74HC32 can't drive BC817 NPN transistor

[Help]

Hello All,

I can't make the BC817 working when i using HC32 to drive it. The transistor Collector is always Low, 0V. The NPN nominal ON and Off is just 0.92us. The NPN will work when I directly use the MCU I/O instead of using HC32.

May i know what the problem is? Why the transistor can't work when using HC32 driver? One more thing, i did tried change the R1 value from 220~500~1K Ohm but still not function. Does anyone can suggest me how to overcome this problem?

Thank you..

[koky]

try with oscilloscope of if you don't have it , try to run at more slow fequency. p.e. 1 sec, so you can use tester to measure every point
or try to decrement value of resistor in base of NPN

[Help]

Hello Koky,

Thanks for your reply.

Ya, i did use oscilloscope to measure it. The Low signal is around 920ns. Ya, i disconnect the output of HC32 and directly connect to MCU I/O pin and write a simple PWM code with almost the same (signal low ~800ns) frequency to drive the transistor. The NPN is working fine. Just that using HC32 output is not working. I did change the new HC32 and 74LVC32 but the result still the same.

I did check the HC32 datasheet. The HC32 output can be source/sink 25mA. I tried used 110Ohm for R1, it still not working.

[dennis78]

Your transistor is in "oversaturation" state, because your base current is too high. For BC817 hfe(min)=100. It's mean R1(max) should be about R1=0.75*100*R2. Because hfe constant varies from model to model in long range 100-600, you should little experiment(For example try for hfe typical 200 or etc.). For several reasons R1 can not be too high and not too low. R1 value should be in range 33k-100k.  For relative fast switching signals you must add capacitance C1 parallel to R1 for fast charging/discharging base electricity. For C1 you can use capacitor 22p - 100p (again little experiment). If this not help try with decreasing R2 to 1k (including previous specified rules).

Does transistor work good in stationary conditions?

I hope this will help to you.

[Help]

Hello Dennis78,

Thanks for your reply and advice.

R1 = 0.75 * 100 * R2

If the R2 = 2K2

Then R1 = ~165K Ohm

Is it correct?

May i know what type of capacitor should i use? ceramic cap?

What do you mean "Does transistor work good in stationary conditions?"

Thanks..

[dennis78]

Hello Dennis78,

Thanks for your reply and advice.

R1 = 0.75 * 100 * R2

If the R2 = 2K2

Then R1 = ~165K Ohm

Is it correct?

May i know what type of capacitor should i use? ceramic cap?

What do you mean "Does transistor work good in stationary conditions?"

Thanks..

I updated previous post. For R1 try 75k or similar (165k is too high for other reasons). Every type capacitor less than 200p should be ok.

Conditions are:
1. Constant (not switching, always same state) logical output from HC '1' -> Vce <0.5v
2. Constant logical outptut from HC '0' -> Vce=5V.

[Help]

Ok, I'm not sure i have <200pF ceramic cap not. Later i will try it, and see what's the result then will feedback to you ya.

R1 = ~75K and R2 = 1K~2K2, is it correct?
 
Thank you..

[dennis78]

Ok, I'm not sure i have <200pF ceramic cap not. Later i will try it, and see what's the result then will feedback to you ya.

R1 = ~75K and R2 = 1K~2K2, is it correct?

Thank you..

Yes, for start try with these values.

[Help]

Dennis78, I just do some study from http://www.kpsec.freeuk.com/trancirc.htm

We can try to use this following formula to calculate the R2 and R1.

Ic(load current) =   supply voltage Vs / load resistance R2
R1 =   (Vc × hFE) / (5 × Ic)    where Vc = IC supply voltage (in a simple circuit with one supply this is Vs)

or

R1 = 0.2 × R2 × hFE

[dennis78]

Dennis78, I just do some study from http://www.kpsec.freeuk.com/trancirc.htm

We can try to use this following formula to calculate the R2 and R1.

Ic(load current) =   supply voltage Vs / load resistance R2
R1 =   (Vc × hFE) / (5 × Ic)    where Vc = IC supply voltage (in a simple circuit with one supply this is Vs)

or

R1 = 0.2 × R2 × hFE

Generally, this is procedure for calculation in statical conditions. Coefficient '5(0.2)' is "technical reserve". Exactly value is not important because hfe varies depending on the type, current, voltage,... , but formula is good way.

[zuisti]

Hi Help;
I think your problem is much simpler!

