Announcement

Collapse
No announcement yet.

Testing Silvias pressure without a gauge

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: Testing Silvias pressure without a gauge

    Originally posted by Lovey link=1143622054/0#14 date=1145576064
    Gday AngerManagement,
    if you decide to ditch the pressurised PF for a non-pressurised number, youll have to rely on the grind and tamp to determine your shot time, not a PF with funny inserts in it :P.

    About the link, its in Robustos post at the top of the thread. Once in the web page, scroll to the bottom and click E models Homepage, no its not a link to a XXX site .
    All the best.
    Had some great luck with and replace the pressurized filter basket with the Krups #0907163. This info was from skybeedoo and Mattyj.

    What was lucky was the assistance I got from the local Sparkys small appliance center at Albion in Brisbane. They have many of the small parts for the home units and while thed cold not find what I wanted when I dropped in, they got back shortly afterwards to me to say they had a pre loved #0907163 filter for $6.

    Now to play.

    Note: As this has now moved OT I will start a new Topic.

    Comment


    • #17
      Re: Testing Silvias pressure without a gauge

      Originally posted by MatthewG link=1143622054/0#9 date=1145364970
      I purchased a portafilter with a pressure gauge (http://www.coffeeparts.com.au/accessories/pfgauge.html) and found my silvia has a pressure of 11-12 bar. Interesting reading on coffee geek (http://www.coffeegeek.com/forums/esp...s/46475?Page=1) about peoples findings of optimal taste at specific bar temperatures. It is unfortunate that the pressure on the silvia cannot be adjusted without taking the machine apart. I did find with a grind, distribution, tamp, and temperature surf that is spot on, some very good shots are still being produced.

      While performing temperature surfing tests I found the temperature stability relatively good - about 1-2 degrees celcius during the shot up to 30 seconds. After 30 seconds the temperature began to rise by a degree or two.

      Does anyone else have any pressure stats on their silvia, done a pressure mod, found a sweet spot, or a link that covers this already?
      Hi Matt,

      I think that youre confusing brew pressure and boiler pressure. What your gauge measures is brew pressure - ie. the pressure of the water going through the puck. This has nothing to do with the temperature, only how hard the pump is working. Ideally, you want it to be around 8-9 bar.

      Boiler pressure is related to brew temperature in heat-exchanger machines. If you go to the resources section of home-barista.com, Espresso Machines 101 will give you a good explanation of how this differs from your silvia.

      Silvia has a thermostat with a very wide band that heats the water to a certain temperature after it falls below another temperature. You can get the temperature that you want by waiting for the machine to cool down after it has finished heating ("temperature surfing"), but youre stuck with the temperature fluctuation in the shot itself. Some people replace the thermostat with a PID controller that can maintain the boiler water temperature at a certain value.

      Unfortunately, cut and paste isnt working on my computer for some reason (!) so I cant post you any direct links, but google is your friend. What an investigation will reveal is that the machines that cost 3x more than the silvia make changing brew temperatures and brew pressures much easier!

      Cheers,

      Luca

      Comment


      • #18
        Re: Testing Silvias pressure without a gauge

        Sorry for taking this thread somewhat off topic.  Luca I am aware of Espresso Machine 101s - maybe not so much the 201s. My post wasnt completely clear - bar temperature should have read brew temperature and I should have kept the temperature and pressure comments seperate.  To be clear I am not referring to HX machines in this post.

        That aside, this thread had me thinking about brew pressure on the Silvia for the first time which is probably why I posted this here.  I have been playing with brew temperature, grind, tamp, etc and have had some success in pulling good shots.  I havent seen much out there on brew pressures on single boiler machines.  I have read the OPV modification for the Silvia, but this will require some surgery and I havent yet opened up my Silvia (yes I have a PID but havent decided whether to install it yet).   When I measured the brew pressure to be between 11 and 12 bar, it seemed a bit high. What I would like to know is whether other Silvia users have measured their brew pressure, and if it is in a similar range.

        Now for temperature.  I have read that the Silvias brew temperatures can range quite broadly due to the thermostat, but I have not seen much on the temperature stability during the shot.  Luca, you mentioned temperature surfing after the heating element goes off.  I have been measuring after the heating element goes on following running water through the steam wand (as the heating element remains on throughout the entire shot). I only found a 1-2 degree difference in brew temperature during the shot (measured by a thermocouple at the bottom of the basket).  Of course my method for measuring brew temperature isnt as good as the scace device, but hopefully the changes in temperature will be accurate enough.

        Ultimately Id be interested in finding out the brew pressure other people have measured on their silvia (using this threads method or through a pressure device).

