Very Puzzling?????

SamMishal

Hi All,
·
This is puzzling me and I cannot figure it out...there is something I am missing???
·
On the Propeller Demo Board (PDM) I can apply a voltage from 0 to 3.3 to the LEDs and they will light up
proportionally. So if I apply 0.1V the LED will be showing very faint light....if I apply 1.5 volts it shows
bright but not full brightness as it would when I apply 3.3 V.
·
This is exactly what I expect......
·
BUT.....on the Propeller Professional Development Board (PPDB) the LEDs do not show any light
until about 2.5V which shows as very dim....and then it is proportional dimness wise to the
voltage applied just as you would expect but all SHIFTED by 2.5 volts...so 2.7 volts is brighter and
3.0 volts is again brighter and full brightness at 3.3V
·
But....WHY????....does it only start at 2.5V and not at 0V ....
·
I am enclosing the Schematics for the two setups of the LEDs on both boards below....
·
This behavior does not make sense according to the electrical calcs.......the only thing I can
think of is that the LEDs on the PPDB require a minimum voltage drop of 2.5 to functions.
This is not like a normal LED.....
·
The LEDs on the PPDB are blue while the ones on the PDB are orange. But I do not see
much difference otherwise......why would the PPDB ones require such a big V drop before
they function????
·
Can someone shed a light on this????
·
·
Thanks in advance
·
Samuel
·
·
·

Rayman

LEDs are diodes and you have to forward bias them to draw current... The light intensity is proportional to current.
The bias voltage of LEDs varies with color... Blue LEDs typically have the highest bias voltage.

Current= (Vapplied-Vbias)/R

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

rokicki

All LEDs require a minimum forward voltage drop before they will pass current; this varies depending primarily
on the color (!) of the LED. 0.1V is not enough voltage to light any LED that I am aware of. Common values
include red (about 1.7V), yellow (2.1V), green (varies, usually about 2.5V), and blue (2.5V again?).

lonesock

rokicki said...
All LEDs require a minimum forward voltage drop before they will pass current; this varies depending primarily
on the color (!) of the LED. 0.1V is not enough voltage to light any LED that I am aware of. Common values
include red (about 1.7V), yellow (2.1V), green (varies, usually about 2.5V), and blue (2.5V again?).

"High-Brightness" (or "HB") LEDs have a forward voltage drop of around 3.25 volts.

Jonathan

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
lonesock
Piranha are people too.

SamMishal

Hi All,

Sorry I forgot to mention that the 0.1V is not actually REAL 0.1V rather it is a PWM signal
as comes out of the Duty Mode in a Counter......(not traditional PWM but quick pulsing
as explained in the Manual...Duty Mode DAC)

Now with the LEDs on the PDB I cna go down to almost 0.1V (DAC) while with the
PPDB I cannot get the LED to light up with any thing less than 2.5V (DAC).....

So what is the difference......Unless the Orange (yellow) LEDs on the PDB have capacitance
while the Blue ones on the PPDB do not have ANY capacitance.....Is that the case

Thanks for the information about the Blue LEDs requiring 2.5V drop to function...I did not
know that ...explains the 2.5V number.......

BUT......if the Yellow requires 1.7V then I should see the same behavior on the PDB and
the LEDs should not show any light till about 1.7V......but that is not the case.....they seem
to have CAPACITANCE and thus are able to maintain the voltage level better.....

Is it the case that Blue LEDs do not have much capacitance?????



Samuel

localroger

Sam, what frequency arey ou doing the PWM at? The diodes don't have much capacitance but the PDB is all surface mount in the wiring to the LED's, while the PPDB requires fairly long jumpers to connect the prop pin to the LED. If you're using a really high frequency that might be introducing enough capacitance to smooth out the voltage peaks.

SamMishal

Hi All,

I have just done a test on the PDB and I can set the DAC to 0.001 i.e. 0.0033 V and the LED
shows a VERY faint discernable light....so these LEDs are able to be activated
at a PWM voltage of down to 0.1% DAC output.

On the PPDB the LED shows the minimal·discernable light at 75% i.e.·2.475V..........

I cannot understand the behavior difference.....?????



Sam

SamMishal

Hi LocalRoger,

I am using the DUTY MODE of the CTRA so that would be ticking at 80_000_000 Hz
I am using a·16bit resolution and thus the FreqA is set to Duty*65_536

It is the PPDB that is not working until almost 75% DAC not the PDB.....so the
capacitance of the Bread Board might actually be STOPPING the smoothing
(does not make sense)....On the PDB the LEDS are wired directly while with the
the PPDB there is the Bread boarding aspect....I am going to try direct connection
and see????

