How Autofill Degrades Shot Temperature Stability - Page 2

[Ken Fox]

This is not what I see on my Vetrano (Fluid-O-Tech rotary pump). The brew pressure drops significantly when autofill is engaged. I just tried a simple test: forced autofill by running the hot water tap, and measured the flow rate. The water volume pumped out in 10 seconds dropped from 70ml to 35ml with autofill on. The Vetrano gauge does not permit determination of the brew pressure change, but the flow rate is cut roughly in half when autofill kicks in. This certainly could ruin a shot.

Fortunately I've never seen autofill engage during a shot, but I plan to keep it available as a convenient excuse: "Yeah, would've been a godshot except for the darn autofill."
________
John

P.S. - Apologies to the OP, I know this is slightly OT.

I don't think this is OT, actually. In part it is another but related issue, and in part it could explain some of what I saw in ways that I don't understand at this point. What I do understand at this point is that autofill SUCKs, for a bunch of reasons, especially when a machine is used in a low volume setting and when boiler water is not used for drinks. On my 11 year old Junior with a sight glass, I find it necessary to manually fill the boiler about every 4th or 5th day. Since both machines have the same boiler size, this should be the same for the rotary, however autofill is so sensitive to small drops in boiler level that it will probably activate at least daily with my usage pattern. Since the boiler gets autofilled at unpredictable times, this unnecessarily often refilling of the boiler does nothing more than degrade shots.

ken

[cannonfodder]

This is not what I see on my Vetrano (Fluid-O-Tech rotary pump). The brew pressure drops significantly when autofill is engaged. I just tried a simple test: forced autofill by running the hot water tap, and measured the flow rate. The water volume pumped out in 10 seconds dropped from 70ml to 35ml with autofill on.

Interesting. I thought that the rotary pump machines had sufficient flow capacity to make it through a mid shot fill. My understanding was that there would be a drop but not as drastic. Learn something new every day.

[cannonfodder]

I don't think this is OT, actually. In part it is another but related issue, and in part it could explain some of what I saw in ways that I don't understand at this point. What I do understand at this point is that autofill SUCKs, for a bunch of reasons, especially when a machine is used in a low volume setting and when boiler water is not used for drinks. On my 11 year old Junior with a sight glass, I find it necessary to manually fill the boiler about every 4th or 5th day. Since both machines have the same boiler size, this should be the same for the rotary, however autofill is so sensitive to small drops in boiler level that it will probably activate at least daily with my usage pattern. Since the boiler gets autofilled at unpredictable times, this unnecessarily often refilling of the boiler does nothing more than degrade shots.

ken

Now the question becomes how much does the boiler level affect shot temperature stability in a heat exchanger machine.

[Ken Fox]

Now the question becomes how much does the boiler level affect shot temperature stability in a heat exchanger machine.

I've played around with that a little in my vibe machine and didn't find any reproducible differences. My PID probes are in the steam column (both machines) which is going to be inequilibrium with the water below. Steam has considerable ability to transfer heat, so whether 80% of the water in the HX is encircled by water and 20% by steam vs. 70 and 30%, I doubt it makes much difference especially in the design of my machine's HX, which has it's water drawn out at the bottom which is where it is hottest (by report from a Cimbali tech with whom I discussed this).

Where it did make a difference was once when I removed the autofill probe from the rotary machine for cleaning off deposits. When I put it back in, it was not positioned correctly and where it was was at a point on the tank where the smallest drop in boiler level would cause the autofill to kick on. The result was the autofill was operating multiple times each day. When I tested shot stability then, I had very poor results. Repositioning the probe correctly resolved that problem.

ken

[miKe mcKoffee]

What I do understand at this point is that autofill SUCKs, for a bunch of reasons, especially when a machine is used in a low volume setting and when boiler water is not used for drinks.

So I asked myself if I'm entertaining and making a bunch of back to back caps', each cap' steaming started while pulling the shot with steaming typically lasting about 35sec for me w/GP2 running 1bar on the Bric', group flush and group flush wiggle after each shot, quick steam purge 'till heater kicks back on before flash flush & go each shot when would autofill kick in and how would it effect shots and production rate?

Started by initially forcing an autofill via hotwater dump. (I never use Bric's boiler hotwater wand for Americanos or anything else for that matter other than flushing the boiler every few days to bring in fresher water to reduce potential scaling) After forced autofill normal long Bric' cooling flush (~9oz, 16 count beyond end of flash)

Pulled 25sec Thermofilter shots (varying my flash flush & go target temp at random). Right after hitting countdown timer starting steaming, roughly 5sec into shot. Continued steaming 12sec after shot completed. Didn't graph temps just watched the Fluke display.

Autofill kicked in right after finishing steaming for cap' number 5, before group post shot flush/flush & wiggle. As normal on the Bric' boiler pressure dumped ~0.5bar during autofill and took ~20sec to come back up to pressure. Forgot to post shot flush/flush wiggle observing the autofill pressure so did right as came back up to pressure.

