Budget BarBot

circle things are RGB LEDs for status, hashed bits are pipe

FYI this was the width of a landscape A3 page so about 420mm wide

10 3.8Kg-cm servos for the clamp-valves :-)

Knocked this up last night from plasticard and spare bits

servo is a HK15138 3.8Kg-cm servo from hobbyking (£2.22)

calculated exerting 5.4Kg at start of travel building to technically infinite at end of travel, either way, plenty of force to clamp the tube properly. bolts used for easy adjustment.

Huzzah! A solution materializes.

after chats with various people, the idea emerges that, as the pump costs the most, that needs to be tackled, eg ONLY HAVE ONE.

use flow control through valves or something to limit what is being sucked up by the pump.

clamp design:

so... Pump ordered, and tube, now to dig through the parts box and see if i can knock something up (clamp-wise)

found this on ebay (http://www.ebay.co.uk/itm/281573370486):

has a much better flow rate of 500ml/min quite good for a peristaltic pump

it seems that diaphragm pumps are much faster, but there are some issues with cleaning, as shown below, the tube can be replaced easily:

this is fantastic for this kind of application, but unfortunately, at £22.99 a pop, that would take the cost of a reagent 'module' up to £50ish

originally had thought one pump, controller and light per reagent in a module, chain-able using I2C so you can add as many as you like. this is great, but turns out might be expensive.

need to find another way of doing this i think.

Testing of this:

Specs are 6V, flow rate (found out after testing) is 60ml/min

this is far too slow, and inaccurate, real flow is more like 45ml/min, this means it would take over a minute to pour a cocktail (one pump per reagent), or over 30 seconds if the smallest measure is 25ml.

this is not acceptable, another solution must be found!

Concept as first envisaged: