Achieving Optimal Brew/Pump Pressure

[Grant]

I guess I have a couple questions in this area.....

How do you know if your pump pressure during brewing is set too high, and perhaps more importantly or applicably to my situation, too low. Too high I think I am familiar with from my history with Silvia. She was very unforgiving of a poor distribution/tamp, and the higher factory pressure would cause easy channeling etc. I did the copper washer pressure mod to Silvia, so have a little experience in the change to a lower pressure and the results. So, I am guessing that having a bit too high of pressure perhaps is not a taste disaster, as good technique somewhat compensates, and as long as you are high, but "close" it would not throw taste off a huge amount.

But, what happens, or what are the symptoms of too *low* of pressure. Having a (previously owned) Bricoletta now with rotary pump, and reading about input pressure effecting output pressure etc., has me wondering if I should be trying to measure what my pump pressure is, and how to do it (cheaply/easily). Now, my technique has not changed that I am aware of, but the one thing I have noticed with the Bric, is that I almost NEVER get channeling or fast pours....even if I have a noticeable bad tamp (angled etc.) I do not have a bottomless PF, but this is just from watching flow color from the spouts.

I am having some taste issues which I cannot nail down (espresso seems harsh, bitter etc), but I have ran through numerous temperature profiles and tried everything I can think of to no avail...extra long cooling flushes, reducing boiler pressure etc. I was thinking temperature was the problem. The first small amount of espresso from the spouts always seems very dark, not oily...almost cloudy/thick dark and black. This dark espresso has a horrible taste.

So, I started to think about what would happen if the temperature was too low? Is that even possible with a rotary pump machine? It is plumbed in, with pressure regulator set to a static 25psi when the pump is off, and the pressure reg shows the pressure drops to about 16psi when the pump is running (cooling flush - no load).

Any thoughts on this...is there a simple (i.e. cheap) way to measure approximate pump pressure, and would pressure have any relation to the harsh taste issues?

Grant

[malachi]

Yes there is an easy way - buy a portafilter-fitted pressure gauge (available from most if not all of the vendors who support this site).

That being said, I have some feedback to some of your points and one piece of advice for you overall.

Advice... I really doubt that the flavour problem in your espresso is a result of brew pressure. There are so many factors that are more likely to be problems (the beans, the grinder, barista technique, brew temp, water, dirty machine, etc). I would suggest looking at the other factors first as they are far, far more likely. If you can scientifically eliminate all of them then, yes, you want to consider brew pressure.

Some feedback...

1 - Applying lessons learned from the Silvia to the Bricc is problematic. They are very different machines and you cannot assume that what worked (and didn't) on the Silvia will apply, and cannot assume that results and diagnostics from the Silvia are directly applicable either.

2 - Changes in brew pressure of 0.5BAR are generally considered easy to taste. Some people (myself included) claim that changes as little as 0.1BAR are in fact discernible.

3 - The easiest way to identify problems with your pressure tends to be through mouthfeel rather than flavour.

4 - Without a naked portafilter, you cannot know that you do not have distribution problems, uneven extraction, etc. Believe me, a lot of very confident people (myself included) have learned this the hard way.




Finally... it would be helpful if you could describe the "taste issues" you're experiencing in more detail.
You say it tastes "harsh" and "bitter". Does it also taste thin? Does it taste a bit ashy? Oily? Do you get both bitter and sour notes? Is the "harsh" note a bitter harsh? A metallic harsh? What is the aftertaste like?
What coffee are you using?

[Grant]

Wow...quick reply Malachi ....thanks! I'll address what I can now...and have to go home later to re-connect my machine to do some more taste tests to respond to some others. We just had new counters installed, so the machine has been out of action for a week.

Grinder is a doserless Rocky. Burrs are very new and razor sharp - less than 10lbs through the burrs. Grinder is cleaned regularly and there is little residual inside. Same beans/grinder makes incredible shots on the Silvia (but Silvia has not been used in a month or so and has been sold, so I can't go back to do some comparisons). Even my wife (who has lousy taste buds) has noticed the drop in quality from Silvia to the Bricoletta.

Bricoletta is spotless clean. I have done numerous thorough cleanings, detergent backflushes and clean water flushes. I have tasted the water running through the machine numerous times, and it tastes good, clean.

