ESP8266 Community Forum

I've recently ended up with two amateur espresso coffee machines, a breville BES840 & Sunbeam EM6910, and after playing with my NodeMCU ESP8266-12E for a little while, I've been thinking about integration (see roughed-up diagram). Primarily I want to be able to turn the machine on remotely to warm up and monitor/regulate thermoblock temp with a K-type thermocouple and SSR, probably through a Java Webserver, and then display temp, pressure and extraction time on a small OLED.

I found the attached schematics for my two machines, and they seem to be simple 8-bit 5v ICs. Further down the track, it could be useful to directly read button inputs or drive the digital outputs of the factory IC with my ESP8266. If I use a logic level converter between ESP8266 digital pins and the associated pins on the standard IC, is there any issue with effectively piggy-backing the factory IC? Items such as pull-up resistors and signal smoothing capacitors would be doing the job for both ICs, and the digital pins of the ESP and IC (a mixture of Inputs and Outputs) would be effectively directly connected.

I'm not ready to replace the brains of the machine all together, but would love to, further down the track. Any advice would be greatly appreciated!

First thing to bear in mind is that the Breville schematic shows a live chassis (no isolation transformer) so anything attached will also be live. That makes working on it a significant safety risk and should only be attempted if you really know what you are doing. One can lower the risk by running the machine through an AC isolation transformer and preferably an RCD trip device. Any deployment must also ensure that no external connection can be made to the ESP8266 whilst machine is powered.

The Sunbeam schematic doesn't show powering arrangements. So you would need to dig into that further.

Having said that monitoring inputs are relatively easy and one could just use resistor dividers to lower any 5V logic down to 3.3V

Outputs are more tricky as you have to deal with conflict between the internal drivers and your logic. You can insert a relay or SSR in the main power path.

The Breville seems to have controls for power, hot water, steam and flow. The last three seem to be pulled up to +5V through 10k resistors. So one can drive these via open collector bjt or preferably n channel mosfets. The power seems to connect +5V direct to P65 so probably is better handled by a mosfet switch or relay in parallel with those contacts.

The Sunbeam looks easier as all the controls are grounded switches going via pull ups into the control logic. I would just put open collector bjts or mosfets in parallel with those so one could just simulate pressing the switches.

btidey wrote:First thing to bear in mind is that the Breville schematic shows a live chassis (no isolation transformer) so anything attached will also be live. That makes working on it a significant safety risk and should only be attempted if you really know what you are doing. One can lower the risk by running the machine through an AC isolation transformer and preferably an RCD trip device. Any deployment must also ensure that no external connection can be made to the ESP8266 whilst machine is powered.

The Sunbeam schematic doesn't show powering arrangements. So you would need to dig into that further.

Having said that monitoring inputs are relatively easy and one could just use resistor dividers to lower any 5V logic down to 3.3V

Outputs are more tricky as you have to deal with conflict between the internal drivers and your logic. You can insert a relay or SSR in the main power path.

The Breville seems to have controls for power, hot water, steam and flow. The last three seem to be pulled up to +5V through 10k resistors. So one can drive these via open collector bjt or preferably n channel mosfets. The power seems to connect +5V direct to P65 so probably is better handled by a mosfet switch or relay in parallel with those contacts.

The Sunbeam looks easier as all the controls are grounded switches going via pull ups into the control logic. I would just put open collector bjts or mosfets in parallel with those so one could just simulate pressing the switches.

Wow, that's an incredible help thanks! The Sunbeam is the better machine, and so that's the one I intend to hold on to and which modifying would be of most value.

I've not worked a whole lot with AC power. I'm a Mechanical & Mechatronics Engineer, but most of my electrical work has just been out of personal interest at home, so I'm not real clued in on AC power circuits and isolation. I assumed that I could piggy-back the factory IC, connecting to the 5V+ & GND for power, and it'd be all good, but maybe not, it sounds!

At the moment I might not need to drive any Outputs, I'll likely just use the ESP8266 to emulate the Power and Manual buttons, then read the Flow Meter signal to measure extraction time. Can I not drive the outputs directly, without MOSFETS, because by the time a 3.3v High runs through a 5v logic level shifter, it'll likely lack the required current?

I was wondering whether voltage dividers would sap current and interfere with the standard wiring, is there any difference in that regard, between voltage dividers and logic level shifters?

I've updated the schematic a little to suit the sunbeam and include the logic level shifter (it's still rough). Greatly appreciate the help!

Based on the schematics I posted in my previous reply, I imagine that T1 on the Sunbeam power schematic is an Isolation Transformer (ie. the 7V circuit is induced by the 240V with no direct connection, and then rectified by the diode bridge)? Whereas the Breville seems to just regulate down the Line Voltage?

The use of an LM7805 in the sunbeam makes me hopeful it should be sufficient to power my ESP8266 circuit as well as the factory IC.