Announcement

Collapse
No announcement yet.

Let's Build a Flat Burr Coffee Grinder

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Let's Build a Flat Burr Coffee Grinder

    Hi Guy

    Twelve or so years ago, my first ‘coffee machine’ was a Braun Espresso Master, which came with a Braun spice grinder. Wow, I was set, and had ‘real’ coffee. The nasty little Braun is long gone, but the grinder lives on as a great spice grinder.


    When the Braun died, I started reading up and visiting forums, over-researching things in my usual aspergers-like fashion. Based on reviews and opinions on this forum, I got a second-hand Rancilio Audrey, which came with a Sunbeam EM0480 grinder. Now I was cooking with gas. While Audrey was a weapon compared to the Braun, the EM0480 was a nightmare – plastic gears, wobbly burrs, wear and tear, needing shims to keep it grinding fine, it wore out in no time – you know the story.

    So, back to the forum, and I picked up a second-hand Rocky grinder of Fleabay, and fitted new burrs. After the EM0480, the Rocky impressed me – steel, brass, aluminum, smooth, and heft, mucho heft. (The mother-in-law will attest to this, she picked it up from Taree, and brought him down to Canberra in her carry-on luggage.) And better yet, Rocky, like Audrey, was pull-apart-able. No silly plastic clips, but real threads and fasteners. They were designed to be repaired - to live forever. Now Rocky isn’t perfect, the adjusting thread is a bit loose, the speed is a bit fast, and the adjusting detent is a little coarse. As per advice on the forum, Teflon tape fixed two of the problems, and Rocky is well loved.

    Now, I’m not fanatical about my coffee, but I am a bit of a tinkerer. And as a retired machinist with too much time on his hands, and a workshop full of machinery, I sometimes start projects for projects sake. So it wasn’t surprising that I kept thinking someone could build a better grinder. All you need to do was attach a couple of big flat burrs to a variable speed geared motor. How hard could it be? More like how crazy would you be to do it!

    Click image for larger version

Name:	tenor.jpg
Views:	1
Size:	19.4 KB
ID:	763111

    This thread will cover the build of the grinder. More to come, , , ,

    Cheers, Steve.
    Last edited by Stavros; 21 August 2019, 02:20 PM. Reason: Formatting

  • #2
    Hi Guys,

    Firstly - the design. Basically, I started with a set of Mazzer Major 83mm flat burrs burrs, two SKF 80mm x 50mm x 16mm deep-groove ball bearings, a 375 watt three-phase electric motor with a 9:1 planetary gearbox, a single-to-three phase variable frequency drive, and 12kg of 150mm and 100mm 6061 aluminum bar.

    Click image for larger version

Name:	IMG_2120.JPG
Views:	1
Size:	159.5 KB
ID:	750477

    Click image for larger version

Name:	IMG_2118.JPG
Views:	1
Size:	136.7 KB
ID:	750478

    I spent quite a few night on the CAD package designing the grinder. Wherever possible, I applied good engineering practices, (typically used for machine spindles and shafts on lathes, pumps,etc), to maintain axial alignment and parallel faces. Initially Iwas planning to mount the rotating burr carrier directly onto the output shaftof the gearbox, as it is fairly rigid, and is supported in a reasonable sized bearing. This would have greatly simplified the design, and made it quite compact.

    Click image for larger version

Name:	IMG_2119.JPG
Views:	1
Size:	194.4 KB
ID:	750479

    However, call me stupid, but I decided touse a completely separate one-piece shaft and carrier, supported between two ball bearings, and connect it to the gearbox with a flexible coupling. This allowed me to use an adjustable nut to preload the bearings to remove any axial and radial play. It would also allow me to fit a different motor/gearbox one day if required.

    Click image for larger version

Name:	Sketch.JPG
Views:	1
Size:	158.0 KB
ID:	750480



    Now just to clarify, I’m not trying to built a commercial grade grinder, or a mega-grinder. It only has a 375 watt motor, with a 9:1 ratio gearbox, so it will grind at a moderate 300 rpm. With frequency control, I can ramp the speed up to a max of 600 rpm. That's pretty much in the ballpark with other domestic grinders. I’m just trying to keep the burrs consistently aligned and spaced apart.

