Larry's Phase Linear 700B

[speakerman1]

The BGs you requested have arrived. I think I will send it after you get the pre done. Is that ok?

Larry

 

[laatsch55]

That will be fine Bud.

 

[speakerman1]

I still am wondering. I'm just wondering. I mean this bass is crisp and there. I am hearing things I haven't heard. The Glockenspiel playing in the back ground . No hiss in the back ground. The Hafler is at less than 1/2. No meter movement. The highs are there on the Violins. Good music to run the range of a speaker on for sure. The highs almost hurt my ears. Jerry would you like a tape to go with the K-2000. Can't erase but I can record. LOL


Larry

 

[laatsch55]

the bass will be even more pronounced when we get done messin with C6. And other stuff.

 

[speakerman1]

OK when my back gets better I'll ship it.

Larry

 

[laatsch55]

Well, couple , three weeks yet.

 

[mlucitt]

Lee,
I was matching Q1 and Q2 transistor pairs for the board I was building last night. Of course, I was going for similar Hfe values but what about the ones at the lower end or the higher end? Have you ever built an amp with the matched pair at the extremes of the scale?

Just curious.

 

[Gepetto]

The lower gain devices will GENERALLY exhibit higher but consistent negative output offset Mark. Higher gain devices will GENERALLY exhibit lower negative but consistent offset.

If you match the devices, you should nearly always have negative output offset. Q1 and Q2 form an NPN input op amp input and you are affecting the op amp input bias current by grading for Hfe.

 

[laatsch55]

Mark I don't have an hfe meter, I do a first order beta match. My Fluke 179 has an ohms function in the diode test mode which measures reistance with a litle current running. So I match the BE- BC junctions on individual trans to within 1 ohm then find a mate that matches that exactly. The closer the BE-BC junctions are in value the more gain you have. It shakes out to culling about 10% unmatchable. It seems like regardless what the redings are there's a matrch for it. Especially the TIS-97's they're pretty close anyway.

 

[laatsch55]

So Joe the mismatch in bias current is what we see on offset?

 

[laatsch55]

Also, Mark and Joe, just bought 82 more 2N1304's , TI's with a 73 date code...1.45 each. So Gentlemen we are well on our way to having at least 300 pairs and possibly more, I'll have to count them finally too see. Let me know when ya need more.
BTW, Joe, what do you think of an Audio Precision One analyzer??

 

[mlucitt]

Here is what I came out with, based on the 28 TIS 97 transistors I had in stock. I realize the numbers are relative and associated with the specific meter I used:
hFe - (number of transistors)
444 - 1
450 - 2
511 - 1
525 - 2
535 - 1
566 - 2
588 - 2
620 - 4
635 - 3
645 - 2
648 - 5
655 - 2
675 - 1

I just ordered more TIS 97 transistors from Mouser and I will see how they compare. I think these numbers represent a fairly average standard deviation although I did not compute it.

Lee I will try your method and see if it gives the same results, I'm thinking the meter I use does the same thing and just outputs a digital value for comparison. The datasheet for the TIS 97 says the hFe should be 250 min. to 700 max. It looks like we are in the ballpark.

Thanks Joe for the explanation but what output characteristics will the amplifier have if the negative offset is higher or lower based on the gain of the current source stabilized long-tailed input pair?

 

[laatsch55]

Are those numbers representative of both BC/BE junctions?? When you ohm in the diode check section you'll be around 708-715 usually. It was hard to get ...say a BE/BC of 712 each, but way easy to get a 712, 713, and easy to pair up. Are those EBC Mark??

 

[mlucitt]

Lee, I do it like this. I turn on the meter. I plug the transistor in the three little holes. I look at the LCD display. I write down the number. I throw the transistor on top of the number I just wrote down. Then I do it all over again.

I have not gotten around to trying your BE/BC and comparing to see if they still match. I know the numbers may be different because I am measuring hFe and you are measuring resistance.

I know what you mean by the numbers. First, they jump up and down a few digits and I try to grab the average number. Then there is never a perfect correlation, so I do like you do - if the number is + or - say "5" then I group those all together and call it close enough. If I end up with more than 2 in the pile, I can go back and retest to get the closest pair.

At the bench, leads hanging down, flat side facing me, left to right CBE.

Like so:
[attachment=0:3fs92uu2]DSC02991.JPG[/attachment:3fs92uu2]

 

[laatsch55]

in my case thje numbers bounce twice then stay. I record the BE jct number then the BC jct number. those are what I match off of . I keep the BE/BC as close as i can then match BE to BE and BC to BC. Yep CBE, the old stuff was EBC. Sure had me chasin my tail the fist few fixes I did 4 years ago. Stick an CBE where a EBC goes and you had rail voltage for offset.

How many RCA-410's ya want!!!! I'm givin em to ya man!!

 

[mlucitt]

I could use 8 of the RCA's for drivers in two amps that I am looking at. Do you need a 10X scope probe? I have three of them.

 

[laatsch55]

Nope, mine is a 1X/10X combo. Leader.Gt a pair of them. So ya wouldn't send 50-410's back??

 

[mlucitt]

I just spit out half my coffee. That would be so cool. My only problem would be convincing the missus that those beatiful TO-3's came from Santa Claus...

I have the combo probes too. Sometimes I like the simplicity of the 1X probe we used in the Navy, the venerable Tektronix P6028. It is a thing of beauty and sailor proof too. Its 10X brother was the P6006, I think.

Anything that you need?

 

[laatsch55]

More hours in the day?? I have shelves of 410's. Bought all those heatsinks and they were LOADED!!

 

[speakerman1]

Winter Solstice today. Have a happy short day. Larrt