Jesse; I use the Smart Espresso Profiler app available on Android/iPhone every day for my espresso and pour-overs. It's actually quite well designed and easy to use, even in the morning when I am half awake. Far better than the apps that come with most scales, including the Acaia Lunar I use with my Flair. And it's free too. I imagine it will soon support the new Decent scale as well if you go that route. (EDIT: FYI you can use the app with only a scale for flow/pour profiling even if you don't use the SEP pressure sensor.)

Peppersass wrote:How do you determine the "flow rate into the puck"? You can't use the scale for that. Without a flow meter, I think the only way to do it is to know how much water it takes to saturate the puck and fill the basket. Then you can calculate the flow rate into the puck from that pre-determined preinfusion water volume and the time it takes for the pressure to start rising, right? If this is how you do it, then how do you know what the preinfusion water volume will be? It varies with dose, grind, basket size, etc.

We used to have a flow meter in the DE1, but we slowly moved to a physics model that tracks pump cycles and pressure, and then calculates the flow rate that way.

Roughly: less flow per pump cycle as pressure increases, and pump manufacturers give you a 2D chart you can calibrate against.

We moved off the flowmeter because they tend to not be accurate under 2ml/s, which is a range espresso usually needs to know about.

However, 2ml/s accuracy is plenty good for preinfusion tracking and so a flow meter will work well to solve the preinfusion-volume question. At 4ml/s, they're typically 95% accurate.

Did I read correctly that you have an Arduino hacked home-built machine? If so, spend a few dollars on an ODE flow meter https://www.ode.it/en/products-hk/produ ... flowmeter/ and hook it up to your water inlet tube.

As an aside, a puck will roughly absorb twice its weight in water. The finer the grind, the more water it will hold. This, btw, is also why very finely ground coffee tends to result in a sloppy puck at shot completion: much more water is held.

-john

decent_espresso wrote:Did I read correctly that you have an Arduino hacked home-built machine? If so, spend a few dollars on an ODE flow meter https://www.ode.it/en/products-hk/produ ... flowmeter/ and hook it up to your water inlet tube.

My machine is Arduino hacked but it's not home-built. It's a GS/3 AV. The Arduino controls a gear pump that replaced the rotary pump/motor. It's also connected to a homebrew interface board I installed in the GS/3 brain box. The board taps into the flow meter, motor on/off, machine on/off and autofill on/off circuits via DC and AC optoisolators. Also has relays for triggering the brew cycle and a bypass solenoid for the needle valve I installed in the brew boiler path for Slayer shots. There's also a pressure transducer on the pump output. The Arduino communicates with an Android app running on a tablet or phone that provides the user interface: sensor readings, pump speed controls, brew start/stop, needle valve bypass control, pressure/flow/weight charts, etc. The Android app also communicates with an Acaia Lunar scale for weight-targeted shots. I can post some screen shots if you're interested.

Bottom line, I can set pump speed and flow rate for pressure and flow profiling. A few things are automated, but it doesn't do profile recording and playback (though it could.) I prefer manual control and didn't want to go down the PID rabbit hole.

The GS/3 flow meter is, indeed, not very accurate at low flow rates because the resolution (ml per revolution) is pretty low. The code can compensate to some extent, but it's not perfect. Nonetheless, I think it's accurate enough for my purposes, which is 1) to set the needle valve to about 3-4 ml/sec for preinfusion (per your recommendation) and to keep the flow rate as constant as I can after peak pressure has been reached.

Our member Assaf did similar modifications to his GS/3 AV (actually more elaborate than my mods and capable of profile recording/playback.) He replaced the GS/3 flowmeter with a higher resolution DIGMESA flowmeter, which helps the accuracy at low flow rates. But I think he still needs some compensation at low flow rates in the code.

The ODE flow meter looks like an easy add-on and it's only $18 on eBay. Do you happen to know the resolution (ml per revolution?) I can find out by registering for an account on their site, but if you know that would save me the trouble.

Peppersass wrote:My machine is Arduino hacked but it's not home-built. It's a GS/3 AV. The Arduino controls a gear pump that replaced the rotary pump/motor. It's also connected to a homebrew interface board I installed in the GS/3 brain box. The board taps into the flow meter, motor on/off, machine on/off and autofill on/off circuits via DC and AC optoisolators. Also has relays for triggering the brew cycle and a bypass solenoid for the needle valve I installed in the brew boiler path for Slayer shots. There's also a pressure transducer on the pump output. The Arduino communicates with an Android app running on a tablet or phone that provides the user interface: sensor readings, pump speed controls, brew start/stop, needle valve bypass control, pressure/flow/weight charts, etc. The Android app also communicates with an Acaia Lunar scale for weight-targeted shots. I can post some screen shots if you're interested.

