...
268 is way too high. How and where are you measuring this? 208 is too low. Call your power company and have them put a recording voltmeter on the line for a few days. It sounds like they may have a problem somewhere.Once the power company solves their issues, your problem may go away.
While I suspect that the cutout voltage on the inverter cannot be changed, I suspect that the 5 minute delay could be reduced without harm.
The start assist / hard start kits are very inexpensive. Google for examples. You should address the sag issue. Forget about the generation part of it when / if you complain to the electric company. They have a load device that replaces the meter temporarily, and the measure the drop when the load is applied. You'll want your inverter locked out while they are playing so as not to "help". Get someone to measure the voltage at the input to your main breaker with a fast recording meter. Look at how low it goes when you put the ~ 100 amp start load from your AC on it. If you go from say
240 to 215 volts for the instant, that would be a 10% drop - way too much for 1/2 load on a 200 amp service. The comment about 268 volts at night is interesting also. I'm thinking transformer issue. The primary side - at most utilities - is regulated within 2.5%. Mention the range you are seeing, and they'll likely put a logger out there, followed by a transformer replacement.Is your transformer pole mounted or on the ground? If pole mounted, see if you can figure out how many houses are on it. Then look at the number on the side of it. 10, 15, 20, 25, 37.5, 50, etc. That is it's continuous capacity in Kva. 200 amps at 240 volts is 48 Kva. It's assumed that no-one is going to consume 200 amps at 240 continuously. The realistic number is something like 10% of that. The real issue is the total load on that transformer.
The inverter has a digitial readout that monitors quite a few things including the wattage and VDC from the the panels as well at the line votage from the grid. I think it should be quite accurate.
Thanks -- I'll give them a call and see when they can come out.
best,
doug
Hmmm. Your inverter may (or may not) be accurate for measuring the AC voltage at the inverter. Of course, where the power company is interested is at your service entrance. When your PV system is running, the voltage at the inverter will be a little higher than at the service entrance, and the voltage at the service entrance will be a little higher than if the PV system was not running. If there is a large difference between the voltage at the inverter and at the SE, then you have a problem on your side.
What you should do (probably before the power company comes out), is measure the voltage at the service entrance several times during the day, with a good meter, and with the PV system shut down, and record it. I suspect that the power company will insist that the PV system be shutdown while they are doing their measuring, so you might as well get your own estimate of what they will find.
You may find it useful to get an estimate of the difference that the PV system makes to the voltage. That can give you an estimate of the effective impedance of the transformer, service entrance cable, feeder cables, etc. all together.
Hi Bill,
It seems like there are $20 ones and ones costing a lot more ($110):
Yes, I'll get them out here to test the line -- I actually just happen to catch it on the inverter meter this morning. I was walking past it when the AC cycled on -- the voltage was low to start (222 VAC) and dropped to 209 momentarily. It then went back up to 221 after the AC cycle up.
Interesting point...about a year ago we lost one leg of our 240 and after several calls, they actually removed the transformer from the pole next to our house and installed a new transformer on a different pole further away from our house, so I probably have a drop line about twice as long as I had before. Because of how they jumpered everyone who was on the old transformer to the new transformer, it is hard to know how many houses it feeds. It is much smaller than the old one, but I assume that is a product of newer technology.
Thanks for all your help.
best,
doug
Ahh, that is the interesting thing about these grid-tie inverters. The inverter has to sync itself to the current line voltage and frequency supplied by the grid. By code, it has to match the line voltage.
On the DC side it will accept anything from 250-600 VDC, and when jumpered for 240 VAC, it will sync with any line voltage between 211 to 264 VAC. By code it cannot increase the line voltage. You are right, though, if the inverter is pumping out 25 amps, it probably will stop the air conditioner from further depressing the grid voltage by making up some of the shortfall.
