New shop! (was: Multiple buildings, ...)

[jwmelvin]

Yea that seems pretty great. Nice progress.

 

[rabler]

That's really cheap. What is the monthly meter charge? Here it is $15.

I’d have to check but that sounds about right. We pay for two drops as the barns are separate from the house.

Power is a membership co-op, and there are several coal power plants as well as coal mines within 50 miles. Big industry in this area. I’d guess we’re on the low end of the electrical cost range.

Lots of oil wells too. I’ve dug up oil pipe, apparently there was a well on our property in the 90’s. Glad it’s gone, rotten eggs sulfur smell around the wells.

 

[woodchucker]

I’d have to check but that sounds about right. We pay for two drops as the barns are separate from the house.

Power is a membership co-op, and there are several coal power plants as well as coal mines within 50 miles. Big industry in this area. I’d guess we’re on the low end of the electrical cost range.

Lots of oil wells too. I’ve dug up oil pipe, apparently there was a well on our property in the 90’s. Glad it’s gone, rotten eggs sulfur smell around the wells.

I think that was a bout of flatulence.. ahhh
How far from you are the nearest 3phase lines?
I am between 2 schools, a middle school (closer about 400ft from my house) and a high school 1/2 mile as the crow flies.. 3phase from middle school is very close But might as well be 2 miles...

 

[rabler]

How far from you are the nearest 3phase lines?

Funny you should ask. About a block in one direction, 1/2 mile in the other. We’re very close to the end of the line (our barn is the end of the line). Lots of big farms around so three phase is very common for grain elevators,etc.

 

[rabler]

I thought about asking about 3 phase. I believe, but am not sure, that 3 phase service would entail a third meter drop, two is enough. The rural electric co-op is a small operation, if I asked I might confuse the whole thing. Realistically I just don't use it for more than a few hours a week, and the RPC does fine.

I have thought about modifying my RPC to be intelligent, to improve line voltage balancing under widely varying loads, that's on the project list. A microcontroller could easily monitor the line voltage a few hundred times a second, and kick in additional capacitance if needed. Use a voltage divider and an analog opto-isolator to safely capture the line voltage on each leg.

Nyquist sampling (120 samples/sec) doesn't really capture transients and voltage changes readily, and gets mathematically messy. Alternatives are to sample each leg around 1000 samples/sec and run through a software peak dectector, or use hardware peak detector circuit (diode) with the right RC time constant to readily pick up 60Hz to within a couple volts margin of error. Software might give some options on additional noise analysis. Easy enough to add on LCD display with voltage for each leg, and measure current with some Hall detectors. Thinking this will be a fun project to try out the new Raspberry PI 2040 microcontroller to build. Could include some data logging features with USB data dump. Might be a good topic for a new thread.

 

[ahazi]

I thought about asking about 3 phase. I believe, but am not sure, that 3 phase service would entail a third meter drop, two is enough. The rural electric co-op is a small operation, if I asked I might confuse the whole thing. Realistically I just don't use it for more than a few hours a week, and the RPC does fine.

I have thought about modifying my RPC to be intelligent, to improve line voltage balancing under widely varying loads, that's on the project list. A microcontroller could easily monitor the line voltage a few hundred times a second, and kick in additional capacitance if needed. Use a voltage divider and an analog opto-isolator to safely capture the line voltage on each leg.

Nyquist sampling (120 samples/sec) doesn't really capture transients and voltage changes readily, and gets mathematically messy. Alternatives are to sample each leg around 1000 samples/sec and run through a software peak dectector, or use hardware peak detector circuit (diode) with the right RC time constant to readily pick up 60Hz to within a couple volts margin of error. Software might give some options on additional noise analysis. Easy enough to add on LCD display with voltage for each leg, and measure current with some Hall detectors. Thinking this will be a fun project to try out the new Raspberry PI 2040 microcontroller to build. Could include some data logging features with USB data dump. Might be a good topic for a new thread.

I don't think the 3 phase electric motor cares much about those little voltage variations/spikes/phase imbalance it simply integrates them like a LPF. If you use a VFD you will get the best ability to fine tune the motor speed and torque even under load. I have a VFD on my lathe and absolutely love and use the speed control while running the lathe, in particular when facing or parting, I increase the speed as I am getting closer to the center.