        Comment


        • #19
          Re: Testing Silvias pressure without a gauge

          Originally posted by robusto link=1143622054/0#0 date=1143622054
          Theres a handy way to establish what the Silvia over-pressure valve is set at, without the use of a gauge. The ulka website http://www.ulka.it/eng/E.htm# says water flow at 9 bar should be 250 ml per minute. Place a blind filter into the portafilter, lift the water return hose from the tank so it will empty into a calibrated container. Turn on the brew switch and time, say, for 20 seconds, noting how much water is pumped into the container in that time. If it works out at 250 mls per minute -- the machine is doing fine, set for ideal extractions.

          Robusto
          Just reporting back after trying this one. My Silvias water flow is 240 - 250 mills per minute, so I guess its OPV is at 9 bar. Thanks again for the tip Robusto.

          - Rob

          Comment


          • #20
            Re: Testing Silvias pressure without a gauge

            Have finally got around to testing my Silvia, and found after 30 seconds around 120 mls, so seems it would also be around 240-250 mls per minute.

            Thanks also for the info Robusto.

            Comment


            • #21
              Re: Testing Silvias pressure without a gauge

              Originally posted by MatthewG link=1143622054/15#17 date=1145660513

              Now for temperature. I have read that the Silvias brew temperatures can range quite broadly due to the thermostat, but I have not seen much on the temperature stability during the shot. Luca, you mentioned temperature surfing after the heating element goes off. I have been measuring after the heating element goes on following running water through the steam wand (as the heating element remains on throughout the entire shot). I only found a 1-2 degree difference in brew temperature during the shot (measured by a thermocouple at the bottom of the basket). Of course my method for measuring brew temperature isnt as good as the scace device, but hopefully the changes in temperature will be accurate enough.

              Ultimately Id be interested in finding out the brew pressure other people have measured on their silvia (using this threads method or through a pressure device).
              Hi, MatthewG -

              I know this response is a little late, but I have a little info to share regarding shot temperatures (apologies, but my data is in F). Depending on shot duration, I have been measuring (avg) 6F-8F temp drops for 20 second shots on fully warmed up PIDd Silvia. If machine is not fully warmed up (i.e. group is still below 186F), drop will be much larger, on the order of 12F.

              But, IIRC, the temp surfing data I read from many years ago showed shot temp rising. Whats up?

              PIDd Silvia will be idling, with heat coil and water nearly in thermal equilibrium, when you pull a shot. The dead time on the Silvia heating system seems to be around 10 seconds (measurable change in boiler temp occurs around 10 seconds after applying full power). So even though the PID begins calling for heat very soon after shot starts, the dead/lag time is so large that effects arent felt during shot. So I think a temperature drop during the shot is inevitable.

              Once the dead time has passed, the heat-up curve for a non-PID Silvia is very steep. If you pull a shot on the heating side of the cycle the coil will still be hot compared to the water and will tend to carry the shot temperature upwards during the 20-30 seconds, partially offsetting the effects of cold water entering the boiler.

              If the above is analysis is even close to correct, then best way to get level shot temps would be to set PID to idling temperature (measured top of boiler) that is a few degrees below some magic temp that is TBD. Hit the steam button, wait 15(?) seconds for temp to rise to the magic temp, then pull shot. Pretty simple procedure, but seems like it would give best of both worlds.

              Anyone tried this who can confirm/deny?

              -- JGG

              Comment


              • #22
                Re: Testing Silvias pressure without a gauge

                There have been many discussions on google alt.coffee regarding temperature at the puck during a shot. The upshot is very stable temperature to a degree or so. Much of the thermodynamics of heating coils, lag, boiler, cold water inrush etc can all be put aside because what keeps the temperature stable is the big chunk of brass metal in the group and portafilter.

                That is obviously a simplistic summary, but in effect, thats what all the Scace et al experminets come down to.

                My PID shows a drop at the top of boiler of around 10 degrees or more during a pour ---- but you can bet your last beans its only a degree down below where it counts during that time.

                Robusto

                Robusto

                Comment


                • #23
                  Re: Testing Silvias pressure without a gauge

                  Originally posted by Lovey link=1143622054/0#11 date=1145529151
                  Gday AngerManagement,
                  the Ulka pump already has a diode inserted, giving it half wave rectification, check out Robustos link and go to the homepage of that site.
                  In this case, why not just reducing the voltage to the pump. That will give it less power and thus lower pressure.

