Spectrophotometer plans and tutorial for turning a mini fridge into an incubator

Skip to first unread message


Oct 1, 2022, 4:04:01 PM10/1/22
to Labs
Hi all, was wondering if anyone had those plans for the spectrophotometer that some of yous made a few weeks ago? also, according to the meet up app, there was supposedly some tutorial going around for how to transform a mini fridge into an incubator?

Jenny Molloy

Oct 2, 2022, 1:43:48 PM10/2/22
to c.t....@gmail.com, Labs

Here is the video for the spectrophotometer, Adrian who led this session was a guest speaker from Canada and has not produced a written tutorial, some participants at the meetup suggested that this would be helpful but to my knowledge, this is currently the main documentation of the project:

The fridge incubator is an ongoing project at the monthly Biomaker meetings so there is no tutorial yet as we (mostly Tony and Richard!) are still figuring out the best approach and testing sensors etc. This is will likely continue at the 19 Oct event:
Hope you can join us!


On Sat, Oct 1, 2022 at 9:04 PM c.t....@gmail.com <c.t....@gmail.com> wrote:
Hi all, was wondering if anyone had those plans for the spectrophotometer that some of yous made a few weeks ago? also, according to the meet up app, there was supposedly some tutorial going around for how to transform a mini fridge into an incubator?

You received this message because you are subscribed to the Google Groups "Labs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to labs+uns...@makespace.org.
To view this discussion on the web, visit https://groups.google.com/a/makespace.org/d/msgid/labs/6cf83888-1dbd-4641-a5a9-9940a4c91cf2n%40makespace.org.
For more options, visit https://groups.google.com/a/makespace.org/d/optout.


Oct 2, 2022, 2:34:44 PM10/2/22
to la...@makespace.org

Hi All,

I am happy to join in and help with the fridge incubator as this probably has a few similarities to my lab cold storage monitor project.

For those interested;

I use an Arduino framework-hardware called nodeMCU, this micro-controller has wifi and will easily connect to the internet. On this device I have installed an open-source IOT/home automation firmware called Tasmota.

Sensor and control data can be either managed locally (via a web interface on the lab wifi) and/or with a cloud-based processor such as Node-red.

All parts of the project are open-source, minimal coding.

A really easy start for the Incubator might be to consider using the Sonoff TH16 with a light bulb as a heat source. The Sonoff can be flashed with Tasmota as well.

I have a Sonoff TH 16 which I could bring to the meetup, and we can see if it could be useful.

Please let me know if this is interesting to you

Best wishes

Patrick Scott


Jenny Molloy

Oct 2, 2022, 4:59:43 PM10/2/22
to Patrick, la...@makespace.org
Thanks Patrick!

The "fridge" is a peltier-controlled incubator that is sold as both a beer fridge and a reptile egg incubator depending on what site you're on so it can already go from about 4 ℃ to 60 ℃. The project has focused on having a programmable control to do custom temperature profiles for biological control (at the moment you have to press the buttons to change the temperature),

Richard and Tony may be able to chip in with more details of their current approach to this and scope for overlapping with the tech you're suggesting!

My very specific use case is protein expression in bacteria, whereby I want the fridge at 37 ℃ for about 6-8 hours to maximise bacteria growth and then to drop to 11-18 ℃ for expression of the protein but this switch happens overnight. hence wanting the temperature profile control. My research group develops manufacturing protocols for proteins in low and middle income countries and being able to get an incubator and controller for <£150 would be excellent. Even very expensive scientific incubators don't offer this. For reference, the same incubator (manual control only) is being rebadged by a well-known scientific instrument brand and sold for labs for nearly 900 EUR.


Richard Hopper

Oct 3, 2022, 4:10:40 AM10/3/22
to Jenny Molloy, Patrick, la...@makespace.org
Hello Patrick, 

Pleased you are interested in helping with the incubator project as well. 

The electronics is basically working, although the parts need better mounting. We are using an H-bridge driver for the Peltier and a PID control algorithm. The approach is described here

We are developing code for the ESP32 microcontroller module which has WiFi and Bluetooth. There is some existing code which could be adapted (linked in the google doc). I intend to start a github page, so the code and details about the project can be shared. This should enable collaborative working. I am interested in other approaches. Perhaps we can discuss further on the 19th. 

Best wishes,


Oct 3, 2022, 2:46:13 PM10/3/22
to Jenny Molloy, la...@makespace.org

Hi Jenny

This idea would require minimal "hacking" of the incubator;

Looking at your Reptile egg incubator, I see that there are temperature set-point selection buttons (up/down).

Assuming the total set-point range is 5C - 60C in 1C increments, there are a total of 55 temperature values that can be set by button-presses.

A method to set a value electronically could be;

1. Zeroing: press/actuate the down button 56 times (in say 3 seconds) - this will result in the minimum set point (from any previous arbitrary set-point).

2. Setting: press/actuate the up button the number of times corresponding to the desired temperature above 5C, so for example if we want 37C , then we actuate UP 32 times.

3. Timing: Now we start a timer for 7 hours, and do nothing else.

4. Timing: At timer remaining = 0 hours, we first goto 1 (Zeroing), then press/actuate UP for 10 time to change the set point to 15C.

