[harrison]

Hi Xiaojun.

You can change that sampling frequency if you want. This is set by the ‘cycleTime’ variable stored under ‘Experiment’ in sysData variable. If you change this then the pumps will indeed change in terms of the loop they run on to achieve a smaller duty cycle. I would perhaps in that case leave the pumps at 60 second dilution frequency…

Why exactly do you want to have less frequent measurements? If you don’t need that much data I would suggest leaving it at 1 minute and then just time-averaging to get better signal vs noise.

[harrison]

Thanks. Also, how “well soldered” do the legs of the microUSB receptacle look? Can you see solder on the legs emerging from both sides of board? Any pictures of this would be appreciated.
In devices I have assembled myself, I have found that these connectors are almost invincible when soldered correctly (i.e. lots of solder). I am sure you are correct that the different cables may be more difficult to plug in, but I am surprised that they would be stiff enough to dislodge a properly assembled connector.

[harrison]

Not sure why that wasn’t working; I have just updated the page so that the link should now correctly download a .zip containing a .stl.
If you make a different version via modification or other changes I’d be happy to put it up there as well, I’m sure people would find it helpful!

[harrison]

This is very troubling, I didn’t realise that the manufacturer had not tested to ensure these work well. I will contact them right away. Is there any obvious difference when looking at the connector at the end of the blue vs black cable?

Here is a link to the blue ones:
https://www.digikey.co.uk/product-detail/en/pimoroni-ltd/CAB0703/1778-1019-ND/6928232

There are also some for sale on eBay in the UK:
https://www.ebay.co.uk/itm/Pimoroni-CAB0703-Flotilla-Cable-50cm/142319665819?hash=item2122e9c29b:g:gcsAAOSw4YdYz5Y0

• This reply was modified 6 years, 4 months ago by harrison.

[harrison]

I am just getting back to this topic, since we have now designed a custom lid which would solve this issue, take a look at the hardware page: https://chi.bio/hardware/

[harrison]

I don’t know, I have not tried!
Personally, I practically never turn the control computers off (I’ve left them on for six months at a time), which is made easier by attaching a 5V power supply directly to them (so they dont need to be plugged into a PC at all times).
If you do figure out some way to do this I’d love to know! We could use a different control computer in the future with inbuilt WiFi, which could solve the issue.

[harrison]

Yes, that is correct. One of the above commands is also making issues, I believe, because of an update to the underlying Debian operating system.
I have just updated setup.sh on Github, and the software setup instructions. It should now run the setup as intended.

[harrison]

I personally use /root/chibio, but ultimately it doesnt matter.
In part of that setup.sh script it should set the folder that cloud9 defaults to as root/chibio, hence why I am suggesting that one.

[harrison]

As in your other thread, they should come with the setup installed from the factory. However, I think it is in a different directory to where I intended.

Like you said, I think a simple fix to setup.sh might resolve this. I will have a look at what step is going wrong there…

[harrison]

The major things you need are glass test tubes, silicone tubing, plastic stir bars. Take a look at the Hardware page (https://chi.bio/hardware/), I have listed ideal specifications and suppliers for these.
Secondarily, you ideally want a computer/PC to plug the devices into, though this can just be a lab PC that is used for other things.

As to the lids, some simple ones come with the devices (made from PCBs and washers). I am currently working on 3d printed lids that will ideally be airtight and autoclaveable, I will post up schematics of these in the next few weeks…

[harrison]

Hi Alice,

That should be no problem. I think the best way to go about implementing that change would be in the PumpModulation function, as this will mean the data/interface in each reactor will reflect its current behaviour, despite the fact that it is using pumps potentially attached to another one.

A rough outline of the changes you might make are as follows:
1) Decide on how you are going to remap the pumps. Perhaps what you could do is for every odd index reactor, use the pumps for the reactor with previous index (i.e. M1 uses P3,4 from M0, M3 uses P3,4 from M2).
2) Add a new “if” statement to the top of PumpModulation. Inside this If statement you should catch each of the cases from step 1 independently. That is, case 1 is if you have an even index reactor, case 2 is odd index reactor. You can sort these by doing something like int(M[1])%2 to take the mod of the reactor index.
3) Set up new variables called MB and itemB which will be defined in a case-dependent manner in these if statements.
4) If case 1 and item==’Pump1′ or ‘Pump2′ then set MB=M and itemB=item. If case 1 and item==’Pump3’ or ‘Pump4’ then maybe just do sysData[M][item][‘ON’]=0 and then a return statement. This is to prevent the even index device from trying to control P3 or P4.
5) If case 2 and item==’Pump1 or ‘Pump2’ then set MB=’M’+str(int(M[1])-1) and itemB=’Pump’+str(int(item[3])+2). Basically what I intend to do here is map (for example) M1 to M0 and Pump1 to Pump3. Thus, we would be using the desired pump boards for the odd index reactor. Also do a similar statement to as above to set sysData[M][item][‘ON’]=0 if Pump3 or Pump4 are called in this case.
6) Now go through the PumpModulation function and update each incidence of setPWM(M,’Pumps’,sysItems[item] to setPWM(MB,’Pumps’,sysItems[itemB] Do not change the “M” in the next argument of the setPWM command.

If you complete the above then the interface should basically work as before, but when you use pump 1,2 on an odd index reactor it routes the digital command to pumps 3,4 of a different reactor. It might be worthwhile implementing the above with print statements scattered throughout to ensure my logic was correct…

[harrison]

Hi Jbe!
The standard lid is open to atmosphere, so you wouldn’t be able to regulate the gas inside. However, the device uses standard test-tubes with threaded 22-400 closure, so you should be able to find some commercially available lid (or make your own) that fits those tubes and allows you to make a gas-tight seal on the tube.
Making a custom air-tight lid is something we are currently considering, but we would likely wait a while until we can have a big demand before ordering them in a large quantity… thoughts?

[Author]

Posts