Continued cap' series with my normal ~35sec start next shot after post shot flush/flush & wiggle. NOTE: This isn't timed anymore but simply using boiler heating timing. Post shot flush/flush & wiggle kicks on heater. While building the PF the heater will toggle on one more time @ ~25sec (on cycle ~6sec normally). Just before flash flush & wiggle I always purge steam wand to force heater on one more time and start shot flash right after heater toggles off. Why? Primarily so machine always at approx' same state starting a shot. And when I didn't quick purge steam before shot flash I'd sometimes not wait long enough during a series and have no flash.

This 2nd series autofill kicked when purging steam wand before shot #7. (still steaming for caps' each shot) Waited the ~20sec to come back up to pressure then proceeded to flash flush & go. If anything the delay caused longer flash before countdown to temp, no noticeable effect on observed shot temp.

Continued simulated caps'. This 3rd series autofill kicked in right after steaming for cap' #7. And I quit.

That was pulling shots and steaming individually for 18 caps' with autofill kicking in after caps' 5, 6, & 7. How much the autofill actually affected shot profile unknown, but seemed to stay in my usual <1f target shot temp range, both initial high and ending low temps during shot, based on visual observation of Fluke. Total production delay ~1 minute caused by autofill after cap'5 and cap' 11. If I'd continued cap' series beyond 18 another ~30sec production delay before proceeding to #19.

Worthless post as far as hard data goes maybe, but as far as real world usage goes possibly beneficial.

[cannonfodder]

So in real world situations, the autofill is little more than annoying provided you allow a few extra seconds for the boiler to recover. Any taste difference would be minor and probably not noticed by most people.

In a high volume business, the delay for recovery, or the blown shots from the autofill engaging during an extraction is a bit more annoying. It would definitely have a negative impact.

The pure empirical results are yes it has does cool the machine and in theory, have a negative impact on shots.

I would imagine the size of the boiler would also play a part in how often the autofill kicks on as well as the impact it has. A smaller boiler would be impacted much more severely than a large boiler.

[Teme]

I find this thread interesting. A few comments and thoughts:

1) I personally do not understand why all machines would not have something in place to prevent the autofill from activating during a shot. This should be pretty easy to engineer and implement at low (or no) cost.
2 At least on a vibe pump machine it is very easy to observe when the autofill kicks in from the tone of the pump. If it does kick in one knows that the shot is a dud.
3) As to the temperature variation in the brew temp on a HX machine in the event of the autofill kicking in, I realise that there are many variables and the effect on brew pressure is probably the most significant (at least on vibe pump machines). It is difficult to generalise between indvidual machines, but could one say if the orientation of the boiler has any implications. The presumption is that on a vertical boiler there is more temperature variation within the boiler than in a horizontal

Br,
Teme

[HB]

On a related note, it's a good idea to introduce fresh water into the boiler before ending a session / shutting down for the day. The side benefit of this practice is less chance of a "surprise" autorefill the next day. This is especially appreciated if the machine is on a early morning timer and your household includes light sleepers (BZZZ-Z-Z! as the pump topped off the boiler on startup was not welcome).

[Teme]

On a related note, it's a good idea to introduce fresh water into the boiler before ending a session / shutting down for the day. The side benefit of this practice is less chance of a "surprise" autorefill the next day. This is especially appreciated if the machine is on a early morning timer and your household includes light sleepers (BZZZ-Z-Z! as the pump topped off the boiler on startup was not welcome).

I agree. I nowadays tend to run some water off from the hot water tap before shutting down the machine in the evening. This kicks in the autofill and brings fresh water into the boiler, therefore reducing the level of minerals in it as well as preventing the wake-up buzz from the autofill in the morning. I suppose this is a reasonable approach or is there another means to the same end?

Br,
Teme

[HB]

I suppose this is a reasonable approach or is there another means to the same end?

It works for me. However, putting a momentary switch to disconnect the water level sensor would allow you to overfill the boiler slightly, assuring that no autofills would occur in the next session. Such a switch would also come in handy when descaling (i.e., when overfilling the boiler with descaler to cover the scale line). Essentially the combination of a manual and autofill.

(How long will we have to wait for Ken "Mr. Thermofilter / Omega Datalogger" Fox to post graphs demonstrating the effects of different boiler levels introduced by this suggestion? )

[cannonfodder]

I make a practice of dumping a couple of cups of water out of the water tap every day or two (I have an 11 liter boiler) to keep the mineral buildup to a minimum. As the water vents out in steam, the mineral content of the boiler will increase and get steadily harder.

Do a hardness test on your tap water, then after a month or two of usage (no water purge just steam) dump some boiler water into a pitcher, let it cool and do a hardness test on it. There can be a dramatic jump between tap and boiler.

I occasionally make hot chocolate for the wife or kids and an occasional tea for myself. I like to keep the boiler water fresh, it can get a stale off taste if it gets old.

[miKe mcKoffee]

So in real world situations, the autofill is little more than annoying provided you allow a few extra seconds for the boiler to recover. Any taste difference would be minor and probably not noticed by most people.

