An idea for simple temperature control in E61 machines

[another_jim]

Sick of flushing the old e61? Of guessing when to stop? The "not with a bang" end to my argument with the WBC measurement standard and my finally tracking down this this elusive TC temperature switch
http://www.dwyer-inst.com/htdocs/temperature/SeriesTCSPrice.cfm (see my post in espresso gear)
has inspired me.

One mounts a solenoid valve on the thermosyphon return line of an E61 (before it gets back to the HX), one also snakes a TC up the pipe so it enters into the chamber in the group and seals it in with a swagelok fitting - it measures the temperature of the group. Then one hooks the TC and solenoid up to the controller. Total parts cost is $100 - $150 depending on the fittings and case.

The operation is like this: The thermosyphon flow heats the group. When the group reaches its target temp, the solenoid valve shuts, the flow ends, and group starts to cool. When it has dropped one or two degrees, the solenoid opens and the group heats up again. If necessary, a fancy solenoid can have adjustable open and closed settings to maintain a minimum flow or restrict the maximum, to slow down the unit's cycle time while maintaining a tight temperature band.

In any case, the net effect is to keep the group, the final temperature control element, at a set temperature ± 1 or 2 degrees. Depending on the size of the HX, this temperature may need to be offset slightly below the desired shot temperature (the unit can be calibrated with this offset). This is a matter for tuning. But the net effect will be to have a ready to go machine after a long idle, not in need of flushing, and set to make shots at the desired temperature.

Not having done it, I have little idea of the absolute accuracy; but I'm thinking that for my practice -- trying to hold a 2C shot window depending on blend -- it'll work just fine.

I'd welcome comments, criticisms, and especially d'ohs before I embark on this.

[AndyS]

Sick of flushing the old e61? Of guessing when to stop? The "not with a bang" end to my argument with the WBC measurement standard and my finally tracking down this this elusive TC temperature switch
http://www.dwyer-inst.com/htdocs/temperature/SeriesTCSPrice.cfm (see my post in espresso gear)
has inspired me.

One mounts a solenoid valve on the thermosyphon return line of an E61 (before it gets back to the HX), one also snakes a TC up the pipe so it enters into the chamber in the group and seals it in with a swagelok fitting - it measures the temperature of the group. Then one hooks the TC and solenoid up to the controller. Total parts cost is $100 - $150 depending on the fittings and case.

The operation is like this: The thermosyphon flow heats the group. When the group reaches its target temp, the solenoid valve shuts, the flow ends, and group starts to cool. When it has dropped one or two degrees, the solenoid opens and the group heats up again. If necessary, a fancy solenoid can have adjustable open and closed settings to maintain a minimum flow or restrict the maximum, to slow down the unit's cycle time while maintaining a tight temperature band.

In any case, the net effect is to keep the group, the final temperature control element, at a set temperature ± 1 or 2 degrees. Depending on the size of the HX, this temperature may need to be offset slightly below the desired shot temperature (the unit can be calibrated with this offset). This is a matter for tuning. But the net effect will be to have a ready to go machine after a long idle, not in need of flushing, and set to make shots at the desired temperature.

Not having done it, I have little idea of the absolute accuracy; but I'm thinking that for my practice -- trying to hold a 2C shot window depending on blend -- it'll work just fine.

I'd welcome comments, criticisms, and especially d'ohs before I embark on this.

Too be honest, since I don't use a HX machine, I've never closely following the thousands of post about group flushing, HX flushing, etc, etc. (This, is spite of the fact that I hail from Flushing, NY). So my comments may be naive:

1. Is it possible that you'll get the group temp on target this way, but still have water in the HX that's highly variable in temperature? And when the water's very hot, will the the cooler group be able to completely compensate?
2. Choose your solenoid valve well. Some have restrictive port sizes and some require minimum differential pressure to open. Of course, a full port modulating valve driven by a PID would be better....
3. When the pump is on, the high pressure water wants to flow towards the group through both the supply line and the return line. Will this reverse flow be a problem for your particular solenoid valve?

