[FallenAngel]

Quote:

Originally Posted by mminutel /img/forum/go_quote.gif Thanks everyone. Here is another question. Should I just get Tangent's packet that has everything or should I order more solid components?

Tangent's set has "solid components"! Chances are you won't save anything by ordering elsewhere considering the shipping.

 

[mminutel]

Thanks very much.

 

[Ace o' Spades]

About it being "hard pressed" to build a PIMETA under $100, it is entirely possible. I was able to keep mine under that by building it inside a Serpac case, using batteries, and not stacking the buffers. Although a bit barebones, it still sounds great to my ears, though like me later you will want to recase it, add crossfeed, add a TREAD, and basically fix everything barebones that you put into it. Moral of the story? You can build a good amp for under $100, but later you will probably want more

 

[mminutel]

There are no OPA132 at digikey anymore so can anyone recommend me another set of chips to try?

-edit- also, the TI rail splitters wont be in until late march so is there another splitter that works well because I would hate to have to spend 2 months waiting.

 

[error401]

Quote:

Originally Posted by mminutel /img/forum/go_quote.gif There are no OPA132 at digikey anymore so can anyone recommend me another set of chips to try? -edit- also, the TI rail splitters wont be in until late march so is there another splitter that works well because I would hate to have to spend 2 months waiting.

You could use the DIP version of the TLE and change your layout (or airwire) instead, as DK has tons of these in stock (the TLE2426CP). It's also fairly trivial (and arguably better performing) to build a discrete splitter with a few resistors and an opamp in unity gain, or a dedicated buffer like the BUF634. Likely more expensive in both board space and dollars than using the DIP TLE though. Unfortunately TI seems to be the only company making anything like this IC, I haven't been able to find a comparable chip from anyone else. If you're comfortable with SMD, Tangent is also selling the 'ADAPTLE' which is a small board taking the SO8 TLE and a 1206 capacitor in about the same space and form factor as the TO92 part, and provides slightly better performance.

You probably want to use the OPA2132 as it's a dual-channel opamp so you'll only need one as opposed to two OPA132. Performance should be approximately identical, though with possibly slightly more crosstalk. You could also try the OPA2134 which is very similar but slightly less stable - for most builds this won't be a problem though if you can't source 2132 (and it's considerably cheaper).

--edit-- Sorry, I assume you're looking for an OPAG now that I actually think about what I'm writing. OPA134 will likely work but I'm not too sure. It'd probably be a better watch for OPA2132 in OPAL/R than one of the AD chips. If you can't source it you might want to troll around the forums for reports of people using OPA2134/134 and see how it works in the OPAG position. Otherwise either choose a different opamp entirely for OPAG or change all of them.

Many other chips will work as well, you're likely to have good luck with any FET input unity gain stable opamp. Don't go for anything that has too much bandwidth.

You may want to read Tangent's OpAmp and Virtual Ground articles, they provide good information on both topics. Or just go with the recommended parts as they're basically available, if you can handle the small redesign required to use the DIP TLE.

 

[FallenAngel]

You're still going to need a single chip for ground, go with OPA134 or OPA227 (slightly warmer) on the TI side.

For Left/Right maybe you could switch gears completly and go with what MisterX suggested here.

I heard great results with this whole output-compensation mod thing, never tried it, but it sounds interesteing.

 

[Ace o' Spades]

I have an OPA2227 in mine. Great all around chip IMO.

 

[tomb]

The TLE's are available from Amb and also at Jameco. Or, as Error401 suggests, you might try a DIP version with the converter that Tangent sells.

You might check the other thread for opamp ideas. No offense, but a 2132 would not be my first choice for a PIMETA. It's perfect for a CMoy, but the PIMETA deserves something better. Some of the AD chips are very nice, but are limited in current - making them unsuitable for a CMoy. That's where the advantage of the PIMETA comes in. Something like the AD823 or AD825 might be inexpensive and a good choice - because the buffers in the PIMETA negate the low current capability.

