You guys are so right with the comments about electronics. It's clear to me that 1) just using a pot works well enough to play around and try things out and 2) obviously the big hit with such manual profiling is a lack of repeatability. "Voltage profiling," i.e. recording and playing back a change of voltage over the shot will give you a basic level of repeatability, although there will be variation of actual pressures from shot-to-shot due to variations in the puck's flow resistance. The next step up is controlling pressure x time using the ramp function of a PID or something similar, where the PID changes the voltage to get the desired pressure profile as shot time progresses. This, I believe, is what Greg Scace has developed. Greg has also speculated about an even more interesting setup, which would vary pressure based on flow through the group--this would allow you to actually simulate the action of a spring lever, where the pressure declines only as there is flow out of the group, rather than as time progresses. This would get you a system that reacted to the shot's flow resistance, which might be more interesting than simple pressure x time-based control.

For my part, I'm waiting for Tom to try his second-tier solution and see if that's adequate before I think about doubling my investment with a PID and tranducer.

shadowfax wrote:Ilya, that's a good question, and one that took me a good bit of time on the phone with Greg to wrap my head around. You're right, if the pump can spin slowly enough to produce low pressure at espresso flow, then you can control the pressure in that range. However, I believe you won't be able to control declining pressure very accurately with a configuration like this. Also, you would really have to buy a very low-flow pump to be able to do this at all, which I made the mistake of not doing. I'm under the impression that even the smallest one has a flow rate that's rather too high to do what I'd assume is possible, if a pump with the right basic flow rate existed. I believe Greg's TMFR pump is a much lower-flowing one than mine (the lowest, Greg?), and he definitely found that he needed to drill the bypass out as I've described.

I hope that helps a little. I'm still not entirely clear on why it's necessary; I'm mostly regurgitating what I've learned from Greg's experience.

The simple answer is that all of the rotary pumps can provide 9 bars to the coffee cake while just loafing along, speedwise. The pressure relief valve is not designed to provide progressive pressure increase. It's designed to bypass lots of water with little or no pressure increase. So if you don't drill out the valve, the thing builds full pressure at low RPM. and the sensitivity (RPM/ pressure) is lousy.You improve the sensitivity a LOT by drilling the hole.

-Greg

When these experiments started, I wondered if a good benchmark for a pressure profile was constant flow. What I mean is that the espresso starts with a drip, then over the course of the shot the flow gets stronger, and by the end it can gush. I don't mean cranking up the pressure to get high flow at the beginning, but dropping the pressure in such a manner that the flow stays constant over the final 2/3rd or so of the shot.

I have no justification for doing this other than it would turn out to be simple and elegant if it were true.

shadowfax wrote:He's suggested that a linear pot is best, and that's what I'm using, but at some point I'd like to wire a logarithmic one to see if it mimics the squared (right?) relationship between motor speed and pressure.

That's the $5 option, of course. You can control the system very elegantly with a pressure transducer and a PID controller that has a ramp function and can read the pressure transducer's signal as psi (preferably) and output 0-5V as the control signal. Note, a cheap Auber PID isn't going to do this, and I believe such a controller is likely to cost hundreds of dollars. I'm hoping to upgrade to something like this in the future; if I do I will post about it. I hope that Tom will also chime in with his own ideas about controlling the pump. For now, however, this post will cover the use of a potentiometer for control.

Relationship of motor speed to pressure is linear, or close enuff to it. A decent first crack control option is the Fuji PXR4. You'll need an electronic pressure transducer as well.

-Greg

another_jim wrote:When these experiments started, I wondered if a good benchmark for a pressure profile was constant flow. What I mean is that the espresso starts with a drip, then over the course of the shot the flow gets stronger, and by the end it can gush. I don't mean cranking up the pressure to get high flow at the beginning, but dropping the pressure in such a manner that the flow stays constant over the final 2/3rd or so of the shot.

I have no justification for doing this other than it would turn out to be simple and elegant if it were true.

It's one idea, and one way of varying grind, while maintaining a certain volume of shot, although not necessarily constant extraction ratio. I recently built a system for Phil and Sebastian Coffee Company in Calgary, and they are trying all sorts of things. their thinking is that fines migration gets locked down when the cake gets saturated, so if you gradually saturate,then increase pressure relatively slowly, you'll use a finer grind than if you don't pre-infuse, and ramp the pressure quickly. So you'll get very different extractions. I'm interested, of course, and I'm glad some other folks are doing all of this exploration, since my time is pretty limited. Seems to me that the Extract Mojo system could be of great use in this.

-Greg

gscace wrote:Relationship of motor speed to pressure is linear, or close enuff to it. A decent first crack control option is the Fuji PXR4. You'll need an electronic pressure transducer as well.

Greg, I'm looking at the PXR4 configurations, and as near as I can tell, they don't offer a 0-5VDC control output option, which you would need to have to output to the pump controller. I think the PXG4 is the next model up, and does offer this output option. So as I understand it the PXR4 isn't quite up to the task of controlling the TMFR pump.

For what it's worth, the PXR4 DOES offer the more-standard 4-20mA DC output, and I believe that in the coming months Fluid-O-Tech is updating the TMFR pump line to have an analog board that takes this input rather than the 0-5VDC that is current. That's good for future adopters, because the PXR4's at least $50 cheaper than the PXG4...