[lennoncs]

Too be honest, since I don't use a HX machine, I've never closely following the thousands of post about group flushing, HX flushing, etc, etc. (This, is spite of the fact that I hail from Flushing, NY). So my comments may be naive:

1. Is it possible that you'll get the group temp on target this way, but still have water in the HX that's highly variable in temperature? And when the water's very hot, will the the cooler group be able to completely compensate?
2. Choose your solenoid valve well. Some have restrictive port sizes and some require minimum differential pressure to open. Of course, a full port modulating valve driven by a PID would be better....
3. When the pump is on, the high pressure water wants to flow towards the group through both the supply line and the return line. Will this reverse flow be a problem for your particular solenoid valve?

for #2
I hear you there; most solenoid valves have poor CV ratings

For #3
The pressure will be on both sides of the valve so the only thing you will need to worry about is the "system" pressure rating not the differential.


drawbacks aside; a very clever idea Jim.

Sean

[terryz]

Theory is good, however ± 1-2 degrees is not reasonable for the money.

The better solution is the Mater pressurestat with a tight deadband located in the proper spot. Then add a flush display.

There is an even better solution by way of a device we will be installing on Malachi's Bricoletta (Soon to be referred to as the Bric House!)

It's a cool idea that we have yet to try on a small boiler HX machine but on commercial machines in testing proved to be a reasonable solution to HX lock up.........

More on that in a couple of weeks.


Aloha,

[malachi]

I'd say that there is going to be an "integrated solution" that incorporates low deadband control (PID or Pstat) with some tweaks to the thermosyphon (the "tuning" that many have started playing with), Jim's grouphead control and real-time brew temp display.

I believe this combination could provide for very good inter and intra shot stability and better yet - control.

oh yes... the Bricc House is gonna be sick!!!

[lennoncs]

Ditch the Pstat...they are a PITA

get a dual loop control from Love controls

1st ch for steam via pressure sensor

2nd ch for valve control in the T-syphon loop (perhaps in the passage to the dispersion screen)
the 2nd loop gives you 2 for 1, brew temp readout and control of group temp.

The system works well in my E61. (different but the same in some respects)


sean

[malachi]

Mmm.... interesting idea!
I like it.

[skyryders90]

Ditch the Pstat...they are a PITA

get a dual loop control from Love controls

1st ch for steam via pressure sensor

2nd ch for valve control in the T-syphon loop (perhaps in the passage to the dispersion screen)
the 2nd loop gives you 2 for 1, brew temp readout and control of group temp.

The system works well in my E61. (different but the same in some respects)

Hmm ... care to share some more details and perhaps some photos?

[AndyS]

Ditch the Pstat...they are a PITA

get a dual loop control from Love controls

1st ch for steam via pressure sensor

2nd ch for valve control in the T-syphon loop (perhaps in the passage to the dispersion screen)
the 2nd loop gives you 2 for 1, brew temp readout and control of group temp.

The system works well in my E61. (different but the same in some respects)


sean

Or could do 1st ch for steam via thermcouple. Cheaper sensor. You know, for those of us who don't have a complete set of analog-out pressure sensors on hand.

[another_jim]

I shot Micheal an email on this, since he's the E61 maven. As usual, the devil is in the details for a project like this.

Andy was right about the choice of solenoid requiring care -- apparently these thermosyphons don't have a check valve, and the pump pressurizes both lines, since the return line provides a bit of free tempering of the initial flow out of the HX. So the solenoid valve has to be not fully losing, but maintain a minimum flow.

I plan to get the TC snaked up the return pipe into, or very close to, the group in the fond hope this will provide a good correlate to brew temperature as well as head temperature (oops, social sciences: by correlate I mean that the reading during brewing can be used to estimate brew temperature closely enough for my purposes -- say within 0.5 C). This is not a necessity for the mod, but would be a nice bonus.

I want to stay away from complex controls. I'm hoping for a cheap mod that can be widely copied and that solves the problem, not perfectly, but up to the good standard of this type of machine. Terry's idea sounds even nicer in this respect, but I have my suspicions about how accurate pressure (or did you mean bulb?) stats would be here.

Faema patented a thermostatic valve for this very purpose back in the 60s, but never produced or implemented it. Given the technology of the time, it may not have worked well enough to be any better than a simple manual valve adjusted to keep the group around the desired shot temperature. (a mod on faema machines, including the re-edition, on commercial Giottos, but not on others -- although I still have that unconfirmed rumor the new expobars have something like this). So the idea is anything but new. It just seems with this device, there's now something out there to make it work at a painless price.

