led MATRIX CURRENT DRAW

[JonSea]

Too late. I see a smoke cloud back behind me where the train has wrecked and all passengers are tragically dead.

You're thinking off the wall but you don't understand the straight forward principles. You're talking about using constant current excitation for uniform brightness but the reason you didn't have uniform brightness is only because you screwed up the 2 component system of one resistor and one LED. If you can't get that right, a more complex (and totally unneeded) solution won't help.

 

[MrDEB]

My test measurments I used one resistor and one LED segment. Are you now suggesting I go back to my orginal plan and schematic BUT use a higher voltage thus a higher current available for the LED segments? For uniform brightness I would really need to use an LED driver? but perhaps driving the anodes with a transistor and the cathodes with a ULN2803?
While performing my test measurments I realize that the 5v to feed the anodes is OUT.

 

[JonSea]

No.

What I am suggesting and recommended is individual LEDs arranged in the pattern I showed above. The right arrow head might have 7 LEDs, each connected by its own resistor to +5v and all 7 connected to ground via a single ULN2003 gate. Same for the left arrow head, and the same for the center bar with many 5 LEDs. 3 port pins control the three elements.

Each LED has a forward voltage around 2.2 volts, and the resistors calculated to provide 25 mA for the LED. 7 LEDs will draw about 175 mA, well within the capability of the ULN2003.

Thie LEDs run at maximum brigtness, circuit control is simple and doesn't require the complexities of multiplexing and it's something you might actually be able to get working. The is no reason that LEDs mounted to a circuit board should be in any way fragile and there are LED options that jare blindly bright.

I specifically recommend against the Kingbright displays you are using and multiplexing. Supplying he required voltage and multiplexing are complications that likely be insurmountable.

No further comments.

 

[MrDEB]

Going with transistors

on the 7 rows (PNP) and the ULN2803's on the 5 colums. I wired up three PNP transistors yesterday and the display looks really nice at 18ma per segment.
In the mean time I discovered a Constant Current Sink device. Looks like it might be the answer to uniform brightness and battery usage drain. There is an 8 channel
version as well.

https://www.electro-tech-online.com/custompdfs/2012/10/CAT4016-DPDF.pdf

http://www.alldatasheet.com/datasheet-pdf/pdf/219592/STMICROELECTRONICS/STP16DP05.html

http://picprojects.org.uk/projects/lc/index.htm

 

[be80be]

MrDeb maybe you'll read this the DS18b20 you sent me i figure black gnd red vdd and yellow data is that right or could you post me a link where it came from

 

[MrDEB]

got your PM but got side tracked
here is a link where I purchased
**broken link removed**

 

[JonSea]

Well, at least MrDEB got Burt and I to agree on something. That's somewhat amazing!

 

[MrDEB]

averting diaster!!!poof

After getting the PIC connected w/ 3 PNP transistors I find the transistors won't shut off?? Not surprising since the base voltage shows 7v (using 8v to power the LEDs)
Looking at mosfets or ?? but needs to be available in a SMD package.
Looking for fix, I find using transistors for switching higher voltages than the PIC output is not wise.
back to the web tomorrow.

 

[be80be]

MrDeb is this for real or what I wounder why we even need transistors for.


What you using a power transistor

Transistor switches are used for a wide variety of applications such as interfacing large current or high voltage devices like motors, relays or lamps to low voltage digital logic IC's or gates like AND gates or OR gates. Here, the output from a digital logic gate is only +5v but the device to be controlled may require a 12 or even 24 volts supply. Or the load such as a DC Motor may need to have its speed controlled using a series of pulses (Pulse Width Modulation). transistor switches will allow us to do this faster and more easily than with conventional mechanical switches.

AND a Picture to go with all that
View attachment 67916

Looked at the data sheet for led-matrix you said your using and there is no reason it doesn't
work with a 6 volt pack. It shows testing with one dot being led on at a time at 20mA and there not saying it takes voltage higher then 5 volts more like 3.6.

Post a picture of this thing hooked up as your testing it.

 

[Mike - K8LH]

... I find using transistors for switching higher voltages than the PIC output is not wise.

You can't use a single PNP transistor or P channel FET driven by a five volt signal as a high side sourcing driver to switch a higher voltage. You need an NPN and PNP pair, or an N-FET and P-FET pair, when used as a high side source driver for switching a higher voltage.

**broken link removed**

 

[be80be]

Mike transistor are used every day to interface uC to higher voltage loads. And from the looks of thing telling how to do so is outside the scope of this forum.


