If you could put a sensor anywhere, and add a valve to put water where you want it, and you could control them all with modern processing, I have a question. How would you design the plumbing and control of the machine from the ground up?

Given all the testing this group has done maybe it is time for the open source espresso project. Just think, a procon pump with a servo adjustment could give you any pressure profile you wanted. A PID with a computer link could be adjusted in real time. The only thing lacking would be a good grinder and operator (self referenced there).

So how would you like the plumbing designed to give the most flexibility for water source, temp, and pressure?

-Cecil

I would be happy with at least an easily adjustable machine. why should I have to take the outer shell off to make simple adjustments to the pressurestat or pump?

I imagine one could design a computer controlled HX machine that could deliver more precise temps with less effort.

It's already been done.

ELEKTRA KAPPA [ and others ] have :
SPECIAL PROGRAMMING - To be carried out solely by the dealer's specialized technician. Carried out using PC + direct connection on electronic control unit and ELEKTRA program. Allows customization of certain machine functions according to important marketing or individual end customer requirements.

cafaIKE wrote:ELEKTRA KAPPA [ and others ] have :
SPECIAL PROGRAMMING - To be carried out solely by the dealer's specialized technician. Carried out using PC + direct connection on electronic control unit and ELEKTRA program. Allows customization of certain machine functions according to important marketing or individual end customer requirements.

That is the point of open source.

Any valve or any variable under the hood available to be adjusted by any home/business user at any time for any reason. Without question. Not "To be carried out solely by the dealer's specialized technician.".

I'm not sure I'd start with an HX machine again, but if you have a lot of time on your hands you can make a flush-free HX that can be programmed for flat or initial-hump profile. Just insert the 2-lb 2-oz reservoir (that's 2 lbs of brass with 2 ounces of water) in the feed pipe, and put some valves in the siphon return, and add some fans, heaters, and controls, and you have a really nice HX.
Here's what the instrumented reservoir could look like:


With basic PID controls (5 in this case, none for the boiler), a test set-up might look like this:


The siphon flow diagram could look like:

The encircled "T's" are the 5 temperature sensors. In this case, T2 is EricS's T/C adapter.
I got away with 30 and 40-watt heaters, and 4-watt fans, using insulation after the last cooling stage.

I suggest putting all controls under a single PLC for best ease of use.

-Ed

Ed, that is totally cool. To heck with holding the boiler stable with a PID controller and praying that the HX loop will mirror the stability you are going straight to the final goal. Just PID the HX loop and let the boiler go bugger off.

-Cecil

CRCasey wrote:If you could put a sensor anywhere, and add a valve to put water where you want it, and you could control them all with modern processing, I have a question. How would you design the plumbing and control of the machine from the ground up?

Given all the testing this group has done maybe it is time for the open source espresso project. Just think, a procon pump with a servo adjustment could give you any pressure profile you wanted. A PID with a computer link could be adjusted in real time. The only thing lacking would be a good grinder and operator (self referenced there).

So how would you like the plumbing designed to give the most flexibility for water source, temp, and pressure?

-Cecil

Pressure profiling pumps have been built and used by a few of us, Andy Schecter being prolly the first, and other systems being built by Jepy, Lino and myself. I'm not sure what you've found on the site so far, but links to posts about mine are

http://www.home-barista.com/tips/...ofiling-t3478.html

http://www.home-barista.com/tips/...pdate-2-t3566.html

http://www.home-barista.com/tips/...pdate-3-t3703.html

I still don't know crap about it from the standpoint of extraction ratios or other simple metrics that could point to some physically different extraction. I continue to use the system. I think it certainly improves consistency and ease of use. I think that the coffee produced on my FrankenLinea is waay better than what I make on my Spaziale S1. The same coffee and water is used in both machines (well same sources, not same molecules - wise ass).

Hope that's useful to you. PM me if you want specific info.

-Greg

Seems like one of the electronically controlled HX machines, like the VBM Domobar Super Automatic, you could automate the flushing sequence. An EricS adapter with a thermocouple instead of thermometer, and a serial bus controller to take the place of the actuator for the group. You could whip up a nice little customizable application with LabWindows or VB (sorry, visual basic). National Instruments thermocouple thermometers are a bit spendy, but you usually get a minimum of four channels for whatever else you want to instrument. And the software can be a little spendy if you don't already have it. The data acquisition rates are plenty fast to be able to hit target temps on the flush down and start of the shot. A flowmeter and you've got adjustable volumetric control.

Of course the coffee bar starts to look like a lab (some of you are already there), and in my case when the wife came home , well it wouldn't be pretty.

decaf_Ed wrote:........Here's what the instrumented reservoir could look like:

Only five PID's??? Whuzzmatta? Run out of funds?

And if you get a jet out of a bottomless portafilter that just happens to hit the PID stack, look out nellie!