CRCasey wrote:With a fixed controller and motor combo you are stuck with the noise. And check the heat output before you stick a motor in a insulated box please.

Heh, thanks for the laugh. As for temperature, I've felt the motor after long sessions and I haven't felt it over room temperature. The controller occasionally gets slightly warm to the touch (not that it makes any noise or needs to be boxed). As Greg said to me before, the pump is extremely power-efficient, and it shows.

Hi guys,

FYI, yesterday I finally figured out why I was getting speed fluctuations with the motor (following advice of a friend). This is due to the fact that the boiler heater which is engaged by the boiler PID draws quite a lot of current when on and that momentarily lowers the current going to the motor. The heater is engaged every few seconds during a shot and if you look at the 'output on' light on the front of the controller and listen to the pump it's easy to understand what's going on.

So there, mystery solved.

Regards,
dsc.

That makes me glad that I configured my box to be powered separately from the machine (indeed, it's actually on a different circuit). Are you planning on living with it, doing what I've done, or re-wiring the Elektra with heavier wire to reduce the voltage (and current) drop under heater load?

Hi Nicholas,

so your pump is plugged in to an actual separate circuit or just separate socket in the kitchen? I believe my whole kitchen is on the same single circuit hooked up to a single MCB (30A or something like that), plus there's a separate circuit for the oven. I can rewire the pump controller and plug it into a different kitchen socket, but I'm not sure if that's going to solve anything. I also doubt changing the cable will solve the problem, the current drop is due to the heater coming on and the heater consumes 8A - 9A no matter what (probably a bit more as there's surge current and whatnots during the initial phase probably). I'm guessing that the motor would be fine with the heater on all the time, but it's only on for like a second every 2-3s which messes things up.

Then again it doesn't make a big difference really, so I might stick with it for now, or just do some quick experiments to see what can solve it.

Regards,
dsc.

dsc wrote:your pump is plugged in to an actual separate circuit or just separate socket in the kitchen?

Mine is a totally separate circuit. Still get a little flutter on the gauge though.

As Bill C suggested, I've removed the 0.6mm gicleur. The pump is showing no signs of cavitation at the higher unrestricted flow rate.



I had my fill of coffee for today, so will try it out tomorrow. I've left the plumbing attached in case its a disaster to revert to the gliceur with minimum fuss.



Cheers, Chris

gyro wrote:Mine is a totally separate circuit. Still get a little flutter on the gauge though.

Ah, I told Tom this privately some time ago, but mine is also on a fully separate (i.e., different breaker) circuit. Unlike either of you, though, I don't notice any real flutter on my gauge when the knob is held steady. Maybe it's a crummy (slow) gauge? I don't know. I do hear pitch changes in the motor as it operates, but again they don't show up on the gauge or as 'pulses' or anything like that of flow change in the 'cone' from the filter. From this (and my tasty espresso, 'profiled' or not) I assume it's negligible at best.

OK, so removing the gicleur has definitely changed things! All the needle flutter has disappeared (not that there was alot, but it was noticeable on the gauge), which is a little confusing as I'm not sure why that would happen.

Obviously the pre-infusion now really needs to be manually controlled, whereas before I was happy with the restrictor and pre-infusion chamber. Now the pre-infusion chamber (downstream of gicleur location) fills up really fast. Response rate to changes on the pot are quicker, as there is no lag effect from the gicleur.

The couple of shots I've pulled this morning have been a little harsh. I was setting 9 - 9.5 bar on the gauge, whereas beforehand I was probably getting 8.5 bar at the head. Hard to change my brain to want to see 8.5 on the gauge!

Pump has been working fine with no funny cavitation sounds that Nicholas experienced. Although as I mentioned in a previous post, I have heard a gurgling sound on occasion if just 'pre-infusing'. Its my belief that the TMFR doesn't allow the flow through of water when not rotating as well as some other rotary pumps. I think it may depend on exactly where in the rotation cycle the vanes stop at as well, as its not always the same.

If I was to process control the pump, this current gicleur-less set-up would be the configuration in which I would do it. Still not convinced that I want to spend the coin on it. The occasional auto-fill (when its not meant to) screws the shot, although this would be mitigated if under process control as the controller would up the RPM to try and maintain pressure.

The fiddler in me wants to automate it, or at least change the manual control to pressure (vs RPM) but the bean counter (I'm not an accountant!) in me says its not worth it.

Cheers, Chris

gyro wrote:Obviously the pre-infusion now really needs to be manually controlled, whereas before I was happy with the restrictor and pre-infusion chamber. Now the pre-infusion chamber (downstream of gicleur location) fills up really fast. Response rate to changes on the pot are quicker, as there is no lag effect from the gicleur.

You might consider a compromise: a larger gicleur size. Something around 1.25-1.75mm might calm things down a little while still allowing good response to pressure profiling.

gyro wrote:The occasional auto-fill (when its not meant to) screws the shot

This bugs me, too. I haven't gotten around to fixing it, but the cure seems pretty simple. Put a normally closed relay in series with the autofill solenoid. If you wire this new relay's coil in parallel with the 3-way's coil, the autofill could not occur when a shot is in progress.

AndyS wrote:This bugs me, too. I haven't gotten around to fixing it, but the cure seems pretty simple. Put a normally closed relay in series with the autofill solenoid. If you wire this new relay's coil in parallel with the 3-way's coil, the autofill could not occur when a shot is in progress.

Indeed a cunning plan, but I think I'll ask Kees about it again. Its not a cheap machine and its obviously a small design error, so really it should be rectified under guarantee.

AndyS wrote:You might consider a compromise: a larger gicleur size. Something around 1.25-1.75mm might calm things down a little while still allowing good response to pressure profiling.

Yup, good idea. I think I recall Bill C mentioning something around 1.2mm might be a good place to start.

Cheers, Chris

Hi guys,

I've got a 1.0mm (or 1.1mm) gicleur and it works well, the cooling dumps are faster and pressure ramp up is faster. I'm getting roughly 130-140ml per 10s, so a bit more than the stock Elektra jet (110ml/10s).

Regards,
dsc.