Fellow board members -- I've cared for my now 10-year old Silvia for a while and gone through a fair bit of manual fixes but this latest problem has me stumped. Over the last year or so, I've had to use cleancaf/backflush more often than not (hard water) and have been successful with this other than the three times that have required a quick 3-way valve cleaning that can get cluttered with calcium.

This latest problem started with appearance of weak extraction. The pump would extract well for the first 10 seconds and then seem to choke up right at the moment when you'd expect it to peak at getting the best out of the coffee. The result is a particularly weak extraction (no body/crema to show for). I suspected the OPV so I disassembled it from the boiler in the hope of cleaning it out. Possibly because of the age and calcium build-up, i wasn't able to take apart the OPV assembly at all. The parts just looked melded together. As a consolation, I put the OPV assembly in Cleancaf for 2 hours and reassembled it to the boiler. I since then read about pressure measurements by putting a blank in the portafilter, running the pump and measuring the output and the return flow hose. While I'm certain that the OPV did return some flow before my (seemingly innocuous) cleaning, now running the pump with a blank portafilter returns no flow at all.

I'm pretty certain the reassembly of the OPV is fine. There is *some* water in the return hose but just not enough to even show a trickle. So should I just bite the bullet and try fitting a new V3 OPV or is there something obvious jumping out at someone?

Thanks for your time and attention.

If you use the "search" feature on this site, it will be amazing. Your subject should be "Silvia OPV" and your author could be me, "erics" or you could leave that blank. There are photos and words to keep you up all night.

I suspect that your pump is very, very tired.

Thanks Eric. I had read up on the forums w.r.t. to the OPV but I'm unsure how a tired pump would compare to a malfunctioning OPV. The tired pump hypothesis looks accurate in that the pump on a blind basket can pull ~70ml after a long rest through the return hose but then ~8ml if I try the experiment a second time. However, how do I rule out a malfunctioning OPV? For example, is it possible to have calcium build-up enough to prevent effective compression of the spring to open up the valve? Sure, if it were only an OPV problem I would always get the ~70ml each time.

If the OPV is stuck closed, the pump tries to pump everything through the boiler and on to the coffee. With a fully functioning pump, the pressures would be very high and not relent.

If the OPV is stuck in a partially open position, you would always see flow in the return tube whether you have coffee in place or not.

Operating the pump with an empty portafilter, you should see about 600 ml/minute coming from the group. Regardless of anything else, you need to be able to disassemble the OPV and this is best done with the assembly off the boiler in your particular case, using a vise to hold the assembly. You should also count on replacing the copper washers which seal the OPV to the boiler. They are available in two dimensions to ensure proper orientation of the assembly when fully tightened.

I had verified that at 0 bar the pump can reach ~640ml, which is right on according to your Ulka curve. The curve is linear so the OPV should in theory relieve the pump from the extra resistance from a full portafilter and the pump should have the same performance with and without pressure, right?

Yesterday, I stumbled upon a posted story about someone using a long wrench and vise to take apart the OPV assembly (can't find the link today and my browser history has already too many hits for OPV . Although I would like to identify the problem without swapping out the pump and OPV, I don't have a vise nor a large wrench, so I may just bite the bullet this time. The newer adjustable OPV is a decent upgrade, however. Thanks for your time.

. . . and the pump should have the same performance with and without pressure, right?

No. The pump flow characteristics with pressure are as shown on the curve you mentioned. At 9.0 bar, the pump flow is 260 ml/min, ABOUT half of which passes through the OPV and back to the reservoir. Naturally, the curves and thus flowrates have normal production tolerances.

I've made some progress.

Unable to disassemble the OPV, I decided to give disassemble the pump as shown here: http://ulkapumprepair.blogspot.com/. The relevant part of the pump disassembly is shown in the picture also available from the pump repair link above. .

During the disassembly, I noticed that I had a wood bit lodged at the end of the rubber piece that hits up against the inset about midway. It's about the size of the rubber piece that sits atop the spring. Here's some perspective, I didn't take a picture when it was lodged.

After the reassembly, the return flow shows ~300 ml/min with a blank portafilter (so ~8 BAR from EricS' diagram). I don't know enough about the inner workings of the pump but I think the presence of a foreign object explains the non-linear output of the pump as pressure is applied. Based on my readings on the forum, I would have expected my 10 year old Silvia to show ~11 BAR since apparently the pumps overshot in earlier versions of the Silvia.

Nevertheless, this is a big step and I can enjoy coffee once again.