# Why does a 2N2222A Darlington pair have 1/2 the hFE as a single one?

I set up a darlington pair out of 2 2N2222As, each with a DMM hFE of 300 or so, and the resulting hFE is 170.

I read on the Interwebs that a Darlington can multiply or add the hFEs of the individual transistors, but this isn't the case (at least with BJTs).

What is the real ballpark hFE calc for a NPN Darlington pair, then?

• "DMM hFE" Question done, let's go home. – Ignacio Vazquez-Abrams Jul 23 '17 at 22:27
• Your process or tools are deceiving you, give it a try in LTSpice – sstobbe Jul 23 '17 at 22:31
• You guys are right, of course. I'm probably way out of the range of this \$5 meter! Thanks! :) – Compassionate Narcissist Jul 23 '17 at 22:41

If you really doubt that the Darlington setup you got doesn't work. Set it up according to this schematic to verify it. The 5MΩ is your finger. The red circle is an LED.

Here's the link to the schematic if you want to mess around.

And as wikipedia will tell you, the resulting Darlington beta will be $h_{fe_{darlington}}=h_{fe_1}×h_{fe_2}+h_{fe_1}$

And in your case it would be $300^2+300 = 90300$

The reason for why your tool gave a wrong answer is most definitive because it is a transistor measurer, not a Darlington transistor measurer. And having such fine equipment to measure that kind of extreme beta is unreal. Also the $V_{be}$ for a Darlington transistor is roughly twice higher than just one transistor, perhaps the tool doesn't take that into account.

• Thanks, @HarrySvensson. That formula was key. You think my crash-test-meter, a XL830L, is to blame for not going that high? :) – Compassionate Narcissist Jul 23 '17 at 22:43
• I doubt anything will go that high. Like, if you set up schematic so your $I_C$ = 1 A, then your $I_B$ should be 11 µA. And if the same meter should be able to measure much worse transistors with beta of, say 10, that means that the meter has to be able to measure all from 100mA to 11µA which is a little bit unreal to me. Also, that's when $I_C$ = 1 A, not every transistor can sustain that => your $I_B$ is even smaller... and your tool has to be even more delicate in order to pick up those small currents. And the range is still enormous. – Harry Svensson Jul 23 '17 at 22:47
• the stats in the Chinese-English instruction paper says that the tester uses 3v as Vce and 10ua as Ib, but it is just as likely it is a random number generator... :) – Compassionate Narcissist Jul 23 '17 at 23:15
• Then I'll assume that your meter can't deliver the estimated 903mA, compared to 3mA only one transistor would give. They probably don't expect people to stick Darlington Transistors into it, and/or it's resolution isn't good enough. And yes, it's part Chinese ;) That kind of answers everything. – Harry Svensson Jul 24 '17 at 0:28