Hello guys I just have a curious question.

say I have 2 LEDs,

:each running with a 330 ohm resitor in series
:each running off the same voltage (5volts in this case)
:both running under a PWM signal (25% in this case)

The only differrence is that the operating frequency for one LED is that it is running at 10,000 hertz and the other is running at 500 hertz.

now for the question: Which one will appear brighter and why???


p.s. I'll try and post the circuit asap, but this is the question simplified in itself.sorry guys

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

ok finally drawn up the circuit diagram on MS Paint. The reason I ask the above question is that I have run out of input and output pins on this microcontroller (16F628). To solve this issue I have marked the needed changes to the circuit in red (the 2 diodes and piezo speaker).

Of course human hearing cant hear above 20khz right?? So if we boost the refresh rate of the LEDs blinking to 10Khz on the speaker we get 20khz. Therefore nothing is audible to the human ear at this rate.

But when we want the piezo to make an audible tone we reduce the refresh rate to 500 hertz and therefore produces 1000hertz at the piezo speaker, well within the human ear range.

But would this change in refresh rate alter the overall brigthtness of the LEDs???

P.S. That RF transmitter is used to activate some other auxillary things to wake me up and transmit the time to other clocks so every clock is sychronised in the household.

Attached Images

dig-clock.gif (10.7 KB, 1168 views)

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

It 'should' make no difference, the overall power level is still the same - it depends really how fast the LED's will respond. I would suggest trying it, and seeing what difference it makes (if any).

Have you developed a suitable transmission format yet?.

I've just started playing with a couple of radio modules, and scoping them shows that at 9600 baud with standard (inverted polarity) RS232 signals the output from the receiver isn't identical to that transmitted. Just like with IR remote control the width of the received pulses is different to the transmitted ones!.

This is obviously the reason for the common use of Manchester coding, which relies on the change from 0 to 1, and 1 to 0, rather than the actual width of the pulses.

Something to bear in mind!.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Hi Pike,
Just try it. Young people and dogs won't like to hear 20KHz.
Your vision doesn't care about refresh rates over about 40Hz.
The piezo has a fairly high capacitance that might slow-down the turn-on time of the transistors, dimming the display at 20KHz. If this happens, use a separate driver transistor for the piezo. Then you could use a low-pass-filter to feed the new transistor.

 

hmm, say it does vary the brightness what can i do about it if that happens.
So far i've only come up with lowering the 25% PWM to lower the brighter frequncy down so they are unnoticable.

Would the above technique work or am i just being too optimistic???
Seeing that I am running the leds off at 25% (maybe lower) does this mean i have to decrease the series current limiting resistor by 4 times???

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

Try it and see! - no point worrying about something which probably won't happen!.

If you do see a noticable difference, simply adjust the PWM ratio to make it the same.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

hahah it worked!!!

The frequency was a bit off at 18khz but who cares atleast i know it will work!!!

Nigel: I'm using a 32bit format which allows time synchronised accuracy down to 1-sec. 1 byte for alarm, 1 for hours, one for minutes and one for secs.

Would i still need to reduce those series current limiting resistors by 4???

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

Hi Pike,
Doesn't the 18KHz bother you? It would drive me and my dog crazy!

If you reduce the LED current-limiting resistors by 4 times, the display would look just as bright as if the 25% PWM was 100%.

 

Depends how bright you want the displays?, you keep refering to PWM, but presumably you mean the multiplexing of the four displays?.

To keep the four displays the same brightness as one non-multiplexed display you would need to increase the current by four.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

audioguru is right... Human eye persistence is just about 1/24 of a second. any changes faster would be unresolvable...

And if you are so very intrested in a uniform brightness LED, I suggest you rectfy your signal and put a shunt capacitor or a pi filter before feeding the signals to your LEDs...Now only the amplitude of your signals shall matter...

__________________
Bharath Bhushan Lohray.
M.Sc. Electronics.

 

Hi Bharath,
You can't "pulse-stretch" the digits of a multiplexed display. If you did, all digits would be the same, a jumble of combined segments.

 

Yeh i know but it was just a test run to prove the principal of frequency- duty cycle output. Same power like nigel said both ways. Anyways my dog is either deaf stupid or just ignorant :?

BTW audio your last post is in the wrong thread :lol:

EDIT: thank you all, particularlly audioguru and Nigel you guys deeserve a medal for your contribution on these forums.

__________________
www.winpicprog.co.uk - Great PIC language tutorials.

 

Hi Pike,
We were talking about your loss of 75% of the brightness caused by the 25% duty-cycle of your display's multiplexing between its 4 digits.

I don't need a medal, I'm always glad to help. Been there, done that!

 

I thought it was a single led that everyone was talking about... Ofcourse, 7segment displays cant be done so without total loss...

Oh bother, I did not see the circuit diagramme...:shock:

Sorry everyone...

But I think uniform brightness may be restored by the use of a monostable multivibrator...(74121) Trigger your Leds by the pulse and let them glow for a definite period of time. keep the timeing RC network to compliment the frequency.

But again it is too much of a trouble...

__________________
Bharath Bhushan Lohray.
M.Sc. Electronics.

 

Hi Bharath,
Too much trouble? You're not kidding! You would need 28 monostable circuits for a 4-digit non-multiplexed display.
Multiplexed displays have their similar segments of all digits joined together so pulse-stretching won't work anyway.
Besides, the display does have uniform brightness because even though the refresh frequency changes, the duty-cycle (25%) doesn't change.