D.I.Y. Op Amps

Very nice looking work winetree

How do they sound?? 

Gary

Just a quick thought from my ears here at the BI Test Kitchen ...

Not as mid forward as my gar2520
Not as clean as my gar918
Not quite as warm/thick as my gar1731

garAM10BC...
A MUCH more polished lower mid range than any of my other offerings.
Should help with mud buildup In a mix.
These have a very nice vintage sound about them that will
work on lots of sources. It's just yet another audio colour
and besides working on 16V rails, its a nice 24v option as well for 51x.


Gary

I think I need a pair of these for my EQP5 plutec eqs.

I am very interested in Gary's DOA too. 2520 and 1731 seem the most common.
But what can you guys report about the garAM10 and gar918.
Highly interested in those. How do they compare and especially: how do they compare to the 2520/1731

thanks for the heads up.. didnt knew there was a DOA tester.
exactly what I need now

I have listened to your OpAmp test you just posted. Definitely good samples and very well recorded.
But frankly, I couldn't hear any big differences.
I never thought that differences were so subtle - it's almost not audible.
Crazy.. I am having second thoughts if there really is a need to make use of different DOAs.

did you ever hear scott liebers red dots?

The Red dot was for years my favorite, the difference between a well built far equivalent is very very small the biggest thing That Scott’s brings to the table is your eyesight and sanity of not having to build the thing, but building som GARs sure will improve your DIY chops.

More or less you don’t need to match transistors on most of the designs floating out there. Easiest is to buy transistors on a reel and use two adjacent ones for the input pair. That’ll be good enough.

The issue with testing is depending on what you're matching hfe may not be all that critical. For an op amp input pair usually hfe in a batch close 'nuff, though some circuits supposedly need minimums even from within gain-specified transistors (B,C etc). hfe changes with current (it is zero at zero and zero at infinite), so to test "properly" you need to test the transistors at the current they're going to be used at. But hfe matching only matters if the external DC resistance is matched.

Vbe matching is much more important than beta for a long tailed pair. Vbe is important for DC offset because Vbe mismatch is multiplied by DC gain.
If you have emitter resistors in your design this forces the input pair operate closer, so matching is less sensitive. Negative feedback wrapped around the entire op amp reduces nearly all of the problems except for DC offset from Vbe mismatch on the input pair.

Both are very temperature sensitive so you need to make sure you match that - even the heat from your fingers will make the readings vary, so you need to use insulating gloves or pliers with insulating jaws. Or tape them down to something, let them normalize, and use probes to test without touching them. And you have to be consistent, because self-heating can also skew the readings. I attached a way to make a test jig to test Vbe. 


On the other hand, most circuits are capacitor coupled so DC offset doesn't really matter a ton, and most op amp applications don't use a ton of DC gain (mic pres excluded). Unless you have something that is sensitive to DC, it's basically audio-transparent (we hear the AC wave, not the DC offset). AND the truth is that even if you perfectly match the input pair, you can still have DC offset due to circuit topology and component mismatch.

So...don't worry about it.