Why PID an HX Machine?

[TimEggers]

...split from Quickmill Anita vs. Alexia vs. Vetrano by moderator...


I see that Chris has a link to a PIDed Anita. I can understand the pros to having a PID on a single boiler machine, but on an HX is it really needed? Doesn't it offset the ability to steam on demand (I know it reduces the need for cooling flushes)? Help me understand the reasoning behind a PID on an HX.

[miKe mcKoffee]

Personally I believe there could be a couple reasons or whys to PID'ing an HX. One could be to simply make boiler temp/pressure changes easier experimenting with different settings for flush versus steam power trade-offs. Also SSR & PID likely more reliable long term than pstat from a failure point of view.

Some have used PID on HX machine to attempt to use their HX machine like a single boiler machine. This may or may not work well depending on the machine and willingness to trade steam power for less flush. However I believe most prosumer HX machines dialed down for almost no flushing won't be as shot temp stable as a single boiler machine like Alexia. HX machines are designed to have their boiler up at steam temp with the boiler heating the HX line! (I don't consider Ken's Junior in the category of most prosumer HX machines }

[Beezer]

I tend to agree that PID'ing an HX is kind of silly. It defeats the purpose of getting an HX in the first place, which is to have the ability to brew and steam back to back. If you PID the boiler and set it for ideal brew temp, it won't steam well. If you set it for steam temp, then you might as well just stick with a stock p-stat and save some money. If you really want that much control over brew temp, get a nice PID'd single boiler like the Alexia, or a double boiler with digital controls like the Brewtus or La Spaz.

In any event, most HX users seem to think that it's not too hard to get decent temp management on their machines by flushing more or less water before pulling a shot. I'm just getting used to my new Anita, but I find flushing to be the least of my problems. Right now, I'm having more trouble getting my distribution and dosing right.

[cafeIKE]

If one plans to PID an HX, one must choose wisely.

A large boiler, heating capacity and massive group are absolute musts.

When I changed from the Solis SL-90, my plan was either dual boiler or an HX PID. In the end, after much data collection and analysis, I chose the Vibiemme Domobar Super Levetta. It's a tad less pro than Ken's Jr, but the project is a complete success. Details here

The machine needs to warm up for 1 hour before use. I made my first sink shot in months the day after the time change because the missus said she changed all the clocks... she forgot the timer on the VBM.

Shot to shot consistency is excellent, better than ±1°C, and no flushing whatsoever is required. I make straight doubles weekdays and lattes for the missus on the weekend. This past weekend we had a dinner party and I made 1 latte, 1 dry cappuccino and 2 macchiato in under 10 minutes, steaming the milk as I pulled each of the 4 double shots. Boiler pressure is a bit on the low side, about 0.7bar, but more than adequate for single drink steaming.

NOTE : This performance may or may not be obtained on other HX machines with smaller boiler, lower heat capacity and less massive group. Choose wisely.

[SLC]

Beezer you just nailed it all in two paragraphs. Well put.

Single boiler, hx, double boiler. They all have there advantages and disadvantages. I the "barista" am the one that has to improve.

With my HX flushing has become a thoughtless task within the routine. Distribution and grind are where I mess up.

Mark

[danblev]

Take a look at http://www.coffeegeek.com/forums/espresso/machines/2272

A HX pstat has a dead band that is usually set to 0.2 bar which translates to 3 deg C. This temperature change affects your coffee. So after getting to know your machine you start temp surfing + flushing routine.

The PID only gives you better control and you need only to work on the flushing routine.

Ken Fox wrote on this topic http://www.home-barista.com/forums/should-i-pid-my-heat-exchanger-t1109.html.
-
Danny

[miKe mcKoffee]

A HX pstat has a dead band that is usually set to 0.2 bar which translates to 3 deg C. This temperature change affects your coffee. So after getting to know your machine you start temp surfing + flushing routine.
Danny

Maybe usually, or rather some pstats have 0.2 bar deadband, some have 0.1bar deadband and Barksdale pstat (which I'm using) has 0.05bar deadband which translates to about 1f.