    Cheers, Steve
    Last edited by Stavros; 21 August 2019, 03:57 PM. Reason: Spelling again.

    Comment


    • #3
      Hi Guys,

      To maintain alignment between the various components, there are mating spigots and recesses in adjacent parts, with just a few thou' clearance. The burrs are also mounted on machined steps and spigots on the spindle and carrier to keep everything in alignment. So, basically, (ignoring accumulated tolerances / clearances), everything should line up, no matter how many time I pull it apart to clean it.

      Click image for larger version

Name:	IMG_2122.JPG
Views:	1
Size:	156.6 KB
ID:	750481

      Click image for larger version

Name:	IMG_2124.JPG
Views:	1
Size:	178.4 KB
ID:	750482

      Burr gap adjustment is basic - by a 95mm
      diameter thread between the body cap and fixed burr carrier. The pitch of the thread is 16tpi, (or about 1.5mm), so 1/4 of a revolution will cause a gap of approximately 15 thou' of an inch. I think this will be fine enough to get the required adjustment. The first time I machined the thread it was not good enough, and I had to start again. Thankfully, the second thread is very smooth with minimum clearance, so there is no wobble or axial play. I must admit that I was a bit slack when it came to a means of locking the adjustment. I should have used a split cap with a clamping arangement, but I'm using a grub screw in the cap which will press on the carrier thread. I will however, use a rubber slug under the screw to prevent damage to the thread.

      Click image for larger version

Name:	IMG_2127.jpg
Views:	1
Size:	296.3 KB
ID:	750483

      The main spindle and rotating burr carrier are machined as a single piece. This should remove any potential problem with play or alignment, like when separate carriers are mounted on motor shafts etc.

      Click image for larger version

Name:	IMG_2128.JPG
Views:	1
Size:	133.3 KB
ID:	750484

      A pair of deep-groove ball bearings support the spindle. They are more than adequate to handle any loads generated by the grinding process, and will last forever. Again, these are readily available and cheap. If I wanted to make it sexier, I could have used a pair of precision angular-contact bearings, (like used in some high-quality machine spindles), but again, way too much $$$. A single lip seal is located under the carrier to prevent coffee grounds getting into the bearing area

      .

      Click image for larger version

Name:	IMG_2072.JPG
Views:	1
Size:	355.7 KB
ID:	750485

      Note that I chose a standard industrial three-phase / frequency drive to run the grinder. Mainly because they are common, and I'm familiar with them. They are easy to set up, and are simple to control - variable speed is by a 10k potentiometer, and run/stop by a low voltage input / switch. The newer 'sensorless vector' drives are very nice - they have nearly 100% torque at any speed. They are also cheap on FleaBay. I did consider a Parker brand servo drive and motor, which I've previously used for CNC applications, but I couldn't justify the extra $$$. The downside of this is a big ugly motor and gearbox - this is never going to look like a neat little kitchen coffee grinder. Maybe I should have just bought a Mazzer Major, , ,

      Cheers, Steve


      Last edited by Stavros; 21 August 2019, 06:36 PM. Reason: Gaa - formatting

      Comment


      • #4
        Originally posted by Stavros View Post
        Now just to clarify, I’m not trying to built a commercial grade grinder, or a mega-grinder. It only has a 375 watt motor, with a 9:1 ratio gearbox, so it will grind at a moderate 300 rpm. With frequency control, I can ramp the speed up to a max of 600 rpm.
        Super impressed with your work Steve...

        The one thing though, that catches my attention, is that the gearbox ratio and the speed numbers you have quoted seem to indicate that the motor you are using is a 2-Pole motor with a nominal ~2,900 rpm rated speed. Given if this is the case, I would have to caution against running this motor at anything above 100% rated speed as this will place enormous stresses upon the integrity of the Rotor. Wanting to achieve an output speed of 600 rpm at the gearbox would require the motor to be able to run at ~5,400 rpm, significantly above its design capability I would think, and safe operation....