Yes, I am sure they will want it disconnected. I have a good old Simpson 260 meter to do some tests with the array turned off. We have a DC-side disconnect and an AC-side disconnect for the inverter. When the utility folks have been out during the installation, they do not want to have anything to do with it. I think some of the folks at the utility see me as an unwelcome presence on their grid. :)
best,
doug
Sort of; in theory. In reality, all electrical components: wires, transformers, etc. all have some non-zero resistance. If the voltage drop across the wire was exactly zero, the current in the wire would also be zero (Ohm's law), and your PV system would be doing exactly nothing. So the voltage at your inverter will never be exactly the same as at the service entrance if your PV system is operating.
How long is the wire from your inverter to the service entrance and what gauge wire is it?
I would suggest buying or borrowing a modern digital meter. The technology really has improved greatly, as has the price/performance.
I can understand their point of view. With any customer owned generation system you have potentially lethal voltages coming from an unexpected direction.
It is only about two feet and is done in 10 gauge -- the inverter was mounted as closely as was possible in March by a professional PV firm.
Say it ain't so -- I love my old Simpson meter. I think it actually beats digital meters for monitoring flucuation rates and trends:
That is one of the reasons why they have required "anti-islanding." If the inverter cannot sync on the grid frequency and voltage, it automatically disconnects from the grid. Which is what it is doing at this moment as we only have 202 grid VAC right now here on the Edison grid in Los Angeles. My panels are baking in the sun, but I cannot reconnect to the grid until the inverter sees at least 211 VAC. It is a pity -- I could be sending 25 amps out but it is 100 degrees right now and poor old Edison cannot meet the air conditioner demand. This is why many PV systems go with batteries, but that adds a great deal of additional cost, maintenance and complexity.We are in an extraordinary heat wave right now that is magnifying my other AC and grid problems I am having. I am fortunate to even have power right now given the current load on the grid.
The PV systems are just different than what the utility workmen typically see and they, of course, have no way of knowing if it was installed right, is anti-islanding, etc. I don't blame them.I had the old analog meter on the house for a month before they came out and installed the digital meter (it was not necessary as the old analog ones run forward and backwards, but I think they can read the digital meters remotely). They came out and said they had to change the meter as it would not run backwards (which it was and was recording the spin-down). It is just new for most of them.
Thank you,
doug
That is pretty close -- shouldn't be any voltage diff. there.
...
Actually, according to your link, 3% of full scale (when the meter was new, if you read it very carefully). That means that if you are measring 240V, +/- 7.5V; if you are measuring 268, +/- 15V.
That seems too low. California ISO is not indicating a shortage now. (
You are right -- DMM's are accurate to .25% or better and that the 260 is only accurate to 2 or 3%). This is a question not a claim -- but isn't 3 percent of 240 is 7.2 V?
I also think Simpson is very conservative in their ratings because of who they supply. Right now I have the Simpson plugged in next to the inverter, and the readings from the 6 mos. old digital meter on the inverter and the
20 year-old Simpson are identical. I guess they both could be wrong exactly the same, but I do have a great deal of faith in that meter. I sent it in for calibration some years ago and it still was in spec.I am not trying to get you to buy one, but when I worked for the railroad as a signalman -- installed and maintained the train signals and the crossing protection (gates at the crossings), we were issued 260's to test and calibrate the signals and gates. Human life depended on accurate readings. Technology has progressed, and I will not argue the point that DMM's are not more accurate, but I would like to believe that the 260 is accurate enough for me to read line voltage.
Great site, thank you. I never knew it existed.
best,
doug
I would ask the inverter manufacterer if they can modify the protection circuits to add a time delay so that it does not kick off line for a momentary problem...
Voltage dip during compressor startup is just a fact of life and the inverter should be designed to deal with it.
Mark
3% of 240 is 7.2. 3% of 250 is 7.5. The accuracy is specified as a percentage of full scale, not a percentage of the reading. So for 240 V you would use the 250 V scale. For 268 V you would use the 500 V scale with an accuracy of 15 V.
