# NPN BJT Transistor HFE Saturation Switch

Discussion in 'General Electronics Chat' started by vandaycalta, Mar 22, 2016.

1. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Thank you to all for taking the time to view my thread. I have been working with NPN transistors as of late, I am just learning the fundamentals here, and had a few things that I was looking for clarification on. I do not have a schematic but hope the provided info will be enough for you to be able to provide some assistance. I am using a fairchild bu337-25 npn transistor, the supply voltage , for base and collector sides, is thru a 5v reg which is giving me 4.97 at the reg. output (the power supply to the volt reg. is set at 10v. from a extech c.v. dc power supply).I am trying to set up the npn to function as a switch. I understand that at cutoff the Vb will be less that Ve and Vc. I also know that in saturation that Vb is greater than Vc and greater than Ve ( greater than the .7 drop threshold). I understand that my Vce will be around .2v at saturation.

On a breadboard I built the circuit up in the following manner:
I used two resistors in parallel(98.9 ohm and 98.2 ohm = 42.7 ohm) to achieve a Ic of 92.3mA. (The current I want in the collector.)
The Beta (Hfe) for the bu337-25 is 160.
I calculated my Ib using the formula: Ib=Ic/hfe which came out to .0005768A
I then calculated my Rb using the formula: Rb=4.97-.7/.0005768 which gave me 7402.9 ohms for Rb.
The closest I could come to the Rb value was 7170 ohms.
This value made my current in the Vb .0005857A which should sat. the npn more. This changes my beta to .0923/.0005857 = 157.58.

After I built this circuit and probed it w/ my meter I found the following values:
Vb: 4.2v, Vbe .733v (it is above the threshold), and Vbc of .607v.
Vce was .125v.
All of which met my requirements of Vb being larger than Vc and Ve.

I made some alterations to the circuit and changed my Rb to 10,475 ohms which gave me a Ib of .000397A(actual meter reading) calculated to .0004009A. I was looking to get my Vce up to the .2v that I had read it had to be. I achieved a value of .206v for Vce. These also met all my requirements. But the beta was 232.49.

I understand that you want to add 5-10% to your Ib to really get the npn into saturation.
In the data sheet I am confused when they give you the beta of 160 as a min and let's say 400 as a max. This I do not understand because I have calculated values out that gave me a Vce of .120 and the beta was around 157.58 which is under 160 on the datasheet so is this still in saturation? DO I try to get the npn into saturation with 106 beta or 400 beta or somewhere in between?

How much Vce should I try for? Is .120 enough?
The graph for the Vce vs Ic doesn't seem to give me the correct numbers vs. what I see on the breadboard.
What is the best way to look at this in terms of saturating the transistor to use it as a switch?

Thanks for your time and help.

-Val

2. ### dl324 Distinguished Member

Mar 30, 2015
3,211
619
Welcome to AAC! Smaller paragraphs and schematics would be helpful.
The fact that transistor beta varies with current, collector-emitter voltage, temperature, etc complicates your calculations.

The general rule of thumb is to start with the beta at the closest current it will be switching and divide it by 10 to determine the base current required to saturate the transistor. You can get away with assuming a higher beta, but beta variations between transistors even from the same lot could cause circuit failures. If you want to be ultra conservative, use 0.1 of the minimum beta.

3. ### GopherT AAC Fanatic!

Nov 23, 2012
5,980
3,706
@vandaycalta

A quick example.

If you use a transistor of gain Hfe of 100 to 200 like a BC337-25, assume that your base current should be 10% of your collector current.

If you want 90 mA collector current and you have a 5V supply you have a 4.4V drop to your base.

Then your base current should be 9 mA. 4.4V/0.009A = 488 ohms, use a 470, 510 or 560 ohm resistor. If you are running into current limits or want to use low power, you can go to half of that amount of current in most cases, use a 1k (especially if your Hfe is well above 100). Don't over think it.

4. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Does the same apply to a NPN Darlington?
Thanks to all for your help!

5. ### GopherT AAC Fanatic!

Nov 23, 2012
5,980
3,706
For a darlington you can use as @dl324 described, 10% of Hfe (generally 0.3 to 3% of collector current).

6. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
l did some testing today to verify what was talked about yesterday in response to my thread. Here is what I found:

Resistance increases, Ib decreases,Vce increases, and Hfe increases with changes.( I was changing base resistance)
I found that even if my Ib was a little over 3x larger than my 10% of Ic I was shooting for, that my Ic was still the same(w/in .1mA). meaning I was VERY well saturated. and was a beta of 3.44.
Even with a beta of 67.6 I still found my Ic the same and a bias current of .0014A, based on a Ic of 94.7mA.

I found my Vce when saturated to be about .04v. I assumed that you would want the Vce to be about .2v but noticed that when you really added(is clamped the right word here?) resistance to the base that the beta would be upwards of 147-214 and the Vce where I wanted but then the Ic would decrease telling me that I was no longer in the saturation region.

Am I right to say that the Vce has to be only lower than .2v but not CLOSE to .2v as I was thinking it HAD to be? (.04v is ok because my Ic has not decreased?)

