Skip to content

Frequently Asked Questions

Enrico Zaghini edited this page Jan 29, 2024 · 2 revisions

Why a mobile remote control? I prefer having the display directly on the diagnostic card!

PicoPOST is not just another display for meaningless numbers. We have added several other readers, so different machines can be diagnosed more easily, and we can avoid mixing different data sources (different addresses) like the cheaper Chinese solutions do.
In order to accomodate all these features, we needed a browsable user interface. Here comes the idea of using a remote control, with a graphical OLED display and a few navigation buttons.

The remote control is wired using a SATA cable and, for now*, you can't fix it to the main PCB without heavy MacGyver-ing. Our idea was to allow more freedom of movement when working inside a case, or with weird board layouts.

* Rev. 8 main and remote PCBs will offer a "hard-mount" option.

Why a SATA cable? It's way too stiff for this!

It's a combination of things.

Conductors and connectors

The remote control requires a few signals for proper operation: I2C data and clock, interrupt for keypad triggers, ground and 5V supply. That's 5 dedicated wires.
There's a few options for carrying these many signals:

  • USB 3.0: connectors are readily available, but easy to mistake for a real USB port, with a chance of damaging one or both devices being connected to it. Plus, the B-side of the plug is relatively chunky and type-B USB3 cables aren't too common.
  • USB-C: same confusion-prone plug, with the additional difficulty of soldering an SMD plug with minuscule pins.
  • 8P8C (RJ45): plugs are very cheap, cables can be easily crimped at home, but it's been used for multiple functions (normally Ethernet, RS-232 ports on some networking equipment, even USB on some UPS machines!). Pre-made cables can be deceiving: having 8 pins doesn't necessarily mean having 8 conductors (only 2 pairs connected, instead of 4)! And not happy as-is, there's the question of crossover cables, carrying inverted signals to the other side of the cable.
  • DE-9 (serial): also readily available and cables can be easily made at home, but it can be confused for a proper serial port, and we don't want + and -12V in our poor Pico...

SATA is as readily available as other solutions, with very cheap (but stiff) cables, or more flexible (but expensive) ones. Additionally, seeing a SATA port all alone at the back of a machine should raise a question that maybe it's not really SATA...

Channel integrity

I2C is a somewhat weak protocol, which needs dedicated extender chips if you intend to use it over long distances.

There's not much control over USB cable lengths, and custom made cables can be WAY too long for their own good. This would cause the remote to not work at all.

SATA seemed like the best option, since it's made for high-speed differential signals and it should be able to maintain good signal integrity over an acceptable distance.

I'm not that good with surface-mount soldering, why didn't you use good ol' through-hole parts?

Some components necessary for interfacing the Raspberry Pico to an old and grumpy ISA bus are available, but they would bring up the final cost for parts quite a bit (talking about the 74LVC parts, mainly). Since we prefer having a somewhat affordable device, we opted to keep costs low by sticking with the more human-sized footprints.
There are no BGA or QFN parts, only SOIC chips and 1206 passives. The smallest parts are the SOT-23 diodes, which are still workable with a good pair of tweezers. Of course, any soldering job is easier with the right equipment and with the right practice, so we suggest starting with a good educational video about this topic, a good soldering iron, tons of flux and a dead board for excersing first.

Can I do a fancy and mount it permanently inside my computer?

Yes, you absolutely can. There's some oddities to take in consideration first:

  • There are no mounting holes for a rear slot bracket, so prepare for a lot of hot-snot.
  • The SATA port is placed towards the rear of the PC, so routing a front panel display could be awkward at best and impossible at worst.
  • Our bundled remote design won't fit either a 3.5", nor a 5.25" drive bay. We may figure out a cleaner solution, if there are enough requests.

Nothing stops you from having a remote cable dangling out the rear, like a tail, but you decide what's best for you... 😄

What display am I supposed to use with the remote?

We're not forcing anyone to buy any specific type of display. As long as it meets these few requirements, it will work just fine:

  • SH1106 or SSD1306 controller
  • 128x32 or 128x64 resolution
  • I2C bus
  • Lives on address 0x3C (but configurable in software at compile-time)

On the remote, you will find 3 connection options:

  • PiOLED, if you have a spare display from previous experiments on a full Raspberry Pi.
  • generic "gumstick" OLED, like the ones you find on 3-packs for relatively little beans on Amazon or other online shops. But beware of pinouts, there's no standard for these, so you may need to bodge some parts...
  • Stemma Qwiic displays, which should also allow for some creative work around the remote.

How do I stop the remote from poking my hands?!

We don't have a 3D model for it yet, so for now I suggest gloves.

Why are the remote config pads for soldering bridges? I wanted pin headers, in case I change my mind!

Once you've chosen your OLED display, you really don't need to change settings after powering it on the first time. These pads are only for the Pico firmware to properly detect and configure your display, so it doesn't display garbage and you don't need to turn your head to read it correctly. They also save a lot of brain activity from the software developer skull.