Question: How can I most simply implement flow control? Using a Fluid-o-tech fg309 (or any BLDC gear motor) can I control flow directly with the simple wiring with a potentiometer, or do I need to control the speed of the pump directly using a PWM controller? I'd like to simply use the pump to do everything rather than have a needle valve and solenoid.

I already have the pump used, in the mail.

Here are some related threads.
GS/3 Mod EP "Chimera" - a fully configurable profiling mod for La Marzocco GS/3 AV
La Marzocco Linea Mini shot timer, automatic backflushing and flow profiling mods


I'd like to set the knob/paddle on a pump speed/flow rate and not have the flow change as the pressure changes. Other users have done this using some coding and sensors. I'm wondering if a manual version could be implemented with the bldc gear pumps without any coding. After preinfusion I could pressure profile watching a pressure gauge. I have a lever now so it would only be used for preinfusion but it would be relatively plug and play for LM machines and other pump machines.

An issue would be knowing the flow rate. Any simple way to connect a flow meter or the pump rpm out (0-5v) to a simple screen with no coding? A flow meter like a DIGMESA nano brass 9N*-01. Unfortunately I am coding illiterate. RPM would be nice, ml/s would be better. Anything repeatable would be good.

It seems you can connect a PWM 0-5v controller where you would wire the potentiometer after the 15k ohm resistor for the bldc motor. There are plenty of cheap alibaba Pwm controllers with built in screens which I believe give you a percentage value. Would these cause the pump to run at a constant flow rate?

Based on the fluid o tech graph it appears the flow rate will change due to pressure at a constant voltage, hence this question.






Thanks anyone who read through these Qs!

Aren't gear pumps having direct flow dependence on the RPMs? If you have a tacho sensor you can precisely figure out how much liquid was pumped through.

There's also a dependency on back pressure. Flow meters are reasonable at moderate rates (10-20% ??), but rather low in accuracy below 1-2 mL/s or so. The DE1 is moderately successful in modeling the flow vs. input over a moderate range of pressures. I know Ray has been working on flow metering for low flow, but I'm not aware of anything commercially available at this time.

I'm keeping my eyes open as the Micra seems like a great platform for a gear pump and flow/pressure management. My current thoughts would be to model the flow as a function of back pressure and speed.

One question is if you really need "flow" as a number, or if pressure and input (voltage or speed) is sufficient. People seem to be able to profile well with a paddle as the control input, with no sense of the actual flow rate.

bostonbuzz wrote:Question: How can I most simply implement flow control?

You'll need a 24V power supply capable of handling the current draw of the pump at whatever maximum load you decide on (more on that later) and a 5V power supply for the speed control (alternatively, you can use a trim pot on the 24V supply and bypass the 5V supply altogether).

You'll want a big potentiometer you can use to control the pump speed and a smaller trim pot to handle the max speed.

With this setup, you'd have both pots running in line from the 5V supply to the speed input on the pump. The trim pot would be either close to the power supply or right next to the pump, somewhere you can set it and forget it, and then you'd put the main pot somewhere convenient to use.

Once you're all setup, you set the main pot to full speed and then turn the trim pot down low and load a blind basket in the machine. Then turn the trim pot up until you reach your desired max brew pressure. That will then allow the main pot to control the pump speed from zero up to whatever that max speed is that gives you your peak brew pressure.

The 24V supply needs a relay in line to the pump, as once the pump sees 24V, it runs. So you'll want to take the signal that turns the existing pump on and use that to trigger the 24V supply to the gear pump.

Once you have that, you're set.

If you don't want to use a 5V supply, you feed the trim pot from the 24V supply and first set it to give you 5V, which you feed to the speed input of the gear pump. You then put the main pot between that signal and the ground, and run through the same process as above to dial in the trim pot to give you your desired brew pressure and then the main pot will give you your pump control in real time. The main pot will resemble the 4k7 pot in the diagram you posted labeled "POSSIBLE ALTERNATE CONNECTION" and the trim pot will resemble the 15k resistor. In this case, you'll want to choose a big trim pot (50k or 100k) so that you get a sizable voltage drop across it even when the main pot (which should be 10k or less) is maxed out. Basically you want to be certain you never apply more than 5V to the pump motor speed input.

Hope this helps!

- Jake

Thanks for the help all. Sounds like I should just stick with a basic voltage-based control of the pump. I'll make a post with all the parts and details when I sort it out. I'm looking for a simple tachometer screen that I can plug the BLDC 24v(-) and yellow (tach) out into and maybe fit into an enclosure with a potentiometer - any ideas let me know.

Jake,

Thanks for your detailed response.

The pump should never be over 100W, so I will pick up a 120W 110AC-24vdc power supply like the one below.
https://www.automationdirect.com/adc/sh ... psv24-120s Getting it to fit will be interesting.

The 24V supply needs a relay in line to the pump, as once the pump sees 24V, it runs. So you'll want to take the signal that turns the existing pump on and use that to trigger the 24V supply to the gear pump.

Gotcha, I was hung up on that since the power supplies have a 3s delay when powered on, so I wasn't sure how to trigger the pump. I have another BLDC pump from a vacuum application that says not to supply control voltage unless the motor controller is supplied with operating voltage, otherwise damage can occur on the motor controller. FoT doesn't say this, but do you think it's safer to put the relay on the 24V leading to the potentiometer circuit instead of the 24V going to the pump?

Either way I'm having trouble finding a mechanical or SSR relay that has a higher AC input (110 VAC in this case from the old Ulka pump) than the DC output, especially since I will probably have the pump on "zero" when it turns on and be slowly ramping it up. Do you have any recommendations for a 110V relay that can trigger 24VDC?

I understand what you mean about the trim and other pots arrangement and setting it for max pressure - good idea. Due to space I decided to go with just the 24V power supply. I ordered some 1% precision 5k main and 100k trim pots, so those should hopefully work, wired like the Fluid-o-tech diagram above.

Again, thanks so much.

bostonbuzz wrote:FoT doesn't say this, but do you think it's safer to put the relay on the 24V leading to the potentiometer circuit instead of the 24V going to the pump?

If you're concerned about this, and running two separate power supplies, I'd just get a DPST relay and switch both at the same time. Setting pump speed to zero will put the controller in a locked rotor state where the motor tries to stay at zero speed. If you do the pot from 24V down to 5V, then just put the relay before the trim pot. That way you can use a single pole relay and kill all power going to the pump.

bostonbuzz wrote:Either way I'm having trouble finding a mechanical or SSR relay that has a higher AC input (110 VAC in this case from the old Ulka pump) than the DC output, especially since I will probably have the pump on "zero" when it turns on and be slowly ramping it up. Do you have any recommendations for a 110V relay that can trigger 24VDC?

Relays are just switches attached to coils. You just need a 120V coil and switch contacts that can handle the expected current. If you are correct that 100W is all that's needed, 10A rating on the contacts would be plenty. 15 or 20A would be just fine and capable of plenty more power if ever needed.

One other thing of note: Depending on what machine you are planning on using this for, you may have 5V available to you from the existing control board. If that's the case, consider using it and a DPST relay to simplify things a bit. While you will still need a trim pot, you won't have to worry about balancing them out as a voltage divider to make sure you don't go over 5V.

Cheers!

- Jake