Soft starting a rotary pump?

[boar_d_laze]

Along the lines of hacking a used, commercial, 1 group with PID, etc., has anyone 'round here "improved" pre-infusion by running some sort of soft start on a rotary pump's motor.

I'm ruminating about a line pressure pre-infusion of about 5 seconds, followed by pump start with TVR (time/voltage ramping) of some sort, from 0 to full in about 3 seconds. I've done a little Googling, and it seems solid state control is wildly expensive, and intended for heavier duty applications. This stuff must exist somewhere at a small and inexpensive level. Plus, there have got to be other ways to achieve the same end.

I am so not an engineer, that if I were an engineer I'd suck.

Thoughts?
Rich

[AndyS]

Along the lines of hacking a used, commercial, 1 group with PID, etc., has anyone 'round here "improved" pre-infusion by running some sort of soft start on a rotary pump's motor.

I'm ruminating about a line pressure pre-infusion of about 5 seconds, followed by pump start with TVR (time/voltage ramping) of some sort, from 0 to full in about 3 seconds. I've done a little Googling, and it seems solid state control is wildly expensive, and intended for heavier duty applications. This stuff must exist somewhere at a small and inexpensive level. Plus, there have got to be other ways to achieve the same end.

One of the functions that a gicleur (aka giggler, jet, orifice, flow restrictor, etc) serves in an espresso machine is to provide the soft start about which you "ruminate."

Of course, the gicleur merely acts as a low pass filter, which is much simpler than electronic motor control.

[boar_d_laze]

Without getting into the functions of reservoirs, flow restrictors, etc., assume arguendo:

(1) I will use an E-61 type head complete with gicleur, etc., but want to to blur the transition between pre-infusion and infusion, while maintaining a (reasonably) large orifice in the flow restrictor. Or,

(2) The group of the machine I buy to hack does not employ an E-61 or other group with similarly effective built in pre-infusion. Take a Cimbali D-21, by way of example, and further assume I'm interested in kludging Ken Fox's tweaks to levels previously unknown.

In point of fact, my train of thought left the station with the idea of hacking Nuova Simonelli like "soft" pre-infusion into a Cimbali which was retrofitted with PID, etc., OR what elese I could do to a non-Cimbali big boiler HX to tweak it as long as I was going to control the boiler temp with a proportionally integrated device.

In other words, do you know how to soft start a low power pump motor without spending 3 bills for a TVR switch?

Rich

[AndyS]

In other words, do you know how to soft start a low power pump motor without spending 3 bills for a TVR switch?

Other than powering it through a 300 foot extension cord, I don't.

[gscace]

Other than powering it through a 300 foot extension cord, I don't.

3-phase with variable frequency drive works. Commercial off-the shelf bits. Money!

-Greg

[john_K]

A delay (on) relay will do exactly what you are looking for. If the machine is plumbed to water AND uses a rotary pump.
The delay relay I used allowed for a variable delay. I had it set for ~ 5 seconds. When you start the extraction mains water pressure is applied to the puck until the delay interval is completed at which time the pump starts. Works great!
john

[boar_d_laze]

Thanks all. Love the 300' cord. Been there, done that, got the t-shirt.

"'Delay on' relay" is a great thought I think. Is the idea to let line pressure do the pre-infusion for a few seconds then have the pump come on without any appreciable increasing torque/time ramp at the pump itself? Or, would torque ramp up over time? Or, could that be accomplished with a series? Or?

Fluid-o-tech makes a mag drive pump intended for espresso, inter alia, that runs at multiple speeds. When I get around to it, I'll ask them if they've got a controller that selects increasing speeds sequentially over a given t.

