Switching Transistor: destroying the switching pulse and creating crud on the Vcc..... Why?

Thread Starter

dr.evil

Joined Aug 18, 2010
80
Gents,
I have a serious problem with my Q2 switching transistor, it's destroying the switching pulse and creating crud on the Vcc. Please have a look at my awesum schematic, you would think I couldn't mess it up, but I did:
Screenshot from 2017-11-26 18-43-01.png
If I yank out Q2, then RED is the signal on where the base was, it's good enough to turn Q2 on. And BLUE is Vcc, both 500nS/div:
Screenshot from 2017-11-26 17-42-31.png
Then I insert a BD244C as Q2: RED Q2 Base, BLUE Vcc, both 500nS/div:
Screenshot from 2017-11-26 17-48-28.png
As you can see the noise on the Vcc is enhanced, and the base pulse is starting to crap out.
Then I put in a B772 as Q2, and then I get this:
Screenshot from 2017-11-26 17-53-11.png
And the base pulse is nearing the Vcc in both shape and form.

So the question is:

1. Why can't I maintain a rock steady switching pulse? And how do I do it?
2. And how do I eliminate the noise/crud on the Vcc?
 
Last edited:

crutschow

Joined Mar 14, 2008
34,280
First, Q2 is upside down.
(Note that the arrow direction in a BJT always denotes the normal current direction through the emitter).

Second, when you connect it properly (emitter to Vcc) you need to add a resistor (could be R1) in series with its base to control the base current.

Third, you also need a resistor in series with the base of Q1.
 

Thread Starter

dr.evil

Joined Aug 18, 2010
80
First, Q2 is upside down.
(Note that the arrow direction in a BJT always denotes the normal current direction through the emitter).

Second, when you connect it properly (emitter to Vcc) you need to add a resistor (could be R1) in series with its base to control the base current.

Third, you also need a resistor in series with the base of Q1.
Q2 was correct mounted in my circuit, adjusted the schematic to reflect same.

Base resistor of Q1 was already present in the circuit, so just added it to schematic.

Do I really need a base resistor on base of Q2? I mean I just want to open the flood gates of Q2 to maximum.
 

Thread Starter

dr.evil

Joined Aug 18, 2010
80
Without it, the only thing limiting the base-emitter current is the resistance/maximum-current capability of the source.
Remember a base-emitter junction looks like a forward-biased diode.
Ok, so I put a 220Ω on Q2 Base, and 1KΩ on Q1 collector, and voila! Problem persists:
Using B772 it looks like this:
Screenshot from 2017-11-26 19-37-00.png
 

crutschow

Joined Mar 14, 2008
34,280
To minimize crud on Vcc, connect a 100nF ceramic cap directly from the Q2 Vcc emitter connection to the R2 ground point, with as short leads as possible.
 

Thread Starter

dr.evil

Joined Aug 18, 2010
80
To minimize crud on Vcc, connect a 100nF ceramic cap directly from the Q2 Vcc emitter connection to the R2 ground point, with as short leads as possible.
Ok, 100nF added and the leads are 3mm long. Definitely less crud, but still no switching pulse present.
Screenshot from 2017-11-26 20-34-31.png
 

Papabravo

Joined Feb 24, 2006
21,158
Depending on the impedance of the "Switching Pulse" source you (have/have not) established a suitable bias point for switching applications. The source has to charge and discharge the base of Q1 through the 330R resistor. For example if you pull the base up to Vcc so Q2 will conduct, then use the low impedance source to pull it low you should get the behavior you are seeking.
 

crutschow

Joined Mar 14, 2008
34,280
What does the input signal look like?

The transistors may be too slow.
Try replacing them with fast saturated switches such as a 2N2369 for Q1 and a 2N5771 for Q2.
(You can often identify a fast switching transistor by its relatively low Vce max rating of typically 15V. But don't use an RF transistor because they are generally not fast when used as saturated switches).
 
Last edited:

Thread Starter

dr.evil

Joined Aug 18, 2010
80
Depending on the impedance of the "Switching Pulse" source you (have/have not) established a suitable bias point for switching applications. The source has to charge and discharge the base of Q1 through the 330R resistor. For example if you pull the base up to Vcc so Q2 will conduct, then use the low impedance source to pull it low you should get the behavior you are seeking.
Papabravo,

The "Switching Pulse" source is the output of a 74HC00, unfortunately I don't know it's exact impedance, but I believe should be around 50-100 ohms. I will try your suggestion and let you know.
 

Thread Starter

dr.evil

Joined Aug 18, 2010
80
What does the input signal look like?

The transistors may be too slow.
Try replacing them with fast saturated switches such as a 2N2369 for Q1 and a 2N5771 for Q2.
(You can often identify a fast switching transistor by its relatively low Vce max rating of typically 15V. But don't use an RF transistor because they are generally not fast when used as saturated switches).
The input signal is 912KHz square with a low period of 121nS.

I believe you could be right, I was concerned about that myself. Thank you for the excellent Vce rule of thumb, very useful.

However the 2N5771 is not suitable for my purpose, as Q2 will later on experience peak current (in the low period of the input signal) of up 4A, so I think I need some medium power switching transistor, can you suggest one? Preferable something that is available at Farnell, because I had to register with them for this project.
 

Thread Starter

dr.evil

Joined Aug 18, 2010
80
What does the input signal look like?

The transistors may be too slow.
Try replacing them with fast saturated switches such as a 2N2369 for Q1 and a 2N5771 for Q2.
(You can often identify a fast switching transistor by its relatively low Vce max rating of typically 15V. But don't use an RF transistor because they are generally not fast when used as saturated switches).
I found NSS40301MZ4 available at Farnell, do you believe it would be suitable?
 
Top