I'm very happy you don't work for La Marzocco. Your hacked GS/3 sounds awesome!

I used to own a GS/3, and at one point, thought about "Decentising" the GS/3 like you've done, so it might talk to my software. As you've found, it's possible to heavily mod it to make it do some very nice things.

The ODE flowmeter in the GS/3 might even be what is there already, as ODE supplies the valves in the GS/3. I don't think the ODE one, if it is different than the one in your machine, will be any better, as all the plastic flowmeters I've seen have an identical design, so that they are interchangeable. We used to incorporate one of this interchangeable flowmeters, and it was 99% accurate (vs ODE's 95%) through some nicer moulding and plastic, but also didn't handle flows under 2ml/s with much accuracy.

There is a super-accurate swiss made flowmeter, I'm blanking on the name (someone here on HB can chime in) that is a few hundred USD$, that can give you accuracy down to espresso rates. I believe it's what the Dalla Corte Mina uses. This video will likely help you figure it out, and also see how it might integrate into your GS/3, as likely similar connections are used:

PIXIllate wrote:Decent is just releasing their new scale. It's bluetooth and will be compatible with apps that should be able to calculate flow rate for you.

As of right now, there are no 3rd party apps for the Decent Scale, as it's just being released (just the de1app).

Full answer here:
Decent Espresso news

as my original post was deleted from here for being deemed too commercial.

decent_espresso wrote:There is a super-accurate swiss made flowmeter, I'm blanking on the name (someone here on HB can chime in) that is a few hundred USD$, that can give you accuracy down to espresso rates.

Thanks! The video helped. I was able to barely make out the embossed product name -- Nano Brass -- which is by Digmesa. It's only about $85 on their website. There's a version with stainless steel fitting for about $200. Don't know if they ship from Switzerland.

Pretty sure this is the Digmesa unit our member Assaf is using. All of Digmesa's flow meters have the same spec for minimum flow rate, 0.050 ml/sec, or if my math is right 0.83 ml/sec. Should be fine for preinfusion flow rates, BUT...

Accuracy for all models is +/- 2%, but there's an asterisk to a footnote saying "Accuracy in the linear range for individually calibrated equipment". Obviously these flow meters aren't linear, which explains why the flow meters aren't accurate at low flow rates unless you calibrate your software for that range.

Assaf found that this is where resolution matters: the more revolutions per ml, the less difference there is between readings at low and high flow rates. The stock GS/3 Gicar flow meter (and Digmesa's flow meter with the same form factor) are rated around 0.45 ml per revolution. This required Assaf's code to have different calibration factors for preinfusion and higher flow rates. The Digmesa Nano Brass flow meter is rated at 0.024 ml per revolution, about 20 times more spins per unit volume. Assaf's preinfusion and high flow rate calibration factors are the same for the Digmesa.

It's not exactly plug-and-play with the Gicar flow meter -- some tubing work required -- but the fittings are the same. However, I'm pretty sure it'll screw up the GS/3's volumetric calculations. Good thing I never use the (flawed) volumetric dosing feature! OTOH, it should also stop the annoying FLOW METER ALARM messages when I grind really fine and preinfuse at a low flow rate.

Peppersass wrote:Thanks! The video helped. I was able to barely make out the embossed product name -- Nano Brass -- which is by Digmesa. It's only about $85 on their website. There's a version with stainless steel fitting for about $200. Don't know if they ship from Switzerland.

Pretty sure this is the Digmesa unit our member Assaf is using. All of Digmesa's flow meters have the same spec for minimum flow rate, 0.050 ml/sec, or if my math is right 0.83 ml/sec. Should be fine for preinfusion flow rates, BUT..

Yes, that's the one!

We liked that flowmeter. I remember, however, that because its impeller spins so much more than a traditional one, that it does cause drag on the water line, and thus lowered the max flow our pumps were able to achieve. You might need to adjust your GS/3 fluidotech pumps if you put this in line, to push water a bit hard, to compensate.

Easy to do with my gear pump. My software has separate speed settings for brew, pre-brew, free flow and backflush. For most coffees and grinds, I run the gear pump at 17% to achieve 9 BAR peak pressure at the puck. Backflushing against a blind basket at about the same speed, 16%-17%, produces 10.5 BAR boiler pressure, same as the GS/3 stock rotary pump. For free flow, 33% gives me the same flow rate at 9 BAR boiler pressure as the GS/3 stock rotary pump, about 450 ml/min.

So if the flow meter slows down the flow, all I have to do is kick the speed up to compensate. Plenty of headroom.

Sweet setup you've got there! With an over-specced gear pump, I don't think you'll find any issues with that flow meter.