Inside A Gene Cafe

Inside a Gene Cafe

Also see:

Andy Cordwell (moc.thgilddo|ydna#!em liame)
Many thanks to HB Steve and to DaveC for their invaluable help.

Please note that none of this information has been confirmed (or denied) by any official source. The firmware version information particularly is based on my observed results with a limited amount of hardware. If you have any additional information, or have observed different results please let me know.

Mechanical Overview

The Gene Cafe is a well designed and constructed appliance. It is easy to service, and where possible, uses commonly available electronic parts.

The case is injection moulded plastic and is accurately manufactured. All parts, including circuit boards, are held in place with a substantial number of good quality screws. Despite the heat, my case shows now visual signs of deterioration after approx 150 roasts.

The heater element is reportedly long life (and I believe has an expected lifetime of 1000 batches or roasts). It is an easily replaceable parts (see my other Wiki article for detailed info on this process). The heater element is voltage specific, and due to its inherent thermal characteristics, must be of the correct voltage. The part is available in 100v, 110, 120v, 220v, 230v, and 240v variants.

Electronics Overview

The Gene works by blowing very hot air through an enclosed drum. There are three separate electrical boards; heater power board, input / display board, and the main controller + firmware chip which collectively control this process.

The controller board contains a voltage specific transformer (see later), while the other two boards are common to all models.

The heater relay board allows the low voltage controller board to safely switch the mains current to the heater element on and off. It uses a high current 12V DC relay to switch mains voltage to the heating element. This means the element is either full on or off and can't be partially on.

The input and display board handles button presses and drives the time and temperature LED segment displays based on signals from the controller board. Electronically its very uninteresting.

The controller board is the 'brains' and determines when to turn the drum motor, when to power the fan, when to power up the heating element and what to display. Its continuously monitoring the state to ensure the roast is continuing correctly and safely.

The controller board has a fair number of unused 'pads' perhaps indicating this board is designed to be used to support additional features and capabilities (looking at the the pad layout I suspect the digital version).

The controller board electronics are powered from a voltage regulator feed by a small onboard transformer. The transformer is voltage specific and is available in 110v, 120v, 220v and 230v (which is used in the UK) versions.

There are two temperature sensors on the Gene, mounted inside the metal conduits at each end of the drum. They both appear to be identical parts and are driven by identical electronics. The drum exit sensor (sensor 1) is used to determine the indicated display temperature and therefore when to switch the heating element on or off. The heating element end (sensor 2) is used to detect overheat (failed fan or imminent fire!) and under-heat (failed element) conditions. Error code E4 will be shown for both sensor 2 conditions.

The fan is can be driven at two speeds, normal and fast (for cool down). The firmware simply adjusts the reference voltage supplied to a standard voltage regular to vary the fan speed.

A similar approach is taken to change the motor speed (in later revisions). Interesting, all boards support dual speed motors, but on later revisions shipped with the necessary component (R24). The controller reads the drum end position sensor to ensure the drum always stops in the same position.

Overall, the electronics are very simple and use easily available parts.

The firmware is a common 40-pin PIC micro-controller, running at 16MHz. The chip is in a socket so can be replaced (but see notes below regarding this). The firmware also has Celsius (AC) and Fahrenheit variants (AF).


Please see my disclaimer above.

The dates shown are firmware revision/build dates (as a run will have the same firmware). Obviously, the back panel dates will be later.

The main control board is virtually identical between revisions, however, there are enough changes that a firmware revision will only work correctly in a board of the same vintage. For example, a rev 5 chip will not work correctly in a rev 0 or rev 3 controller board.

Firmware Rev 0 (unmarked - Nov 05)

If the unit comes from a 230v batch (see back panel), the element may rapidly switch on and off from around 200 degrees. This may make is impossible to reach 2nd crack within reasonable period of time. I believe this was the initial firmware release.

Firmware Rev 3 (June 06)

Element switching is very significantly reduced and may be non-existent.
Drum rotates at half speed when on its final revolution before stopping.
Motor drive electronics tweaked (possibly from later versions only).

Firmware Rev 5 (Dec. 06)

Drum now marked 'Max Dry Processed' rather than the cryptic 'Max Brazil'
Possibly motor gear assembly modified (unconfirmed).
Internal cable routing improved (this may have occurred later in v3 also).