Generator surging (hunting) when charging sytem thru stacked SW 5548s.
Options
Comments
-
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.
God Bless you Chris, hope that was an allergy sneeze.
-
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »I caught our XW doing this this morning.
Does not seem to be caused by loads. Loads vary much less then the draw on AC2.
Mine does it all the time when "Gen Support" limit moves lower than "Charging Percentage" limit. Doesn't depend on the generator. It was doing this with grid too.
Your Honda handles this real well. Generac moves between 59 and 60 and I can hear it.
Like John, I managed to do some "better" settings, but they're below generator capacity. Perhaps, this will make generator last longer. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.Does not seem to be caused by loads. Loads vary much less then the draw on AC2.
This is only the third time that I recall since we've owned the inverter that it has started the generator for low DC volts. So I haven't had much of an opportunity to observe it doing this.
I think what I concluded is that the range "burner" was causing a couple hundred watts fluctuation in the load. I got Gen Support amps set to 17 (17 x 240 = 4 kW). The actual continuous rating of the generator is 3.6 kVA. It will do 4 kVA for 30-60 minutes and maintain 60 Hz and 240V, so that's what I run it at for Gen Support because those run times for overload are rarely more than 30 minutes.
The XW will limit the AC input amps on the highest loaded leg to about an amp below the Gen Support setting by varying the draw for the charger. With the fluctuating load it was trying to use the charger to regulate the AC2 amps to about 16 on L1 because that leg had a slightly higher load on it than L2. But it can't adjust the charger fast enough to keep the load on the gen constant. So the end result is that a load that's varying by a couple hundred watts during battery charging causes easily double that amount of swing (~3,300-3,700) on the generator.
Without the transformer for our little 4 kVA generator it's a disaster. The legs would be severely imbalanced and probably hardly nothing left for battery charging because it would regulate based on the highest loaded leg. So if L2 only had 3 amps load and L1 had 16 it would only allow ~2.3 kVA input from the gen and hardly no battery charging (because of the high load leg) instead of average ~3.5 kVA with the transformer and we get battery charging out of it.
That's my theory on it.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »The XW will limit the AC input amps on the highest loaded leg to about an amp below the Gen Support setting by varying the draw for the charger. With the fluctuating load it was trying to use the charger to regulate the AC2 amps to about 16 on L1 because that leg had a slightly higher load on it than L2. But it can't adjust the charger fast enough to keep the load on the gen constant. So the end result is that a load that's varying by a couple hundred watts during battery charging causes easily double that amount of swing (~3,300-3,700) on the generator.
Yes, it certainly cannot adjust charging fast enough, and often overshoots.
I suspect that the loads are not really varying. Rather, the measurements are varying. When the charging starts, the load measurements get much worse compared to pass-through or inverting modes. I observed, when 100A charging begins, load W goes up 200-300W, but load VA goes down the same amount, so VA becomes substantially lower than W
. As the current goes down, it improves.
I have two theories:
- there's some interference from charger on output current sensors.
or
- they don't really measure output power, but rather calculate it as a balance between input and charging.
Perhaps these measurement bugs make your loads look jumpy and cause charger hunting. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.load W goes up 200-300W, but load VA goes down the same amount, so VA becomes substantially lower than W. As the current goes down, it improves.
Actually, that completely normal. The Volt-Amps, or Kilovolt-Amps is watts corrected for Power Factor. I've found the XW's measurements to be quite accurate, actually, when compared to my high-end Fluke meters. My Fluke's have maximum value capture and a bunch of other features that can be used to diagnose generator surge problems, etc.. And the XW agrees with those meters every time.
I made another video because the inverter started the generator for Load Amps this morning in Gen Support mode, then after it shut off I made a clip showing the huge difference the auto transformer makes. As soon as the video gets done processing I'll upload it and post it - you'll find it interesting to see what a dramatic difference it makes for your generator.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »Actually, that completely normal. The Volt-Amps, or Kilovolt-Amps is watts corrected for Power Factor.
Power factor should be below 1. In this picture it is 1.65. This cannot be accurate. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.
Something is wrong there... VA is always equal to, or greater than watts:
Power = Volts * Amps * Power Factor (PF ranges from 0.0 to 1.0, but typically 0.6 to 0.95-1.00 depending on loads).
is there an alternate power source involved here? (generator, AC mains, etc. that is supplying some power, or floating the battery bank)?
