BC550C hFE matching

[TunaHawk]

I need to hFE match these transistors the BC550C.

My DVM has the hFE capability it seems. So I plugged it in as see in the pic. And I get a reading of Hfe 006, most of the BC550C I have in a bag are measuring 006, with a 4 units measuring at 004. You can see in the pic the trans and the reading. So am I doing this right?

[IMGDEAD]http://s31.postimg.org/fsvzckcvv/DVM.jpg[/IMGDEAD]

 

[nigelwright7557]

It should be hfe of min 100.

 

[TunaHawk]

I flipped it around on the DVM. Now it reads 585. Is that about right?

It should be hfe of min 100.

 

[JMFahey]

Yes.

You were originally measuring it upside down , swapped collector and emitter.

It's still NPN and can be used in a limited way, but Hfe is *horrible*.

Next time when you post a picture, do not *hide* socket labels behind the transistor body , how do you expect an answer?

 

[TunaHawk]

Here is a another BC550C in the pic measures hFE 574

The other Transistors, measure as follows. I picked just any old one out of the bag.

BC557 = hFE 538
BC547C = hFE 581

The DVM is new. So are these really a bad grade of transistors, and should I order from another vendor. Which I believe are authentic not fake. Or is my Cen-Tech DVM giving me a false reading?

Yes.

You were originally measuring it upside down , swapped collector and emitter.

It's still NPN and can be used in a limited way, but Hfe is *horrible*.

Next time when you post a picture, do not *hide* socket labels behind the transistor body , how do you expect an answer?

 

[jan.didden]

What's your problem? These are very good Hfe numbers.

Do you ever look at data sheets??

Jan

 

[TunaHawk]

So nigel and JMfahey say the numbers are poor. Jan who a smart dude. Says the numbers are very good. Confused


The datasheet for the BC550C shows hFe min 110 max 800

I get that, but where do I get 110-200, is that some kind of mil spec, or?

 

[jan.didden]

So nigel and JMfahey say the numbers are poor. Jan who a smart dude. Says the numbers are very good. Confused


The datasheet for the BC550C shows hFe min 110 max 800

I get that, but where do I get 110-200, is that some kind of mil spec, or?

No they didn'say that. Read it carefully - they were referring to your wrong way of measuring.
It can be DIY but you still need the old brain ��

Why do you want max 200?

 

[Mooly]

A lot depends on which manufacturer and whose data sheet you look at. These are ancient devices and now available as a 'generic' part that probably varies somewhat from the original devices.

Your hFE values are 'excellent' in as far as the limited info of your test set up goes. The original data sheets showed '270' as typical with no maximum and curiously, no minimum either.

Also, small signal current gain varies dramatically depending on the test current. So we need to know what value of collector current your meter is working with to give a more informed answer.

Simple answer... they are fine

 

[erik777]

http://www.farnell.com/datasheets/1733168.pdf

 

[jan.didden]

Higher or lower is better as a rule of thumb?

Higher of course. The whole idea is that it amplifies Ib as much as possible.

Jan

 

[Mooly]

Higher or lower is better as a rule of thumb?

It doesn't always work like that

Any circuit should be designed to operate correctly with best and worst case values of the device in any location in the circuit. If a particular gain or gain group of device is needed for a specific location, then the circuit should state that.

 

[AndrewT]

bc550c are usually rated as 400 to 800 for hFE in datasheets.

If you need less than 200 for hFE, then the b grade and c grade are the wrong devices.
a grade might meet you target.
or buy mps42, or 2n5551 they have much lower gain than c grade

BUT !!!!
the hFE of af a DMM is not accurate. The DMM sends a nearly fixed and nearly constant very low value of base current into the DUT (Device Under Test) and reads off a voltage drop across a current sensing resistor.
Your circuit may operate at a very different Ib or Ie and then your DUT may have a different hFE.
This is what Mooly is referring to

So we need to know what value of collector current your meter is working with to give a more informed answer.

 

[jan.didden]

Then again, I do know a lot of circuits that would not work due to too low Hfe.
I do not know of any circuit that would not work because the Hfe is too high...

Jan

 

[erik777]

So if someone needs to match some PNP or NPN's.
Would something like this Trans Analyser be helpful instead of the typical HFE slot often found on most DVM's ?

http://www.farnell.com/datasheets/1733168.pdf

I mean this seems more robust than some of the DIY circuits that people use to match transistors, right?

 

[Mooly]

Ideally, transistors should be matched under operating conditions that are similar to what they would see in the intended circuit. So again, that brings us back to knowing the conditions of the test set up.

A DVM can 'match' hFE but it is doing so under the specific operating conditions of the test meter. In fairness, that comes pretty close for things such as matching pairs for input stages and so on.

Like so many things, you can refine and improve on test procedures... it all depends what you are looking for.

One note of caution (and this is something that I have no proof or experience with) is that if a tester reverse biases a junction (say B-E) and the transistor is a 'sensitive' device such as a high gain RF transistor then I believe that the noise figure of the device can become degraded. As I say, I have no first hand experience of this but the mechanism for degradation is well known.

 

[erik777]

Great stuff Mooly!

Ideally, transistors should be matched under operating conditions that are similar to what they would see in the intended circuit. So again, that brings us back to knowing the conditions of the test set up.

A DVM can 'match' hFE but it is doing so under the specific operating conditions of the test meter. In fairness, that comes pretty close for things such as matching pairs for input stages and so on.

Like so many things, you can refine and improve on test procedures... it all depends what you are looking for.

One note of caution (and this is something that I have no proof or experience with) is that if a tester reverse biases a junction (say B-E) and the transistor is a 'sensitive' device such as a high gain RF transistor then I believe that the noise figure of the device can become degraded. As I say, I have no first hand experience of this but the mechanism for degradation is well known.

 

[JMFahey]

My "horrible Hfe" comment applies to reversed transistor and measured value of 6 (six).

500 something is **excellent**
To boot, although datasheet suggests 400-800 is acceptable, those are actually within less than 10% ... incredibly good.

Something else: although devices in general are non linear, if they match at any reasonable current value such as applied by a multimeter, being same batch, same process implies that they should track quite well at other values, after all they are "same DNA" so if used at low current values such as found in preamps and similar small signal level applications, they should remain close.


Of course, if you want closest possible you'd need to specify monolithic devices, doubt it's the case.

 

[googlyone]

Spending idle hours matching hFE can be therapeutic, but is not necessarily the most critical thing you might want to be watching.

You might also ant to consider:
- What is the Vbe matching of the devices?
- At what current are you testing
- Are the devices the same temperature when testing...

In my experience a check that the hFE is "decent" and more than the basic design value, then you are good. MAsive or matched hFE is not necessarily in itself the most important thing.

Pay attention to thermal matching of devices where used for example in differential pairs / current mirrors / log amplifiers etc.

 

[davidsrsb]

Pay attention to thermal matching of devices where used for example in differential pairs / current mirrors / log amplifiers etc.

Once you get to exotica like log amplifiers, you have to start worrying about thermal gradients in the pcb layout. Not much point carefully matching if one device is close to a hot regulator or output transistor