• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Fast switching PNP transistor request

Status
Not open for further replies.

James77

New Member
Hi all,

Please would someone give some guidance as to which PNP transistor I should be looking for when switching PWM at 4MHz, ON signal is LOW (0V)

All the transistors I have found so far have rise times of 300 + ns, I need around the +/- 60 ns mark.

Background: I am going to drive a series of 3w LED's using a TLC5947, via a PNP transistor then an NPN MOSFET, then the LED.

Thanks
 

Hero999

Banned
At what collector current?

I imagine any RF transistor will work, try BFT92 assuming this is just a low current gate driver.
 

crutschow

Well-Known Member
Most Helpful Member
you won't notice any switch at such a high frequency
If you are referring to visual flicker, the eye can not see any flicker above a couple hundred Hz.
 

whiz115

Member
If you are referring to visual flicker, the eye can not see any flicker above a couple hundred Hz.


of course not.. that's what i'm saying...

I don't know what's the point switching on and off at such speed a led that emits visible light... i think even on data transmission with IR the switching is usually no more than 30-50 KHz
 

Hero999

Banned
LEDs have a limited turn-on/off time anyway.
 

crutschow

Well-Known Member
Most Helpful Member
LEDs have a limited turn-on/off time anyway.
What limit are you referring to? For frequencies up to a few MHz the switching time of an LED is mostly determined by the drive circuit, not the LED. The primary limitation is carrier lifetime which is usually less than 100ns for standard LEDs.
 

whiz115

Member
What limit are you referring to? For frequencies up to a few MHz the switching time of an LED is mostly determined by the drive circuit, not the LED. The primary limitation is carrier lifetime which is usually less than 100ns for standard LEDs.
5 MHz is not little... even if it is possible.. still i don't think leds emiting vissible light to have the apppropriate design for such fast switching..
 

Hero999

Banned
What limit are you referring to? For frequencies up to a few MHz the switching time of an LED is mostly determined by the drive circuit, not the LED. The primary limitation is carrier lifetime which is usually less than 100ns for standard LEDs.
It depends on the LED, if it's a white LED it'll be slower because the phosphor takes time to stop emitting light.
 

James77

New Member
Thanks all for your input...

The reason for the 4MHz switching is the Texas Instruments TCL5947. This chip is a 12bit 24 channel, pwm LED driver that has an internal OSC at 4MHz. The chip's specs are a 100ns rise time and a 60ns fall.

Posing a potentially stupid question; How would one drive an LED using this chip otherwise? The TLC5947 output's LOW when on which means I need an PNP transistor connected to an NPN MOSFET right...? Or, is there a better way?

Thanks again
 

fernando_g

New Member
The white LEDs do indeed have a turnoff time, as HERO999 has correctly stated, the phospors have persistency.
 

Hero999

Banned
What's more the phosphor is yellow so the apparent colour will probably change at high frequencies, there a again the phosphor might take time to emit light too so this could cancel the persistence.
 

fernando_g

New Member
This is also very true, the color index for white LEDs is only guaranteed for certain conditions.
 

Space Varmint

New Member
Transistors switch faster in emitter follower configuration. I have placed an LED in the emitter lead (no resistor) and collector straight to vcc use voltage divider bias on the base to control the current level. Try 10K to start with. I achieved very satisfactory results.
 

James77

New Member
Hey all,

Below are the rise and fall times for the outputs.

SCLK↓ to SOUT 15 ns
BLANK↑ to OUT0 sink current off 20 ns
Propagation delay time OUT0 current on to OUT1/5/9/13/17/21 current on 24 ns
OUT0 current on to OUT2/6/10/14/18/22 current on 48 ns
OUT0 current on to OUT3/7/11/15/19/23 current on 72 ns

So I still am in need of a fast switching transistor and I not? Or, can I use a transistor that does not switch as fast as the TLC5947?
 

Nigel Goodwin

Super Moderator
Most Helpful Member
What FREQUENCY is the PWM output?, you're really looking at the wrong spec.

You might try posting the link to the datasheet, that number isn't found on TI's website.
 

fernando_g

New Member
As Nigel mentions, you are looking at the wrong spec.

From reading the datasheet the 4Mhz is ONLY for the internal clock, which is used to drive a 12 bit PWM output timing. Thus, the actual output is more like 1000 Hz which is 4 Mhz divided by 2^12
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top