Chocolate Alchemy’s DIY Winnower

Okay, for those of you completely chomping at the bit, here is the first draft of an easy to build and operate Do It Yourself (DIY) winnower.  It’s made from 2″ and 3″ PVC fittings and pipe, a Champion Juicer and a common Shop-Vac.  You will need to work out a support structure for it if you want to build one now, but that too will come along at some point in the next month or so.  It will handle both roasted and raw cocoa beans.  Roasted behave much better, but in either case, the better the bean preparation, the better the separation.  It will produce 1 lbs of nibs in just over a minute and will average 77-79% recovery when ‘tuned’ properly.
I have not yet detailed every last bit of information on how to assemble it, but  I wanted to get it out there.  I will be talking at length as to how and why it works, how the inspiration came (it’s from 15 years as a chemist working with quadupole and ion trap mass spectrometers – how’s that for a small tease?).There are virtually endless configurations that can work – and of course virtually endless variations that will not work.  If you can build it exactly as is, perfect.  If you feel you must modify something, PLEASE try not to change the pieces in yellow.  Their dimensions, sizes, lengths, configurations, etc WORK.  But as you will notice, there are LOTS of places you can customize for your needs.That’s it for now – here it is:winnower-v1-no-notes.JPGAnd here are the more detailed plans.


Feel free of course to e-mail me if you have questions.  If you build it, you will have to ‘tune it’ by adjusting the two valves to regulate the air flows.  Off hand, as an estimate, the valve on the left can be all the way open for raw nibs, and about 1/3 close for roasted nibs.  The valve on the right is adjusted to regulate the total flow and is closed until husk stops falling out of the nib exit.  Again, I’ll detail tuning in greater depth, and email if you want more information now.
Finally, and please pardon my soapbox (stepping up as we speak), but PLEASE don’t ask me why I didn’t design it this way or that way.  I can’t answer that negative question. The base answer is I’ve designed and built something that works with parts that are off the shelf, and/or very simple to make if you have a mind to.  And I don’t fix what isn’t broken.   Is it perfect?  No.  Is it $35,000? No.  Is it $5,000? No.  Am I a little defensive – yeah, maybe a little.  In almost every case the answer as to why this or that isn’t in there is cost and ease of making it.  This is an elegant, simple design that works.  Are there places for improvement?  Yep, and they will most likely add cost because they will be more complex.   (stepping down).  But right now for about $50 in PVC, plus a Champion (which you may well have), and  a Shop-Vac (again that you may have), plus a couple hours time, you can have an automated working winnower.
That all said, feel free to ask why I designed it a certain way (do you note that difference?).  I am more than happy to discuss why it is designed as it is and how it works.

7 Responses to “Chocolate Alchemy’s DIY Winnower”

  1. All hail to the Alchemist. Thank you John. I was very happy to see the results. Even with raw beans it’s doing a great job. Well done!

  2. so, does this mean no sieving of the cracked beans?

  3. Awesome work John, my fan just burn out on my winnower so this is great timing!
    Question – can the 45 deg coupling where the nibs exit be straight?

  4. Alison,

    In theory, yes, it can be straight. I have simply found I like the deflection created as it goes into my collection vessel. What it can not be is too long (or you get drag which changes the air flows) or restricted (for the same reason).

  5. Thanks for answering questions John, I’m building it as I write this! Is it crucial to have the 2″ ball valve at the same height as the deflector? My “Y” section of pipe doesn’t have any space to put in the hole, it widens immediately for a joiner so I was wondering if the ball valve could be lower than the deflector?

  6. Alison,

    “Is it crucial to have the 2″ ball valve at the same height as the deflector?”

    Well, I want to say no for a couple reasons, but can not 100% back it up.

    I’ve actually been thinking of trying to moving it down so that the top of the discriminator hole (the 2″ ball valve) is below the bottom of the wye. In the drawing above that would put it into the joint, but I see no problem with that. Although the arrangement above works, I’m noticing on roasted beans, I need a larger hole than I previously needed, and I surmise that some of the draft is in-effective be bypassing along the top of the hole. By moving it down, it would all come into play and I could get the hole back down to where I had it.

    I’ve been thinking about it because that is how it was in my previous non-pvc model, but I moved it slightly to accommodate the valve geometry.

    I would no move it too far down as experience says that is bad, but some I think would be fine if not better.

  7. Hi John, Just wondering if you had any updated to this design. I just purchased the parts to build and a support structure, but figured I would ask.

    Any luck with moving the discriminator opening lower than initially designed?

    Also, is there any specific material the deflector need to be made out of?

    Thank you.

Leave a Reply



Articles and other Helpful pages

Alchemical Formulations