How close should readings be matching gain BC109B

[jabez]

I have a good size quanity of BC109B's. I have been testing them to make sure they are all okay so far they are all working fine. I was just wondering how close in % wise do the readings between transisitors have to be before you can say they are matched in gain.  Like would a reading of 360 & 380 be close enough or does it need to be closer?

[petemoore]

Best test, try 'em out.
 Generally any NPN si 'works'
 Some may exhibit lower noise characteristice, and if High gain is called for, a low gainer will probably sound 'limp' in comparison, in a gain stage.
 for a *standard FF, matching transistors is recommended, especially to get the rolloff of gain at guitar volume backoff, Low Leakage also [Si's are generally low leakage]...these gains fall between around 65hfe-150hfe [those would be the 'wide span #'s], 70hfe-130hfe are the numbers given int the technology of the Fuzz Face article, GEO...you can put anything that'll bias in there of course.
 Anyway, I've not had need to match gain of Q's above 150hfe, and most of the time if it matters, a socket and swap affair will tell me.
 In the envelope control of the Nurse Quacky, Hfe makes alot of difference, changing the sweep range and effect alot from one Q to the next.
 I got a little more Octave mileage swapping Q's in the SuperFuzz...I didn't measure anything...just swapped alot, then swapped some BCxxx's...left it where I liked it.

[jabez]

Well I was testing these BC109B's to see where they were at gain wise so if someone wanted some around a certain gain or matched in gain I could give them what they need. I got about 250 to sell. On my meter they all range from 325 to about 470 with most of them between 355 to 425. I just wanted to know what was considered to be close enough in gain to say they are matched in gain. is there any other test I should do?