mauijer,

I was hoping you were going to say, "hey Steve, I just ordered the Verslab so I'll let you know how it goes!"

Seriously, my decision is between these grinders:

Versalab M3, immediately modified by Terranova (great choice - many users are very happy)
EK43 (legendary grinder, but need to be good with thin mouthfeel plus 1:2.5 or 1:3 output volumes. Oh yes, wife would hate it )
Compak K8 (predictably good, great company, kinda ugly)
Titus (Hard to beat if one can come up with the $$$$)

The EG-1 doesn't make my list because it is not available, nor would there be adequate experience with it even if I knew it could be purchased tomorrow

mauijer wrote: Steve if you purchase the m3 I would love to hear your impressions and see if you find the claims of improvement to be accurate. I am in the same position as you: had good experience with Baratza but very interested in the m3 or frank hybrid as an upgrade. I have been hovering over the purchase now button daily so if you wouldn't mind pulling the trigger so that I can live vicariously through you I would be grateful lol

mauijer wrote:I've been following this thread and while I don't pretend to know what half that data means, I think the bottom line and process is interesting and it's good to know people are working on solutions. Pitty VL didn't keep the better motor.

Let me try to explain. In a DC motor torque rises linearly with current. So let's measure current at a certain RPM:



I marked the important maximum current (average) at 2.8A.

Now let's look at the Bodine (motor maker's) specification graph. Current is on the right, and there is a dotted line that marks the current / torque relationship (which is linear). I marked the previously measured 2.8A in red. I drew a line down to the torque (green line). We get 160 oz-inches.



Based on the above I'd expect the speed to drop to about 300RPM (1200/pulley ratio) which it doesn't. It does drop to about 500 (I have to measure some more), but it is much higher than the graph would lead us to expect. Which tells me that this motor is actually very conservatively spec'ed. Bodine are playing it safe and under spec'ing their motors. It would make sense since their motors are used to run conveyor belts and may be operated outside in direct sunlight on a hot blistering day.

Is the motor adequately spec'ed? Duty cycle (on to off ratio) should average the current to 1A maximum (current box marked in purple on the graph). So 2.8A/1A = ~3 ratio. If I grind 20 seconds, and then pull 25-30 seconds, wipe etc. and get to 40 seconds between subsequent grinding - I am safe: 20sec/(20 + 40) = 3.

Motor is safely spec'ed.

I was asked in a PM what the trimmer settings were on the 0866 for those that don't have access to a scope or are worried about connecting an earthed scope to a non-isolated supply like the Bodine or Versalab ones are (Be very careful - either use an differential isolation probe or Ch1-Ch2 differential mode).

So here is a picture of trimmers. The wiper is the flat angled area on the trimmer.

Many thanks for the picture Assaf.

To conclude the project and ameliorate some of the concerns (as expressed by nickw and others) - I added a thermal re-settable fuse on the DC circuit.

It is a German Made ETA 1.5A unit I sourced for cheap on eBay. I wired it in series to the motor. I tested this for a few grinding sessions - it didn't pop. Stalling the rotor - it popped within 5-7 seconds, protecting the motor.



Note these fuses are spec'ed for 48VDC and 250VAC. Theoretically - the 48VDC rating isn't enough for the 130VDC of the motor. But reflect that the 130VDC is actually a chopped up rectified AC (in fact - forward current goes low enough to turn off the SCR in the controller). So in reality - it is an undulating current that goes down to 0 and thus will extinguish any arc. See oscilloscope screenshot of the stalled motor below (this oscilloscope diagram is for switches 3,4,5; at switches 2,5,7 and trimmer as in the pictures above, the current is substantially higher ~4.5A rms but shape is the same).

I am learning quite a bit about grinders via this ad-hoc project. Via PM I am working with ericc on his Versalab and getting different results - which ends up sending one down rabbit holes one wouldn't fret over otherwise.

I am now approximately 3pm on the torque pot. So what happened?

After all these discussions I decided it is time to realign the grinder - but this time I had the nerve to tear the top apart. 2 hours later, with a head-visor lens and feeler gauges my grinder is now "perfectly" concentric and parallel ("perfectly" as to my ability to gauge - not with a run-out gauge etc.).

