I am attempting to provide a low cost and adjustable night/day mode for a set of panel LEDs.
I have I/O outputs free but only one PWM.
Microcontroller I/O will be used to switch each LED on/off.
The LED On/Off control signals will be fed through AND gates with the second input taken from a PWM output.
In this way the output level of all the LEDs may be controlled (to a single level) using a single PWM.
Will the current capability of the AND gates be similar enough so as to give uniform LED intensity?

 

hi Pete,
Welcome to AAC.
What type of AND gate do you plan to use.?
eg: TTL, CMOS etc.
E

 

CMOS I would imagine as we will have lots of AND gates driven by the PWM. Not fussy though, whatever works!

 

hi,
Do you have a specification for the panel LED's ie: voltage, current, colour.?
E

 

Apologies. Thanks for checking all the factors.
Up to 20 or 25 mA full brightness I would say. Its intended for a bright reflective/dazzle environment (marine). We are free to choice voltage but 3.3V and 5V available. Colour will be white or company colour of 470nm (off blue).
I have driven RGB LEDs using an AP3156 and 3 mA seems quite bright in the office.
The mechanics have not been decided though and guess heavy frosting to give uniform button lighting. So more current may be best.

 

hi,
Look at this LED colour/voltage chart.
White LED's will be marginal with a 3.3V logic, on load.
E
Also Blue....

 

Thanks, Who says you cannot teach an old software engineer new tricks
So 5V CMOS then.
The micro is 3.3V so then I need level shifters and cost....
Sounds like this idea does not fly in white and also seems tight at 470nm (3.2V).

Would the current/intensity have been uniform if the forward voltages had worked out (i.e. each AND drives the same)?

 

hi,
If you are asking about apparent uniformity of White versus Blue, I would say the Blue would not appear as bright, taking this plot as a guide.

E

 

hi,
Using a AND gate is so that you have PWM on one pin and a 'On/Off' control on the second pin could be realises by using a transistor + components.
Eliminates the need for additional level shifting.

E
EDIT:
A draft sketch posting of the project, showing the numbers of each type of LED and wiring pattern would helpful.

 

One possibility, PWM for each R, G, B channel. I got lazy and only wired up Red.
Part can handle the rest of the design. I picked 8 of each color as an example.
Other parts can handle more I/O, more total LEDs.

So you would have control, on/off each LED, and all Reds on one dedicated PWM,
same for G & B.

Lastly would have to look at allowed total rail currents, to not exceed them. That is
in datasheet.



Cypress PSOC 4 design. If you need A/D, monitor supply, temp blah blah blah part can also do that.

This is a single chip solution unless external power drivers needed to handle
LEDs, depends on how many you are going to use and their current ratings.

Regards, Dana.

 

Thanks.
In summary then:
The AND gates will each provide the same current and so we have uniform brightness (when using same colour) controlled with a single PWM channel.
I will need to use CMOS ANDs because of the current needed on each AND input.
I will need 5V logic if using white LEDs.
Thanks again, I learnt a lot!

 

I will need to use CMOS ANDs because of the current needed on each AND input.

CMOS AND's (or for that matter CMOS logic in general) is a voltage input, with very
small currents needed to charge/discharge the input gate C (typically 2 - 5 pF).

Regards, Dana.

 

hi Pete,
When using 3.3V White/Blue LED from a 5V supply, you should have a series resistor in order to limit the current thru the LED.
Are the White and Blue separate LED's or are they an RGB type.?
E

 

Thanks,
The white and blue are presently made up using separate LEDs (ASMB-MTB1-0A3A2) - we were playing with colour schemes.
Our prototype drives 3.3V. The datasheet gives forward in range 2.8V to 3.5V.
Does that mean:
a) acceptable driving range i.e. that 2.8V will always switch on but don't go over 3.5V or it will fry
or does that mean
b) you may need 3.5V to switch on?
I would assume a) as otherwise there is no stated maximum voltage but why then 'typical' value?

 

hi,
This spread in working forward voltage occurs in all general purpose LED's.
Some users buy in quantity and select groups/batches that have the same Vfwd.

LED's will not self limit the current they can draw from a higher than a Vfwd supply, so a series resistor is required.

E

 

Thanks. Yes I will have a current limiting resistor. We used a charge pump before but are switching to PWM so we will need the resistor.

I am still not quite clear on how to read the datasheet. Is your reading that some of our LEDs may not switch on until they have 3.5V applied?

 

The LED may start conduction at just below the specified Vfwd, but the light output will be very low.
Perhaps this graph will show it better.
E

 

hi Pete,
This plot is for typical White LED's.
E

 

Great. So if I want more current then I am likely to need a higher voltage.
The datasheet is telling me that I may need 3.6V to achieve 20mA.
Looking at the plot though I am likely to achieve 5mA or better at <3.3V however.
It just depends on the LED.

 

hi,
My next step would be to get a small quantity of White and Blue LED's and do some empirical test on that batch.
The problem is that light intensity is a subjective effect, especially in different ambient lighting conditions and the transmission quality of the planned frosting diffuser.
One way to match the White/Blue apparent light intensity would be by choosing a different resistor value for the White and Blue LED's
[thats is setting the operating current with a series resistor]

E