5. Timing: Now we start a timer for n hours, and do nothing else.

6.Timing: At timer remaining = 0 hours, we don't change the set-point, we send a notification to Jenny's smartphone, and switch on a "cycle-complete" light. and do nothing forever.

Tasmota firmware offers the ability to switch a relay, and switch it quickly (pulse) like a button-press for example.

We could have two relays, one wired across the contacts of the UP button, the other across the DOWN button.

We can write Tasmota rules to actuate the relays, the correct number of pulses and use a timer to create your temperature time profile.

The advantage of using Tasmota rules, is that all the processing is local, no IOT needed. The disadvantage is there is no user interface to speak of.

Off-course you don't have to run the process on the Tasmota device, you could undertake all the control either locally in a PC/SBC or in the cloud using something like Node-red.

Tasmota can also read a IO input (called a switch), this could be used to start/stop/reset you profile.



Oct 4, 2022, 1:27:19 PM10/4/22
to Richard Hopper, Jenny Molloy, la...@makespace.org

HI Richard,

Thanks for your email. Yes Im happy to help, its an interesting project. It looks like you are planning to re-design most of the control electronics, quite a task!

Here's a quick idea:

You know that you can buy temperature controller at a low price; 

https://www.ebay.co.uk/itm/154706490808   14 pounds

These have both heating and cooling modes, a user interface etc..

You could get two of these and relays to control your heater and cooler devices, and use two programmable time switches to select temperature controllers.

You program the set-point temperature you want for the first period, and then you program the timer to switch on the 1st temperature controller for that period, then switch off.

For the next period you program the set-point temperature you want into the 2nd controller, and set the timer to turn on the 2nd temperature controller at the same time the first one turns off.

You will have:

1) fully programmable two temperature control system.

2) user interface.

3) Low cost in time and effort, I recon you could build it for GBP 70 in parts and one afternoon wiring it up.

4) No coding.

Programmable time switches:  11  pounds



I will come meet you and share ideas on 19th.

See you then


Richard Hopper

Oct 4, 2022, 2:35:49 PM10/4/22
to Patrick, Jenny Molloy, la...@makespace.org
Hello Patrick,

There are a few different ways to do this and maybe the approach I have been looking at could be simplified and it would be good to test other approaches.

The cost of the current system is around £25, basically the H-bridge and ESP32 module + temperature sensor.

An advantage with customised control is that you can tune the PID parameters to improve the temperature stability. Basically every system will ideally need tuning, often at different temperature points. I am not clear if you can do this with a cheap off-the-shelf controller? However, it is a good question as to what temperature stability is needed?

The cheap controllers use a single pole relay and do not have bidirectional current control for the Peltier to enable both active heating and cooling but maybe some others out there do.

I suspect PWM switching of the Peltier might enable better stability than simple relay switching which is slow to respond but perhaps switching speed is not critical given the large thermal mass of the system?

There are existing code modules for the PID controller and to interface the sensor. I would need to look further at communication. Perhaps this could be done over WiFi or USB? Sending serial commands from a USB connected computer should be straightforward using Python. You would just need to send temperature points and maybe some PID parameters for tuning.

I think it would be good to test other approaches as well, if you can help with this. Mimicking the pressing of the push buttons seems like the easiest option for setting the temperature, if we can rely on the stability of the controller. 

Kind regards,



Oct 8, 2022, 6:40:38 AM10/8/22
to Richard Hopper, Jenny Molloy, la...@makespace.org
Yes ok interesting.

Looking forward to meeting up with you again.

Do you have the device at the makespace?. I agree with you on the mimicking button presses..I'm quite sure this can be achieved with tasmota and a Sonoff or other esp8266/esp32 board.

You might be able to read the led display by tapping into the driver IC..

I assume the original machine has a control and stability specification.. I'm not if you want to improve the thermal performance and stability.

I found that it matters where you place the temperature sensors as door opening draughts can be quite an issue..I use a thermo well (a buffer like a aluminium block in which I place the thermocouple).

I have used the Inkbird temperature controller with great success.. I used one on a coffee machine with a sold state relay..the PID control tunes itself after a while as it 'learns' the thermal profile of the device and environment. My coffee machine would stay within 0.5C of setpoint..They are £6 on ebay..

You could switch your H bridge with a double pole relay.




Oct 8, 2022, 1:11:22 PM10/8/22
to Richard Hopper, Jenny Molloy, la...@makespace.org

Hey, I just can across this which might be a fun way to run the incubator:


I have a Sonoff TH to donate to the project, I will see if I can get it flashed before we meet up


Richard Hopper

Oct 10, 2022, 2:58:10 PM10/10/22
to Patrick, Jenny Molloy, la...@makespace.org
Hello Patrick, 

It looks like a great idea for process control.
The incubator is in Makespace and I could show you this and what we have done. I am likely to be in Makespace on Saturday afternoon, after helping with the biomaker move, if you are around then.  More approaches and ideas, like the ones you mentioned, could be tested. 

Best wishes,
Reply all
Reply to author
0 new messages