[podo98]

...merged similar topic with prior thread by moderator...



I have read many of the posts and I am still a bit confused about the advantages of PID'ing a Quickmill Andreja. Can someone explain how the PID improves life over a non-PID'd machine. What are the pros and cons? Does it matter less because the Andreja is a HX machine versus a machine like the Silvia or the Alexia?

Brett

[gscace]

I tend to agree that PID'ing an HX is kind of silly. It defeats the purpose of getting an HX in the first place, which is to have the ability to brew and steam back to back. If you PID the boiler and set it for ideal brew temp, it won't steam well. If you set it for steam temp, then you might as well just stick with a stock p-stat and save some money. If you really want that much control over brew temp, get a nice PID'd single boiler like the Alexia, or a double boiler with digital controls like the Brewtus or La Spaz.

In any event, most HX users seem to think that it's not too hard to get decent temp management on their machines by flushing more or less water before pulling a shot. I'm just getting used to my new Anita, but I find flushing to be the least of my problems. Right now, I'm having more trouble getting my distribution and dosing right.

Hi:

Be careful when you generalize. There is are advantages to PID control of hx machines and they are the same as for double or single boiler machines, namely eliminating temperature oscillation within the deadband and providing a convenient way of raising or lowering brewing temperature. Your third sentence in the first paragraph is just plain wrong. A great example of this is the Simonelli Aurelia.

I would also say that most people think it's not too hard to get decent temp management with hx machines because they haven't had extensive experience with something better. Get your hands on a La Marzocco GB5 or Synesso and spend some time with it and you will hate the water dance BS. Re your sentences 2 and 3 pf paragraph 2, once your technique is dialed in, reproducibility in other aspects of your system will become limiting. I think that once you experience the ease of control that modern industrial controllers give you, you will change your mind.

-Greg

[Ken Fox]

Hi:

Be careful when you generalize. There is are advantages to PID control of hx machines and they are the same as for double or single boiler machines, namely eliminating temperature oscillation within the deadband and providing a convenient way of raising or lowering brewing temperature. Your third sentence in the first paragraph is just plain wrong. A great example of this is the Simonelli Aurelia.

I would also say that most people think it's not too hard to get decent temp management with hx machines because they haven't had extensive experience with something better. Get your hands on a La Marzocco GB5 or Synesso and spend some time with it and you will hate the water dance BS. Re your sentences 2 and 3 pf paragraph 2, once your technique is dialed in, reproducibility in other aspects of your system will become limiting. I think that once you experience the ease of control that modern industrial controllers give you, you will change your mind.

-Greg

What he said.

And, pressurestats by their very nature are inconvenient to adjust. If for no other reason, the ease with which I can change the boiler temperature of my two PID'd machines, is enough to justify the expense and aggravation of PIDing them. P-stats are so inconvenient to adjust in almost all cases, that they are set and forgotten. With a PID I can bias my machines towards the production of milk drinks or straight shots, and just as easily bias it in the other direction for my next drink or set of drinks.

ken

[PaulTheRoaster]

They're also silent and last forever.

[miKe mcKoffee]

They're also silent and last forever.

True PID silent but in application usage may or may not be the case. Depends whether using SSR or mechanical relay. Nothing lasts forever, computers fail all the time.

[PaulTheRoaster]

True PID silent but in application usage may or may not be the case. Depends whether using SSR or mechanical relay. Nothing lasts forever, computers fail all the time.

Yeah, yeah. Also be sure to put some heat-sink compound on the heat-sink on the SSR and mount it such that the heat-sink can do its job.

[cafeIKE]

What Greg said.

What Ken said, except that with planning the aggravation is less than it takes to descale. With all the bits in hand, actual PID installation and boiler insulation took less than two hours. HX and boiler descale is a half day process and much more of a PITA.

If you choose your HX machine wisely, you can have your latte and espresso too.