Upon grinding with the newly aligned grinder - the grinder stalled and the resettable fuse popped (Torque pot was at 1pm)! It mattered little what the grind setting was. Moving the pot to 3pm resolved the situation*.

The observation: The aligned grinder requires more torque than the non aligned one.

My cockamamie theory: In an aligned grinder the gap is uniform, which results in a more uni-modal grind and uniform friction 360 degrees the burr. in a non-aligned grinder the grinds get "released" wherever the gap is wider, reducing the overall friction and required torque.

Conclusions:
1. Burr alignment can be deduced by measuring grinder power draw (stator current, in the case of PMDC motors)!
2. Ericc's grinder was (originally) better aligned than my grinder and hence the torque required was higher.

*Note: Someone may question if increasing current is wise. To which I will restate the physics of DC motors - the current is a function of the load ONLY! The only way to reduce current is to reduce load (add less beans; roast darker; reduce burr size; mount burrs so they are not aligned...). The torque pot will only set the maximum current available for the motor - which should be above the highest torque that could be exerted by real beans. Above which point it will stall. So more torque will require more current. The thermal fuse works well to protect the motor if the current is excessive (either stall, or protracted grinding over time - like grinding rice).

AssafL wrote:The observation: The aligned grinder requires more torque than the non aligned one.

My cockamamie theory: In an aligned grinder the gap is uniform, which results in a more uni-modal grind and uniform friction 360 degrees the burr. in a non-aligned grinder the grinds get "released" wherever the gap is wider, reducing the overall friction and required torque.

I think that's the case too. Not ejecting the grounds from the axial gap, would actually grind the boulders down further, and thus require more power.

Thanks for all the work on this. At some point I'll probably add a new controller and thermal fuse to mine, for the benefits of grinding light roasts.

I also found once I aligned my grinder, the belt slipped less. I believe the more uniform burr gap results in a more constant load - And less loading up of the burrs, which can result in belt slippage.

Nick,

Once again, your girlfriend's traction control analogy illustrates nicely!

Reminds me of a snowy hill that is steadily climbed by a car when all 4 tires "catch" (4WD) and you gingerly feed the gas pedal to get up the hill using a higher (torquier) gear vs the car sitting at the bottom of the hill easily spinning its tires (in 1st gear).

With Assaf's help via this thread and PM I have also modified my M3.

I purchased the correct Bodine 0866 Controller from a seller in the US via eBay.

Changing out the original Versalab version for the Bodine was straightforward and no trouble at all, due to Assaf's precise and detailed instructions posted earlier in this thread.

I did have to buy a 10K ohm POT for Speed control as the controller was not supplied with one.
I purchased this locally, it was only a couple of pounds.

I drilled the rear plate just above the IEC socket and mounted the speed controller there.
At the lowest speed the grinder runs at 64 RPM and at the top end it is 640 RPM, this is of course in the UK and running on 240V 50Hz.

I was initially running the grinder at 350 RPM and suffered no belt slippage or stalling, result.
Now i have the grinder running at 500 RPM, apparently the speed setting that they arrive from Versalab.
It's interesting to be able to alter the speed and i will have to put aside a day sometime to see if there is any taste difference.

Following Assaf's further modifications i bought an E-T-A 1.5Amp circuit breaker and installed it in series into the white wire from the motor so as to protect it and at the same time upped the power fuse to 3.15 Amp.

I am now absolutely thrilled with my Versalab M3, which i have now owned for more than 8 years.
Many other grinders have come and gone but the Versalab is a constant.

DISCLAIMER

I can highly recommend this modification for anyone who is competent, don't forget to completely disconnect the grinder from the mains supply before opening it up and putting your fingers inside, and keep them out when reconnected to the mains supply, especially whilst testing.

IF YOU DO NOT FEEL COMPETENT GET A QUALIFIED ELECTRICAL ENGINEER TO DO THE WORK FOR YOU.

Excellent work Assaf, Thank you.

Thank you for the kind words.

As per a previous post - were it not for EricC's installation (current consumption) I would not realize my grinder needed further alignment.

H-B is great when we can share these ideas that take a perfectly good product and make it great. Cheers!