        Mal.

        Comment


        • #5
          Good point Mal, I may have to contact the motor manufacturer and see what they recommend for the maximum speed. I should also do they same for the gearbox, it may not appreciate too high an input speed.

          Edit: Just looked up the technical specifications for the Varvel gearbox, and the maximum input speed is 6000 rpm. Lets see what the motor is good for.

          Edit: I contacted the Royce Cross Group, who distribute the Chinese made TEFC motors. One of their technical guys was familiar with frequency drives. He seemed more concerned with the frequency/voltage/current issues than the mechanical limitations. Once I told him it was on a little coffee grinder, he laughed and said no more than 100Hz on a two pole motor.
          I also went off to the WEG site, who have a lot more technical documents available, and a table (see below) listed the maximum speed at 7200rpm. I would imagine that WEG motors have a better electrical and mechanical build quality than a $120 Chinese import, so for the time being, I will limit the overspeed to 70%.

          Click image for larger version

Name:	IMG_2137.PNG
Views:	1
Size:	103.8 KB
ID:	750492

          Cheers, Steve.
          Last edited by Stavros; 22 August 2019, 03:23 PM. Reason: Add information.

          Comment


          • #6
            Hi Guys,

            The machining of the parts is by bread-and-butter means – turning in a lathe and milling in a mill. I am relatively fortunate here, as over the years I’ve accumulated a number of quality English, German, Swedish and Japanese machines. Even though some are up to 50 years old, they are in good condition, and will hold tight tolerances and give good finishes. (No Chinese machines here – if there was a machinery snobs site, I would be a founding member.)

            When starting with bar stock, there is a lot of meat to remove. I think 60% of the aluminium ended up as swarf and chips.

            Click image for larger version

Name:	IMG_2067.JPG
Views:	1
Size:	209.6 KB
ID:	750488

            Everything was machined with more care than usual, with critical bores and faces mostly completed in single settings. I machined the discharge path in the body today. It is at a tangent to the outside diameter of the grind chamber, and exits the main body at the centre. I will probably end up fitting a chute from a Rancilio Rocky, so made the final width of the port to suit.

            Click image for larger version

Name:	IMG_2115.JPG
Views:	1
Size:	158.6 KB
ID:	750487

            Well, so far, so good. Everything is coming together OK. I made a couple of temporary nylon bushes slightly smaller than the bearings, which allowed a quick and easy assembled of the parts to double check the design. I didn’t want to use the ball bearings as they are an interference fit in the body, and I want to install them once only, after everything is complete.

            Click image for larger version

Name:	IMG_2083.jpg
Views:	1
Size:	216.9 KB
ID:	750489

            I still need to make the thrower fingers on the spindle, these will be made from a fancy plastic called PEEK, which has a combination of mechanical strength, resistance to chemicals, wear, and temperatures up to 260°C. These will locate in slots in the spindle flange, and will be secured with M4 cap screws. The material is ordered, and should arrive next week. I made the height of the grinding chamber 20mm - about the same height as the burr set plus a little bit. Hopefully this will limit the amount of old grounds which will accumulate in the chamber.

            Click image for larger version

Name:	IMG_2132.JPG
Views:	1
Size:	146.4 KB
ID:	750490

            I had a chat to a few anodising companies yesterday. As there are none in Canberra, I'll have to use one in Sydney or Melbourne. (If anyone can recommend a company I would appreciate it.)I was advised that before I send it, to make sure that I've got all of the machine marks and scratches out of it, as anodising is a pretty unforgiving finish, which will show up all the imperfections. I was also warned about acid from fingerprints, which can leave invisible pitting which makes the final anodising blotchy. Woops, I probably messed that up.

            I did run into a problem with the Mazzer burrs. After they were bolted to the spindle, I did a trial assembly, and found that they were not quite parallel. It turns out that there are some ugly little burrs at the back of the three mounting holes, enough to stop them seating flat on the spindle. I'll need to stone these off tomorrow and check it again.

            And finally, this is the last thing a coffee bean will ever see - Arrrrrrrrrrrrrr, munch, munch, munch, munch. Bwahahaha!

