EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: petert on September 04, 2020, 01:11:29 am

Title: Understanding forced hFE in saturated BJT
Post by: petert on September 04, 2020, 01:11:29 am
Hello,

When using a BJT (NPN) as a switch, there is no classical hFE, but you still have to assume a reasonable Ic/Ib ratio to put a transistor into saturation, while keeping the Ib current as low as possible (to not waste energy/create needless heat), and choosing the biasing transistor accordingly.

Reading
https://www.reddit.com/r/AskElectronics/comments/7z3cvn/what_is_the_difference_between_beta_and_forced/ (https://www.reddit.com/r/AskElectronics/comments/7z3cvn/what_is_the_difference_between_beta_and_forced/) and
https://electronics.stackexchange.com/questions/207374/forced-beta-too-small (https://electronics.stackexchange.com/questions/207374/forced-beta-too-small) or this
https://www.physicsforums.com/threads/transistor-current-gain-in-saturation-mode.865392/ (https://www.physicsforums.com/threads/transistor-current-gain-in-saturation-mode.865392/)
it seems 10 is a common value for forced hFE.

A video showing the base resistor calculation for saturation mode:
Code: [Select]
https://www.youtube.com/watch?v=hEVLJc4R9JE

When looking at this datasheet (https://www.mouser.com/datasheet/2/149/BC547-190204.pdf) for example, in Figure 4 Ic/Ib = 10. But when looking at the table on page 2, the values for VCE(sat) suggest a ratio Ic/Ib = 20.

This raises a few questions:

In other words, I am confused how the forced beta values were obtained. "Normal" hFE could be measured for a transistor in the active/linear region.
Title: Re: Understanding forced hFE in saturated BJT
Post by: Doctorandus_P on September 04, 2020, 02:45:39 am
The whole "forced beta" is bull. Just forget that nonsense.

Hfe is a transistor parameter. When the transistor is used as an amplifier (standard 4 resistor circuit) then you can use it in some calculations. When the base current is increased then the output either clips at one of the voltage rails, or dissipation of the transistor gets so big that it lets out the magic smoke.

Hfe is somewhat constant within limits.
One of the limits, is that Hfe decreases with high current (which is listed as the maximum current for a transitor).

But there is no substitute for doing some measurements yourself.
Grab a few transistors and a breadboard and measure it yourself.
Also look at the extremes of operation of the transistor. Does Beta change? Does it burn out?
(TO092 trannies cost a few cents each. Get a mixed box from Ali / Eba / China.)
Title: Re: Understanding forced hFE in saturated BJT
Post by: David Hess on September 04, 2020, 03:41:05 am
Quote from: petert on September 04, 2020, 01:11:29 am
If there is no constant Ic/Ib ratio (table vs. figure), how can you make a measurement at Ic/Ib=10, as in Figure 4?

They make that measurement by setting the base current to 1/10th of the collector current which is what a forced beta of 10 means.  If I was making a test circuit to do it over a wide range of collector current, I would probably use a current mirror but that is not representative of real world applications.

Quote
People speak of setting the Ic/Ib ratio to a forced beta, how exactly is this done? (The figure shows a varying range of Vbe and Vce, so it can't be those voltages controlling Ic/Ib.)

Vbe reliably falls within a small range so the base current is set just like any other circuit which drives a current through a diode.  Typically a voltage source with a medium impedance is used but sometimes constant current drive is used.

Quote
10 is a common value for forced hFE.

10 is a common value but it depends on the transistor and operating conditions; values from 20 to 50 are also possible.

High performance applications control the base current to hold the transistor just out of saturation like a baker clamp would.

Title: Re: Understanding forced hFE in saturated BJT
Post by: petert on September 04, 2020, 03:52:27 am
Quote from: Doctorandus_P on September 04, 2020, 02:45:39 am
But there is no substitute for doing some measurements yourself.
Grab a few transistors and a breadboard and measure it yourself.
How does that help me if I want to design a circuit that works with different brands of transistors and remains compatible in future?
(Assuming I want to switch a circuit on and off, going from saturation to cut-off region, not amplification in the active region.)