I'm not obsessively fastidious about ±0.1° temperature variation or perfect foam every time. What I desire is damned fine espresso without a lot of hoopla. The missus desires her latte sans CLACK. The Vibiemme HX with PID delivers on all accounts. I can walk up any time, pull a shot for her latte, steam milk and pull my double with no flushing at all.

For a dinner party, two button presses put me back in full HX mode for frufru production with out the need to partially disassemble the machine, adjust the pStat, wait until the machine stablilizes, adjust the pStat, wait until the machine stabilizes, etc. AND no guest asks, "What's that CLACK?"

Just the lack o' CLACK is reason enough to PID an HX, regardless of the setpoint.

[Ken Fox]

Yeah, yeah. Also be sure to put some heat-sink compound on the heat-sink on the SSR and mount it such that the heat-sink can do its job.

The SSR could certainly fail at some point, but they are readily available and cheap; just check ebay.

If the controller is mounted inside the machine and exposed to heat, all bets would be off as to its longevity, although I know at least one person with such a setup who has had no problems with it.

The SSR does not necessarily need to be mounted onto a "heatsink," per say, if what one means is a thingie called a "heatsink." In some cases, the case material of the espresso machine can serve as the heatsink. This is exactly how I mounted the SSRs in both my machines; facing downward on the bottom casing of my Cimbali Juniors, screwed in and with heatsink compound on the interface. I sanded off the paint on the bottom case, to expose bare metal, at the point of attachment.

I arrived at this solution after considering whether or not I had room for a formal "heatsink" in this location were I to mount the SSR there. So, I called the MFR. of the SSR and spoke with a very knowledgeable tech support person who steered me to online materials showing the needed heat dissipation of the SSR I was using. He felt my machine almost certainly had enough metal on the mounting surface to dissipate the heat, but he suggested that in order to be certain, that I put a thermocouple under the contact point to measure the SSR in actual use, to insure that the bottom of the SSR was not getting too hot, e.g. that it was evacuating its heat appropriately through the case. I did this and found temperatures considerably below what was considered acceptable for this SSR, during periods of maximum SSR usage, such as during the boiler heatup phase and after evacuating water and steam out of the boiler and then refilling it rapidly, calling for nearly constant "on" boiler element conditions.

Of course, a Cimbali Junior is a fairly massive machine and other machines' cases may not have the same heatsinking capability; one would have to test it.

Finally, it goes without saying that the terminals on a SSR are electrically "live." Anyone contemplating an installation such as I have done will need to purchase a plastic SSR top cover, to protect the SSR terminals from water splashing should there be any problems with the machine which could make this a potentially dangerous PID installation. Also, I'd suggest that using a GFCI outlet is more or less mandatory.

ken

[OkcEspresso]

I have a PID on my CMA (Laurentis) single group machine with a 5 liter boiler. Combine that with a PID (for monitoring purposes only) attached to a thermocouple that is mounted inside the group (just above the giculeur) and I have an espresso making platform that is ugly as hell but is verifiably predictable and calibratable. With the thermo in the group, I could probably surf the temp the same way I do with the PID. However, as has already been declared in this thread, adjusting the boiler temp with the PID is a couple of button presses, with the pressurestat it is several screws and cooling the machine down.

[cafeIKE]

Of course, a Cimbali Junior is a fairly massive machine and other machines' cases may not have the same heatsinking capability; one would have to test it.

Stainless is a poor heat conductor relative to aluminum or carbon steel, so choose the SSR location carefully to ensure long life if the SSR mounting location is stainless.

In the initial installation of the Omron G3NA-220B SSR, it was mounted to the stainless steel bottom of the machine. This location was chosen because, in the event the stainless case provided insufficient dissipation, there was room to mount a heatsink on the outside.

