Advice please

N

nickpatts

Member
Hi All,

Background:

I have been playing around with a 8x8 Red/Green Led matrix (common anode) module, just driving the red leds at present, via two 74hc595 and 1 uln2803a, one 595 provides + for common anodes (rows) other is paired with the uln to sink to ground for the columns, using a 20x2 and using the hspiout command and getting great results running at set freq m64 I am not getting any flicker but due to the 1k resistors the leds are quite dim, at the moment I am playing it safe as although my sink end (uln2803a) can handle 500ma the 74hc595 i believe can only handle a 70ma max and with a potential situation of 8 leds being lit at once, I would like to include something extra to provide a higher current to these leds.

Advice:
I like using the 74hc595 chips and would like to still use these as i have them to hand as I have seen the higher rating shift registers, componant count is not an issue so is there something i can construct after the 74hc595 to be able to source a higher current, like how i have on the sink side of things.

Future:
Eventually once happy I will also add another 74hc595 and uln2308a for the green columns, then move onto a full rgb led matrix.

Image:
Image attached to give an idea of current layout.
*note. I have not included all the wired links etc just for a rought idea purposes.

Current Code:
just my test code for lighting a few of the leds animated between two sets with a for/loop just for delay.

Thank you for your time having a look over this.

Kind Regards

Nick
 

Attachments

hippy

hippy

Ex-Staff (retired)
This is the main problem with multiplex displays. It may be possible to overdrive the current through each LED to boost brightness using the theory that if it's only on one eighth of the time even eight times over current when averaged out is it's within limits. For the drivers though they may be sinking and sourcing max current all the time when the LED's are all on.

The best approach is probably to use high-side drivers for the anodes, similar to using the low-side drivers of the ULN2803 for the cathodes. I don't have any recommendations for doing that I'm afraid.
 
N

nickpatts

Member
Thank you Hippy,

Just to add looking at them TPIC6B595 high power shift registers may be an ideal solution or as there are only 8 rows control that directly via the 20x2 pins into a fet or something?( this would bring the code shifted out down and may help in future with more of the 595's daisy chained, this is really just a hobby so value your ideas and maybe see a few different views on how someone would control these 8x8 matrix modules themselves.

Nick
 
Last edited:
S

SgtWookie

Member
Well, the 74HC595 certainly is convenient as a latching shift register, but you're right - it has pretty meager current source/sink abilities. You're even going to have trouble using one to properly saturate a PNP transistor for sourcing current, unless a 50mA limit is OK.

How many of these LEDs do you want to light at one time, and what is your desired LED current?

Keep in mind that your red LEDs are quite likely to have a different Vf than the green LEDs. What are their specifications? I'm looking for typical Vf @ current.

Although the ULN2003A is rated for up to 500mA current, the practical limit is more like 350mA. Note that Vce increases a great deal when you get near 350mA.
 
N

nickpatts

Member
SgtWookie;130726]Well, the 74HC595 certainly is convenient as a latching shift register, but you're right - it has pretty meager current source/sink abilities. You're even going to have trouble using one to properly saturate a PNP transistor for sourcing current, unless a 50mA limit is OK
Click to expand...
Hi SgtWookie

spec from datasheet

red 1.6-2.0v 20ma typical 1.8v
green 2.2-2.5v 20ma typical 2.3v

max leds lit at any one time is 8.

as for desired current nothing set in stone really, something I would have to play with to really to evaluate brightness over current.

Just a hobby though so no real application as of yet, only controlling one of these modules at moment with the intention of adding the green leds next (currenlt only controlling red leds).

Nick
 
Last edited:
S

SgtWookie

Member
OK. The TPIC6B595 is an open-drain configuration; it can sink current, but cannot source it.

The ULN2803A is an open-collector configuration; it also can only sink current.

However, they DO sell source driver arrays that are roughly complementary to the ULN2803A.

UDN2981A, UDN2982A and UDN2987A are a few to look at.

Now your challenge is: which of those (if any) would be most suitable for your situation?

Find the datasheets for them, and read the specifications. Note packaging and the levels they were designed to work with. Also note what their Vce is over the range of Ic that they might have.

You'll learn a lot more by doing the research yourself instead of us doing it all for you.
 
N

nickpatts

Member
ok something to look into thank you, I'll have a good read up on them chips

If you refer to my original post I have not asked anyone to do anything for me, I was only seeking advice / past experiance, my current software and hardware is working fine be it limited in brightness.

Regards

Nick
 
W

westaust55

Moderator
nickpatts said:
Hi All,
Future:
Eventually once happy I will also add another 74hc595 and uln2308a for the green columns, then move onto a full rgb led matrix.
Click to expand...
Have a look at this thread where I have done an 8 x 8 RGB LED matrix
http://www.picaxeforum.co.uk/showthread.php?t=10836&page=4

Multiplexed and using some transistors for current ratings.
Series resistors take into account the differing forward volt drop of the different LED colours
 
You must log in or register to reply here.
Top