Bleeding pressurestat lever machines. (Especially 2nd gen La Pavonis)

A haven dedicated to manual espresso machine aficionados.
User avatar
rpavlis
Posts: 1799
Joined: 12 years ago

#1: Post by rpavlis »

I hate to make posts that sound like lectures for a chemistry laboratory project, but, in reality, making espresso is really a laboratory project!

I gradually came to the conclusion that many temperature control problems, especially with pressurestat equipped machines, stem from incomplete bleeding. These problems are especially acute with second generation (1974-2000) La Pavoni machines because air that is trapped in the space above the water in the boiler gets quickly swept into the space above the piston in the group, and it stays there until the next shot when it gets forced back into the boiler. The temperature of this space is thus pcooler with incomplete bleeding because the total pressure has to be the same as the boiler, but the partal pressure of a lot of air makes this area--the top of the group--cooler. (This reduces the partial pressure of water vapour, and hence temperature.)

If it were to retain the same partial pressure all the time, that would be really a good thing; however as the boiler liquid volume falls, the total amount of non-liquid volume increases, and this results in lower partial pressure for air through the whole system including the group. Thus the group temperature rises as more shots are made from this cause as well as from conductance of heat through the flange. When the system is thoroughly bled, air is absent, and the temperature in the top of the group on these machines is identical to boiler temperature.

Another assumption that is often made with all pressurestat equipped machines that have a vacuum release valve is that this valve also provides complete bleeding! It most certainly does NOT. The valve closes when water vapour begins to condense rapidly in the valve, this is NOT when all air has been eliminated. It is a PARTIAL bleed.

Incomplete bleeding will have consequences for other machines that are probably less acute.

HOW TO EVALUATE COMPLETENESS OF BLEEDING:

This can be done with any type of espresso machine with a steam valve:

Obtain a tall jar like the ones frequently used to sell olives and similar items. It needs to be about 15cm tall and 5 or 6 cm in diameter.

Fill the jar half to two thirds full of water. When the machine is up to pressure and the pressurestat turns off the heating element put the jar under the steam tip so that the steam tip is at least 3 cm below the surface fo the water. Open the steam valve. Any air in the system will form bubbles that reach the surface of the water. Water vapour will completely condense a short distance beyond the steam tip. If you leave the steam valve open, the air will be eliminated in perhaps five seconds, and bubbles will no longer reach the surface.

Now raise the group handle! With many group designs the air in it will be at least partly forced back into the boiler. A flood of air bubbles will appear in the water in the jar! If you raise and lower the handle a couple of time air should no longer appear when you do this.

With second generation La Pavoni machines with pressurestats thorough group bleeding will prevent erratic temperature problems. It is best first to raise the group handle just short of enough to release steam and water through the group, and then hold the handle in that position during the bleed process. After perhaps 5 seconds of bleeding, raise the handle momentarily fully, and then drop it just enough to stop water emission. Do this two or three times and the whole system will be purged. One can follow this with the olive jar technique. Once you know what it takes to give a complete bleed, you can simply put an empty container over the end of the steam tip during the bleed.

I have used the "olive jar technique" on a very recent Micro Casa a Leva, and even though it has a vacuum relief valve, there is still quite a lot of air in the space above the liquid water unless it be bled. However, the design of these machines makes thorough bleeding less critical.

Note that older La Pavoni Europiccolas and other machines with two heating elements tend to constantly vent steam, and any air in the boiler gets swept out. With these machines when group handle use pushes air into the boiler, it too gets swept out. However, group handles still need to be raised and lowered to force air into the boiler so that it can be swept out!

User avatar
LDT
Posts: 242
Joined: 10 years ago

#2: Post by LDT »

Robert, I tried your procedure on my new (December 2014) Elektra Mcal. I did get bubbles that soon disappeared as you indicated, but leaving the steam wand on and lowering the lever did nothing, at least visible at the steam wand. Perhaps this is normal?