He lighting a Led matrix wanted to light five at a time that takes almost 20 to 25 volts if he is right about the readings he said he took by lighting one dot.

Now if he does what he been saying I'm sure that the he will blow a led.

Now there is no reason he can't use a handfull of NPN or the UIN2803 and a high side driver or make one using some npn

The problem is in knowing how to put the parts together which could be as simple as this View attachment led_matrix.gif

Or as hard as this

 

[MrDEB]

am LOW side switching 7v and the PNP will not shut off (base goes HIGH). WHY you ask, this link states the issue, the 7v is keeping the base high.
http://www.w9xt.com/page_microdesign_pt8_pnp_switching.html
looking at an example Jon used http://digital-diy.com/forum/genera...help-yes-mosfet-most-likely-t2328.html#p14039.
Don't want to get to complicated with multiple transistors. The LED segments won't light up very bright until I hit them with a min of 18ma but with the ULN2803 on the cathodes and a Vf of 4 its kinda iffy with 5volts.
Going to do more research etc. but the mosfet may be the ticket.

 

[JonSea]

I really hate to feed MrDEB's fantasy circuit design but.....

Read what Mike said carefully. Controlling a PNP transistor or a P channel FET requires a control voltage equal to the supply voltage. This is reality of the situation.

You might look at a TD62783APG high side driver. It's the high side equivalent of the ULN2803.

MrDEB you really should read and heed the advice given here.

 

[be80be]

Medeb there no way to tell what your doing you can pick parts that would
Work if you read the page you linked it tell how to make a high side driver
and it would be nice to no what part your really using

 

[Mike - K8LH]

I agree it's difficult to figure out exactly what you're trying to do. If I'm interpreting post #72 correctly it seems you're trying to use a PNP transistor as a low side "sinking" driver to switch a higher voltage load. Yikes! If that's the case, you really need to take a closer look at the tutorial you linked to. It shows how to use an NPN or N-FET as a sinking driver (part 7) and how to use an NPN & PNP pair or an N-FET & P-FET pair as a source driver (part 12) in order to switch higher voltage loads.

I should mention that, in addition to the Toshiba TD62783APG 8-bit parallel 'source' driver IC member JonSea mentioned, the Micrel MIC5891 8-bit serial-to-parallel 'source' driver IC is particularly well suited for driving the higher voltage multi-LED-per-segment displays which are becoming popular. The MIC5891 has separate pins for the five volt logic supply (VDD) and the high voltage display supply (VBB). Of course you can always tie VDD and VBB pins together for standard displays that run ok on five volts. Please let me know if you'd like to see an example design (six pin interface, one 5x7 anode-row/cathode-column matrix display, one MIC5891, five N-FET sinking column drivers, 20% duty cycle, full brightness 100-ma "peak" / 20-ma "average" current per LED, and full fade-to-black PWM brightness control).

 

[be80be]

He could use a NPN as emitter follower which is used in most high side drivers to let you use low power to turn on the gate but the part count will go way up it takes 3 transistors and a handful of resistor to make a good high side driver. I would use a UDN2982 http://www.datasheetcatalog.org/datasheet/allegromicrosystems/2981.pdf

 

[MrDEB]

The UDN2982 looks like it should do the deed. It appears to be the direct opposite of the ULN2803. I am going to look at Jons suggestion of the TD62783APG.
And yes I am trying to follow ALL the advice given. I realize why the PNP in a LOW side driver failed. The tutorial I linked to was found after I attempted to try the configuration out. I found the base staying high as the tutorial stated it would. Lesson learned.

 

[MrDEB]

The two parts seem very similar. I think we have a winner!!
this would cut down the parts count considerably.
For clarity I will draw up a schematic of why the PNP failed to work BUT curious (a mute point now) but would the mosfet IRLD014PBF work as well but it increases the part count vers the TD62783APG or the UDN2982.

 

[JonSea]

This is truly a waste of words, but....

The MOSFETByou listed is an N-channel device. It could take the place of the ULN2003.

If you meant to refer to a P-channel MOSFET, as Mike clearly explained, it can't be directly controlled with a 5 volt micro pin. Please show Mike some respect and read what he wrote.

 

[MrDEB]

I read what Mike posted but was curious since you used an IRLD014PBF in the same basic setup.
I plan to go with the high side driver TD62783APG. Did some research and fond this https://www.electro-tech-online.com/custompdfs/2012/10/LEDs_Tutorial_Rev_3.pdf
It explains what I am basically doing.
Now to figure between Digi-key or Mouser (shipping issues) Jamco or Allied don't carry.