Water entering the Bric is softened and filtered through a 20 micron carbon filter. Again, have tasted it and is clean and crisp.

Beans are fresh roasted, I have my own roasted with my i-Roast (medium milk chocolate in color, just into 2nd crack), and also purchased some other from a local roaster (his are much darker and oilier). Actually this is interesting because out of frustration, I even picked up some starbucks espresso beans to try, and while the crema was no where near as thick or dark or plentiful (actually, it was very pale and thin), the "harsh" tastes I will expand upon later when I do some more taste tests were greatly diminished. i.e. I am perhaps overextracting somehow from the fresh beans, but the older, weaker starbucks beans to a lesser degree if that makes any sense.

Last night, I did a roast of "Espresso Choco" beans from the Green Beanery where I get my beans from....here is their description...."You'll do a double take on this rich blend of superb beans from Yemen and Guatemala, added to a quality Brazilian base. No, we haven't added Lady Godiva chocolate to the brew, just the best beans going. Roasting recommendation : Avoid very dark roasts, to maintain the nuance in this blend."

These have always made a great espresso on the Silvia I have been unable to reproduce on the Bricoletta. I will try some again tonight to taste test.

I have tried various lengths of cooling flushes to try the coffee over a range of temperature settings. Nothing seems to help in the temperature department, so while I am thinking this is not a temperature problem, I may be totally out of whack. I have reduced my boiler temp to bounce between about .95 - 1.15 or so. as I was getting a huge amount of hissing and steaming during cooling flushes (7-8oz's needed). After about 4oz now it settles down at the lower boiler pressure.

I think I will try to convert my single PF to a bottomless and see what I can see from there. I am pretty handy with a dremel, so that should keep my costs down in that department.

In general taste summary, I want to say that the taste seems to have an acrid, combination sour and bitterness. Just horrible tasting. Not really burnt, but I guess the term "ashy" you used might be a good description. Even in milk, the "harshness" still comes through. I will post back more taste notes later.

Grant

[malachi]

I'm going out on a limb here, but my first guess is going to be that this is caused by a combination of brew temp and barista technique.

Some suggestions:

1 - in order to give us a chance to better diagnose the problems, it would be really helpful if you could get a little bit of a commercial blend that we're familiar with. I know it's going to be a pain, but this would eliminate one chunk of guesswork on our part. Maybe Brad from Wicked in Vancouver could set you up with a little Black Cat? If you were able to work with a coffee we've all used, we can not only eliminate this guesswork but give you practical advice for that blend.

2 - if you'd be willing to try the "flush and go" technique to see what the taste result would be, that would really help. I'd suggest, given your machine, that you try building your shot and once tamped etc do as follows: flush until water stops sputtering, count "5 mississippi" (don't rush it), insert portafilter, immediately extract. Taste and record. Repeat with a "4 mississippi" and a "6 mississippi" and then let us know what you get from all of these.

3 - it would also be great if you could describe your shot pulling methodology (sequence of events, measurements including time and volume, etc.).


The naked portafilter conversion is going to be really helpful for diagnosing issues with technique. Let us know what you get.

Finally... I know some people here (Dan) think I'm crazy, but I don't like the stock Bricoletta portafilters at all. If you can source a Marzocco portafilter it is worth doing. I doubt this is in any way related to your issues - but it's something I consider worthwhile in any event.

[HB]

Finally... I know some people here (Dan) think I'm crazy, but I don't like the stock Bricoletta portafilters at all.

To put some context around Chris' comment, see The Fiorenzato Bricoletta - A Pro's Perspective:

The stock basket is fine, but there may be something not right about the stock portafilter. Now, to be fair, I cannot guarantee that most people would care or even notice the difference. And without testing a few other Bricoletta portafilters, it's not clear if this is a design issue or just something about this particular portafilter.

As an editor, I felt his original text about the portafilter speaking to him at night would introduce credibility issues. The paragraph above was our compromise. OK, joking aside... can I assume chopping the stock Bricoletta portafilter to bottomless would excise the demons reported above? Or is the stock basket also suspect?

[malachi]

basket is totally fine (tested it in other portafilters).

