Category Archives: Uncategorized

Detailed Comparison of High Speed Voltage Level Shifting

This is a detailed comparison of Voltage Level Shifting from 3.3V logic to 5V logic, specificly for driving Addressable pixels.

In this test I used an ESP8266 set up to transmit data at different frequencies over the SPI port.

The top 5 screenshots use a single SK9822 pixel.

The bottom screenshots use the SN74HCT245N.

5 Diferent SPI frequenies were tested: 1Mhz, 5Mhz, 10Mhz, 15Mhz & 20Mhz.

The SK9822 pixel is consistently over 4.5V BUT at 15Mhz and UP drops to 4.4 and 3.7 which becomes useless. However the The SN74HCT245N is consistently between 5.3-5.3V even at 20Mhz. I did not test at a faster rate because the SK9822 pixel has a maximum “imaginative” rate of 30Mhz, although driving strands longer than 100 pixels becomes impossible without serious RF insulation. The APA102 pixel, on which the SK9822 is based, has an maximum rate of 20Mhz Which is almost impossible to reach.

Project ideas

This could be an interesting wearable?


Experimenting with cheap Power Banks

I recently tried this product:


its an SBEC you can get details here 

I wanted to get 5V with high current out of simple 10-20Ah power banks like these

For this test I have 4 power banks. Each one has 5 x 18650 LION batteries in parallel, I opened up each one and soldered a cable onto the GND and +V Rails that run to the batteries:20160515_142749

I then sealed them back up and added JST plugs onto each battery. Thi sway they can be removed and recharged individually using their own charging cables:


I then Connected 2 modules in series and connected 2 of these in parallel via a switch to the SBEC:


For the moment i have not had anything taking out more than 6A from the SBEC. Ill perform more tests as I test more LEDs at the same time!




A POV Project and some updates

Working on a Persistence of Vision project for a friend.


1st results looked promising so I went full steam ahead:



Some new tech for a new project


Received a huge bunch of new power banks for an exciting project ill be doing in January 2016! Cant say much more at this time. These are 5V @ 20,000mah and they were VERY cheap. $24AUD each! Have tested one and it worked rather well!


One of them powered the Disco Wig for 11 hours

Power distribution to large numbers of LEDs

power dist

I hacked this guy together last night and it works perfectly.

The GND and 5V cables go to 6 x 10Ah 5v Power Banks. That gives the suit 12A @5v MAX draw before they all cut out! This is enough for me.

Each switch controls power to 2 power taps. The power bank are connected to this block and all powered on. I then sequentially turn on the taps. None of the LEDs light up randomly at all!! and everything works smooth as!

Too much power


These bad boys are the solution to my power drain issues. Each 2 power tap cables will go to one of these. They will be sequentially cycled on to prevent the LEDs from randomly going HIGH


Wow Suit Gen 3 – Production Update

This was a busy weekend! I managed to complete all the panels for the 3rd Generation Wow Suit and have everything wired up ready to go. Software and LED Map is also done…


My 1st run time attempt ran into several issues

  • Each panel has at least 1 extra power conduit as well as the main daisy chain connection from the previous panel. What this means is that when power is applied, its applied to all panels at once. This causes the pixels to power up randomly and drain a TON of power at once. I currently have the suit plugged into several high output 5v supplies plugged into mains. They cut out as soon as you apply power. This is going to be a massive problem come portability… I’m thinking of setting up switches to each panel and slowly apply power to each panel as i power the suit on…but this is annoying.
  • About half the suit is not getting enough power….Although this is just a thought. The left part of the suit begins to stutter as if its under-powered. I currently have 6 x 2A @ 5v and even running at super low power is faulting.. I have 2 things i may try. One is to reduce the SPI frequency, i remember this issue with the Galileo and the other Pi. I’m running SPO at 10MHz now, I’ll reduce it down and see if that helps. The other is to push more power. I have a couple of PC power supplies that can push out 14A @ 5v. That has got to be enough!

Hopefully reducing the SPI frequency will fix these issues. If so, this presents another problem. The refresh rate of the suit is going to be slow.. Gen 2 was already approaching a limit and this one has over 2200 Pixels, which is a 9600Byte frame..