I am trying to find a reasonable minimal current amplification, that can be safely assumed to be given over a larger range of transistors (for switching).
Title: Re: Understanding forced hFE in saturated BJT
Post by: petert on September 04, 2020, 04:00:35 am
Quote from: David Hess on September 04, 2020, 03:41:05 am
Vbe reliably falls within a small range so the base current is set just like any other circuit which drives a current through a diode.  Typically a voltage source with a medium impedance is used but sometimes constant current drive is used.
Setting Ib makes sense, but I am not sure about Ic. Sorry if I am dense here, but how can you be sure Ic will be 10 times Ib? Isn't it the transistor and its amplification properties that will control Ic? How could you set it "manually"? Or do you just assume Ic is at least 10 times Ib, and then limit the current over Ic to be sure it doesn't exceed 10 times Ib?
Title: Re: Understanding forced hFE in saturated BJT
Post by: David Hess on September 04, 2020, 04:50:05 am
Quote from: petert on September 04, 2020, 04:00:35 am
Setting Ib makes sense, but I am not sure about Ic. Sorry if I am dense here, but how can you be sure Ic will be 10 times Ib? Isn't it the transistor and its amplification properties that will control Ic? How could you set it "manually"? Or do you just assume Ic is at least 10 times Ib, and then limit the current over Ic to be sure it doesn't exceed 10 times Ib?

Since the transistor is assumed to be operating in saturation, only the supply voltage and load impedance control the current.

The design procedure is to find the worst case collector current given the load impedance and supply voltage, and combine that with the worst case transistor hfe, to find the minimum required base current to guaranty that the transistor remains in saturation under all conditions, and then include some safety factor.  Commonly that means a forced beta of 10 but it could be several times higher in some applications with the right transistor; high gain transistors can have a worst case hfe of 100 or more at useful collector currents.

If the collector current is poorly defined, then a more complex drive circuit might be used which does not rely on force beta.
Title: Re: Understanding forced hFE in saturated BJT
Post by: Berni on September 04, 2020, 05:20:23 am
Quote from: petert on September 04, 2020, 04:00:35 am
Quote from: David Hess on September 04, 2020, 03:41:05 am
Vbe reliably falls within a small range so the base current is set just like any other circuit which drives a current through a diode.  Typically a voltage source with a medium impedance is used but sometimes constant current drive is used.
Setting Ib makes sense, but I am not sure about Ic. Sorry if I am dense here, but how can you be sure Ic will be 10 times Ib? Isn't it the transistor and its amplification properties that will control Ic? How could you set it "manually"? Or do you just assume Ic is at least 10 times Ib, and then limit the current over Ic to be sure it doesn't exceed 10 times Ib?

Well the transistor is NOT a current source. It can't just magically push the current trough no matter what. Whatever is supplying current to the collector pin has to give it to it. If you apply 0V to the collector the current is 0 mA no matter what state the transistor is in. If you apply a fixed 10V to the collector then the transistor can draw whatever current it wants.

The whole point of this is to characterize the transistors performance as a switch.

The gain of BJT transistors deteriorates at high currents, so as a result higher base currents are needed to keep the transistor in saturation at high collector currents. All of this lets you get an idea of how much base current you need when operating as a switch and how much collector voltage drop will the transistor have in such conditions, this drop helps you calculate how much power is dissipated and hence how hot the transistor will get.

So this "forced beta of 10" just means you will be providing it 1/10 the base current compared to the collector current you are trying to switch This amount of base current is enough to keep most transistors well into saturation, so it represents the best case voltage drop that this transistor can give you, so with it giving you the lowest power loss possible. You want to know this because it tells you how big of a heatsink you need to strap to that transistor to keep it from blowing up and melting.
Title: Re: Understanding forced hFE in saturated BJT
Post by: petert on September 04, 2020, 05:11:01 pm
Thanks!
Quote from: David Hess on September 04, 2020, 04:50:05 am
If the collector current is poorly defined, then a more complex drive circuit might be used which does not rely on force beta.
What circuits would that be for a transistor used as a switch?
Title: Re: Understanding forced hFE in saturated BJT
Post by: David Hess on September 04, 2020, 07:34:11 pm
Quote from: petert on September 04, 2020, 05:11:01 pm
Quote from: David Hess on September 04, 2020, 04:50:05 am
If the collector current is poorly defined, then a more complex drive circuit might be used which does not rely on force beta.