[Grant]

I'm going out on a limb here, but my first guess is going to be that this is caused by a combination of brew temp and barista technique.

Stretching it a bit classing me as a barista....but thanks.

Some suggestions:

1 - in order to give us a chance to better diagnose the problems, it would be really helpful if you could get a little bit of a commercial blend that we're familiar with. I know it's going to be a pain, but this would eliminate one chunk of guesswork on our part. Maybe Brad from Wicked in Vancouver could set you up with a little Black Cat? If you were able to work with a coffee we've all used, we can not only eliminate this guesswork but give you practical advice for that blend.

I'll see what I can do here...I will see if I can figure out how to get in touch with Brad. Anyone have an e-mail address?

2 - if you'd be willing to try the "flush and go" technique to see what the taste result would be, that would really help. I'd suggest, given your machine, that you try building your shot and once tamped etc do as follows: flush until water stops sputtering, count "5 mississippi" (don't rush it), insert portafilter, immediately extract. Taste and record. Repeat with a "4 mississippi" and a "6 mississippi" and then let us know what you get from all of these.

This is the technique I use now....I will document some exact wait times and taste results. I read the Bricoletta review on site here about a dozen times before I made the purchase, and I believe that was the recommended.

3 - it would also be great if you could describe your shot pulling methodology (sequence of events, measurements including time and volume, etc.).

Easy to describe....I do the exact same every time.

- Grab one level "scoop" of beans from container (works out to roughly 17g) and dump into Rocky.
- Remove PF from machine and dry basket with towel.
- Grind 1/2 the beans into basket/PF, distribute around a bit, grind remainder of beans in basket/PF including using air tight lid on the Rocky to push through any residual grounds (so there are a minimal amount in there for next time).
- Distribute evenly in basket and level off, and then apply 1 solid/even tamp (Reg Barber 58mm flat bottom steel tamper) to 30lbs (or so) - damn digital bathroom scale.
- Push cooling flush programmed button (about 4.5 oz right now I think)
- Blow/brush remainder of grinds from lip of basket - I usually don't worry about the few loose grounds on inside edge of basket.
- Insert PF, and push "double" button - programmed to extract 2oz including crema. This would be about 4 seconds or so after previous step.
- Watch for color and volume out of spouts, but if I do not interrupt it early, the controller stops at 2oz.

The naked portafilter conversion is going to be really helpful for diagnosing issues with technique. Let us know what you get.

Finally... I know some people here (Dan) think I'm crazy, but I don't like the stock Bricoletta portafilters at all. If you can source a Marzocco portafilter it is worth doing. I doubt this is in any way related to your issues - but it's something I consider worthwhile in any event.

Like I believe is mentioned, I think I will try a bottomless first, which should eliminate any issue with PF I hope, and I can do it myself for free. I never use the single spout anyways except to house my blind basket for cleaning.

[malachi]

OK
I was not aware you were using an auto version of the machine. More clarity here now...

Two things.

1 - flushing by volume is less effective than flushing by time. If you could do me a favor.... after the machine has sat idle, hit flush and time from the end of the water sputtering. Note that time. Repeat when you've pulled a shot or two. Let us know what those two times are. From the time and your boiler pressure setting, we can probably make a reasonable guess on brew temp.

2 - your 2oz extraction... what is the total extraction time on those shots?


I'm going to predict that you're going to see some issues when you use the naked portafilter. You'll probably be able to remedy these by switching from grinding by weight to dosing by volume and by working on distribution. Speaking of which... how do you distribute? Are you using a technique like the Schomer quadrant distribution or the Stockfleths Move or are you doing something less regimented? Also, you'll probably want to experiment with a two-stage tamp (light tamp to preserve distribution, hard tamp with no-weight polish to finish).

[Grant]

First off Malachi....thanks for taking the time to work with me on this....I really appreciate your time and expertise in this.

OK
I was not aware you were using an auto version of the machine. More clarity here now...

Two things.

1 - flushing by volume is less effective than flushing by time. If you could do me a favor.... after the machine has sat idle, hit flush and time from the end of the water sputtering. Note that time. Repeat when you've pulled a shot or two. Let us know what those two times are. From the time and your boiler pressure setting, we can probably make a reasonable guess on brew temp.

