[acvtre]

Quote:

Originally Posted by MisterX /img/forum/go_quote.gif If you are using a RK27 pot simply connect a wire from one of the terminals on the switch to one of the "S1" pads on the circuit board. Then connect another wire from the other switch terminal to the other "S1" pad on the circuit board (It doesn't matter which switch terminal is wired to which pad with a SPST switch (nobody is going to call a wambulance if you install your switch upside down)). If you are using a RK097 pot you won't need to wire in a switch...

Ok so everything is right, since I've got a RK27 pot but.....nothing works. Not even the led lights up. What do I have to check?

 

[MisterX]

What pads on the circuit board do you have the power supply connected to?
May be more direct to post a couple of pictures so we can double check the wiring....

 

[acvtre]

Quote:

Originally Posted by MisterX /img/forum/go_quote.gif What pads on the circuit board do you have the power supply connected to? May be more direct to post a couple of pictures so we can double check the wiring....

The pictures will arrive in the afternoon. I've got a battery pigtail, red in the V+ pin black in the V- pin.

 

[MisterX]

Do you have the diodes in the right way around?

 

[acvtre]

I've just senn that I haven't soldered the buffer chips, cn it be a problem?
However the diodes are okay.

 

[tangent]

The power LED should still light up with no buffers soldered in.

Post pics and voltage measurements between IG and all the power pins in the circuit.

 

[sgtroyer]

Hi folks, I'm working on planning an amp / dac to sit on my desk at work. I don't intend to include battery power since it's not going to be portable. With that in mind, would the canonical AD8620/8610 op-amps still be the sensible choice, or should I "upgrade" to something like the AD843? From my understanding of tangent's op-amp reviews, the big advantage of the 8620 is that it works well with batteries. Low voltage requirements, low current draw. Since neither of those are really a consideration with a wall supply, I'm inclined to try out the 843, if it might sound a little better. The price difference is pretty small.

On the other hand, has anybody even tried the 843 in the pimeta v2? I'm not sure how much pathclearing I want to do on my first attempt. After all:

Quote:

Originally Posted by tangent /img/forum/go_quote.gif Noobs don't make good pioneers.

Any thoughts?

 

[tangent]

Quote:

Originally Posted by sgtroyer /img/forum/go_quote.gif has anybody even tried the 843 in the pimeta v2?

I'll get around to it someday.

Quote:

I'm not sure how much pathclearing I want to do on my first attempt.

Well, this question is a little different from the post you quoted me replying to. You can't break the amp by using the wrong op-amp. You just have to be prepared to do tests and have a fall-back plan if it doesn't work. Do a current measurement test before plugging headphones in, start with cheap phones you can afford to fry, and only then step up to the good ones.

It might be nice if you had a current-limited power supply, but lots of people roll op-amps without one.

If you're not prepared to buy multiple sets of op-amps or are unwilling to do careful testing, best wait for someone else to be the pioneer.

 

[sgtroyer]

Hmmm, might be fun. I don't have a current limited supply, just the el-cheapo variable supply built into my breadboard. I mean, it's definitely current limited, just not in the way that you want. But I can bring it up gradually and keep an eye on the current draw as I do so. I've got a scope, not much of a scope, but probably enough for audio debugging. I assume the failure mode here is oscillation? Drive a square wave and look for ringing on the output?

Other than the fun of messing around and forging ahead, is this worth it? Is the AD843 "better"?

The fact that no one has done this yet makes me wonder if anyone really sees any value in it.

 

[cobaltmute]

Quote:

Originally Posted by sgtroyer /img/forum/go_quote.gif Other than the fun of messing around and forging ahead, is this worth it? Is the AD843 "better"? The fact that no one has done this yet makes me wonder if anyone really sees any value in it.

Just because no one else has used it doesn't make it a bad chip. Yes you'll burn a few dollars in chips, but then you'll know if the chip works for you.

You can check the Op-amp thread for what some people will do for different opamps.

 

[tangent]

Quote:

Originally Posted by sgtroyer /img/forum/go_quote.gif I assume the failure mode here is oscillation?

