[syhprum]

It would seem to be a simple matter to install an auto transformer on the MRI machine to boost the voltage to the cooling motor or to demand that the manufacturers install a motor appropriate for UK mains supply, I understand that making a modification to the machine would require dispensation from the makers but if they had any sense they would get one of their technicians over double quick if you explained why it was out of action and you were losing money because a unsuitable motor had been fitted.
This would seem to be a much less drastic response than what you had suggested.
it is my experience with the Germans that they consider all non German speaking countries as primitive and third world so if have any telephone discussion of this problem it would be best to find a native German speaker to assist.     

[William McC]

Also you might want to have a tech check the pressures and sub-cooling differential. A slight over charge, will not be noticed, until you get a very hot day. With R-410A the amount it takes to overcharge the system is almost comical. The system will run with much lower RLA Run Load Amps if you properly charge it. A lot of techs do not know you have to run the unit on second stage cooling, if the unit has a second stage in order to check sub-cooling. Many techs do not know how to do a sub-cooling test. You can just google it, there are a couple of old timers that explain it well. To simulate a hot day we often cover the condenser with cardboard or plywood, to simulate a very hot day.

If it is a package unit with multiple compressors and they share a condenser cooling air discharge both compressors should be running for any testing.

The new Field Piece HVAC gauges are pretty cool and have temperature sensors built it. But you can do a sub-cooling with regular old fashioned gauges and a thermocouple attached to a multi-meter. Sometimes I cheat and use the infrared temperature sensor.

It is remarkable how the amperage goes way up on very hot days. One other trick that is pretty cool is installing a small bypass damper, a barometric damper. This allows your evaporator to run a bit cooler which also lowers your amperage substantially. It will also cause the output air from the evaporator to be a bit cooler. As long as you do not freeze over your evaporator it is a very efficient thing to do. It only cuts down the CFM a little bit. It is less efficient technically, however it is often much more practical and useful.

If you are running a rack system with hermitic or semi hermitic compressors, with a chiller, condenser, and chilled water piped to your air handler, you can check the cooling tower water flow to the condensers, and the water output temperatures, lack of enough cooling tower water flow will cause high head pressure and higher RLA. If you are getting air output that is not cool enough from your air hander, you might try lowering the fan speed in the air handler, that circulates the room air. This can sometimes be accomplished by a pulley change, or by other controls. That will definitely lower your RLA.

Sincerely,

William McCormick

[Bored chemist]

The energy used by a cleaner is the product of the power and the time taken. A 10KW vacuum cleaner doesn't mean you can walk round the room in a 20th of the time it takes to walk round with a 500W one

It took a while
https://newsthump.com/2017/03/29/dyson-begins-work-on-new-10000-watt-vacuum-cleaner-for-proud-brexiters/?fbclid=IwAR3IS7AfhZeoPK5uAKBj-tNtsl-FLeeGxXcNXLv3fb-bJVT_BIPOU-6S9_M

[SeanB]

For the street light circuit the first lights used carbon arc lamps, and they were run in series, so as to have easy control of the current, as the arc is almost constant voltage irrespective of current, but light output depends on current. Thus you regulate the current in the string. Later on incandescent lamps replaced the carbon arc, selected so as to have the same lamp current, so you could simply retrofit the lamp and not do any adjustment, and they also have a special socket with a thin paper insulator between 2 contacts. The paper will withstand around 150VAC but will break down and fail, closing the contacts, when the lamp burns out, keeping the series string intact.

Series strings won out because they save on expensive copper wire, as you needed to have an extra conductor to carry the current, and with only 1 wire on the original light circuit you saved a lot of money, as you could either run the wire in a loop, cut in where needed to put a lamp, and still have them work, or you could simply put a ground mat at the far ens and use ground return for the one wire. One thin wire rated for a single lamp current of around 2A, and cheap ceramic insulators for 10kV on the supply side, and in the lamp units, to pass this through. Only later on, when they started to have houses supplied on the same poles, did the lighting manufacturers remove the constant current transformers and go to parallel lamps fed off a thicker wire that carried the lighting current. Constant current also was used with discharge lamps, which is why the lamp powers were odd things like 325W and 125W, because at those currents the lamp would operate like the original carbon rod, but with longer life and higher light output. No ballast required, just screw the new lamp into the exact same incandescent E40 socket, and same paper cutout, for triple the light output and 20 year life.

The killer for the series circuit was energy efficiency, you could have up to 60% of the input power as heat in the constant current transformer, but they are still used as airport approach lights, simply because they make monitoring for failed lamps easy, and a broken cable is easy to detect automatically. However even here they are rapidly being replaced by LED, as the lower running cost is a big factor, and you can integrate monitoring into the LED units as well, though you need to replace all cabling with new.