I am not completely sure what you mean here. I have programmed my flush volume to end just as the sputtering stops after a long idle time at my current boiler pressure setting - it is currently progammed to flush about 4.5 oz. I have tried lower boiler settings, and then I would also just re-program the button to flush less (for example).

So is what you are asking is for me to do the following?:

- Let machine idle for some time....(let's say 20 minutes)
- Prepare the shot and hit flush button.
- Time from end of water sputtering until flush stops (but this will be zero as I have programmed the button to end right when sputtering stops).
- Prepare another shot, and then when ready hit the flush button.
- Since Hx will not be as overheated as before, the flush will be longer than needed, so record the time from when sputtering stops until flush ends?
- Repeat a couple times and document?

So, is what you are after the time that a programmed flush (in this case, 4.5oz or so), extends past the point where the sputtering stops? Or did I totally not understand your note?

What happened last night after getting things set up again (with my 30 hour old beans), was prepare my shot (machine had idled for a long time), hit the flush button (water dance just settles as the flush ends), counted "5 mississippi", and then pulled my shot. The shot was quite good actually, for the first time in a long time....not great, but good. I still have had better on my Silvia. But lots of rich thick crema and was quite drinkable. No "harsh" or acrid tones that have been common. But a short time later (about 2-3 minutes) I pulled another double to make my wife a cappa, and the horrible taste was back. That time, I manually flushed using the "*" button until the sputtering just stopped, counted "5 missisippi" and then pulled. Shot "looked" OK by appearance, but taste was very off.

2 - your 2oz extraction... what is the total extraction time on those shots?

With my Rocky set to about "8 clicks" above the burr touching point, the 2 oz shot comes out at about 20-25 seconds. If I move my grinder up one notch, it decreases to about 15 seconds or even less if i keep my tamp constant. One tighter on the grinder, and it stalls if I keep my tamp constant. (Yeah I know...time to upgrade grinders for more control).

I'm going to predict that you're going to see some issues when you use the naked portafilter. You'll probably be able to remedy these by switching from grinding by weight to dosing by volume and by working on distribution. Speaking of which... how do you distribute? Are you using a technique like the Schomer quadrant distribution or the Stockfleths Move or are you doing something less regimented? Also, you'll probably want to experiment with a two-stage tamp (light tamp to preserve distribution, hard tamp with no-weight polish to finish).

My son is a real video hound, so I will get him to record some of my shots, and I will post them to my WEB server so you can watch the results. I hope to convert the single spout PF to bottomless this weekend.

I am using a slightly less regimented routine, but more a combination of both I think. I push grinds into the emptier areas near the edges...roughly quadrant based, and then use a circular levelling. I will try some two level tamping as well and see what difference it makes.

Grant

[malachi]

OK.

You need to flush longer.
If you're only flushing until the water stops sputtering, you're probably brewing at a very high temp. The sputtering is from water at or above boiling temp.

Example...

On the Bricc as I had it set up (1.1BAR) - if I were to flush for a full "5 mississippi" count after the water stopped sputtering, I would be at a brew temp of 200F. In general, almost all espressos are best brewed somewhere between 196F and 203F. Based on what you've said, I would guess that your brew temp is in the 205F-208F range.

In other words... you don't flush until the water stops sputtering, stop the flush and then count "5 mississippi" but rather flush for a "5 mississippi" count after the water stops sputtering.

Does that make sense?

[DavidMLewis]

- Grind 1/2 the beans into basket/PF, distribute around a bit, grind remainder of beans in basket/PF including using air tight lid on the Rocky to push through any residual grounds (so there are a minimal amount in there for next time).
- Distribute evenly in basket and level off, and then apply 1 solid/even tamp (Reg Barber 58mm flat bottom steel tamper) to 30lbs (or so) - damn digital bathroom scale.

While I don't think this is your problem, I believe you may be able to improve a bit. In my experience of a similar doserless grinder (NS Grinta, heavily modified), I found it quite hard to achieve a really even distribution grinding right into the portafilter and manually trying to fill in the gaps. I've switched to grinding into a small container (the bottom of a Zass Turkish grinder, as it happens), then putting my hand over it and shaking to fully homogenize the grinds and dissipate the static charge. This also allows me to distribute by tapping with my finger on the container as I move it around the basket. The result, at least in my case, has been much more even and much more repeatable.