Ariel

 

[rabler]

I don't think the 3 phase electric motor cares much about those little voltage variations/spikes/phase imbalance it simply integrates them like a LPF. If you use a VFD you will get the best ability to fine tune the motor speed and torque even under load. I have a VFD on my lathe and absolutely love and use the speed control while running the lathe, in particular when facing or parting, I increase the speed as I am getting closer to the center.

Ariel

Ariel,
Short transient, you are correct. Shifting load demand, not so. Unbalanced three phase results in degrading performance in three phase motors. Classical run capacitors used in RPCs only balance the voltages properly at one load. Issue with transients is getting reasonably accurate analog to digital sampling to measure voltage in the presence of transients and other power line noise. A full analysis would be to essentially run a spectrum analysis via Fourier transform or DFT. That’s way overkill, but would be interesting given possible power line noise sources like VFDs.

I use several VFDs, typically for machines up to 3 hp. I have machines with 10HP main motors such as my K&T 3K mill, where VFDs get pricey. My big lathe has four motors (main, rapids, hydraulic, coolant) VFDs just wouldn’t be practical. I have a couple of surface grinders that I'll eventually rebuild, both of them have two motors, spindle and hydraulic. Long term I’d like to add a cnc VMC. All of that is covered by one RPC. I've had to replace the Delta VFD in my smaller mill to the tune of over $400, Grizzly wanted over $900. Price out a decent VFD that will turn a 10HP motor on single phase input.

Being a hobbyist doesn’t always mean small.

 

[ahazi]

Ariel,
Short transient, you are correct. Shifting load demand, not so. Unbalanced three phase results in degrading performance in three phase motors. Classical run capacitors used in RPCs only balance the voltages properly at one load. Issue with transients is getting reasonably accurate analog to digital sampling to measure voltage in the presence of transients and other power line noise. A full analysis would be to essentially run a spectrum analysis via Fourier transform or DFT. That’s way overkill, but would be interesting given possible power line noise sources like VFDs.

I use several VFDs, typically for machines up to 3 hp. I have machines with 10HP main motors such as my K&T 3K mill, where VFDs get pricey. My big lathe has four motors (main, rapids, hydraulic, coolant) VFDs just wouldn’t be practical. I have a couple of surface grinders that I'll eventually rebuild, both of them have two motors, spindle and hydraulic. Long term I’d like to add a cnc VMC. All of that is covered by one RPC. I've had to replace the Delta VFD in my smaller mill to the tune of over $400, Grizzly wanted over $900. Price out a decent VFD that will turn a 10HP motor on single phase input.

Being a hobbyist doesn’t always mean small.

I have no personal experience generating 3 phases out of an RPC. I used 3 phase that "came from the wall..." starting many years ago as a teenager with my very first drill press that had a 3 phase motor. These days I am using VFD.

Your idea for a "research project" sounds interesting but will be very time consuming. From my experience the RPi is not suitable for any real time work. Don't ask how I know that... Check out RPi with LinuxCNC, the issues of RT kernel pops up immediately.

A friend of mine that had similar requirements to yours for a high capacity 3 phase source just bought this:
PHASEPERFECT 240V Digital Phase Converter PT020
expensive and very capable.

Ariel

 

[rabler]

From my experience the RPi is not suitable for any real time work. Don't ask how I know that... Check out RPi with LinuxCNC, the issues of RT kernel pops up immediately.

If you google RP2040, despite being made by Raspberry Pi, it's not a linux CPU, it's a stand alone dual core microcontroller, and is even supported in the Arduino environment.

 

[ahazi]

If you google RP2040, despite being made by Raspberry Pi, it's not a linux CPU, it's a stand alone dual core microcontroller, and is even supported in the Arduino environment.

You are right, it is a microcontroller and it does not run Linux, can probably do the job. I am not certain but you might need to stay away from MicroPython if you want some predictable performance or if you need real multithreading.

Will this chip be able to do your RT FFT calculations with enough resolution fast enough?

Ariel