The HC32 is a "Quad 2-input OR gate"
But what's about the other input of the OR gate what you are using?
It must be connected to ground or to the other (used) input. See the truth-table.
(If it is floating or connected to the VCC, the output is always high).

[solutions]

Use a FET and stop wasting your time trying to source and sink current in an HC gate

[SteveyG]

As dennis78 asked, if you switch the transistor on and off slowly, does it work, or is it the fast switching which is causing the problem?

[hate]

dennis78: I really would like to hear what's the theory behind 'R1=0.75*100*R2'.

Here is my calculations:
1- For a .7V Base-Emitter voltage drop, the value of R1 is calculated as follows; R1=(Vsupply-Vbe)/Ib. Ib is the base current to switch the transistor ON. For R1=2k2, Ib=(5-.7)/2.2=~2mA seems good.
2- To push the transistor into saturation (if that is desired), the Collector-Emitter voltage should be down to the saturation voltage. From the datasheet that's about .1V for 200mA collector current at 25 degrees so R2=(Vsupply-Vcesat)/Icsat. Icsat is the saturation current desired and can be approximated by Ic=Ib*hfe. Considering a 2mA base current the collector current will be minimum 2*100=200mA and inserting this into the equation for R2 gives R2=(5-.1)/.2=24.5 Ohms. That's the minimum value to force the transistor into saturation. But using a resistor that low is not very efficient so a resistor appropriate for the desired collector current is mostly used. Say for a 5mA collector current R2~=5/.005=1000 Ohms. You can ignore the collector-emitter saturation voltage for values that low.

After all this I too think this is not ur problem as the values seem to be quite normal. I agree with 'zuisti' here that u should check the other input for whichever gate u're using on the hc32 IC. It should be connected to logic low (GND in ur case) to get a pulse on the output.

Regards...

[dennis78]

dennis78: I really would like to hear what's the theory behind 'R1=0.75*100*R2'.

Here is my calculations:
1- For a .7V Base-Emitter voltage drop, the value of R1 is calculated as follows; R1=(Vsupply-Vbe)/Ib. Ib is the base current to switch the transistor ON. For R1=2k2, Ib=(5-.7)/2.2=~2mA seems good.
2- To push the transistor into saturation (if that is desired), the Collector-Emitter voltage should be down to the saturation voltage. From the datasheet that's about .1V for 200mA collector current at 25 degrees so R2=(Vsupply-Vcesat)/Icsat. Icsat is the saturation current desired and can be approximated by Ic=Ib*hfe. Considering a 2mA base current the collector current will be minimum 2*100=200mA and inserting this into the equation for R2 gives R2=(5-.1)/.2=24.5 Ohms. That's the minimum value to force the transistor into saturation. But using a resistor that low is not very efficient so a resistor appropriate for the desired collector current is mostly used. Say for a 5mA collector current R2~=5/.005=1000 Ohms. You can ignore the collector-emitter saturation voltage for values that low.

After all this I too think this is not ur problem as the values seem to be quite normal. I agree with 'zuisti' here that u should check the other input for whichever gate u're using on the hc32 IC. It should be connected to logic low (GND in ur case) to get a pulse on the output.

Regards...

Ib=(Vb-Vbe)/R1; Ic=(Vcc-Vces)/R2; Ic=hfe*Ib;  Vb=5V,Vcc=5V. Vbe,Vce<<5V; -> Vbe=0,Vces=0

Ib=Vb/R1,Ic=Vcc/R2,  Saturation condition(approximately): Ib*hfe>=Ic -> R1min=hfe*R2

hfemin(from dataseheet)=100 -> R1min=100*R2, but maybe hfe can be smaller, and because Vbe=0 in previous calculation, I get for reserve coefficient 0.75 -> final result R1min=0.75*100*R2

I hope this explain...

[hate]

Ok fair enough. But I think Ic=Vcc/R1 should be Ic=Vcc/R2. Also assuming Vbe=0 and a smaller hfe than the minimum are quite bad approximations in my opinion. And transistor needs a minimum base current to switch ON especially in higher frequencies. Considering a 1kOhms pullup (which I think is a very low value for a pullup) resistor, the base resistor turns out to be R1=.75*100*1=75kOhms. Then Ib becomes (5-.7)/75~=57uA which is also very low for a base current to turn the transistor on. Just my opinion.

Regards...