[lennoncs]

Hmm ... care to share some more details and perhaps some photos?

I will post as soon as I take some photos this week.

remember; the brewtus is a dual boiler setup, so I use the system to control a circ pump in the thermosyphon loop to bring the group up to a temp much closer to the boiler temp. when I need to drop a degree or two I use the cooling loop of the control to reverse flow on the thermosyphon to drop group temp using the relatively cooler water from the bottom of the boiler. there is not a lot of capacity to really "cool" the group but to drop a degree is not a big issue. the "cooling" feature is more to give me a bit more agility with adjustments and was never intended as anything other than "I could do it ...so why not".


Sean

[Zendel]

There is an even better solution by way of a device we will be installing on Malachi's Bricoletta (Soon to be referred to as the Bric House!)

It's a cool idea that we have yet to try on a small boiler HX machine but on commercial machines in testing proved to be a reasonable solution to HX lock up.........

More on that in a couple of weeks.

Aloha,

How much longer?

[erics]

There is just so much fun in resurrecting an old post - especially when it talks about a group temperature sensor mounted using a Swagelok fitting

Some alternative solutions would be:

Install the Faema E-61 Legend thermosyphon control valve (p/n 2211-135 817-SX OR 2211-135 816-DX) for about $127. Unfortunately, some disassembly, re-soldering, and adapting is required.

Install a Swagelok brass needle valve (properly chosen) for about $30 to control the thermosyphon flow.

Install a fixed orifice teflon disk as Jon did with his Expobar - http://www.home-barista.com/forum...-pulser-t2789.html

Something tells me that the days of the cooling flush are slowly drawing to a close.

[RegulatorJohnson]

hello my name is jon.

Install a fixed orifice teflon disk as Jon did with his Expobar

i have a fixed orifice teflon disc in the thermosyphon in my expobar pulser. i really like it, however without erics TC adapter i would be not so happy. i think it is a vital piece of the puzzle for brewing espresso.

Something tells me that the days of the cooling flush are slowly drawing to a close.

i think the cooling flush is here to stay for me, until brewtus. ahhh someday.

i like the idea of being able to reduce the water to a specific temp. i wish i started flushing at 208° instead of 215°. eric showed me that the temp at the TC adapter is about 4° hotter than at the group. i guess i cold turn down the p-stat but we all know that story, i like to steam milk. i found a good spot with minimal flushes but maximum steam power.

i can choose a temp to brew at and flush to a specific temp on erics tc and usually can get what i want. it like adjusting the temp on the fly, its fun to experiment and see what temp each coffee tastes best at. or i can mess with my wife and make her a really hot or really cold shot and see if she notices, sometimes its an eww thats terrible or sometimes the opposite.

things have really smoothed out for me once i put the restrictor in there. an adjustable one (manual or PID/electronic) would be cool but i think it would be over engineered and simplicity sometimes wins.

dont forget to get yourself an official Eric S thermocouple TC adapter.

fun times. lets have some coffee.

jon

[j7on]

Something tells me that the days of the cooling flush are slowly drawing to a close.

I know, i am drawing up an old post...

What do you mean? Even if you slow the thermosyphon down to a standstill and get the grouphead temp to "perfect" you would still need to cool the thermosyphon loop water with a cooling flush.

[erics]

No, you wouldn't. It is absolutely possible to adjust the pstat of certain machines such that one can simply walk up to the machine and pull a shot as evidenced by the below graph of a Vibiemme Domobar Super.

Note that the US market Vibiemme machines are fitted with a ~2.8 mm diameter restrictor in the upper thermosyphon line. Another idea (unproven) I had is shown by the schematic below which is representative of the Quickmill hx machines. The temperatures are real but the actual construction never took place - those magic stumbling blocks - time, money, space constraints. In this schematic, the solenoid valve would control grouphead temperature by shunting the thermosyphon flow through finned piping in lieu of the group.

Expobar machines (Office, Pulser) can EASILY be retrofitted with a 2.50 mm restrictor in the thermosyphon circuit to dramatically reduce the flush requirements. And, if a 2.50 mm restrictor reduces the flush, maybe a 2.20 mm orifice would eliminate it?