Mini Volumetric Dosing Not Accurate?

[Endo]

Feel free to prove me wrong, but I'm pretty much convinced now that the Mini volumetric control does not work accurately for low-flow shots like ristrettos.

I first suspected this when I discovered the Mini flow fault did not work. It suggests the flow meter is still registering flow even with a blind basket in place. If you look at the flow diagram on page 19 of the manual, it seems impossible since the flowmeter is just before the group and there should be no flow this direction when the group is choked and 100% of the flow is going back through the OPV and returning to the pump. (LaSpaziale told me the fault logic is the same for both Mini and the regular S1...so they are stumped too....so they say).

All I can think of is that perhaps the pressure fluctuations in the flow (the same ones that cause the group pressure meter to vibrate) also cause the flowmeter to turn slightly and register some flow.

In any case, I've learned to live with it since it's only really a problem if you are trying to make a very tight ristretto ( say < 1 oz). In which case it may cut off at half the volume of a regular 1.5 oz espresso. But as Oton says, normally nobody would think to use volumetric flow to shut off a ristretto anyway.

I set my single dose cut-off to 1 oz (30ml), and I only use it for warming flushes and cleaning shots. I set the double dose button to 2 oz (60 ml). I use this for all my coffee shots, and I just turn it off manually 99% of the time when it starts blonding. I only really use the cut-off as a fail-safe if I get distracted by a phone call or kids, in which case my Latte gets cut-off at 2 oz and my drink is usually still fine (no overflow).

[symbology]

I too have noticed inconsistencies with the dosing feature. Not sure what the root cause is. I will add that based on the diagram the flow meter is before the brew boiler. It is possible that some of the water that has flowed through the flow meter ends up residing in the brew boiler and never makes it to the puck.

Once dialed in, it should be good enough for my wife to use, which is my largest concern.

[Bitches_Brew]

I set my single dose cut-off to 1 oz (30ml), and I only use it for warming flushes and cleaning shots. I set the double dose button to 2 oz (60 ml).

i set mine opposite. the button with the "large cup" i set for 2.5-3oz.(not exactly sure, never measured) the button with the 2 small cups i set for about an ounce to warm my cup. i set it up this way because the large cup button if first in line and that is the button i use the most. i also use it for scrubbing the group to make sure i have enough water for a good clean.

i have never utilized the volumetric dosing while brewing a shot, but i do like the feature for cup warming. so accuracy is not really a problem for me.

[Endo]

I think I may have stumbled on the reason why the volumetric dosing is inaccurate.

I measured the the volume of the first blank water shot after warmup and it was about 20ml less than all the following shots . It got me thinking.....could it be related to the woosh on the regular Vivaldi? I'm thinking the first shot has less volume because the water is measured as it passes trhe flowmeter, but it doesn't make its way into the boiler since there is air trapped in the vertical copped tube the must be filled on the first shot only.

Just a theory, but interesting. I'll see if it repeats tomorrow morning.

[Louis]

Endo,

Reviving an old thread.

I've recently begun to pull shots by weight instead of volume. By doing so, I've noticed that the flow meter seems quite accurate. From my experience (I haven't logged anything to say for sure), it is accurate to about ±1 ml (±1g).

When I test the flowmeter using final espresso volume, it varies greatly, due to the changing amount of crema.

I don't know if this is related to your issue with the Mini...

[Endo]

When I test the flowmeter using final espresso volume, it varies greatly, due to the changing amount of crema.

Perhaps as you say the amount of crema is misleading. Then if you combine this with different absorption of water depending on blend and freshness, this might explain it.

After taking apart my flowmeter and looking at the flow path closely, I have a hard time seeing how it could be inaccurate.

Still, something has always bothered me about the blocking alarm not working when I backflush. But I suppose it might not be related to what I percieve as inaccurate dosing.

[Louis]

Still, something has always bothered me about the blocking alarm not working when I backflush. But I suppose it might not be related to what I percieve as inaccurate dosing.

Could the "vibration" of the water (vibe-pump) in the flowmeter be enough for the flowmeter to report something back to the controller, leading it to believe water is still going through? This is purely hypothetical, especially as the rotor seems to have only two "magnets, so a 180° rotation would be needed for the flowmeter to report flow?

This would not be the case with the rotary pump, which maintains a constant pressure, so no water oscillation in the flowmeter...

... if this makes any sense at all?!

[Endo]

Could the "vibration" of the water (vibe-pump) in the flowmeter be enough for the flowmeter to report something back to the controller, leading it to believe water is still going through?

This was my original guess as well. But as you say, now that I see it takes a 1/2 turn to register flow, I'm not so convinced any more.

Then this post appeared with the melted flowmeter rotor! This got me thinking, could the flow be reversing during backflush somehow, causing both a flowmeter indication (I assume reverse would register that same as forward flow) and the occasional melted rotor in extreme cases?

[chas]

You wouldn't think they'd use a type of plastic that was melted by water of that temperature.

[Louis]

You wouldn't think they'd use a type of plastic that was melted by water of that temperature.

The plastic type was probably chosen for completely different reasons and I'm sure the flowmeter specification says that it should only be fed with cold water... Something that was not supposed to happen happened.

[Endo]

You wouldn't think they'd use a type of plastic that was melted by water of that temperature.

The plastic type was probably chosen for completely different reasons and I'm sure the flowmeter specification says that it should only be fed with cold water... Something that was not supposed to happen happened.

I thought it looked like Teflon, which has a high melting point of over 300C. Even if it was regular plastic, the melting point is usually at least 150C as a minimum. Since the resettable overtemp on the brew boiler trips at 130C, I'm a bit confused how it could be melted.

Is it possible it was melted by friction? The black cover looks pretty fried. Maybe it was rubbing against that?

[chas]

I was actually wondering if the sensor itself might have shorted/overheated and conducted through the housing.

[Endo]

I was actually wondering if the sensor itself might have shorted/overheated and conducted through the housing.

I don't think so. The pickup/sensor simply sits on top of the black cover. No place to conduct heat, and no heat indication on top of the cover. The heat definitely came from within.

It definitely looks like it was melted by a hot liquid. If it was rubbed away through friction as I suggested earlier, I would think you would see small burrs of plastic on the edges where it rubbed. Besides, I don;t think it spins fast enough to generate enough heat, especially when emersed in a liquid.

On the other hand, I really can't see how the water could get above 130C with all the safety stuff in place. If it did...he's lucky it backflowed to reduce the pressure rather than turn into a boiler bomb (like this one) !

http://www.youtube.com/watch?v=fUkjXGfCLIM

It must be that the flowmeter rotor is made of a low temp moldable plastic to ease manufacture (like the stuff that comes out of a glue gun). To test this theory maybe someone can do some melting tests (toaster oven, etc) on a detroyed rotor.