Go Down

Topic: Boosting Logic Voltage (LED Strips on Teensy) (Read 7693 times) previous topic - next topic

Grumpy_Mike

It seems that these chips have internal pull-up resistors.
No.

How would that boost the logic voltage?
You would need a pull up resistor on the drain. The signal would then be inverted.

Qdeathstar

Well. A lot to consider.

I worked on a schematic based on what we had talked about with the 74LC14... I think I have it right, but i'm not sure about the resistor values or capacitor values, what type of capacitor, ect....

I tried looking at this... but it is a different chip..

http://www.adafruit.com/product/1779


It looks like a ceramic capacitor and from what I can read, 100ohm resistors... why 100 ohm?

Here is the schematic I have so far:

A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Grumpy_Mike

The 74HCT245 is an octal non inverting buffer, so you would only need two out of the eight buffers in the chip. It can also be a switched bi directional buffer but you don't need that. It is a more modern chip but otherwise offers no particular advantage.

Connect the unused inputs of the 74LS14 to ground for stability.


Quote
It looks like a ceramic capacitor
What does? It is normal to use a ceramic capacitor for decoupling and you should put one across the power pins of the 74LS14.

Quote
why 100 ohm
It is not too critical. It is there to soak up reflected signals and damp down any standing wave that could build up. Those two things are actually the same thing.

Qdeathstar

#18
Jul 13, 2015, 06:17 am Last Edit: Jul 13, 2015, 06:18 am by Qdeathstar
The 74HCT245 is an octal non inverting buffer, so you would only need two out of the eight buffers in the chip. It can also be a switched bi directional buffer but you don't need that. It is a more modern chip but otherwise offers no particular advantage.

Connect the unused inputs of the 74LS14 to ground for stability.
Ok

Quote
What does? It is normal to use a ceramic capacitor for decoupling and you should put one across the power pins of the 74LS14.
In the picture of the octoWS2811 I was using as a reference, there is a ceramic capacitor there tooo, I just wanted to make sure I should a ceramic one. Should I use a 100uf capacitor?
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Grumpy_Mike

Quote
Should I use a 100uf capacitor?
Well you can't buy a 100uF ceramic capacitor.
It needs to be 0.1uF or 100nF ( they are the same thing )

Qdeathstar

Can the decoupling capacitors be connected to the ground plane? I can't see a way to route them like this:

https://cdn.sparkfun.com/assets/2/6/7/7/e/51965e31ce395f412a000000.png

(just an example)

I can route the power, but the grounds just connect to my ground plane... Is that oK?
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Wawa

It seems that these chips have internal pull-up resistors.
No.
Why would I say that without reading the datasheet first...
https://frack.nl/w/images/3/33/LPD8806_datasheet.pdf
Leo..

Qdeathstar

#22
Jul 14, 2015, 06:19 am Last Edit: Jul 14, 2015, 06:21 am by Qdeathstar
I am using a 1.5amp voltage regulator to step down from 12v to 5v... I think that my current draw is at most around 100ma (35ma for the teensy, 6ma for the Heterdyne 433mhz rx, and I think I read a max of 20ma for the 74ls14? (Icch?). So that adds up to 61ma, so 100ma should be plently..

On adafruit, it says that the voltage regulator can drop 2w without a heat sink.  I found this formula for wattage drop:

(InputVoltage - 5V) * AverageCurrentInAmps  (12 - 5) * . 1 = .7

So I shouldn't need a heat sink.



Please take a look at my schematic and let me know what you think...
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Wawa

I'm lost.
Why termination resistors. What is the clock speed.
Are you planning to go twice around the block with Cat-5.
Will there also be 150ohm termination resistors at the end of the cable.
And an amplifier, because your signal would have dropped to 2volt.
Are you planning to drive one strip, multiple strips, series, in a star.
??
Leo..

Qdeathstar

#24
Jul 14, 2015, 01:42 pm Last Edit: Jul 14, 2015, 01:48 pm by Qdeathstar
I'm lost.
Why termination resistors. What is the clock speed.
Are you planning to go twice around the block with Cat-5.
Will there also be 150ohm termination resistors at the end of the cable.
And an amplifier, because your signal would have dropped to 2volt.
Are you planning to drive one strip, multiple strips, series, in a star.
??
Leo..
uhh..,  this is the third board I have designed and the first two where just moving Arduino pins around.. I only know what I interpreted from Google searches and these forums...

I'm not sure what you mean by termination resistors, but I assume you mean the ones for data and clock. I thought grump said I needd the for the 74s14. Clock speed, I don't know.. I think I read that lpd806 is 400hz, but I won't be updating them more than 10 times a second....

I plan to have <1ft leads between the screw termina and the strip.

should there be 150 ohm termination resistors at the end of the strip? Im not sure, but I don't think I need an amp.. I think the 6808chips boost the signal.. in testing I was able to connect three 5 meter strips together no problem straight from an arduino... 

I was hoping to connect 6~8 strips together in series... around 40 meters max
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Wawa

Termination resistors are used when cable reflections are expected.
High switch frequencies and long cables.

If your teensy is 30cm away from the first IC, there is no need for that.

And the LED chips are not boosting the signal, they are buffering the signal.
You only drive the first IC. That IC buffers the data, and passes it on to the next one.

Try two small mosfets or two NPN transistors with 1k base resistors.
Takes 10 minutes to hook that up on a breadboard.
Make sure you reverse your output pin code.
Leo..

Qdeathstar

Are you using the mosfets instead of the 74ls14 to get from 3.3v to 5v or in addition to it?
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Wawa

Connect a 5volt supply to the LED strip. Nothing else.
Measure the clock and data voltages in respect to LED strip ground.
You might find 5volt on the pins, because the LED IC is already pulling those pins up to it's own supply.

If so, all you have to do is switch those lines to ground, with a transistor.

Mesure that first, and report back.
Leo..

Qdeathstar

The strips I'm using are 12v.


I measured from ground to data in and ground to clock in with just the 12v power supply connected and got 0 volts dc.

I verified 12v between ground and positive.
A creaking creeping shadow
stiff against the freezing fog
glares at a tickless watch.

Time has failed him -- all things shall pass.

Wawa

So you're NOT using the LED strips with the LPD8806 drive chip, mentioned in previous posts.

That chip has a maximum rating of 5.5volt, and would blow up on 12 volts.

??


Go Up