and feed it to a simple LM7809, LM7909 complementary 9V regulators.

polarity mistakes and overvoltage usually means kaboom.

why are you using a noisy LMC7660?



This post has been edited by LittleGhost: Apr 13 2010, 10:46 PM

You still have to solve the issue of caps blowing up don't you?

Also, I don't think the LMC7660 counts as a regulator. It's more like a buck boost isnt it?

bro gu wak chai, still hanging around MMU?

Let's hang out EDIT AT CHCHYONG89's PLACE.

I have a few good things you can try.

This post has been edited by LittleGhost: Apr 18 2010, 09:23 AM

we have tons of shit to do too Next week is nice.