                  Comment


                  • #24
                    Re: Testing Silvias pressure without a gauge

                    Gday tk and welcome to CoffeeSnobs [smiley=thumbsup.gif],

                    Not sure whether youre familiar with the operation of this type of pump and its use of a half-wave rectified DC input, but reducing the voltage beyond a certain minimum point will in all likelihood result in the pump itself just stalling and then provide no output at all. A far better method is to use a Pressure Limiting Valve that bleeds excess pressure back to the reservoir and also serves the purpose of smoothing out the pressure pulses experienced by the coffee puck during a pour. Its a very simple system and when properly adjusted is also very reliable.

                    Mal.

                    Comment


                    • #25
                      Re: Testing Silvias pressure without a gauge

                      Originally posted by Mal link=1143622054/15#23 date=1174203217
                      Gday tk and welcome to CoffeeSnobs [smiley=thumbsup.gif],

                      Not sure whether youre familiar with the operation of this type of pump and its use of a half-wave rectified DC input, but reducing the voltage beyond a certain minimum point will in all likelihood result in the pump itself just stalling and then provide no output at all. A far better method is to use a Pressure Limiting Valve that bleeds excess pressure back to the reservoir and also serves the purpose of smoothing out the pressure pulses experienced by the coffee puck during a pour. Its a very simple system and when properly adjusted is also very reliable.
                      The vibratory pump is a fairly simple contraption. From the diagrams it appears that when current rises a piston is electromagnetically moved against a spring that forces the piston back when current falls.

                      The reason I suggested adjusting the voltage (to control the current) is because it is a lot easier to do this electronically (similarly to the PID). Also, you just made me realize that outside of replacing the pump with a rotary one, driving the pump with a sawtooth (or an inverse sawtooth) could make a huge improvement in pressure stability, especially since the spring resists compression linearly.

                      Comment


                      • #26
                        Re: Testing Silvias pressure without a gauge

                        Originally posted by tk link=1143622054/15#24 date=1174318364
                        especially since the spring resists compression linearly.
                        Quite apart from a range of other factors, a spring coil does not compress/extend in a linear fashion beyond a small proportion of its overall length (as per Hookes Law), it does so exponentially over its total length. Other things to keep in mind is the resonance of the reciprocating mass including the spring. While it might be possible to control the output of the pump to an output pressure basis of control, unless the full stroke of the pump is maintained flow rates will vary enormously with input frequency. It would be an interesting experiment to undertake though so keep in touch with how it turns out,

                        Cheers,
                        Mal.

                        Comment


                        • #27
                          Re: Testing Silvias pressure without a gauge

                          Originally posted by robusto link=1143622054/15#21 date=1148582183
                          There have been many discussions on google alt.coffee regarding temperature at the puck during a shot. The upshot is very stable temperature to a degree or so.  Much of the  thermodynamics of heating coils, lag, boiler, cold water inrush etc can all be put aside because what keeps the temperature stable is the big chunk of brass metal in the group and portafilter.

                          That is obviously a simplistic summary, but in effect, thats what all the Scace et al experminets come down to.

                          My PID shows a drop at the top of boiler of around 10 degrees or more during a pour ---- but you can bet your last beans its only a degree down below where it counts during that time.

                          Robusto

                          Robusto
                          On the botticelli with the watlow 43 kit, I get a drop in the extraction of between 3-4 C. Maybe effect of larger copper boiler? Probe on top of boiler.
                          Brett

                          Comment


                          • #28
                            Re: Testing Silvias pressure without a gauge

                            Originally posted by telemaster link=1143622054/15#26 date=1174385232
                            On the botticelli with the watlow 43 kit, I get a drop in the extraction of between 3-4 C. Maybe effect of larger copper boiler? Probe on top of boiler.
                            When you say "drop in the extraction of between 3-4 C" Brett, do you mean that this is the difference in temperature between the Brew Water and the location where your feedback t/c is located, or is this the gradual change in temperature for the duration of a pour measured on top of the coffee puck?

                            If its the former, then thats ok and a surprisingly small difference compared to a Silvia and what I used to measure with my Mokita.
                            If its the latter, then that is quite unusual..... Never measured this amount of droop/rise during a pour on the Mokita. Worst case I ever measured was a droop of about 1.5C and I think that was because I had the PF out of the machine for too long while while setting up the Fluke and t/c. Typically it was always less than 1.0C.

                            Mal.

                            Comment


                            • #29
                              Re: Testing Silvias pressure without a gauge

                              Originally posted by Mal link=1143622054/15#27 date=1174391791
                              Originally posted by telemaster link=1143622054/15#26 date=1174385232
                              On the botticelli with the watlow 43 kit, I get a drop in the extraction of between 3-4 C. Maybe effect of larger copper boiler? Probe on top of boiler.
                              When you say "drop in the extraction of between 3-4 C" Brett, do you mean that this is the difference in temperature between the Brew Water and the location where your feedback t/c is located, or is this the gradual change in temperature for the duration of a pour measured on top of the coffee puck?