What circuits would that be for a transistor used as a switch?

A baker clamp does it be shunting the base drive into the collector as the collector voltage drops.  This suggests monitoring the collector voltage to control the base current.

Some ICs place another transistor in series with the collector, and then use the base current of the second transistor to scale the base current of the first.  This works because the current gain of both transistors will match.

Linear and switching bipolar ICs both implement anti-saturation circuits but for different reasons.  Switching regulators do it to lower storage time.  Linear regulators do it to prevent excessive base drive at dropout which would lower efficiency.
Title: Re: Understanding forced hFE in saturated BJT
Post by: T3sl4co1l on September 04, 2020, 07:51:26 pm
Also a Darlington, which delivers only as much base current as needed, and draws it from the collector (output) circuit; downside, the extra transistor in series increases voltage drop, and you can read the wording of this very explanation, as a feedback loop where, if collector voltage drops, well then base current drops, and so it's another kind of anti-saturation circuit.

Of course the intent of circuits like Baker clamp, and other more specialized anti-saturation circuits, is to get a usefully low voltage drop, operating the power transistor just shy of saturation, at exactly enough hFE to do the job.

When you have an unknown load, yes, the best you can do is either: use an anti-saturation circuit like these, or use Ib for the maximum design Ic and just deal with the lower efficiency and slower switching you get at lower Ic.  (Or shrug and use a MOSFET.)

FYI, a typical Baker clamp circuit looks like so:

(https://www.seventransistorlabs.com/Images/BakerClamp.png)

Interesting thing about using the diode-strapped transistor, is you can potentially get the Vbe to track better.

The resistor ratio R2/R3 sets Vce(sat), and of course the resistance (including R1) depends on how much base current you need (and R1 / (R2+R3) depends on what "logic level" is).

Can even do both at once, like so: https://www.seventransistorlabs.com/Images/BakerDarlington.png (https://www.seventransistorlabs.com/Images/BakerDarlington.png) (left side; right side is just an attempted parasitic model: it seems the circuit can be prone to oscillation, but it's hard to make the simulator agree), the advantage being that the first transistor doesn't need its collector tied to output, again, allowing lower Vce(sat) than a regular Darlington.  Notice the diode drops of D2 and Q3 B-E can effectively cancel out, giving Vce(sat) ~= Vbe(Q4), and that can be adjusted if you add voltage drops to the circuit (say by increasing V(D2) with a little resistance).

Tim
Title: Re: Understanding forced hFE in saturated BJT
Post by: petert on September 14, 2020, 11:48:22 am
Some other useful links:

http://electronicsbeliever.com/category/tutorials/bjt-tutorials/ (http://electronicsbeliever.com/category/tutorials/bjt-tutorials/)


especially:
http://electronicsbeliever.com/how-to-know-if-a-transistor-is-saturated/ (http://electronicsbeliever.com/how-to-know-if-a-transistor-is-saturated/)
http://electronicsbeliever.com/how-to-saturate-a-pnp-transistor/ (http://electronicsbeliever.com/how-to-saturate-a-pnp-transistor/)
http://electronicsbeliever.com/how-to-select-a-transistor/ (http://electronicsbeliever.com/how-to-select-a-transistor/)
http://electronicsbeliever.com/transistor-operation-mode-determination/ (http://electronicsbeliever.com/transistor-operation-mode-determination/)

http://electronicsbeliever.com/npn-transistor-principles-practical-uses/ (http://electronicsbeliever.com/npn-transistor-principles-practical-uses/)
Title: Re: Understanding forced hFE in saturated BJT
Post by: dietert1 on September 14, 2020, 03:48:33 pm
Another simple way to control base current near saturation is using a schottky diode between base and collector. It works as a negative feedback regulator to keep the transistor in a region with reasonable hfe and with normal switching behavior.

Regards, Dieter