AssafL wrote:Just a note - I think you meant the expansion valve.

Correct. I often make that mistake!

AssafL wrote:Also - since the expansion valve will prevent the pressure for exceeding 12bars under all circumstances (given that the throughput of the pump is somewhat limited - at least to the flow that the expansion valve supports) - why is the caution important?

Turns out the expansion valve setting that limits boiler pressure to 12 BAR when the boiler reheats after a shot does not limit a high pump speed from exceeding 12 BAR. I don't know why. Must have something to do with the expansion valve design. Maybe a difference in handling a slow passive pressure increase due to reheating versus an active pressure increase due to a high pump setting?

All I know is that if I set the expansion valve to max out at 12 BAR when the boiler reheats, and then set the pump speed too high, the expansion valve will start whistling and pressure will exceed 12 BAR. Not good. If I set the expansion valve to limit pressure to 12 BAR when the pump runs at high speed, reheat pressure drops way below 12 BAR and idle pressure drops below 6 BAR. Also not good (or at least not right.)

I gave the caution for this reason, and also to point out that in my setup there's only one safeguard against over-pressure. The stock pump bypass valve provides a second level of protection.

AssafL wrote:I am not sure I understand. If the pressure feedback sensor is connected to the brew boiler - what does it matter if the needle valve is after it? If the flow is slow - the gear pump will be slow as well...

I don't have a pressure feedback sensor. My setup is strictly manual. My impression from a Slayer owner was that the gear pump runs at a constant speed, settable by the user but nominally 9 BAR -- even during pre-brew. I've asked that owner how Slayer deals with over pressure when the needle valve is switched in. Waiting for an answer.

EDIT: He got back to me. Slayer does throttle back the pump speed during pre-brew. He thinks it's proportionate to the restriction seen by the pump. They also cap the speed during backflush. I can see how throttling can be done with feedback from a pressure transducer in the group, but it would have to be before the flow restrictor to show any pressure during pre-brew. I thought the pressure transducer came after the flow restrictor so it can read pressure at the puck, which is zero during pre-brew. Maybe zero pressure during pre-brew means reduce pump speed, but the reduction wouldn't be proportionate to the restriction. Maybe they have a second transducer between the pump and flow restrictor or between the pump and needle valve?

In my case it would certainly work to reduce gear pump speed. Similarly, I have to set a pretty low speed when backflushing to keep the boiler pressure the same as it would be with the stock pump, around 10 BAR. But great care would be needed to get this right manually during a shot -- not reducing the speed when the needle valve switches in could have bad consequences. I think it would be better for my setup to simply shut off the pump whenever the needle valve is switched in. Might require a delay to ensure that the motor stops before the needle valve solenoid activates, but that's not rocket science. The downside could be a delay in the ramp up after pre-infusion, but I doubt that would be an issue taste-wise.

Another edit to my previous post. Further discussion with my friend revealed that he was talking about the pressure decrease as the puck resistance decreases, which is what happens when you don't have a bypass valve to set constant pressure. It's more like a lever profile.

We're still discussing how Slayer deals with possibly exceeding the gear pump's max pressure rating when the needle valve is closed way down during pre-brew. It may not be a big problem. I have to pull back on the speed when backflushing against a blind basket, which is the worst case. That speed setting wouldn't be high enough to reach 9 BAR on a typical shot. But perhaps as long as the needle valve is open at least a little the speed setting wouldn't need to be cut back so much.

I may be able to simulate the effect of the Slayer needle valve on the gear pump with the needle valve on a home-made PF pressure gauge I made years ago but haven't used since I modified my Scace to make it into a Scace II.

Peppersass wrote:... The gear pump may be more or less expensive now, and might be available from other vendors (including LM USA -- which I still can't confirm because their parts store website is still down!)...

That LM parts page is not coming back, just FYI. It's dead in the water until they rebuild something new.

TomC wrote:That LM parts page is not coming back, just FYI. It's dead in the water until they rebuild something new.

Yeah, yesterday LM USA told me it would be back in the coming months. Sad. It wasn't the easiest website to use and had some annoying glitches, but it was better than having to email parts orders. That said, it was probably a bear for them to maintain.

LM USA emailed me the current price list in Excel format. I was able to copy/paste what I believe are the necessary parts from the latest GS/3 parts catalog, cross reference them to the prices, and compute the total price tag for converting an AV to an MP: $2,770. .

Much higher than I expected.

This is the cost for all of the parts in the MP that are not also in the AV, which mainly includes the group parts, logic board, button decals, input fixture and exhaust fitting. Without the logic board it would still be $2,140, and would require building an interface circuit between the microswitch and front panel buttons, as suggested by Assaf.