                              If its the former, then thats ok and a surprisingly small difference compared to a Silvia and what I used to measure with my Mokita.
                              If its the latter, then that is quite unusual..... Never measured this amount of droop/rise during a pour on the Mokita. Worst case I ever measured was a droop of about 1.5C and I think that was because I had the PF out of the machine for too long while while setting up the Fluke and t/c. Typically it was always less than 1.0C.

                              Mal.
                              Hi Mal
                              this was the drop in temp during the pour as registered on the readout of the PID. Thermocouple is attached to a washer which goes under the earth connection quick connect on the top of the boiler.
                              Brett

                              Comment


                              • #30
                                Re: Testing Silvias pressure without a gauge

                                I think all this discussion of boiler and shot temperatures is rather too simplistic. Even the effect of the group mass on the temperature is usually overstated in my experience.

                                Heres a few facts based on the thermodynamics and fluid dynamics of a boiler. Ill assume a PIDed boiler for simplicity.

                                1. A PIDed boiler is not in thermal equilibrium, but in what is called a steady-state. That means the input of heat from the element precisely matches the total loss of heat from the hot metal surfaces so that the temperature map inside the boiler remains constant in time.

                                2. The temperature map inside a boiler can show variations of up to 10-12 (or even more) degrees between the top of the boiler and the bottom. Thats because the hot water cools by heating the walls of the boiler. The cool water then sinks to the bottom of the boiler. So the water temperature inside the boiler is by no means constant. You can see that by noting the PID set points are usually greater than 100 oC. That means the water inside is even hotter. Way hotter than recommended brew temperature.

                                3. Most single boiler machines draw water from the top of the boiler, where it is hottest. This water then flows down into the group, which is colder, and is replaced by cold water that is usually introduced into the top of the boiler.

                                4. The degree of mixing of the cold make-up water will often determine how stable the brew temperature is. If there is too much mixing, the brew temperature will fall, while not enough mixing can cause a rise in temperature. This is a very dynamic effect and will depend critically on the boiler design.

                                5. Without any mixing, a slightly falling brew temperature can occur because the water is first taken from the top of the boiler, where it is hottest, depleting the hot layer, which is replaced by cooler water from below.

                                6. In no case is a flat temperature profile stable. It has to be carefully designed into the machine.

                                With this understanding of boiler dynamics, I was able to tailor the brew profile according to how I brewed the shot in my Faema Family. By placing my PID sensor into the group I could determine the group temperature. Then I drew water from near the bottom of the boiler, where it was coolest (near group temperature). So I set the PID temperature to the brew temperature and could dial in the shot temperature pretty close to the set temperature. However, if I operated the machine straight from idle, Id get a shot that started at the set temperature and increased by about 1 oC during the shot, as it progressively drew hotter water from higher up the boiler. If the shot was slow enough, I could flatten this to about 0.5 oC or even less. But the trend was for rising profiles. Also in a series of shots, the group would heat up, increasing the average shot temperature with each successive shot.

                                An alternative method proved to be very repeatable. It involved using the boiler response to a flush. By flushing for about 1-2 sec, I could inject some cold water into the top of the boiler. This water would then drop to the bottom of the boiler, mixing with hotter water on the way down. The PID reading would initially increase, as the flush caused the water pushed through the group to increase (ie as the bottom water near the group is depleted, it has to draw hotter water from above). So Id wait for the cold water to register on the PID reading and when the PID had fallen back to the set point Id pull the shot. This gave me nearly flat profiles to about 0.2-0.3 oC every time and didnt cause the group to heat up from successive shots.

                                This latter mode is very dynamic. It depends on how much cold water you introduce into the boiler. Too little and youd get rising profiles, and too much gives falling profiles and somewhere in between could even give drooping profiles. But shot-to-shot repeatability was about the best Ive ever seen.

                                What Id like to emphasise is that the operation of these machines is not simple. In most cases, the walk up from idle shot will be different to the next shot. If this were not the case, then companies like La Marzocco and Synesso would be wasting their money designing machines that are supposed to be stable under all operational regimes.

                                For home use, most shots are probably walk-up-from-idle shots, so the dynamics of the boiler is moot. The stability of the temperature is also not too relevent as no one knows whats best anyway.

                                Bottom line is that PID control works on these single boiler systems. But not the way most people think it works.

                                Comment

                                Working...
                                X