Best,
David

[Grant]

OK.

You need to flush longer.
If you're only flushing until the water stops sputtering, you're probably brewing at a very high temp. The sputtering is from water at or above boiling temp.

Example...

On the Bricc as I had it set up (1.1BAR) - if I were to flush for a full "5 mississippi" count after the water stopped sputtering, I would be at a brew temp of 200F. In general, almost all espressos are best brewed somewhere between 196F and 203F. Based on what you've said, I would guess that your brew temp is in the 205F-208F range.

In other words... you don't flush until the water stops sputtering, stop the flush and then count "5 mississippi" but rather flush for a "5 mississippi" count after the water stops sputtering.

Does that make sense?

Aaahhh....that makes more sense. I am totally screwing up the flush! I will try some longer flushes (sputtering PLUS 5 mississippi) and experiment from there.

While I don't think this is your problem, I believe you may be able to improve a bit. In my experience of a similar doserless grinder (NS Grinta, heavily modified), I found it quite hard to achieve a really even distribution grinding right into the portafilter and manually trying to fill in the gaps. I've switched to grinding into a small container (the bottom of a Zass Turkish grinder, as it happens), then putting my hand over it and shaking to fully homogenize the grinds and dissipate the static charge. This also allows me to distribute by tapping with my finger on the container as I move it around the basket. The result, at least in my case, has been much more even and much more repeatable.

I will play a bit in this area as well as while I am sure there is a lot of room for improvment, but when I cut the bottom off the PF, that may become more apparent. I do not think this is my problem either in this case though. Keeping beans, grinder, distribution, and tamp constant across 2 machine - going from a PID Silvia (a very unforgiving machine to bad technique, but incredibly temperature accurate with the PID) where the espresso was EXCELLENT to the Bric, resulted in an immediate degradation in espresso taste/quality. I am 99% sure it is the machine...or more accurately, how I am USING the machine that is the problem...hopefully my resolving the total failure to understand cooling flushes properly will fix all...I miss my chocolate bomb cappa's!

Grant

[malachi]

Cool.

Let us know what transpires.

Assume that you're going to need to experiment with the timing and that you'll be looking at flushing somewhere between "2 mississippi" and "8 mississippi" to get the target temp for a specific batch of coffee.


On a (final) side note.... most espresso is best at between 3 and 5 days out of the roaster. 30 hours is a bit fresh for most coffees if used as espresso.

[CraigA]

1 - in order to give us a chance to better diagnose the problems, it would be really helpful if you could get a little bit of a commercial blend that we're familiar with. I know it's going to be a pain, but this would eliminate one chunk of guesswork on our part. Maybe Brad from Wicked in Vancouver could set you up with a little Black Cat? If you were able to work with a coffee we've all used, we can not only eliminate this guesswork but give you practical advice for that blend.

I'll see what I can do here...I will see if I can figure out how to get in touch with Brad. Anyone have an e-mail address?

Hi Grant,

Brad Ford
Distributor for Canada

Intelligensia Coffee & Tea, Inc.
Canadian Offices
1399 West 7th Avenue.
Vancouver,
B.C.
V6H 1B8

Tel: 604-630-2435
Cell: 604-319-3065
Fax:604-630-2466

http://www.intelligentsiacoffee.com

bford@intelligensiacoffee.removethis.com

The same address as above for Brad's Wicked Cafe too!, & the Tel # there is 604-733-9425.

[Grant]

Just wanted to post an update on what is happening and what I have done in the last week or so.

As a "sub-project", I wanted to address any distribution/tamp concerns so I went out and picked up a bi-metal hole saw that fit perfectly inside my single spout PF. Wow...it cut through so fast and clean I couldn't believe it. 30 seconds with a dremel sanding drum, and about 2 minutes with some steel wool, and I now have a bottomless PF I am quite proud of....not bad for $20. Looks quite "commercial".

