japa_fi wrote:With PID and silvia stabilized to 109c, I'm 99% sure the water inside is also 109c or pretty close. After all, the water is surrounded with metallic surface (=boiler) that is 109c.

Fortunately, it's not. There is ABOUT a 10 F delta between the boiler top surface and the water. See here:

http://www.home-barista.com/forum...out-pid-t4691.html

And the reason that it is not is heat loss as the top portion of Silvia's boiler is devoted to steam.

erics wrote:Fortunately, it's not. There is ABOUT a 10 F delta between the boiler top surface and the water. See here:
http://www.home-barista.com/forum...out-pid-t4691.html

erics wrote:The first graph shows the normal variation in boiler water temperature and grouphead temperature over a three hour period using the stock 100 C button thermostat.

The graph in your first image shows that the 100C t-stat turns on at 195F (ish) and off at 255 (again, -ish) of boiler water temp, but doesn't show the relation between that and the surface of the boiler where the t-stat gets its 'data'. And, (if I'm reading the graph correctly) the temperature starts to decrease as soon as the t-stat turns off, and there is no (or very, very little) overshoot.
'Zat pretty accurate interpretation of your data?

The graphs do not show the point at which the thermostat "clicked on/off" but I would say that the element was switched on around 195F (boiler water temp) and switched off at 245-250 (boiler water temp). There is considerable overshoot - I just don't know exactly how much.

Silvia has been returned to her stock PID state and the thermocouple has been removed from the boiler water but maybe I'll edit this later on with typical boiler surface overshoots.

EDIT - I have always had a switch installed such that I can cut power to the heating element. If I take Silvia's boiler surface temp from 195 to 245 (as indicated on the PID), it overshoots to 261, i.e. a click/clack thermostat would do a 16 degree overshoot. So, based on these numbers, I would say this particular Silvia (when in completely stock fashion shuts the heating element off when the boiler water temp is AROUND 240 and the boiler surface temp AROUND 230.

erics wrote:I would say this particular Silvia (when in completely stock fashion shuts the heating element off when the boiler water temp is AROUND 240 and the boiler surface temp AROUND 230.

So, a stock Silvia with a 100C t-stat is turning off when the t-stat/boiler surface is at 230F? Can that be right?

In a Finnish coffee devoted site an user has performed some measurements on his new Silvia (100C tstat).
For temperature measurement K-type thermocoupler was used (0.15 mm) with Fluke Hydra 2620A. Elements were checked against boiling water and found to be accurate. In his measurements (picture) the thermostat turned of at 112C and the boiler temperature kept on rising up for 10c more during the next minute. The TC was attached to the side of the boiler. Picture and some measurements available on this image

If you understand Finnish, background etc info available at http://www.sinijari.fi/linkit/mycoffee/tempera.htm

japa_fi wrote: the thermostat turned of at 112C and the boiler temperature kept on rising up for 10c more during the next minute.

If a 100C turns off at 112C, and we assume that the water is also at 112C as this page claims that it is, what temp does the 110C t-stat turn off at? My Finnish is rusty (I only learned a few curse words form Keke, Mika, Mika, Kimi, and now Heiki,) but I can't tell if the turn on/of points on the graphs on the "this image" page were where the brew t-stat was cycling on, or where the steam switch was thrown to add heat. OTOH, those looked like the more temp stable shots...

The time people could save by just spending a couple hundred more on a machine..

Well, if this were about saving time, I could just get coffee at any of the Starbucks I pass every day. Or I could just quit drinking coffee -- best time savings ever.

Psyd wrote:If a 100C turns off at 112C, and we assume that the water is also at 112C as this page claims that it is, what temp does the 110C t-stat turn off at? My Finnish is rusty (I only learned a few curse words form Keke, Mika, Mika, Kimi, and now Heiki,) but I can't tell if the turn on/of points on the graphs on the "this image" page were where the brew t-stat was cycling on, or where the steam switch was thrown to add heat. OTOH, those looked like the more temp stable shots...

It is not mentioned in the text, but looking trough the other graphs I come to conclusion that those turn on points are controlled by thermostat. The thermostat seems to turn on just below 90c (there is one exception at approx 79c, perhaps measuring error).

I would assume 110C stat goes 10 degrees higher than the 100-tstat, though I have no evidence to back it up.

The discussion is quite interesting and has gone quite analytical. Or what would be the word to use when another person has joined in with calculations on how incoming cold water effects brew temperature? Ie. calculating mass flow, temperature differences, etc. Even created an excel sheet with macros to model up the events. By the limited amount of measurement data his calculations and model are matching quite well.

It has provided some interesting (to me at least) finds. For example, if incoming water from the reservoir is warm, a shot may be ruined. Ie. reservoir water temperature of 40c was measured from Silvia that had been on for long time. Now when brewing and the heating element is on during the brew, water temperature raises during the brew some 4 degrees centigrade. That can ruin otherwise perfectly "surfed" shot. On the other hand if the water in the reservoir is 16c during the brew with heating element on, then the temperature is steady. I still have some doubts about the model as certainly all the brass have some effect on the temperature changes too.

Though I don't have to worry about those as I have PID in my silvia.

Psyd wrote:If a 100C turns off at 112C, and we assume that the water is also at 112C as this page claims that it is, what temp does the 110C t-stat turn off at?

Hello all,
My intention to make the measurements, kindly shown earlier by japa-fi, was to get some insight feeling, how the actual temperature of the water reacts. After reading the many reports, I was somewhat baffled.

I suspected that most of them were made with 110 V Silvia as the comparisons of cycle times to Jim Gallt's measurements were quite different. Then I found out that there indeed is a big difference with powers 800 W / 110 V vs. 1100 / 230 V

The measurements of the idle Silvia shown in the picture showed that the measurement from the top of the boiler were giving the water temperature quite correctly at the end of the cooling cycle, where it took 70-80 seconds for the temperature to change with 1 degC.

But, during the heating stage the temperature with the 230 V model increases with ca. 0.6 degC per second, and the wall of the boiler induces a lag to the temperature. The heat was switched off by the thermostat as the boiler surface temperature was around 112 degC (those values were quite constant, but some other tests have shown also more variance).

After that the boiler surface temperature increased 10 degC during a minute, which I concluded to be due the thermal lag of the boiler wall. The temperature falls at the top level 2 degC, which would mean that the boiler water temperature at the time of switching off would be around two degrees higher than shown on the surface.

The impact of steam hadn't occurred to me, which might change the the situation. However, wouldn't the temperature still be constant, until we would reach to the superheated region?

I do hope I am able to have my web pages on coffee in English as well.

(ed. 12/31/07 Now they are also in English)