Advertisement
User avatar
rpavlis (original poster)
Posts: 1799
Joined: 12 years ago

#3: Post by rpavlis (original poster) »

The MCAL does not have the space above or around the piston under steam pressure like the La Pavonis do. In fact there is a small hole that allows air to be in this space. The MCAL group thus does not have to be bled for transfer of heat to the group. Heat is transferred to the group by conduction, that is why the metal between the group and the boiler is so thick. That is also why the boilers on these typically is kept at a somewhat higher temperature than many others.

Retained air in the space above the boiler, of course, lowers the temperature of the boiler a bit. I have An MCAL from the end of 2013, a year older. If I bleed the boiler through the steam tip on it the temperature typically goes up a bit over 4 Celsius degrees, which is significant. My boiler pressure gauge reads 1.2 bar when it is up to pressure. (About 2.2 bar absolute, corresponds to 396.5K or about 123.C if no air be in system. However the temperature of my boiler without further bleeding is about 4.5 degrees lower, or 118.8C. This corresponds to 1.91 bars vapour pressure. Thus nearly 0.3 bars are due to retained air, or about 14% of the vapour is still air when only the bleed from the vacuum relief is used.)

The thermal arrangement here is very different than any of the La Pavoni lever types or Olympia Creminas.

If you steam milk it is best to steam the milk first, as this gets the air out of the boiler, and gets the group up to temperature faster. I normally do not leave mine on for long periods, so I normally bleed for about 5 seconds through the steam wand after it gets up to pressure unless I steam milk.

(Another thing MCAL owners need to know is that the open piston on top means this area is dry. Salts in water that get by the seals will thus evaporate and concentrate so that they can do more damage. Chloride is bad in any espresso machine, but I think much much worse in an MCAL. I normally use pure water with 100 mg potassium bicarbonate per litre instead of any bottled or tap water. That results in about the least corrosive pH, and it also reduces the tendency of many coffees to produce sour espresso by its buffering action.)

User avatar
baldheadracing
Team HB
Posts: 6225
Joined: 9 years ago

#4: Post by baldheadracing »

rpavlis wrote:I hate to make posts that sound like lectures for a chemistry laboratory project, but, in reality, making espresso is really a laboratory project!
I love these posts! :mrgreen:
rpavlis wrote:... Fill the jar half to two thirds full of water. When the machine is up to pressure and the pressurestat turns off the heating element put the jar under the steam tip so that the steam tip is at least 3 cm below the surface fo the water. Open the steam valve. Any air in the system will form bubbles that reach the surface of the water. Water vapour will completely condense a short distance beyond the steam tip. If you leave the steam valve open, the air will be eliminated in perhaps five seconds, and bubbles will no longer reach the surface...
Doh! :idea:

Like most things brilliant, so obvious once you read it 8) .

I always wondered how much 'false pressure' I should be bleeding off. (My MCaL is an older model with no vacuum breaker). Now I know!

Thanks!
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada

day
Posts: 1316
Joined: 9 years ago

#5: Post by day »

Will try it when u get back home and see:)
Yes, i you per this on an iPhone

BramJ
Posts: 9
Joined: 13 years ago

#6: Post by BramJ »

Nice post!
This explains some diffirences in temperature I have been noticing lately with temperature probes (one on the outside of the boiler, one of the group)

day
Posts: 1316
Joined: 9 years ago

#7: Post by day »

Alright. So, I tried this...and I am totally blown away :) I will post here for others that might have similar problems:

1. Temperature. Before I was pulling at 86-90 C on my External Probe and was able to get the right temperature at puck (via both taste and a drilled portafilter with probe. I havent gotten to internal probe in to check yet, but at the same external temperature reading it is clearly a bit undertemperature. It doesnt exactly taste bad, but a slightly sour note and noticeably cooler at the same External probe temperature. Cant wait to recalibrate my pull timing and see what I can do now. Not sure how that works, but I will have to revise my temperature calculations by several C.