            Click image for larger version

Name:	IMG_2135.JPG
Views:	1
Size:	133.1 KB
ID:	750491

            Well, time for a beer. Cheers, Steve.

            Comment


            • #7
              Awesome work mate, super interesting read and looking forward to seeing the end result.

              Comment


              • #8
                This is awesome, love this thread. Can't wait for the next updates

                Comment


                • #9
                  That’s amazing, great work!

                  Comment


                  • #10
                    That's some real fine machine work there. Will be following this closely!

                    Vector drives are amazingly cost competitive. Loved them to bits in a past life. Will you be putting an anti static wire across the tangential port?


                    Would love to see the enclosure you put around it!

                    Comment


                    • #11
                      Go cold anodising in a grey or dark green. It would look very military in the latter. Cold anodising is very hard wearing.

                      Comment


                      • #12
                        Hi guys,

                        What's a static wire? I'll have to google that. Mind you - the entire grind chamber, spindle etc are metal, and they are physically bolted to a motor/gearbox, which is earthed, so would this do the same thing?

                        An enclosure? What an odd concept - who would want to hide the 'industrial chic' look of the grinder? Well, probably my wife, I suspect she will be horrified when Rocky is replaced with this eyesore, even though the coffee machine and grinder live in a sneaky laundry, mostly out of sight.

                        To be honest, I haven't given it much thought as I have had target-fixation of the construction, not the end use! I could knock up a simple plywood box for the time being. It could house the motor and drive, with just the end of the gearbox sticking out with the grinder on top, and a run switch on the front. It may cut down some of the noise too. Some time this year I want to tidy up the laundry, and replace the free standing tub and shelf with a proper laminate kitchen bench and cupboards. I could put a 70mm hole in the bench, and stick the gearbox up through that, with only the grinder body visibly on the bench. Now that would look swank! Hmmm, that's odd - I just googled "coffee grinder hidden in bench", and it seems no one else is doing it, , , I wonder why? Perhaps they are sane.

                        Cheers, Steve

                        Comment


                        • #13
                          Originally posted by Stavros View Post
                          Hi guys,

                          What's a static wire? I'll have to google that. Mind you - the entire grind chamber, spindle etc are metal, and they are physically bolted to a motor/gearbox, which is earthed, so would this do the same thing?
                          The only source of triboelectric potential is the PEEK, if there is earthed metal in the path after the PEEK it should dissipate.
                          Last edited by Lyrebird; 22 August 2019, 09:29 PM.

                          Comment


                          • #14
                            Even an earthed grinder (most already are!) will do this:




                            What's a static wire?
                            It's become known as a static wire but it's really just something to lower the velocity of the coffee exiting the shute.

                            Mazzer's patent can be seen here:
                            https://patents.google.com/patent/US6948668

                            " The grid reduces the velocity of the ground coffee, or powder, so that the powder does not accumulate upon, or adhere to, the walls of the funnel. "

                            An even simpler solution was done by DesignedByCoffee a few years ago:
                            https://coffeesnobs.com.au/grinders/...-grid-mod.html

                            I suggest you wait and see what sort of mess your grinder makes before working out how to best fix it.

                            Great work so far too!

                            Comment


                            • #15
                              Originally posted by Stavros View Post
                              Edit: I contacted the Royce Cross Group, who distribute the Chinese made TEFC motors. One of their technical guys was familiar with frequency drives. He seemed more concerned with the frequency/voltage/current issues than the mechanical limitations. Once I told him it was on a little coffee grinder, he laughed and said no more than 100Hz on a two pole motor.
                              I also went off to the WEG site, who have a lot more technical documents available, and a table (see below) listed the maximum speed at 7200rpm. I would imagine that WEG motors have a better electrical and mechanical build quality than a $120 Chinese import, so for the time being, I will limit the overspeed to 70%.
                              Good stuff Steve...

                              Wasn't referring to Chinese imports by the way. Plenty of standard build induction motors are only tested to 120% of nameplate rated speed and is normally the absolute maximum allowed.

                              Mal.

                              Comment

                              Working...
                              X