End result of the bottomless action is that I am not seeing any "serious" issues in the basket...at least from what I can tell. I seem to be getting a fairly even distribution of espresso across the entire basket on almost all shots. I occasionally get some "flat" spots in the basket, and some early blonding in some areas, but as soon as we can figure out how to get some video, I will try to post some shots for some comments and analysis.

The longer cooling flushes have made a BIG difference in the shot quality and taste. The off flavours have pretty much disappeared, but the repeated "wow" shots I was so used to on the Silvia have yet to appear. But, with some more playing, it is getting there.

I have gotten one e-mail from Brad, and we are arranging to have some coffee shipped over here. I really hope it isn't as good as everyone says it is....I would hate to have to make it a habit. Thanks for the info Craig.


The previous week of "playing" has resulted in a couple questions though....

1) How do you time the 2nd or 3rd shots if you are pulling multiple shots. For example this AM, I get up, prepare everything, pull the cooling flush (up to about 7 seconds now past water dance ending), and and then load and shoot. Then, I dump/clean the basket, re-load and tamp the basket...but then am unsure how long (if at all) I should flush for. If I manually start a flush, I do not see any water dance or hear any boiling/hissing, but am unsure if I should flush a couple ounces anyways? Time to clean the basket, re-load etc. is probably about 45 seconds - 1 minute past the end of the first shot.

2) Bean age...I am thinking that the green beans I have are probably starting to get a bit old as well. I think they are going to have a 1st birthday soon...should I be thinking of tossing them into the compost? Looking at the numerous (hundreds) of threads on bean age and keeping greens, there seems to be no set "rule"...everything from 6 months to 2 years is "OK" depending on who posted the note. Any thoughts on this?

Grant

[malachi]

1 - flush by time not volume. Seriously. if you're finding your coffee tastes best if you flush a seven count from when you hear no sputtering, then always flush the seven count.

2 - it depends on storage conditions. i'd suggest roasting your green in small batches to a light cupping roast and cupping each coffee. see if they've faded or have become damaged.

[Grant]

1 - flush by time not volume. Seriously. if you're finding your coffee tastes best if you flush a seven count from when you hear no sputtering, then always flush the seven count.

Perhaps I am not asking this correctly....perhaps I should ask about rebound times instead. So let's say that from idle, I flush to 7 seconds past the point of the hissing/water dance to get my ideal shot with my current coffee. This is easy to duplicate every time - if the machine has been idle. But when pulling repeated shots, let's say each one every 45 seconds after the initial water dance shot, the machine might not even return to a point where it will hiss and sputter, so won't 7 seconds be too much of a cooling flush in this case?

If I am using the end of the water dance as an reference point for how long to flush, how long do you flush when the machine does not recover long enough to get back to the "overheated" point - i.e. you have no point of reference to start timing.

Looking at the various articles, a 45 second period between shots produces a small temperature hump that can increase the complexity of the shots with at least some coffee (taken from "How I Stopped Worrying and Learned to Love HXs" by Dan Kehn), so if I am pulling repeated shots at 45 seconds, so I even need to do any flush?

Grant

[krazycranes]

that's the beauty of an e61 group head. 9+ lbs. of chrome plated brass has good temperature stability. 1 minute or longer and a short flush 1-2 oz. should be fine. Let the taste be your guide. When the off flavors start coming back, time to flush.

Kevin

[Grant]

Man...I am just starting to think this machine is very temperature unstable....I can do something one day that seems to work, and then the next can't duplicate it, but then something else will work. I am almost positive it is temperature related.

Will a simple styrofoam cup temperature test give me close enough results to see if the machine is out of whack somehow? I was thinking of trying the following - will the below work to give me any close temp reading? (I have a very high quality digital probe thermometer which reads very quickly and accurately as confirmed on my PID Silvia many times).

1) Cut a styrofoam cup short so it holds exactly 2oz before overflowing.
2) Insert probe into cup
3) Insert cup into PF
4) Pull cooling flush (emulating pro method style flush/pull)
5) Immediately put PF/cup directly under group and pull shot until it just overflows (2oz).
6) Read temp.

..and if this will give me a close measurement, what should be the approximate temp in the cup to give me the needed temp in the basket?

Grant

[malachi]

I hate to say this, but it's rarely the machine.

What seems to be the trouble?