Some excellent combo's I've tried that work well:

L&R: AD8620 GND: AD8610 (this one is the PIMETA standard)
L&R: AD8066 GND: AD744 (my favorite - but note AD8065 will not work in GND)
L&R: OPA627 GND: OPA627 (expensive, but really, good)
L&R: OPA637 GND: OPA627 (same as above)
L&R: OPA2107 GND: OPA627, AD744 (cheaper version of the 627/637's)

IMHO, the OPA132 would make a great Ground chip, but I'm not sure you can get one that's not SOIC, which could be difficult on your first try. Some of the others above are only available in SOIC, but you can buy them pre-mounted at Cimarron Technology, Amb, and Parts Connexion.

One last thing - I have tried all sorts of combinations with the AD843 and AD845 - I can't get them to work in the PIMETA without huge offset issues.

 

[FallenAngel]

Personally, never really liked the OPA2227 in Pimeta. Was pretty nice in Sijosae Class-AB, but in Pimeta I enjoy the AD8066/AD8065 (I'm lucky it worked for me in Ground) combo and OPA2107 the most.

 

[tomb]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Personally, never really liked the OPA2227 in Pimeta. Was pretty nice in Sijosae Class-AB, but in Pimeta I enjoy the AD8066/AD8065 (I'm lucky it worked for me in Ground) combo and OPA2107 the most.

I thought the AD8066/8065 worked fine - OK, but a little lean, subdued, and bass light - sort of like the 8620/8610 ('cept they have good bass). I since found out that the lack of bass and general strength was the Gnd chip oscillating. I put other chips in that I knew were unity-gain stable, and they sounded completely different - much more like a very, very fast 627. I only settled on the OPA744 because I didn't want to waste a more expensive chip.

 

[FallenAngel]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif I thought the AD8066/8065 worked fine - OK, but a little lean, subdued, and bass light - sort of like the 8620/8610 ('cept they have good bass). I since found out that the lack of bass and general strength was the Gnd chip oscillating. I put other chips in that I knew were unity-gain stable, and they sounded completely different - much more like a very, very fast 627. I only settled on the OPA744 because I didn't want to waste a more expensive chip.

Nice, guess I'll try to put an OPA627 in there I've like 4 left since a trade for a few of my silver IC's

 

[tomb]

Good luck - could be just me, but the difference was profound. I tried the 627 in gnd, first - just because I was rolling from OPA627's in LR, anyway. Then I saw another thread about oscillation resulting in anemic bass. With the 627 in Gnd, the AD8066 was powerful, with no lack of bass - completely different than before. Once I knew the difference, I put the 744 in there since it was cheaper.

Amb's M3 misled me to a certain extent because the AD8065 in all three channels is specifically listed as an acceptable combo. However, I believe there's a little local feedback loop in the Gnd channel on the M3, but it's not there in the PIMETA. Anyway, the above is what worked for me - YMMV, of course.

 

[mminutel]

I can use SOIC amps though right? tangent says that there is a pad under each DIP socket but im not sure exactly what he means. Also, I am still not sure about the rail spliiter. There are none from TI in stock until late March but I would hate to put it on hold

.

 

[error401]

Quote:

Originally Posted by mminutel /img/forum/go_quote.gif I can use SOIC amps though right? tangent says that there is a pad under each DIP socket but im not sure exactly what he means. Also, I am still not sure about the rail spliiter. There are none from TI in stock until late March but I would hate to put it on hold .

If you're comfortable soldering them, yes. The footprint of an SO8 opamp is just small enough to fit between the holes of the DIP8 footprint. To support both, all you need to do is add the pads and traces (and use a double sided board, the layout is much easier if you put the SOIC chips on the bottom). The only issue is opamp rolling; you could use BrownDogs for that if you want to change opamps later and want to use SMD chips. It looks like this:

Since you're buying from Tangent anyway, and seem comfortable with SO8 (they're really not bad at all, the space is the same as the pad width, unlike with harder packages and their tiny spacing), why not get one of his ADAPTLE boards? It'll end up costing you $5 more than the TO92 part, but you do get slightly better performance as well.

 

[mminutel]

Is there a difference from the AD8066ARZ and the non Z version? They have the AD8066ARZ in stock but not the AD8066AR. Thanks guys. I am almost ready to finalize my order.