Street lighting also went more to local photocells, so as to limit the step load of turning on all of a city lights all at once, instead the load slowly ramps up over a half hour or so as the sun sets, easy to follow with power plant output. As well going to LED drops power use, your 400W HPS/MH/MV  or 60W SOX main road light now is 135W to 220W, a saving in power, and side street lighting went from 125W/80W MV/MH/HPS or 36W SOX (mostly the UK only, with the monochromatic red sodium light of low pressure sodium) being replaced with 36W LED fixtures.

But as to voltage, it is all fixed by the original inventors, and the insulation they had determining the voltage at which it would not break down. 120VDC for the first Edison generators, as the rotors would break down at higher voltage, with the shellac and cotton insulation they had. This then got fixed in stone, as subsequent systems had to be compatible with the old system, and the UK and rest were later, with improved insulation that allowed over 200VDC output, and 240VDC by simply using improved insulation. UK originally used mercury arc rectifiers, so had 2 240VDC lines in the street, with one half of the street getting positive with respect to the grounded common, and the other half negative. For resistive loads not a worry, but the new fangled radiograms needed to be set to which side of the street you were, and later on needed to have a rectifier tube put in for AC use. By the time TV came into use AC mains was the norm, only a very few small areas were still DC ,and they got changed out rapidly.

As an anecdote, across the road is an old tramway substation, where there is still the old mercury arc rectifier room, separate from the transformer control gear hall, and one empty bay where the one transformer was. Mostly empty, all gear long gone, except for the bolts left in the walls, and the cut off remains on the floor. Cables are still in the ground, never removed, there was still the one DC feed head till 2009 outside.

[Petrochemicals]

[evan_au (OP)]

Huh...246V

The video doesn't play for me, but are you asking why the voltage reads 246VAC instead of 240VAC?

The mains voltage is not perfectly regulated, but can vary up and down by ±10% (although +20% might shorten the lifetime of some old-fashioned incandescent lamps).
- In particular, if there are a lot of solar panels in your street, the voltage can increase during the day.
- The power transformers in the street were designed for delivering power from a centralized power sation to distributed consumers. The turns ratio in the transformer is not ideal for delivery of distributed solar power back into the central grid.
- Solar panels are designed to reduce their power output when mains voltage gets too high in the street

Another factor that can affect the reading on a multimeter is if the voltage is not sinusoidal - for example if it has significant harmonics.

[Petrochemicals]

Huh...246V

The video doesn't play for me, but are you asking why the voltage reads 246VAC instead of 240VAC?

The mains voltage is not perfectly regulated, but can vary up and down by ±10% (although +20% might shorten the lifetime of some old-fashioned incandescent lamps).
- In particular, if there are a lot of solar panels in your street, the voltage can increase during the day.
- The power transformers in the street were designed for delivering power from a centralized power sation to distributed consumers. The turns ratio in the transformer is not ideal for delivery of distributed solar power back into the central grid.
- Solar panels are designed to reduce their power output when mains voltage gets too high in the street

Another factor that can affect the reading on a multimeter is if the voltage is not sinusoidal - for example if it has significant harmonics.

No its an American showing you that 2 phases ammount to 246v I believe

[alancalverd]

Also worth noting that mains voltage is specified as RMS, and a simple voltmeter such as the one shown can't actually measure RMS - you have to calculate something based on peak or rectified time average and an implicit form factor.

[Bored chemist]

I love the way that everyone assumes that the meter is accurately calibrated.

[alancalverd]

They are usually accurate to +/- 1 digit, but don't necessarily display what you are expecting for the reason given above. The basic voltmeter chip is a DC instrument so for AC you add a bridge rectifier and a 1.41:1 voltage divider to give an estimate of the rms voltage, assuming the waveform is sinusoidal and the rectifier forward voltage is negligible.

[Bored chemist]

They are usually accurate to +/- 1 digit...

...When they leave the factory

The basic voltmeter chip is a DC instrument so for AC you add a bridge rectifier and a 1.41:1 voltage divider to give an estimate of the rms voltage, assuming the waveform is sinusoidal and the rectifier forward voltage is negligible.

I think you forgot something.

[vhfpmr]

They are usually accurate to +/- 1 digit, but don't necessarily display what you are expecting for the reason given above. The basic voltmeter chip is a DC instrument so for AC you add a bridge rectifier and a 1.41:1 voltage divider to give an estimate of the rms voltage, assuming the waveform is sinusoidal and the rectifier forward voltage is negligible.

Digital multimeters use dual-slope integration ADCs because the integration smooths noise and ripple, and because the integration period can be made a multiple of 10ms for further rejection of mains related hum. They are also independent of integrator time constant and the clock frequency. Like moving coil analogue meters, DMMs (that aren't true RMS) are sensing the DC average of the input, which for a rectified sine wave is 0.637, and the form factor is therefore .7071/0.637 = 1.11. To measure RMS values of waveforms other than a sinusoid, you need to divide by 1.11, and then multiply by the form factor for the waveform in question.

