Driving neopixels using the glowbit library on a pico

Hi Murray
That should do it. It seems to have OP Amps built-in and apparently supply 8mA which should be plenty. Core description mentions driving NeoPixels with 3.3V logic. This may be one of the ones I spotted earlier.

I would still do some measurements with the current converter and take notes. That information could be valuable in the future.

I have a LMC6484 on hand for a current project so for my own info I am going to do a check on frequency capability when wired as a voltage follower.
Let us know how it works out. I expect others are having similar problems.
Cheers Bob

3 Likes

HI Bob,

I also found a note on a different level shifter:

“The 10k pullups do make the interface a little more sluggish than using a TXB0108 or 74LVC245 so we suggest checking those out if you need high speed transfer”

Murray

3 Likes

Hi Murray

Yes, as the source impedance is relatively high (10kΩ) and it has whatever the target device has in the way of input capacitance time constants come into play increasing the rise time so square waves start to resemble sine waves when the frequency gets high enough. That is a simple way to put it.
Cheers Bob

3 Likes

Hi Murray and all,

I redid the circuit on a breadboard and wasn’t getting any fluttering of the colours like I was with the FET driver Logic level converter.

I’m keen to see how you go as well!

2 Likes

Guess what I have just ordered :slightly_smiling_face: along with a few other bits.

Based on the testing I did with Bob’s encouragement, and a scope, I came to the opinion that a main part of the issue was the level shifter … so as soon as the bits arrive and the re-re-re-build is done I will report back.

3 Likes

Been away in Adelaide but I have found an interesting book:
“Custom Raspberry Pi Interfaces - Design and build hardware interfaces for the Raspberry Pi” by Warren Gay ISBN-13 978-1-4842-2405-2 (paperback) ISBN-13 978-1-4842-2406-9 (electronic)

Chapter 2 is all about 5v-3v and 3v-5v interfacing, and he makes specific reference to the speed of the interfaces. He talks about the 74HCT family of chips for this purpose (specifically the 74HCT255, it is an 8 channel device and I think it is the big brother to the 74HCT125 4 channel device available from Core under SKU:ADA1787)

I am expecting the 74HCT125 to arrive on Monday next for me to trial…

2 Likes

Hi Murray
For your application you only need to go in one direction which may make your solution that much easier.
Cheers Bob

2 Likes

Hi all,

Now I am REALLY confused. I have received the 74HCT125 chips, and put 1 into my testboard. I then ran the test Glowbit script.

If I connect the stick to 5v, and drive the data in line directly from the Pico - works properly.

If I connect the stick to 5v, and drive the data in line through the level shifter, I appear to get an ‘off by one’ error in the display, and I get flickers!

See the script and videos below.

https://drive.google.com/drive/folders/1Blv0ycdG1DXr9zcNaKohSVYwqZMITy0I?usp=sharing

Haven’t tried the neopixels or 3 watt leds yet - kind of annoyed actually

2 Likes

Hi Murray,

Would it be possible to grab a still image/a diagram of how you have everything connected to the logic shifter?
(For my 74AHCT125 - Quad Level-Shifter (3V to 5V) - 74AHCT125 | Adafruit ADA1787 | Core Electronics Australia)

Here’s how I had everything connected on my setup

Also which logic shifting IC did you grab? Just to be sure what’s measured on the V pins of the GlowBits and LLC?

EDIT: Swapped out the pinout diagram after Bob mentioned that it was incorrect

1 Like

Hi Liam
I think this may be a different IC to what Murray has. I looked up the data sheet for 74HCT125 and I get the following pin out
image

With the following internal block diagram

There seems to be some confusion here.
Cheers Bob

2 Likes

Hi Liam and Bob
Bob is correct … herewith the device and my connections - oops covered the label for the enable - should be 1OE* (* instead of the overbar)

HCT175

2 Likes

Hi Murray,

If you trace out your GND’s are they all linked at some point?, it looks like it may OC somewhere.
I used the GND pin of the 74HCT125 as the references (5V supply and a diode connected to the Vsys pin of the Pico) GND then went to the (74HCT125, Pico, 5V supply and LED strip)

1 Like

Hi Liam
Not very careful are you. That is not the same IC you had here a few minutes ago. Powerful tool that editor.
This is exactly how confusion and mistakes occur.
Cheers Bob

2 Likes

Hi Bob,

Pulled the trigger a bit quick on the post, I thought I could get in before anyone saw it :sweat_smile:
Strong agree, tryna do too much at once :smiley:

1 Like

Hi Liam,
Power on testbed is from the USB connection. Ground is linked from Pico ground pin 38 to the protoboard ground line, 5v is from Pico VBUS pin 40 to the protoboard power line.

74HCT125 and leds power from the protoboard busses

Murray

2 Likes

Hi Murray,

Same connections here,

Running the code straight from: How to use WS2812B RGB LEDs with Raspberry Pi Pico - Tutorial Australia

With modifications to lines 8-10:

NUM_LEDS = 34
PIN_NUM = 17
brightness = 0.1

It’s a shame you couldn’t pop by a local makerspace to get it nutted out super quick!

I do recall there being some GlowBits with moisture damage aaaaaaaages ago, I haven’t heard of any issues on the forum, if you plug in one of the Neopixels (WS2812B) strips does that work? (I’m trying to track down some GlowBits now :smiley: )

1 Like

Hmm are you using VBUS or VSYS ??

Just digging out some neopixels now and sorting out the connections

2 Likes

Hi Murray,

Can confirm Vsys - just tested with a GlowBit and its blinking well.

What voltage were you measuring on the 3v3 Pin of the Pico?

UPDATE: I measured 3.24V with my cheapie DMM, this number is in spec to run the Pico
(Also added an edit note to the previous reply :smiley: )

1 Like

Gotta be quick

2 Likes

Hi Murray
Bit hard to figure what could be wrong here, obviously something is. Your connections and Liam’s seem to be the same.

Have you still got that oscilloscope. If it has 2 channels connect channel 1 to level shifter input and channel 2 to output. Trigger off channel 1 and you should see what the shifter does to the data signal. Trying to fix remotely is proving difficult.
Cheers Bob

2 Likes