Yes. My comments about frying headphones are because rail-to-rail oscillation looks like really really loud obnoxious music to the cans. Beware that an amp that doesn't oscillate without phones plugged in could still take off into oscillation once you do plug something in, due to capacitive loading and such. Cans are a complex network of Rs, Cs and Ls.

Quote:

Drive a square wave and look for ringing on the output?

That would be good. You want to do that anyway, if you have the tools for it, since that helps you pick optimal C6es. And that in turn is an option for making these chips work if they don't, initially: increasing C6 can rescue an amp that won't work with C6 left out or using its low default value.

Quote:

The fact that no one has done this yet makes me wonder if anyone really sees any value in it.

*Shrug*. You've read my review....I like 'em. They're not my absolute favoritest chips in the whole wide world, but they're certainly worth trying if you've got the cash and time.

Quote:

Originally Posted by cobaltmute /img/forum/go_quote.gif Just because no one else has used it doesn't make it a bad chip. Yes you'll burn a few dollars in chips, but then you'll know if the chip works for you.

True, though a trio of AD843s is a significant chunk of change. Pizza dinner tonight or rollin' chips? Tough choice.

 

[sgtroyer]

Quote:

Originally Posted by tangent /img/forum/go_quote.gif Beware that an amp that doesn't oscillate without phones plugged in could still take off into oscillation once you do plug something in, due to capacitive loading and such. Cans are a complex network of Rs, Cs and Ls.

Wouldn't the buffers mostly isolate the op-amp from the complexities of the load, though? It still feeds back, I guess, so something ugly could happen.

Quote:

Originally Posted by tangent /img/forum/go_quote.gif True, though a trio of AD843s is a significant chunk of change. Pizza dinner tonight or rollin' chips? Tough choice.

They aren't cheap, but neither are the AD8620/8610. By my math, once I've bought the necessary brown dog adaptors, the 8620/8610 solution is $26.65 from digikey (and out of stock), while 843 is $32.98 for three op amps and an adaptor. So the price difference is all of $6.33. The AD843 is actually cheaper for OPAG than AD8610 because, being dip, it doesn't need an adaptor.

Of course, that math all hinges on them *working*. If I fry them, or have to get a replacement because I can't make them work, it looks more expensive.

Speaking of expensive, I didn't realize how fast those square wave tests you're using are. I figured they were well within the audio band and I could just use a soundcard or my spiffy new, yet to be built, DAC to drive the amp, and watch it on my 5MHz heathkit scope. To drive the kind of square waves you're using, I would need an actual signal generator. And a better scope wouldn't hurt either.

Both of which I want, so having an excuse is okay, but it makes the cost of op amps the least of my problems.

Any advise for signal generator shopping? 20MHz or so? I know Tektronix is a favorite brand for scopes, is there a corresponding popular choice for sources?

 

[Beefy]

Quote:

Originally Posted by sgtroyer /img/forum/go_quote.gif The AD843 is actually cheaper for OPAG than AD8610 because, being dip, it doesn't need an adaptor.

Why not just solder the SOIC-8 OPAMPs to the bottom of the board on the SMD pads?

 

[tangent]

Quote:

Originally Posted by sgtroyer /img/forum/go_quote.gif is there a corresponding popular choice for sources?

Agilent makes me happy, but I can well understand if they make you very unhappy after you see the prices.

My previous sig-gen, a B+K 4017A, was flaky as a box of Wheaties. 1.8% THD on the "sine wave" output, square waves that ring pretty badly, low build quality... It's back in the box now. Anyone want it?

 

[sgtroyer]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif Why not just solder the SOIC-8 OPAMPs to the bottom of the board on the SMD pads?

Fair enough, I guess I could. I just like the idea of sockets in case something goes wrong, or I change my mind.

Quote:

Originally Posted by tangent /img/forum/go_quote.gif Agilent makes me happy, but I can well understand if they make you very unhappy after you see the prices.

Agilent makes me happy, too. I used to work there, so I've got a soft spot in my heart for anything Agilent or HP. But $1200 for a signal generator? Little bit out of my price range. I think I'll watch ebay for an old HP. Maybe I'll look for an old HP bench supply to keep it company. Then get a Tek scope so they can fight.