Rich

[DickC]

"'Delay on' relay" is a great thought I think. Is the idea to let line pressure do the pre-infusion for a few seconds then have the pump come on without any appreciable increasing torque/time ramp at the pump itself? Or, would torque ramp up over time? Or, could that be accomplished with a series? Or?
Rich

For a low cost variable speed drive, consider a surplus tradmill motor (DC 130V) and controller. (Not a good fit for in-machine mount, however.) I use one on my 1x42 belt sander, and it works great. Soft start too. Motor and controller both for about $60-70 from SurplusCenter. (Navigate to DC motors and then go for the pdf of catalog page 112.) A somewhat more expensive controller will include pots you can adjust for ramp-up, top speed, etc. You will need to do a little machinist-style hacking to fit the drive to your pump, or hire it done. These folks are also a great source for relays, switches, and many other fun items.

Regards, -Dick

[starry]

This is probably a dumb idea, but I'll present it anyway . Maybe it could be done by pulsing the pump. A programmable electronic timer/relay from Grainger or McMaster will let you program the duration of on AND off, and will repeat or not. The pulses could be some small fraction of a second and spaced a larger fraction of a second apart obtaining the desired low pressure and volume for the preinfusion. It could be done for about $50.

[JonS]

This is probably a dumb idea, but I'll present it anyway . Maybe it could be done by pulsing the pump. A programmable electronic timer/relay from Grainger or McMaster will let you program the duration of on AND off, and will repeat or not. The pulses could be some small fraction of a second and spaced a larger fraction of a second apart obtaining the desired low pressure and volume for the preinfusion. It could be done for about $50.

I don't think this is a dumb idea at all, a basic Pulse Width Modulation scenario, similar to what you see in those variable speed controllers for drills (although there's no actual automatic ramp up on those) I'd had a similar thought myself when reading this thread, but I let it slip by because immediately my mind turned to thoughts of microcontrollers and home-brew circuitry, something I figured would be out of the question for the original poster. If you can get something off the shelf that'll do the trick, then it might well be a winner.

One thing that's not clear to me in respect of a rotary pump used in an espresso machine: as far as I understand it, the pumps are over-specced for espresso brew pressure, and it's the inbuilt bypass valves that effectively limit this to what's required. If your pump is basically magnifying incoming line pressure, and achieves 9 bar with ease, I wonder what degree of variability in brew pressure you could actually get by slowing the motor down. Or rather, what proportion of the total speed range you can apply to the motor will actually provide a noticeable reduction in pressure. And whether it'd actually be worth the extra effort over a basic line pressure fed pre-infusion scenario as provided by a delay on relay ? I guess you'd have to try it to find out

I keep threatening to add a delay on relay to the pump motor in my Iberital, which lacks a pre-infusion strategy. But, with a good grind, and the headspace provided by a smaller dose, I get a good dwell time and I don't have a problem with the way the pours develop. I don't have a problem as yet with how they taste, which would be the other argument for adding pre-infusion, but I'm still learning to get the best out of the machine, which is relatively new to me.

Jon

[cannonfodder]

The LaMarzocco GS3 uses pump pulse preinfusion. On my Faema, I installed a variable time delay on make relay inline with the brew button. When I hit the brew button, the solenoid would open and preinfuse for 6 seconds (IIRC) then the relay would trip the pump. The trick is getting the preinfusion/timing correct. You want to saturate the puck top to bottom side to center before the pump engages. The relays can be had on eBay for just a few dollars. If you decide you do not care for the results, simply remove the relay.

[AndyS]

One thing that's not clear to me in respect of a rotary pump used in an espresso machine: as far as I understand it, the pumps are over-specced for espresso brew pressure, and it's the inbuilt bypass valves that effectively limit this to what's required. If your pump is basically magnifying incoming line pressure, and achieves 9 bar with ease, I wonder what degree of variability in brew pressure you could actually get by slowing the motor down. Or rather, what proportion of the total speed range you can apply to the motor will actually provide a noticeable reduction in pressure.

Read this. (Greg Scace's writeup).

You'll need to bypass most of the pump's outlet back to the inlet (as Greg did). Otherwise, the pump will end up rotating very slowly (to keep the pressure within the proper range) and you will get a gnarly, pulsating flow (as I did <g>).