True RMS meters are only really helpful if you don't know the waveform, and have a resistive load. In the common case when you have a load running from a rectifier and capacitor input filter you have neither, and a more accurate estimate of power can be had by multiplying the peak voltage by the mean current if you don't have a power meter.

±1 digit on a 3½ digit DMM would be an accuracy of ±0.05%, which is tall order even for a good quality instrument like a Fluke. The 1 digit on digital instruments is in addition to the basic accuracy, of 1%, or whatever.

[Bored chemist]

Alan nearly got the right answer.
If you put a bridge rectifier and a capacitor in the circuit you can measure (pretty nearly) the peak voltage and that is 1.141 times the RMS value (if you have a sinewave).
Oops!
Typo
1.414 times the RMS value (if you have a sinewave).

[vhfpmr]

Alan nearly got the right answer.
If you put a bridge rectifier and a capacitor in the circuit you can measure (pretty nearly) the peak voltage

No, non True-RMS meters are mean sensing, not peak-sensing, and this is usually stated explicitly on good quality meters.
Putting smoothing capacitors in meters would make them susceptible to large errors caused by noise spikes.
https://en.wikipedia.org/wiki/Form_factor_(electronics)

and that is 1.141 times the RMS value (if you have a sinewave).

1.414

[Bored chemist]

No, non True-RMS meters are mean sensing, not peak-sensing, and this is usually stated explicitly on good quality meters.

The easy way to get an average is still to use a capacitor. (or the mass of the needle in an analogue meter)

The value you use (and that of the shunt resistor) is dependent on the frequency range- which can make things awkward.

[alancalverd]

No, non True-RMS meters are mean sensing, not peak-sensing, and this is usually stated explicitly on good quality meters.

The average value of a full-wave rectified sine is 0.637 V_peak, which is not easy to sense or sensible to display.

[vhfpmr]

No, non True-RMS meters are mean sensing, not peak-sensing, and this is usually stated explicitly on good quality meters.

The average value of a full-wave rectified sine is 0.637 V_peak

Yes, I know:

Like moving coil analogue meters, DMMs (that aren't true RMS) are sensing the DC average of the input, which for a rectified sine wave is 0.637, and the form factor is therefore .7071/0.637 = 1.11. To measure RMS values of waveforms other than a sinusoid, you need to divide by 1.11, and then multiply by the form factor for the waveform in question.

which is not easy to sense or sensible to display.

It's trivially easy to sense, you just leave the smoothing capacitor off the rectifier, and meters don't display mean level, they are calibrated in terms of RMS.

"Mean Sensing, Calibrated RMS":

TM3B.png (680.75 kB . 883x402 - viewed 2035 times)

No, non True-RMS meters are mean sensing, not peak-sensing, and this is usually stated explicitly on good quality meters.

The easy way to get an average is still to use a capacitor. (or the mass of the needle in an analogue meter)

The value you use (and that of the shunt resistor) is dependent on the frequency range- which can make things awkward.

A capacitor directly on the output of a rectifier will give you peak, or something close to it, because the rectifier charges it but doesn't discharge it. That's what rectifiers do, only conduct in one direction.

Smoothing at 0, 40ms, & 120ms:

rect.png (23.57 kB . 1150x514 - viewed 2131 times)

[Bored chemist]

A capacitor directly on the output of a rectifier will give you peak, or something close to it, because the rectifier charges it but doesn't discharge it. That's what rectifiers do, only conduct in one direction.

Thank you for restating my point.

In that graph of yours the voltage drops between each peak.
Why is that?
Is it because of current flowing into the meter?
Would that be dependent on the effective resistance of the meter shunting the capacitor?
Would the extent of the "droop" depend on the frequency of the supply as well as on the effective resistance and the capacitance?
Have you also just shown more or less what I said here?

The value you use (and that of the shunt resistor) is dependent on the frequency range- which can make things awkward.

[vhfpmr]

The easy way to get an average is still to use a capacitor

If you want an average reading instrument you need to leave the capacitor off. Put a capacitor on the output of a rectifier, and you'll have a (close to) peak reading instrument, because the charge and discharge currents are dissimilar, as shown in the plot above. If you need to smooth the ripple with a capacitor and still retain the mean sensing, you need a buffer between the rectifier and the smoothing.

[Bored chemist]

If you are using a mechanical moving coil meter, you can use its mass to smooth out the reading.

If you are using a digital meter then you need to so something more complicated to get rid of aliasing errors.
The easiest way to do that is with a capacitor or some such to roll off the high frequencies.