2. I was having a big problem with pulling correct timing and weight. Most of my shots looked like a blond gusher very early into the shot-starting with a thin tail but very quickly opening up into a funnel. Even some quite drinkable shots, I was constantly dealing with that funnel and had tried every remedy I could read and was about ready to post a video asking for help. That is now fixed. Before, as a result of that, it was really hard to diagnose my distribution and tamp because the cone was constantly too big, but now I can clearly see the impacts of my distribution and tamp, and the pull is actually readable with excellent timing and mouse tail/etc.

3. Not infrequently the crema was quite bubbly even when clearly not burned/boiling. That is now fixed.

4. Before using this technique to purge, I was never, not once, able to get the grind to give me both the proper weight and timing, and also start to have dripping on the edges during the preinfusion. I was leaving it up for 10 seconds and never getting any edge dripping, unless the grind was far too coarse. Now, every single shot, when dialed in correctly, is giving me all three-some dripping forming long the edge at about 5-6 seconds preinfusion, proper weight and timing.

There is some variation in how much air has to be purged, but lifting the handle and purging has made the second greatest impact on my ability to use the lever machine (a distant second to fixing some kind of alignment issue with my grinder).

Thank you for the excellent explanation :)
Yes, i you per this on an iPhone

Advertisement
User avatar
drgary
Team HB
Posts: 14347
Joined: 14 years ago

#8: Post by drgary »

I've experimented with a Cremina and a 1st generation Europiccola modified so that it has the 3rd generation heating element and a pressurestat. The overpressure valve on both are set so they don't constantly vent steam when the machine is up to pressure. I found that there's no spongy pull with the Cremina when it's had time to warm up and the lever has been raised for a sufficient time that droplets are landing in an even distribution across the bottom of the cup (using a bottomless portafilter). The same is true for the 1st generation Europiccola, and I expect it would be true for ones with the dual heating system.

I haven't tried this with a 2nd generation La Pavoni because I don't have one. I will post here about whether this works with a 3rd gen machine in a few minutes.

Added: A little more experimenting to do, but it's promising. I raised the lever for a long preinfusion. No droplets hit the bottom of the cup, but the pressure is set lower on my 3rd gen machine. A couple of pumps yielded droplets and the first pull without any purging was much less spongy than usual. I believe I choked the machine so air couldn't easily exit the coffee puck and be replaced by water. I will try later with less coffee, a cold start and a long preinfusion using a bottomless portafilter for feedback about even distribution of coffee droplets. (A stock portafilter should be sufficient if the droplets start, pick up speed, and then get slower as the coffee cake gets saturated.)
Gary
LMWDP#308

What I WOULD do for a good cup of coffee!

User avatar
drgary
Team HB
Posts: 14347
Joined: 14 years ago

#9: Post by drgary »

Reducing the coffee cake slightly so that droplets came through did not entirely eliminate the spongy pull on my 3rd gen machine on the shot from cold start or on the second shot; neither did using Robert's purging technique. Thus shot size was slightly smaller and the feel of the pull wasn't as pleasing, but shot quality was fine. The variable of air trapped in the group makes it harder to control shots in 3rd gen machines, although mine is tuned with the pressurestat turning on at 0.8 and off at 1.0 bar and with a heat break gasket, so I rarely get burned shots.
Gary
LMWDP#308

What I WOULD do for a good cup of coffee!

wkmok1
Posts: 272
Joined: 10 years ago

#10: Post by wkmok1 »

I am stumped as to why my second shot is less spongey. I bleed false pressure and practice late lock-in (after letting about a tablespoon of water out). I can't see having some residual air on top of the piston being the cause either. That trapped air may make group heating less effective. But, I have a thermometer strapped to the group and use half-pumps to reach my target group temp (lately 90C for Coava David Mancia). I use a cold PF to cool the group for the second shot. First and second pulls should have the same initial conditions. I have tried grinding 2 doses and pulling the second one first. No effect.

Will try keeping the steam valve open while lifting the lever tomorrow.
Winston

Post Reply