"Just like rolling tubes!  I think you might get a best combination by mixing parts.  Suggest OPA2134 in first two stages and then 2604 for output.  Or something like that.  First stage is most important for noise and cartridge matching/interaction.  You don't want a bipolar type here.  Output stage needs to be powerful.

    The squeal is an RF oscillation.  It screws up everything, even in the lower registers.  You'll get dc bias offset and all kinds of issues.  The circuit board was not designed to handle the faster opamps.  But by adding localized supply decoupling, it might be good enough.  Maybe, there are still nasties caused by the sockets and added trace lengths to the feedback resistor, plus added capacitance on the inverting node.  Only a very fast scope will tell.

    So try this, 0.1uF capacitors mounted to the adapters.  Or solder right to the opamps pins.  Run both supplies to ground.  That's two caps per stage.  Value is not critical, 0.01uF is fine, but caps need to be low impedance type such as ceramic or poly films.  The same type as used on the main circuit board will work fine."
    - Jim Hagerman

---

OPA2134's are soldered into positions 1 & 2. The last output position is slotted and used for rolling.




OPA2604
This was the first opamp installed at U3 in the finished Bugle. It seemed very detailed over the Cambridge 540p. In comparison it excelled in all areas, bass, detail, organic midrange, top end cohesion, leading and trailing transients, calmer, on and on. It kicked ass. Now this is the preamp overall, I suppose, but you have to start talking about this opamp somewhere. This opamp ran in the preamp for 3+ months. In a direct comparison to the 2888P it turns out that this opamp is warmer, grainier, slower, and thicker. Duller, in a word.




OPA2228
This opamp ran in the preamp for several months. I put the 2604 back in for a side of an LP then dropped the 2228P back in. The 2228P was more open, lively & dynamic. Not that much more detailed if at all. I had already been noticing much more bass extension on the 2604.This was the same with the 2228 but there was more ambience overall. This 2228P has some verve that the 2604 does not have. Tasty. It was faster, and sweeter, but the detail was never overly forthcoming. Sweetness is it's forte. Very non-analytic. 


Burson discrete opamp
Increase in detail but bass dropped out, very thin, organic whole and presence disappeared, magic all gone. Damn, I had high hopes for this. Full bodied sound is back with the OPA2228P back in.


AD825
More open, more layered than the OPA2228, or rather you can see instruments as separate when they play in unison - this is nice. Bass is a different, more in control and deeper, whereas the OPA2228 was warmer, fuller, fuzzier. I like both. Voices are more 'nuanced'. The top end is 'crisper' and more defined, cleaner, and more extended. There is more bite which is a step back in 'romance' but the midrange tone is better in that respect. The overall Richness is down but tone is cleaner, differentiated, and so there is actually MORE colors at the cost of this Richness thing.

After about 40 hours of break-in the detail and layering are far better than the OPA2228. The crispness has diminished and the whole thing is smoother.

This delivers more definition, adds air to the midrange. I like the bass on either equally. I got these AD825s from Reference Audio Mods.




LM4562
Transparent.

No Romance, just music that was previously colored by each opamp in play. Doesn't seem like anything added or taken away. Transparent.
Less surface noise and open-ness which makes it dynamic by just reducing the noise floor.

I placed this opamp in the Pro-Ject HeadboxII headphone amp and the detail in combination with the Bugle jumped up. Transparent is the best description.