During boiler heating, the metal in the SSR would exceed 40°C, the derating corner temperature. Adding the mating Y92B-A100 aluminum heatsink on the outside bottom of the machine reduced the worst case temperature to below 33°C. At operating load, the metal is barely warm to the touch. Note : This location requires milling the heatsink upper surface to mate with the bottom of the machine.


Heatsink Location

Regardless of the location chosen, always make sure there are drip loops, i.e. low points below the terminal, in the wiring to ensure any stray moisture cannot run down the wire into the connection or SSR.


Omron GN3A-220B load graph from Omron technical document D20G3NA0305.pdf

[erics]

The Quickmill Anita (for sure) and probably(?) many other hx machines in its class are already equipped with a 25A SSR. The pstat switches the SSR and the SSR does the job of handling the heating element. Any machine that has a Sirai pstat probably uses it to switch the heating element directly.

I once suggested to someone that in lieu of installing a PID on his hx machine that he simply crank down the pstat to 0.6 Bar territory and see how he likes the results. Well, here are the results on Anita:



I wasn't necessarily shooting for 0.6 bar (but that's where it ended up) rather I was going for ~ 198 F group temperature. Steaming capability is still there at 0.6 bar but obviously not quite the same as 1.2 bar. Machine recovery, back to 198 F group temp, after pulling a "real" shot this AM, was a little longer than desirable - about 20-25 minutes. As others have said, the PID'ed hx does give you a lot of control when trying out different beans/bean blends and most definitely it will provide very repeatable results which I view as a most desirable attribute.

[Ken Fox]

Stainless is a poor heat conductor relative to aluminum or carbon steel, so choose the SSR location carefully to ensure long life if the SSR mounting location is stainless.

stainless is also relatively expensive, and although it is commonly used on outer machine panels, it may not be commonly used in the sort of location (bottom internal casing) as I was describing in my machine. I sincerely doubt that the panel I mounted my SSR on is made of stainless. It was painted black, something you would be unlikely to do with stainless but would do with something like plain steel or possibly aluminum. Also, it was easy to drill through. Stainless is a bitch to drill through, which would give me pause as the installation would likely end up being a PITA.

ken

[LeoZ]

I wasn't necessarily shooting for 0.6 bar (but that's where it ended up) rather I was going for ~ 198 F group temperature. Steaming capability is still there at 0.6 bar but obviously not quite the same as 1.2 bar. Machine recovery, back to 198 F group temp, after pulling a "real" shot this AM, was a little longer than desirable - about 20-25 minutes. As others have said, the PID'ed hx does give you a lot of control when trying out different beans/bean blends and most definitely it will provide very repeatable results which I view as a most desirable attribute.

this sort of relates to my thread that youve been really helpful in. I havent had time to experiment further, but, at 0.7bar, some coffees are great, and some are terrible. HX inconsistency, well, rather, its tight deadband of temp/pressure, seems just too limited for my tastes. for now, ill try to work with it, but long term, like you said above, try a double boiler or so.



anywayyyyy..onto the meat of my post.

you chose your boiler pressure based on a group temp of 198. do you find that group temp ideal to pull shots without a flush, or, just a really short flush?

from what ive seen at 0.7bar, group temp is around 205(ish) and the water appears to start at a fairly low temp, around 155, then creep up to 198ish. im assuming it stabilizes around there, cant recall.

but, again, without a longish flush, youre still going to get temp instability. itll either drop at boiler pressures of 1.1 bar, or raise at 0.6 or 0.7 bar. your chart shows youre still flushing around 4-5oz in those 6 seconds or so until temp stability is maintained, or, are you actually pulling the shot from a cold start, and just letting the first amounts of coffee to be extracted cold?

that brings up another issue, if stability isnt maintained across the entire time of the shot, isnt final outcome affected?

so, start colder and increase, or start hotter and decrease. in the end, does it matter?

i think what im driving at, is that there must be an ideal pressure/temp crossover point where stability between grouphead temps and water temps are close enough that flushes are small and stability is maintained, and, of course, the results tastes good.
question is, what the hell is it!


lots of questions, huh?!