How much is beta really important in using the transistor as a switch? The 10% of Ic for Ib really puts one in a good spot as your range of swing(?) for going out of the sat. region has more wiggle room?

Val.

7. ### #12 Expert

Nov 30, 2010
16,248
6,744
I don't think so. A lot of transistor datasheets only promise 0.3 volt in saturation.
Then there is the 4401 which promises to get below 0.1 volt.
Look at the Vce/Ib curves and see that gain drops like a rock as Ib changes from 1% to 10% of Ic.

Gain is not a stable number. It is different for several transistors with the same part number. It changes with temperature, collector current, and collector voltage. The goal in designing with transistors is to make designs that find the exact gain of a transistor to be almost irrelevant.

File size:
489.6 KB
Views:
2
File size:
251.4 KB
Views:
2
8. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Thank you for your response. #12

I guess I am still confused. My test data showed me that for my transistor, npn fairchild bc337-25, https://www.fairchildsemi.com/datasheets/BC/BC337.pdf, that with a 10% beta as my guide and seeing a Vce of .04 compared with a beta of 5.5 and Vce of .038, compared to a Beta of 3.4 w/ a Vce of .037, a beta of 15.52 w/ a Vce of .043 , all these maintained the targeted Ic of 94.7mA, even w/ a beta of 49.7 and a Vce of .062 my Ic was still 94.7mA telling me it was still in saturation. Like I said, if I TRIED to make the Vce close to .2v w/ results of beta of 147.87 w/ a Vce of .115 but this decreased my Ic to 93.9mA as such a beta of 214.3 and a Vce of .178v gave me a Ic of 92.8 which confirm to me that I am not in saturation.

Thanks again.
Val

9. ### #12 Expert

Nov 30, 2010
16,248
6,744
So what?
What circuit are you working on that needs exactly 0.2 volts Vce?
Are you trying to force the published typical values be exactly what you can measure?
Do you think the gain should be the same at 0.2 Vce as it is at 5 volts Vce?

10. ### ronv AAC Fanatic!

Nov 12, 2008
3,281
1,232
I think what has you confused is the data sheet is not real good to understand your transistor.
If you look at Vcesat spec it only has a maximum and not at your current. So most transistors will be better.
What saturation means is that there will not be much improvement in Vce if you increase base current. This number of 10 is kind of a rule of thumb for all transistors. In reality if it is a high gain transistor using 20 might be close enough. If you want to set a certain Vce using beta is not the way to do it as there are large variations with different transistors as well as things like temperature.
Here is a data sheet with a little better specs and some graphs so you can get an idea of the relationships.

11. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4

1. No circuit. I am just trying to learn about npn bjt transistors
2. NO. I was just stating the results that I saw when I measured the circuit.
3. No. The gain will not be the same at .2v and at 5v. and I proved that by my testing.

My confusion is with Vce. Is my transistor saturated if, as my testing shows in my previous post, the Vce is roughly .04v.

12. ### ronv AAC Fanatic!

Nov 12, 2008
3,281
1,232
Sorry, I forgot to post the link.
https://www.fairchildsemi.com/datasheets/BC/BC547.pdf

13. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4

I agree with what you say about the Vce not changing much with increases in base current as that is exactly what i saw in my testing. The thing that is really stumping me is that with all that being said my Vce is not even close to .2v(or the .7 max value as stated in the data sheet for the bc337-25 that I have.) rather it is around .04-.038.
I was not trying to set a Vce of .2v but rather using it as a reference or guide to assure me that I was in the saturation region.

Val

14. ### #12 Expert

Nov 30, 2010
16,248
6,744
I believe you have found that this transistor is much better at saturation voltages than what is typical for this part number, and you are really stumped that this transistor exceeds its guaranteed minimum quality. What do you think we should do about this?

GopherT likes this.
15. ### dannyf Well-Known Member

Sep 13, 2015
1,767
356
Unimportant at all

Set ib to be 10 percent of Ic is a standard rule for high current switching. You can go lower for smaller current switching.

16. ### ronv AAC Fanatic!

Nov 12, 2008
3,281
1,232
Yes, I understand.
Your problem is that the transistor has no spec where you are running it. The only spec it has is that every one you buy will have a vce(sat) less than .7 V at collector current of 500 ma and any temperature between -55 and +150C.
So the only thing the spec tells you is that Vce(sat) will always be below .7 volts if you drive it with an Ib of 1/10 Ic.

17. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Am I right to say that this transistor is in saturation based on my findings?
I have a lot of 10 of these bc337-25 npn trans. I will do a cross the board test and record my data.
Are my findings common?

Val

18. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Right, that's my thinking! Thats what I thought the datasheet was telling me. It just seems that with all ive read and watched vieod wise that you have to hace about .2v Vce but I know that wasnt the case w/ my datasheet and my hard testing of the npn bu337-25.

Again confirming my results of .04v +/- are correct for being in saturation.

Thanks.
Val.

Nov 12, 2008
3,281
1,232
Yes.

20. ### vandaycalta Thread Starter Member

Mar 22, 2016
53
4
Thank you to all for all your help.
Painful at times but always fun to learn.

Val