It's always bothered me that some KA ProLine grinders are wholly unsuitable for espresso and some users report great shots. A recent thread about problems after changing Mazzer burrs started the grey matter cogitating.

I don't remember the numbers, but when I measured the KA burrset, I was less than enthralled with the tolerances and went so far as to have the burrs precision ground perfectly [±1/10th] flat and parallel. Sadly, still sub par espresso. I considered machining the carriers, but decided to stop losses and buy the MC4.

A new, unused set of MC4 burrs was measured. One burr is ±2/10ths. The other is ±4/10ths. The lower carrier is ±5/10ths. Upper carrier not measured. [For the art majors, a tenth is 0.0001 inches. There are 10 in 1/1000th or 10,000 to the inch. The average human hair is ~40/10ths.]

Depending on how the burrset is installed, the tolerances could zero out or sum to as many as 11/10ths stacked up. Big Deal? Consider that typical espresso range adjustment on the MC4 is 2 teeth or 2% of a rotation. The thread pitch on the MC4 burr carrier is 1mm or 0.03937 inches. 2% of a rotation is 0.0007874 inches or ~8/10ths. Typical shot adjustment is 1/2 tooth or ~2/10ths.

This sketch may help illustrate.

( * indicates the burr rotated 180° in the 2nd and 3rd columns )

- With perfect carriers and burrs, the burrs are always in alignment.
- With imperfect carriers and burrs, when the errors cancel, the burrs are also always in alignment.
- With perfect carriers and imperfect burrs, the gap between the burrs varies with rotation.
- With imperfect carriers and imperfect burrs, the gap between the burrs could vary dramatically with rotation, depending on how the errors sum.

Bottom line, if after changing burrs the espresso goes to hell in a handbasket, verify the burr tolerances. If they are satisfactory, try rotating one burr at a time, making notes on the cup change.

This is a very interesting thread. Lets take a look at how on the KA Proline holds its burr in place.

The back carrier is fixed in place, beans comes down the hole and are advanced by the auger you see in the second picture.



The adjustable burr is held in place by an aluminum (or zinc) burr carrier with a steel shaft in the end (pictures taken during a recent cleaning, hence the coffee)...notice the tiny spring at the end that pushed back on the burr. This is very different from the other grinders I've seen in which the entire burr carrier is adjusted by the threads on the carrier.



The adjustments of the carrier come from the advancement (or retreat) or the bearing in the middle of this.



=-=-
Mark

Now IF everything aligns, the beauty of the system is the adjustment threads do not need to be cut onto the burr carrier. In addition, to clean the system, back off the two large bolts on the side of the black adjustment nob and the setting you have is not changed (well in theory, we are talking about such small changes that nothing goes back together that accurately).

The downside, the system as it ships from KA has massive slop in the adjustment gear and you can see the adjustment nob wobble during grinding. Adding Teflon tape to the plastic screw eliminates the wobble and greatly increases the repeatability.


You might notice the setting is at 6 2/10, (step-less mod on the KA, without this useless for espresso). All the beans I've ground typically are within what used to be 1 step. The other grinders out there have a large planetary gear and you adjust a small gear against the larger gear for very fine adjustments.
=-=-
Mark

Ian,

What alerted you to this discrepancy in the first place? I've wondered about such carrier alignment issues with my Super Jolly due to the sound the burrs make when they just start touching. It just sounds like the rubbing isn't constant. Other than that sound, I've had no reason to question the grinder...but if the prospect is a more consistent grind, I'd like to check it out.

Cheers,
Paul

The seed was first planted with the two KA Prolines that did not live up to expectations and milling the burrs had minimal effect. Later, Mr. Kehn disparaged the MC4 as "one of the clumpiest" grinders extant, but I haven't found it to be the case. Later a new MaxH owner claims it's clumpier than his Mini. Still later, a fellow changes Mazzer burrs and his shots are awful.

Something else must be at work.

Being a techno geek who worries about picosecond jitter, microns and the like, I did a bit of decimal shifting after I'd mic'd the burrs and the carrier.

What's not clear is whether parallel,eccentric or some intermediate alignment of the burr faces are the cat's meow. If the particle distribution is inherent in the burr design, then one would surmise parallel.

I've wondered if some of the 100 or higher tpi precision threaded adjusters would be useful in fully tweeking out a burr set. I was thinking of a equilateral triangle made up of these to fine tune a burr set once in place.
http://www.rdmag.com/ShowPR.aspx?PUBCOD ... monCount=0

or perhaps (warning pdf)
http://www.newport.com/file_store/PDFs/ ... Screws.pdf

My main concern is the force on the grinders is likely stronger than these could withstand. Still an intriguing thought.

Mark

Hi,

I understand that when talking about error cancellation you only meant a stable, non moving burr state? because with the lower burr turning during grinding you will go through a range of errors, from 0 to maximum error being the sum of both burr errors.

Cheers,
dsc.

For simplicity sake, errors from the bearing, spindle, motor mount and any non-linearities in the fixed burr mount that could change the orientation of the fixed to rotating burr are lumped together as 'carrier'.

mhoy wrote:I've wondered if some of the 100 or higher tpi precision threaded adjusters would be useful in fully tweeking out a burr set. <snip> My main concern is the force on the grinders is likely stronger than these could withstand. Still an intriguing thought.

A very intriguing idea from a research perspective. A dynamic measurement device would also be required to take measurements in situ.

Think I'll go brew a pot of drip

Perhaps this could be hacked into http://www.brandsonsale.com/ht-001348.html to create a dynamic readout in a similar way that the digital calipers have been used in http://www.hackaday.com/2007/03/11/digi ... er-modding.

A reference dot on the spindle could then be shown vs the dial indicator results in a plot.

Mark