....I just did.....no difference.....still needs 75% to activate

Very PUZZLING??????

Sam




Post Edited (SamMishal) : 10/27/2009 10:50:19 PM GMT

Lawson

This is really simple. On the demo board, the compact layout and surface mount parts is passing the DUTY mode counter output directly to the LED almost un-modified. So when the counter is set for 0.1 volts average, the LED sees the instantaneous voltage which has spikes high enough to light the LED. On the professional dev. board hooking the LED's up with jumper wires and a breadboard results in enough inductance and capacitance that the LED sees something like a one micro-second average of the DUTY mode output. That little bit of averaging is enough to keep the 12.5 nano-second pulses from the counter from lighting the LED.

Lawson

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?

localroger

Like Lawson said. I bet if you temporarily switch the clock to RCSLOW you'll see the PPDB LED's start to light up a bit at low duty cycles. An oscilloscope would also confirm this (I'd check, but my PPDB is at work and I'm at home now.)

SamMishal

Thanks Guys....

My thinking has been all wrong.....now I see....INDEED...the capacitance actually DOES in fact hinder the
the LEDs from seeing the quick 3.3V since the capacitance is smoothing out the voltage and thus the
LEDs only see the average smoothed voltage and since the PPDB LEDs need 2.5V drop then that is what they
will have to see from the smoothed out voltage.

But on the PDB the voltage is not smoothed out and thus the LEDs will have an average brightness.

So the PPDB LEDs never get to see the 3.3V, they only receive the averaged voltage due to the capacitance...

I was thinking the other way around.....I now understand....THANKS A LOT....

It makes sense.....now....

The PPDB capacitance has also caused me trouble when I was trying to get the Microphone Sigma-Delta ADC
to work....Ugh...

This has been a very interesting thing....and all this at 80MHz....imagine the effects even with a PCB when the
frequencies are in the Giga Hz....how do these Cell phones ever manage to work???

·

Regards

Samuel

Phil Pilgrim (PhiPi)

You can always use one of the PWM objects in the OBEX instead. That way you have control over the frequency.

-Phil

localroger

Sam, one of the great things about the Propeller is that it gives us 80 MHz core speeds without having to run signals of such high frequency significant distances to memory chips and peripherals. I remember back around 1980 what a PITA it was to get a circuit to stop ringing at 4 MHz. We take these high frequencies for granted nowadays but a lot of engineering goes into making them work right, such as sandwiching the signal traces between ground planes in multi-layer circuit boards and careful separation and route length control. The cellphone or PC mobo manufacturer who will sell hundreds of thousands of units can afford to get all those details right but for us hobbyists it's a major pain, and the Prop gives us great speed while keeping the really hard signals on the chip unless we really want to let them out.

SamMishal

Phil Pilgrim (PhiPi) said...
You can always use one of the PWM objects in the OBEX instead. That way you have control over the frequency.

@Phil,

Yes but the DUTY mode has advantages over the NCO PWM method
1- You can set and forget it while with the PWM mode you need to reset the Phs register periodically
2- It is easier to do ....just set the Frq register to a multiple of a scale depending on the resolution
3- The code required to update the Duty level is MUCH less thus giving SPIN code a lot more bandwidth to do
··· more time constrained actions.
4- When updating the Duty level @80MHz you get an almost instantaneous change, while with the PWM
··· method you have to wait until the PWM PERIOD finishes to update the level

Actually the Duty mode is perfect for what I am trying to do....the trouble is that I was trying to use the
LEDs as a quick and dirty way of verifying that the DAC is working while developing the code....too lazy to
go and get my Voltmeter and connect it·etc..... I guess I need to use my DMM more often....



@Localroger,

I did try the PPDB with the RCSlow speed and it WORKS....just like you suggested...it IS the high speed that
causes long leads to manifest capacitance beyond what you might expect.....

localroger said...
the Prop gives us great speed while keeping the really hard signals on the chip unless we really want to let them out.

ABSOLUTELY.....imagine if we had to wire up 8 μCs with shared RAM and at 80MHz.......endless trouble
from all aspects.....if it can be done at all....

....The Propeller is GREAT....no doubt about it whatsoever.....

Thanks all for the help

Samuel


Post Edited (SamMishal) : 10/28/2009 3:34:05 AM GMT