In any case, there's no way to do this for Rich's $300 limit unless you are very good at trolling eBay.

[boar_d_laze]

I think my question is pretty well answered -- for the time being at least. We'll see what the future brings.

I wish I'd read Greg's thread before I asked the question. In fact, some of the impetus for the question involved an old thread I read on another forum regarding fines loss, solubility, dispersion, etc. At any rate, it got me thinking about a very soft pre-infusion and the effect that would have on puck integrity. Apparently, "asked and answered." The remainder had more to do with the idle thought, "What will Ken Fox do next when he realizes how long it's been since he last tweaked his Junior?" than anything else.

You guys may not realize your celebrity status and the amount of hero worship directed your way. Unfortunately, I had a slow summer and can't afford to pay for autographs.

I dig the "jitterbug" solution. While the ripples could be easily tamed, it begs a bigger question: Why go to a "smooth" pump if only to mimic a rough one requiring downstream remediation. A vibe unit comes with a softer ramp ab necessito, and the noise can be minimized with mounting and sound insulation. Cheap too.

I'm not sure if I could put together the required electronics from scratch or not. Close call. When I said I'm not an engineer, I didn't mean I didn't know which end of the hammer to hold. My education is more theoretical than practical is all. Not to mention decrepit. If your first instinct is to breadboard does it mean your old? On second though, semi-serious circuit design might actually not be that close a call. At least I know Lee deForest didn't really invent the "audion."

If and when I do start this project, or any other requiring a lot of electronic kludging, e-bay will definitely be a source, not to mention Frys, Grainger, and a bunch of other places that know me well.

Rich

[JonS]

Read this. (Greg Scace's writeup).

You'll need to bypass most of the pump's outlet back to the inlet (as Greg did). Otherwise, the pump will end up rotating very slowly (to keep the pressure within the proper range) and you will get a gnarly, pulsating flow (as I did <g>).

I think that answers it nicely Thanks, Andy

If your first instinct is to breadboard does it mean your old?

Not too old, just "of a certain age", I think Or at least I don't consider myself as being old !

Jon

[gscace]

I think my question is pretty well answered -- for the time being at least. We'll see what the future brings.

I wish I'd read Greg's thread before I asked the question. In fact, some of the impetus for the question involved an old thread I read on another forum regarding fines loss, solubility, dispersion, etc. At any rate, it got me thinking about a very soft pre-infusion and the effect that would have on puck integrity. Apparently, "asked and answered." The remainder had more to do with the idle thought, "What will Ken Fox do next when he realizes how long it's been since he last tweaked his Junior?" than anything else.

You guys may not realize your celebrity status and the amount of hero worship directed your way. Unfortunately, I had a slow summer and can't afford to pay for autographs.

I dig the "jitterbug" solution. While the ripples could be easily tamed, it begs a bigger question: Why go to a "smooth" pump if only to mimic a rough one requiring downstream remediation. A vibe unit comes with a softer ramp ab necessito, and the noise can be minimized with mounting and sound insulation. Cheap too.

I'm not sure if I could put together the required electronics from scratch or not. Close call. When I said I'm not an engineer, I didn't mean I didn't know which end of the hammer to hold. My education is more theoretical than practical is all. Not to mention decrepit. If your first instinct is to breadboard does it mean your old? On second though, semi-serious circuit design might actually not be that close a call. At least I know Lee deForest didn't really invent the "audion."

If and when I do start this project, or any other requiring a lot of electronic kludging, e-bay will definitely be a source, not to mention Frys, Grainger, and a bunch of other places that know me well.

Rich

Actually I ain't no spring chicken (but I refuse to act my age). My first reaction when I wanna do something researchy is to see what off the shelf stuff is available. Usually it ain't building the apparatus that's the research project. The research is facillitated by the apparatus.

-Greg