[dennis78]

Ok. I corrected the mistake in formula (fast typing...). I all time talk about minimal current, but enough for saturation margin. I calc R1 for guaranteed saturation, but no for too high base current. I told our friend Help (If this is reason for his problem) to put capacitor C1 value about 100p parallel to R1 for fast charging/discharging base electricity. I think 75uA is enough for small general purpose transistor like BC817. I told him he must do little experiments with this starting values.

[hate]

I think 75uA is enough for small general purpose transistor like BC817.

That depends on what frequency the transistor is switched. As a simple example a MOSFET having an input impedance in order of teraohms on DC may need 2-3A of gate currents to switch at MHz frequencies. I still think a 75uA base current is not enough to switch at frequencies other than DC if the transistor is not a 'superbeta' version. 'Help' needs to test it.

Regards...

[dennis78]

It's one of the reasons why I told him to put C1=100pF parallel to R1. Xc=(1/(2*Pi*1Mhz*100pF))=1.6k<<75k -> Ib~Ub/Xc=5/1.6k=3mA>>75uA!!!!!! in switching states.

Of course, this is very approximate calculation, but enough for analysis.

Good night!

[Help]

Hello dennis78, Hate, SteveyG, pjmelect, zuisti

Thanks for your all discussion and i get some hint from the discussion.

Sorry, for never reply for so long... because i'm spent too much time on doing testing until yesterday nite. Finally i get the thing work. May be late by to day I will capture the waveform with using scope then we analysis the result again ya.. and see why the BC817 can't saturate.

Thank a lot...

[hate]

It's one of the reasons why I told him to put C1=100pF parallel to R1. Xc=(1/(2*Pi*1Mhz*100pF))=1.6k<<75k -> Ib~Ub/Xc=5/1.6k=3mA>>75uA!!!!!! in switching states.

Fair enough. Good for current consumption, bad for frequency response but a nice way to decrease dissipated power.

Regards...

[Help]

Re: 74HC32 drive BC817 NPN transistor
Dear All,

Actually i'm using 8051 MCU to control the driver. Following is the practical result and the waveform.

CCT

Wave A WA
Wave B WC
Wave C WB
Wave D WQ

8051 port current 15mA

R1 = 2K2, R3 = 2K2, WC = WA = 1.875mA
WAVE11.JPEG
* No waveform on Wave D (WQ).

R1 = 10K, R3 = 2K2, WC = WA = 1.875mA
WAVE15.JPEG
Red waveform (WA) - Waveform getting better
Yel waveform (WQ) - Transistor is not fully saturation, <400mV. HC32 not enough current to drive the BC817 if R3 is 2K2??

R1 = 10K, R3 = 1K, WC = WA = 1.875mA
WAVE19.JPEG
Red waveform (WA) - Waveform getting better
Yel waveform (WQ) - Transistor is saturated and able to trigger my following circuity

[hate]

There is something strange in ur graphs. In 'wave11', the inverted output of the hc32 'wq' doesn't behave like the output of an OR gate for the corresponding inputs (wb, wc). Also why are u using the transistor in the input of hc32? For inverting purposes or for buffering? If the latter, u don't need to use that as the hc32 is a CMOS IC and doesn't need much current to turn on. Can u pls provide a graph of just wc, wb, wd signals. Also what are u trying to achieve with this circuit? Maybe there may be some other way. ;/ Pls also provide what 8051 IC u're using and to what pins wc and wb are connected.

Regards...

PS: You don't need to thank each and every reply, this forum is also for helping each other not just for sharing, that's a routine.

[Help]

Hello Hate,

Ya, you are rite. May be is because i'm using single probe to capture/save the waveform and the trigger level. My scope only have 2 channel.

Ya, for inverting purposes. Because this A15 pin is either enable external SRAM or IO and the output of WQ is to enable the IO. I'm using AT89C51RE2. WB and WD is not connecting to any where just for Input signal of HC32 and drive the transistor. WC is connected 8051 /WR pin. Ya, no problem latter i will capture the waveform/grah signal for you..

PS: Is from my sincere. ya, i wish my problem can be solve and also help our member up in future if they facing the same problem.

Best Regards,...

[hate]

I didn't get the part how u enable disable SRAM and IO. Currently WC or WA both can change the output (WQ) independently. Are u sure u don't need an AND gate there? Whether u need AND or OR gate u can use the 2 diode versions (link below). It'll be more practical and easier to analyze.
http://www.play-hookey.com/digital/experiments/dl_or2.html

Regards...