Adapting PC Joystick

For the discussion of all hardware related topics.
Post Reply
Posts: 12
Joined: Wed Feb 19, 2020 2:07 pm

Adapting PC Joystick

Post by moose »

Hi all,

I've been working on converting a PC joystick to work with the Dragon, having picked up an old Quickshot Skyhawk with a 15 pin connector for very little on ebay. I've now got it working, but given that I didn't find much info on this when I googled before I thought I'd outline the steps here - apologies if this is old hat.

Old style 15 pin PC joysticks are generally analogue, but rather than using the pots on the axes as voltage dividers they instead form the R in an RC circuit. The pc feeds 5v into one side of the pot, then times how long the internal cap takes to charge up. In practice this means that only one side of the pot and the wiper are connected, at least on the joystick I was working on. This means a simple 15 pin to Din adapter isn't enough on its own to make the stick work.

Inside the stick there are the pads needed to connect ground to the pots, but the ground wire vanishes into the body of the stick to reach the fire buttons. I spliced into this and bodged connections back to the X and Y pots. This resulted in a working joystick, but with X and Y upside down. Heading back in I desoldered the +5v and newly bodged ground connections, swapped it all round, and got a joystick that seems to do all the right things. I also took the centring springs out for that authentic Dragon experience.

I've had a few oddities with the stick which might down to the pots being higher resistance than the usual dragon pots (they measured on my dodgy meter at about 130k Ohms end to end), but may equally be the hack soldering job. A bit of extra reflow seems to have sorted it, but we'll see how it stands up to a full Mudpies session.

Posts: 245
Joined: Fri Apr 10, 2009 1:40 am

Re: Adapting PC Joystick

Post by prime »

Yeah Dragon/Tandy and BBC analog joysticks do theings the correct way a potential divider, which is wy you never have to calibrate them because they are in effect self calibrating. However to read them you need to be able to read the voltage level which means some sort of DtoA and a comparitor.

The IBM way of doing things means you get a train of pulses that can be counted digitally and the space between pulses counted (or number of puleses per time division) to get the position, so no DtoA is needed, however this is a lot less consistent and so requires the calibration / adjustment.


Post Reply