Problems driving a large display with a transistor

  • Thread starter Thread starter insignia96
  • Start date Start date
  • Tags Tags
    Transistor
Click For Summary

Discussion Overview

The discussion revolves around the challenges faced in driving large multiplexed 7-segment displays using transistors. Participants explore issues related to circuit design, specifically focusing on the behavior of PNP and NPN transistors in the context of controlling LED segments, as well as the implications of voltage levels and current limiting in the circuit.

Discussion Character

  • Technical explanation
  • Debate/contested
  • Experimental/applied

Main Points Raised

  • One participant describes a problem where segments of the display are partially illuminated when they should be off, suspecting issues related to the transistor's base voltage.
  • Another participant suggests that the emitter-base voltage (Veb) is likely causing the transistor to remain on, recommending the use of an NPN transistor instead of a PNP for better control.
  • It is proposed that the base voltage needs to be pulled up to 21V to turn off the transistor completely, with suggestions to modify the circuit accordingly.
  • Some participants emphasize the importance of using current limiting resistors to prevent fluctuations in LED brightness and ensure consistent operation.
  • There is a suggestion to increase the power supply voltage to 24V and adjust resistor values for optimal performance.
  • A participant inquires about using simulation software to analyze circuit characteristics, indicating a desire for tools to better understand the circuit behavior.
  • One participant acknowledges their limited experience in circuit analysis and shares their revised circuit design, which incorporates feedback from others, ultimately reporting successful operation of the display.

Areas of Agreement / Disagreement

Participants express differing views on the best approach to resolve the display issue, with some advocating for the use of NPN transistors while others discuss the necessary voltage levels for proper operation. The discussion includes both agreement on the need for current limiting and differing opinions on circuit design specifics.

Contextual Notes

Some participants mention the need for specific voltage levels and resistor values without fully resolving the implications of these changes on circuit behavior. There are also references to the participants' varying levels of experience, which may influence the technical depth of the discussion.

Who May Find This Useful

Readers interested in electronics, particularly those working with LED displays and transistor circuits, may find the insights and troubleshooting approaches shared in this discussion beneficial.

insignia96
Messages
4
Reaction score
0
Background:
I am building a giant clock with four, multiplexed, 7 segment displays. The problem is that the displays are five inches tall, each segment consists internally of 10 LEDs wired in series and then three chains of 10 are wired in parallel. The datasheet for this chip is here (PDF WARNING). It lists the max continuous forward voltage as 21V and the max cont. forward current as 60 mA. The display is common anode and it is being driven by a bcd to seven segment chip (active-low outputs, for this type of display).

Here is an exact schematic of the part of my circuit i am using to drive the displays:
http://img338.imageshack.us/img338/8499/pnpsim.png

The reason why it is doing this is still eluding me, but when segments are supposed to be off they are sort of half on.

http://img14.imageshack.us/img14/3874/1215112141.jpg
(Picture of the problem, sorry for the quality, my cellphone was the only camera available)

The problem is very well defined in this picture, in case it wasn't obvious it is trying to display a five and the b and e segments, which should be off, are still half on.

I originally thought this might have been due to the transistor getting turned half of, hence the 10K pull-up resistor in the schematic, but that didn't help.

Also, it may be useful to note that when measured with a multimeter IRL and in sim software the node at the transistor's base reads ~19V. I am almost 100% sure this is related to the problem, but I don't know why the transistors don't just fry up if they are getting 19V across the base.

One more pic that may be useful:
http://img217.imageshack.us/img217/2796/1215112141b.jpg

It shows all 7 transistors (one for each segment), the gray wires lead to the 7447 chip's outputs and the clips are clipped to the proper pins on the display. Also I have checked repeatedly and no clips are touching. This happens with any number I display. FYI the breadboard's power rails are connected to ground and 5V.

Thanks in advance,
Isaiah (insignia96) :)
 
Last edited by a moderator:
Engineering news on Phys.org
Unfortunately, your Veb (emitter base voltage) is almost always going to be greater than 0.7V, and it will almost always be on, regardless of whether you attempt to put 5V or 0V at the base. To use the PNP to switch, you'd need to have a voltage close to V1 (21V).