Sorry. It's PXG4. I had a middle-aged moment. You can generate a voltage output with a PXR series controller by running the current through a standard resistor, then connecting the Fluidotec pump controller across the + and - leads of the resistor. Remember that V = IR, so if you need 5 V max, the proper resistor is 5 / .02 = 250 Ohms. The real problem with the PXR series is that the update rate is too slow, and the ramp / soak is not sophisticated enuff.

-Greg

Greg, thanks for the info.

I just wanted to mention, I screwed up on ordering the correct pump model. The TMFR really only comes in SS anymore--they've more or less stopped making the brass model, from what Shawn at F-O-T has told me. However, they have several different pump models, ranging from the TMFRSS031 to the TMFRSS201. the 031 model is the slowest-flowing pump, at a basic flow rate of 30 liters per hour, or about a half liter per minute. This corresponds to just about the perfect water debit for an espresso machine, although it's still considerably higher than the flow rate you would expect for a shot, which is more like 150 mL per minute or so, about 8-10 liters per hour. Apparently in my zealous overconfidence, I just ordered the TMFRSS154, which is one of their highest-flowing pumps (150 liters/hour). It wasn't until after I'd bought and received the pump that I talked to greg and he expressed some surprise at the model I'd chosen, and said he'd gotten a much slower-flowing pump. He thought I could overcome this issue by simply drilling a bigger bypass, which is indeed more or less true, but it's arguably a waste and definitely not something you should look for if you're buying one for yourself. In my (mock) defense, I made this stupid error because I was too lazy to do the necessary flow calculations (which is stupidly easy to do), and I just trusted that the 154 model was the one I wanted--that's the model that my Elektra's pump is, which goes to show you how crazy the guys are that select these pumps for espresso machine manufacturers.

[EDIT] As an aside, you may notice that the pump model number ends in 1 in my examples, but the pump I bought ended in 4. The last digit of the model number refers to the bypass valve mechanism. 4 is a balanced bypass, a special design that is intended to minimize the effect of inlet pressure on output pressure when the pump is operating. This is great for a standard rotary pump, but it will NOT work for the drilling application that defeats the bypass spring--in fact, the balanced bypass has a funky plastic shaft with a spring in it. Every time I look at it and try to understand what it does, I get confused. Suffice to say, it's not what you want. Get the standard bypass, which is the model numbers ending in 1 (031, 051, etc.) Thankfully, F-O-T actually doesn't make the SS model with a balanced bypass, so Shawn sent that to me from a brass one (as you can see in the first picture in the thread). Subsequent photos of my installation show the stainless steel standard bypass assembly installed, for comparison. [/EDIT]

In their defense, I suspect that commercial manufacturers choose such crazy-high flow-rate pumps to minimize the pressure drop that you experience if the autofill kicks in during a shot. The higher the flow rate, the smaller that dip ought to be. So they're not totally crazy for doing that, but it can cause users some grief. Heck, I had to install an accumulator in front of my Elektra's pump because it used to cavitate during autofill. I guess I should have taken the hint!

Anyway, this will all end pretty well for me, I think. I'm going to send my pump back to have its liner and pins replaced--it's just those two parts that distinguish a high-flow pump from a low-flow pump. However, if you snag one of these, don't make the mistake I did and grab yourself the 30 or 50 l/h model... I am going for the 50 l/h setup, per Eric Svendson's suggestion.

shadowfax wrote:However, if you snag one of these, don't make the mistake I did and grab yourself the 30 or 50 l/h model.

Oooops, had the same one for a couple of weeks already! Just curious as to why you are changing it now that you have it up and running? Do you think you will get a better level of control on a lower flow model, or it is another reason. I've just got it in the back of my mind it being very noisy at the higher RPMs and thinking a higher flow pump run at lower RPM would be quieter, or is the flow such that even on the lower flow models it still operates at a low RPM despite the 'lost flow' through the drilled bypass valve? (ie I have no idea how much of the advertised flow rate is 'unavailable' due to drilling the bypass)

Shame I can't plug it into the PXG4 I've put in my roaster!

Cheers, Chris

Well, just a couple of potential problems, I guess. At first I thought the pump was rather noisy--as I mentioned it's almost alarmingly loud at 3500 rpm--but after chatting with Greg, it sounds like this is intrinsic to the motor spinning at 3500 rpm, about double what a normal rotary pump spins at (~1750 rpm). So that's not likely a real concern.

On the other hand, when I first installed the system, I had the setup like this: filter -> check valve -> regulator -> pump -> 2nd check valve. I found that when changing the speed, especially doing so fast, can really thrash the needle on the regulator gauge. I was also seeing big jumps of backflow (as rising pressure on the regulator) when the pump turned off. Greg suggested that I reverse the regulator and check valve so that the check valve would absorb this and buffer the regulator. I've done this (as shown in the photos), and it does help, but it still kind of bugs me. The other thing is that, with such a high-flow setup, even though the excess water pumped is probably quite safely spun in a circle, it's still moving that water, and still has to do 5x more work. It strikes me that this is unnecessary and adds extra wear and tear to the parts. I also wonder if there's some part of this I am not grasping that could cause problems; it seems safer to just use a correctly sized pump.

For me, it's probably going to cost about $10 total to get the parts switched out, maybe a bit more, and I was already planning on sending my other bad pump (from the Elektra) into F-O-T to get refurbished (which they do for about $40!), so that's a non-issue for me to piggyback and save on the shipping. At this point I am just curious if it can resolve some of the quirkiness I've seen. If it does, I'll be able to post about that.