That's my take on it, as far as it pertains to me and my Bricoletta.

... the blown shots from the autofill engaging during an extraction is a bit more annoying. It would definitely have a negative impact.

Fortunately that one's not an issue with the Bric' 'cuz autofilling during a shot, especially at the start of the shot, would highly likely cause a sink shot.

I would imagine the size of the boiler would also play a part in how often the autofill kicks on as well as the impact it has. A smaller boiler would be impacted much more severely than a large boiler.

Of course with larger sized boiler it might be expected autofill would also be bringing in more cold water during autofill so recovery might be as much or likely more impacted taking longer to recover than a smaller boiler if the same heating element power. For example let's say one machine has a 3.5L boiler and the other 1.5L and both have 1500w heaters. Let's also say they both autofill when down to 50% filling to 75% full. It would take longer to heat the 0.875L cold water added to the 3.5L boiler than the 0.375L cold in the 1.5L. (Have no idea what typical boiler autofill deadband is, just made up an example) But then since the larger boiler would refill less often so maybe an overall wash recovery impact wise in production. But two boilers of equal size but different power heating elements the lower power would indeed highly likely be more severely impacted.

[cannonfodder]

All things being same, the guy with the most horsepower wins.

[RapidCoffee]

Interesting. I thought that the rotary pump machines had sufficient flow capacity to make it through a mid shot fill. My understanding was that there would be a drop but not as drastic. Learn something new every day.

Yeah I know - but this makes sense, at least in hindsight. The pump setting doesn't change, but during autofill the flow is directed to the boiler as well as the group head. Stands to reason that the group head flow will be reduced, unless the pump somehow "knows" it must push twice as much water. Apparently my pump's not that smart.
________
John

[Ken Fox]

What I do understand at this point is that autofill SUCKs, for a bunch of reasons, especially when a machine is used in a low volume setting and when boiler water is not used for drinks.

So I asked myself if I'm entertaining and making a bunch of back to back caps (snipped)

Worthless post as far as hard data goes maybe, but as far as real world usage goes possibly beneficial.

In my opinion, and this is just my opinion, fine temperature control is more or less irrelevant in milk drinks, so if you are making cappa after cappa after cappa, what the recipients of these drinks are going to notice is how good was your milk frothing, not what temperature the shot of espresso was made at.

I think that temperature management would only be important if you were making some milk drinks and some straight shots; fluctuations in temperature, if large enough, would be noticed in the straight shots that were part of the series.

At this point in my understanding of this phenomenon on my machine with autofill, I know only that it happens. I don't know whether it is due to autofill occuring during shots or simply between them; what I do know is that bypassing autofill reduces variation in inter shot temperatures. It could be that shutting off or modifying when autofill comes on will effect shot temperatures obtained for shots that go into milk drinks also but I think the real impact would be in the straight shots that are part of the series.

ken

[Ken Fox]

It works for me. However, putting a momentary switch to disconnect the water level sensor would allow you to overfill the boiler slightly, assuring that no autofills would occur in the next session. Such a switch would also come in handy when descaling (i.e., when overfilling the boiler with descaler to cover the scale line). Essentially the combination of a manual and autofill.

(How long will we have to wait for Ken "Mr. Thermofilter / Omega Datalogger" Fox to post graphs demonstrating the effects of different boiler levels introduced by this suggestion? )

Of course you wouldn't know how far you were overfilled in a boiler without a sight glass and your slight overfills each day could easily end up with a fully topped up boiler over time. Having had an overfilled boiler for a while after my input solenoid failed, I can tell you that the machine functions poorly in that state.

I have informally tested varying water levels in my vibe machine with a sight glass and at least on my system with its large heat exchanger I did not find big shot temperature stability differences. Nowadays I just fill the boiler up in between the "min" and "max" levels on the sightglass and try to leave it more or less in that position, which works well for frothing as well.

I actually much prefer the Fluke 54-II datalogger, but since I don't have the software for downloading its data, I use the much les fun to use Omega for those tasks.

ken

[miKe mcKoffee]

In my opinion, and this is just my opinion, fine temperature control is more or less irrelevant in milk drinks, so if you are making cappa after cappa after cappa, what the recipients of these drinks are going to notice is how good was your milk frothing, not what temperature the shot of espresso was made at.

I think that temperature management would only be important if you were making some milk drinks and some straight shots; fluctuations in temperature, if large enough, would be noticed in the straight shots that were part of the series.

At this point in my understanding of this phenomenon on my machine with autofill, I know only that it happens. I don't know whether it is due to autofill occuring during shots or simply between them; what I do know is that bypassing autofill reduces variation in inter shot temperatures. It could be that shutting off or modifying when autofill comes on will effect shot temperatures obtained for shots that go into milk drinks also but I think the real impact would be in the straight shots that are part of the series.

ken

I don't disagree. I ran test series of caps' test only because it would force autofill faster than shots only or mixed shots and caps'. In real life entertaining scenarios I'm more often pulling straight shots or mostly straight shots versus all caps.