It all depends upon what you are trying to do. It probably is (or with careful use, can be) accurate enough for most uses.
I just noticed what appears to be a revision date "11/2/04" at the bottom of the spec sheet. Do they still make it?
Yes, industrial supply houses like Grainger and MSC sell them. I think McMaster-Carr also:
Hand-made in the US, phenolic cases, they really are a joy to use.
best,
doug
M Q,
For what it is worth, Simpson is owned by a native American tribe and has a pretty interesting history:
best,
doug
Doug,
I've seen several posts in this thread where the possibility of reducing the "dropout" voltage of the inverter is mentioned.
But... what about reducing the amount of *time* the inverter is offline? Surely it isn't necessary for the unit to decouple for 5-10 minutes just due to a 250 mS droop from a motor starting?
I really think you need to contact the dealer or manufacturer of the inverter. I can't imagine they've never seen this problem before!
Eric Law
AC cycles on. I have 200 amp service to the
inverter is an SMA 6000U with a minimum grid
either momentarily depresses line voltage below
to maintain the sync (either one of these
wait mode monitoring the state of the grid).
that Edison must have connected it to the
with a larger gauge drop?
start" so it does not depress the line
array to go off-line for five or ten minutes
believe it is good for the inverter, either.
You need to reduce your cooling costs.
You can do it with an upgrade to a cooling system that captures the earths cooling system -geothermal.
Or if you don't want to spend that kind of money, you can buy several small very high efficiency wall units with EER of 10.7 to 11.5 and put them in the rooms where you spend most of your time - especially the bedroom. You can cool a very large bedroom with 5 amps @ 110v with a
6,000 btu window or wall unit,and leave the rest of the house alone since you are not using it at night. Same goes for the other rooms you move to during the day.It should cut your cooling bill easily in half.
...
Wow. The price performance is worse than I thought. For $40 you can get a cheap BK Precision 2703B digital meter with an AC accuracy of 1.5% of reading + 4 (least significant) digits. That would be 4V for a 240V reading or 4.4V for a 268V reading. And you don't have parallax issues adding to your error like you do with an analog meter.
DC accuracy for the BK meter is .5% of reading + 1 (LS) digit.
Accuracy of an analog meter increases toward midscale. The greatest inaccuracy is at full scale (either way - 0 or 500). I would think a
260 would give very accurate readings measuring 240VAC and 268VAC on the 500VAC scale, approaching the accuracy of a DVM.I used a Triplett 310 for decades before moving to Fluke DVMs. The early DVMs were a bear to use on AC - most measured in peak to peak voltages and you had to do your own calculations for RMS. Now I don't think I could live without my Fluke 77.
Just my 2/100$.
Even if the analog meter was exactly accurate, you would loose a percent or so in readability due to paralax between the needle and the scale unless the scale has a mirror (as some do).
The Fluke 77 actually doesn't measure true RMS. Most meters (analog or digital) do not. They just measure the P-P voltage and scale the reading as if it were a sin wave. I have used a number of analog meters over the decades. I now have a Fluke 75 and Fluke 87 III (true RMS) and would never want to go back to analog meters. If you are measuring a true sin wave, true RMS meters tend to be slightly less accurate.
I think all good analog meters (all Simpson's anyway) have mirrored scales, so parallax really is not an issue. I always do mid-scale readings on the
260.Using an analog meter probably takes a little more skill and time to be as accurate, though. It is pretty hard to misread a DMM. If I were making my living with a meter in my hand and time was money, I would use a DMM.
As someone who occasionally needs to use a meter, and someone who appreciates Starrett mics, S-K wrenches, Klein hand tools, etc., I truly appreciate a quality, hand crafted tool. There is just something special about the experience of using a quality analog VOM compared to reading a digital display. Sort of like watching a movie or reading the book. The book takes more time, but it is far more engaging. :)
Won't argue, though, DMM's are cheaper and faster...but sometimes that really does not matter.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.