Bear in mind this is the cost for converting an AV to the old MP, not the new Strada-like MP. Since the conversion kit hasn't been released yet, I don't know the kit price or which parts will be included. It's possible using the kit plus any parts not included could be cheaper, but I wouldn't count on it. I don't believe there would be any savings in converting an AV to an MP with the old Strada mod, and likely it would be more expensive.

FWIW, it took me a while to figure out how they get the leads from the micro switch out of the group and into the machine. It's not shown in the parts catalog diagrams and there are very few photos on the Internet showing that part of the machine. Turns out there's a small metal tube leading from the dry part of the group along the goose neck and into the lower half of the front panel. I can't see how the tube exits the group, but hopefully it's through the plastic ring and not the metal cylinder that's common to both the AV and MP and would have to be drilled for the tube. In some photos it looks like the tube enters the front panel in the gap that's usually covered by the square gasket, but in others it looks like there's a hole drilled in the panel below the gasket. The part number for the lower front panel is the same for the AV and MP, so if the tube enters via a hole I have to assume that the part is missing from the catalog and price list or the hole is drilled at the factory and perhaps by LM USA when shipping a replacement. In any case, if a hole is required it would be better to drill in the AV panel instead of buying a new front panel. But I can imagine some difficulty getting the location of the hole exactly right for the tube unless it's bendable. I didn't see a rubber grommet or similar gasket to prevent the tube from abrading against the hole.

In thinking this through, I would want to examine the inside of an MP, or at least see some good photos, before undertaking such a conversion. There's a fair amount of documentation on the group head assembly, but the diagrams in the parts catalog don't show how the micro switch wire is routed through the machine to the logic board (the wires aren't in the parts catalog and would have to be spec'ed and sourced separately.) Also, the connection between the new Y-fitting for the input water to the heat exhanger shows the wrong kind of fitting, so it's not clear to me exactly how the connection is made to the heat exchanger in the MP.

Even though I already own a GS/3 AV, the cost of conversion is way too high. I could probably sell my GS/3 and buy a new MP for roughly the same money or less. A better option for me is to upgrade my gear pump profiling to incorporate a switchable needle valve to extend pre-infusion time beyond the 10-15 seconds I can do now.

AssafL wrote:That would also explain why LM had to use the vent spout for the 3 way in the MP and EP versions.

Not sure what you mean. I don't know the exact configuration of the valve, but in addition to using the exhaust to vent pressure after pulling a shot, I think they use the vent to release excess pressure when the brew valve is partially open. Otherwise, the low flow would cause superheating of the input water in the heat exchanger, causing loud banging noises when it hits the cooler boiler water. As you know, I get those sounds when I reduce the flow below about 3 BAR for pre-infusion.

I believe this is why the AV pump comes on during "pre-infusion" and the MP pump comes on almost immediately after pushing the paddle to the left. The higher flow provided by the pump is required to avoid superheating, but also must be vented when flow is restricted at the puck in the MP.

On protecting the gear pump, further thought leads me to believe a bypass or expansion valve or OPV between the pump and needle valve may not be necessary. It's possible that as long as the needle valve is open, even a little, the gear pump may not be in danger as long as the speed is capped. It's certainly capped during backflushing. I have to do that, too, with my gear pump. Turns out the pump speed I have to use for backflushing at 9-10 BAR isn't dramatically different from the speed to reach 9 BAR when pulling a shot. It's definitely too slow to reach 9 BAR with a puck, but it's the setting for a blind basket. With the needle valve open, the max speed would increase and might be fairly close to what's needed for a regular shot at 9 BAR. I plan to test this with the needle valve on a crude PF pressure gauge I built years ago before modifying my Scace to a Scace II.

AssafL wrote:I still don't understand how they control the 9bar during the pour. I think they have a pressure sensor on the boiler like the Stradas do but am not sure of it.

Do you mean the MP? No feedback loop on that machine.

If you mean the Slayer, they don't need a pressure sensor to control 9 BAR during the pour. The pump runs at a constant speed set by the user. That's nominally 9 BAR at the puck when max pressure is reached. It runs at the constant speed while the needle valve is in the loop, and continues running at that speed when max pressure is hit and pre-brew ends (which I think can be manually or by a timer.)

The pressure slowly declines as the puck resistance decreases because they don't use a mechanical bypass on the pump and there's no feedback loop to keep the pressure constant. My gear pump does the same thing -- it drops to about 7-8 BAR by the end of the shot. Automatic ramp down. Evidently, some people love this idea because it's similar to a lever, and some hate it.

However, I was also told that the pump speed appears to be capped when you backflush, so they might have a pressure transducer on the boiler to detect a rise above either the max pressure set by the user or some other level they consider the max, in which case they throttle back the speed. I have to lower the speed on my gear pump when backflushing in order to hit the same max pressure at the boiler as the rotary pump, about 10 BAR.