Measuring hFE for BJTs
Mar 21, 2006 at 6:23 PM Thread Starter Post #1 of 11

RLCrisp

New Head-Fier
Joined
Mar 19, 2004
Posts
29
Likes
0
I'm going to be builidng a dynalo soon, and I'm trying to find the best way to measure hFE for transistor matching.

I'm also going to be buying a multimeter soon, and I see many have the ability to measure hFE. Is this an accurate method?

My other method revolves around what hFE actually is. For lack of thinking what else it could be, is the following true? Forward Current Gain = hFE = Beta? If so, I can just set up a circuit to measure current gain (although the multimeter method is easier).

Thanks,
Ryan
 
Mar 21, 2006 at 7:45 PM Post #2 of 11
You might want to check out the dynalo construction thread for some details. As for description on hfe matching, matching using the multimeter built-in functionality isn't very good, mainly because you need to source the transistors with the appropriate current under the same conditions as the amp. For detailed methods of transistor matching, you can read the dynahi transistor matching thread given here:
http://www6.head-fi.org/forums/showt...istor+matching

Essentially all you need is the right linear pot, a 9V battery, a breadboard, and a multimeter... Hope this helps!
 
Mar 21, 2006 at 10:35 PM Post #3 of 11
I just got my new DMM today and it using the built in tester (2.8V, 10uA)

These transistors have good linearity so it shouldn't be a problem. Others have used hFE testers and then tried the other way with close results.

My problem is my 1015 are from 208 to 255. The 1815 seems to be clumped at 241 to 248 (most around 248).
Hmm..
 
Mar 21, 2006 at 11:48 PM Post #4 of 11
Quote:

Originally Posted by Paragon
I just got my new DMM today and it using the built in tester (2.8V, 10uA)

These transistors have good linearity so it shouldn't be a problem. Others have used hFE testers and then tried the other way with close results.

My problem is my 1015 are from 208 to 255. The 1815 seems to be clumped at 241 to 248 (most around 248).
Hmm..



As for linearity, it is true that you shouldn't have much trouble, however, throw in thermal effects and DMM hFE tester won't get you accurate measurements under stabilized thermal conditions. This is primarily due to current flow across the output transistors to be high compared with what DMM hFE tester's use, which will result in higher temperatures on the transistors. I normally let the transistors stabilize before taking a final reading for hFE.

I've never had much problem matching dynalo transistors since I purchased 4x as many as I needed, the dynahi trannies on the other hand are a b_tch to match!
 
Mar 22, 2006 at 12:07 AM Post #5 of 11
I'll have to compare and see how big of a difference it makes and just to see how accurate the meter is.
 
Mar 22, 2006 at 2:47 PM Post #6 of 11
I've heard quite a bit that I need to get transistors from the same gain class. Is this something I should specify when I order them or if i order them all from the same place should this be assumed?

Don't really know why two transistors with same part # would have a different gain class, or what gain class is.
 
Mar 22, 2006 at 3:56 PM Post #7 of 11
You have to ask for it. One distributor may have several gain classes available and they won't care what type they send you unless you ask for it.

BTW, gain class would refer to different intervals where the transistor's hFE gain may be within. For example, for the 1815 transistors have 4 different gain classes:
O: 70~140, Y: 120~240, GR: 200~400, BL: 350~700

Some gain classes overlap as you can see.
 
Mar 22, 2006 at 4:36 PM Post #9 of 11
Quote:

Originally Posted by RLCrisp
So ideally you would want to ask for transistors from the range class with the smallest range (i.e O for 1815).


I don't think the actual gain range really makes a difference (correct me if I'm wrong oh DIY gods
smily_headphones1.gif
), you really care only about the matching of the gain classes between complementary transistors, e.g., 1015 and 1815. Only so that you can get hFE's between a pair of transistors with equivalent values when you do matching.
 
Mar 22, 2006 at 5:28 PM Post #10 of 11
Right. I'm just figuring it will be easier to match hFE on transistors that will have only 70 A/A of gain range than 350 A/A. I would imagine you to need a lot more class BL (350-700 A/A) transistors to find 2 or more that are close than if you had class O (70-140 A/A).

Of course, I could be wrong (tell me if I am).
 
Mar 22, 2006 at 5:51 PM Post #11 of 11
Quote:

Originally Posted by RLCrisp
Right. I'm just figuring it will be easier to match hFE on transistors that will have only 70 A/A of gain range than 350 A/A. I would imagine you to need a lot more class BL (350-700 A/A) transistors to find 2 or more that are close than if you had class O (70-140 A/A).

Of course, I could be wrong (tell me if I am).



really depends on the transistors. The spread is usually uneven, but depends on your experiences
smily_headphones1.gif
. Good luck!
 

Users who are viewing this thread

Back
Top