diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   Transistor Beta Numbers & Matching Q? (http://www.diyaudio.com/forums/solid-state/79707-transistor-beta-numbers-matching-q.html)

Zero Cool 17th May 2006 03:12 AM

Transistor Beta Numbers & Matching Q?
 
I am working on a 1976 vintage Audio Research D100A power amp and i have some questions about Beta matching output devices.

This amp uses 6 pairs of output devices plus a pair of drivers and another pair of transistors for Biasing. (wierd amp design).

Each transistor has a Beta number marked on it. The negative half has 3 output's marked "14" and 3 marked "39"

The positive half has 2 marked "17", 2 marked "43" and 2 marked "19"

The drivers and the Bias transistors are all marked "72" and are noted that they must be matched.

But for the output transistors. How close is close enough? all of the transistors are factory parts, so i can only assume that they were originally installed that way, or were installed by the factory that way during a repair of some sort.

17 & 19 i would imagine are close enough, but is 17/19 & 43 close enough??

How about the half with 14 & 39?

And what about the differences between each side?

I really dont want to have to replace all the transistors at $18.00 each!

This thing is such a PITA to work on that i only want to do this once!


Zc

Eva 17th May 2006 03:23 AM

Matching is a follows:

(3*14+3*39)/6=(42+117)/6=159/6=26.50
(2*17+2*43+2*19)/6=(34+86+38)/6=158/6=26.33

I thought that it was obvious that what matters in an output stage is the overall current gain of all the paralleled devices.

Zero Cool 17th May 2006 03:39 AM

OHHHHHHH now i get it! DUH thank you Eva!

Zero Cool 17th May 2006 03:42 AM

Ok, so now, how close is close enough & not close enough. lets say out totals were 26.50 and ??? at what point does the spread become unaccaptable?


Zc

Eva 17th May 2006 04:03 AM

Most people has been happily listening to class B amplifiers with over 2:1 gain mismatches between output stage halves for ages, and they don't complain at all, so who cares :D It seems quite strange to see such advanced matching practices applied to a 1976 product.

I would try to get close to 25, that seems to be the target value, together witn 72 for the drivers for a total gain of 1800.

Which are the gain figures of broken transistors and available replacements?

Zero Cool 17th May 2006 04:10 AM

One Beta 17 output device failed. I was able to buy a replacement Beta 17 device directly from A.R. and One Driver failed. I bought a matched pair of drivers from A.R. so i should be ok.

I was just concerned that i would have to replace ALL the transistors because the numbers were all different. I missunderstood how the numbers worked.

Zc

P.S. one capacitor in the V-reg for the front end shorted out sending Un-reg to the driver & first output device in line before the fuses blew.

Eva 17th May 2006 04:16 AM

Could you post some internal and external pictures of that strange 1976 thing? I'm starting to become curious about it :)

Zero Cool 17th May 2006 10:16 PM

2 Attachment(s)
Here are the schematics. Note these is not standard BIAS arraingment. isntead they use 2 constant current sources to bias the output stage.

Zero Cool 17th May 2006 10:17 PM

2 Attachment(s)
And here is the power supply.

I will take some photos of the unit and post them later.

Zero Cool 17th May 2006 10:19 PM

Oh and BTW, i talked with A.R. today and the beta numbers actually have a "1" in front. they shorted the number to fit. so a Transistor marked 17 is actually a beta of 117 so the numbers are as far off as they seem anyway.



Zc


All times are GMT. The time now is 04:49 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2