I'd suggest using an NPN, a 2n3904 (the complement of the 3906) or the 2n4401 (higher switching current). Connect the emitter to ground, the base the same, and the collector to your LED. I would ALSO use a ballast resistor to control the current going to the 7-segment, instead of spit-balling it with the applied voltage, which will usually fluctuate (and as you've probably observed, a little bit of fluctuation in the forward voltage changes the current a fair bit). Unless you want characters that require more segments (e.g. 8) to be dimmer than those with less (e.g. 1), you should probably have a resistor going to each segment.
 
Your base of the transistor is always much lower than the emitter voltage in both the on and off states, just when your base voltage is in the "on" it is less lower than when its "off" so that the light is less bright, but its still never off because the base never is brought back to 21V to stop current flow.

You need your base pulled up to 21V for it to be off, and then simply sinking it to ground will turn on the transistor. You will either need to use an NPN or change your circuit a bit, and like the guy above said, you should always current limit your LEDs with a resistor unless you know the internal resistance of your voltage source.
 
Also if you do the KVL of the circuit, you will see that 21Volts minus the LED drop minus the emitter-base junction will give you around 19V, and so that is just telling you the transistor is on. When the transistor is off, the base should be atleast about 20.5V to 21V (ideally 21V).
 
Okay, so based on your previous posting, you're probably 14 or 15 and probably don't have a whole lot of experience / knowledge of circuit analysis.

I'm assuming that you have a microcontroller or microprocessor supplying the '5V' or 0V on the left--generally, this is represented with an open circle and a little note above it, reading 0 or 5V, not a constant 5V supply.

If you have a power supply, I'd recommend bumping the voltage up to 24V and using 150 ohm resistors to control the current. 1/4 W resistors will suffice.

As mentioned previously, use an NPN, either a 2n3904 or a 2n4401, whichever is available, and connect it as per the previous message (connect the resistor in the branch connected to the collector and not between the emitter and ground). I'd get rid of the resistor R1 going to ground, and use a 4.7 kohm resistor for R2. According to my calculations, that should ensure that the transistor turns on and fully saturates when you apply 5V.
 
Can we use software like B2Spice to see the input and output characteristics of the any particular component and also the input and output characteristics of the entire circuit? Is there any other better software which can help me do this?
I want to measure the voltage across the variable resistor which is connected in parallel at the output of the amplifier and then I want to measure the voltage across the input resistor which is connected in series at the input of the amplifier. Can I use this software?
 
MATLABdude said:
Okay, so based on your previous posting, you're probably 14 or 15 and probably don't have a whole lot of experience / knowledge of circuit analysis.

I'm assuming that you have a microcontroller or microprocessor supplying the '5V' or 0V on the left--generally, this is represented with an open circle and a little note above it, reading 0 or 5V, not a constant 5V supply.

If you have a power supply, I'd recommend bumping the voltage up to 24V and using 150 ohm resistors to control the current. 1/4 W resistors will suffice.

As mentioned previously, use an NPN, either a 2n3904 or a 2n4401, whichever is available, and connect it as per the previous message (connect the resistor in the branch connected to the collector and not between the emitter and ground). I'd get rid of the resistor R1 going to ground, and use a 4.7 kohm resistor for R2. According to my calculations, that should ensure that the transistor turns on and fully saturates when you apply 5V.

Well #1, you guessed right! I am 15, but the last post I made is a bad meter of my current skill level. I was younger and hadn't taken any electronics/physics courses yet. You are also correct in assuming that I don't have much circuit analysis experience.

#2, on the problem, I re-wired the circuit with 2n3904 transistors, the 1k resistor between the input signal and the transistor's base, the 10k pullup from the base to 5V (in retrospect, this should probably be a pulldown resistor, but I can change that later) and then the collector goes to the display through a 150 ohm resistor. I had always planned on using current limiting resistors, but I figured that for quick testing it wasn't a big deal, however they are in the circuit now. Also I obviously had to use a 7448 (common cathode/active high) so the driver works properly.

Anyways, it does work! Thanks to everyone who posted!
 

Similar threads

How Can I Use a Transistor to Drive a Relay with a Negative Output?
  • · Replies 2 ·
Replies
2
Views
3K