-BillNear San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.is there an alternate power source involved here? (generator, AC mains, etc. that is supplying some power, or floating the battery bank)?
There's no problem. You have an imbalance there. The XW measures volt-amps and calculates watts. Only with a balanced polyphase system can you have a direct accurate relationship between real and apparent power. Any imbalance with a split-phase or polyphase power system is not defined. So the XW applies a programmed PF curve to that formula. The PF is only accurate with a completely balanced load.
So when looking the XW's meters, forget what it says on the watts side. The VA side shows actual measured and there is sometimes a slight computational delay, or the display showing rounded off figures while the XW's internal computer is using decimals out to 10 places, or whatever. So the VA reading will sometimes lag by ~2 seconds behind what it shows for amps and volts on the display, per leg.
I've verified this with my Fluke meters several times.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »There's no problem. You have an imbalance there. The XW measures volt-amps and calculates watts. Only with a balanced polyphase system can you have a direct accurate relationship between real and apparent power. Any imbalance with a split-phase or polyphase power system is not defined. So the XW applies a programmed PF curve to that formula. The PF is only accurate with a completely balanced load.
If you look at it as two separate legs, then PF certainly would apply to each leg. The data that I got from XanBus for each leg separately show exact the same pattern.
If they were applying PF curve here, the PFs on their curve would certainly be between 0 and 1, not 1.65.ChrisOlson wrote: »So when looking the XW's meters, forget what it says on the watts side. The VA side shows actual measured and there is sometimes a slight computational delay, or the display showing rounded off figures while the XW's internal computer is using decimals out to 10 places, or whatever. So the VA reading will sometimes lag by ~2 seconds behind what it shows for amps and volts on the display, per leg.
If there's a delay, it wouldn't really matter. It shows this picture without changes for quite some time. The discrepancy is consistent over the whole range of loads, and is directly proportional to the charging current.ChrisOlson wrote: »I've verified this with my Fluke meters several times.
I have a Fluke-287, but it doesn't measures power. So, my plan is to set up exact measurements to figure out what exactly is measured wrong.Something is wrong there...
It certainly is. When I sent this picture to Xantrex tech support, they told me that it's because of low measurement accuracy - 5% of full scale (+-300W), but it's so consistent that I seriously doubt that.
I took it when chaging from Grid long time ago. No other powers applied. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.If they were applying PF curve here, the PFs on their curve would certainly be between 0 and 1, not 1.65.
Well, that would be a negative power factor, which is not abnormal either. Like Bill said it can be caused by power from another source being returned to the source being measured (like AC Mains power). This can even be caused by reactive loads.
The main thing to remember is that "watts" or kilowatts" are typically only used in the US and a couple other countries on earth that use a 60 Hz current system. Anywhere else you go on the planet, when it comes to generator and inverters, etc., they are all rated and measured by kVA because the kVA rating does not have to deal with Power Factors and it represents the actual capacity of a generator to the loads, etc..
Another consideration could be your inverter having a calibration error? Because ours is usually dead on:
Attachment not found.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.
Here's that video that I snapped this morning showing how the transformer enhances generator performance on a split-phase power system with a leg imbalance. Without the transformer, our little generator is not allowed to charge batteries much because it's being amp-limited based on the high leg amps. With the transformer it increased the output of the generator by about 1 kVA and that power can then go to charging batteries.
Basically it makes the most of your gen run time by not allowing the generator to run partially loaded. Putting the transformer between the gen and the inverter does not work. It balances the generator, but the inverter still amp-limits the generator based on its high leg amps. Putting the transformer on the output of the inverter balances both so the inverter "sees" a lower load on the higher loaded leg.
[video=youtube_share;IdMYUUh5j2s]
http://youtu.be/IdMYUUh5j2s[/video]
People are always reluctant to buy these transformers because they $400-500 and they think they don't need it. So they only get bought, primarily, for stepping up 120V to split phase to run well pumps. But they got more uses than that.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »Well, that would be a negative power factor, which is not abnormal either. Like Bill said it can be caused by power from another source being returned to the source being measured (like AC Mains power). This can even be caused by reactive loads.
Negative is below 0. 1.65 is above 1. It is opposite of negative. And my loads are mostly resistive. VA is just a touch higher than W when inverting.ChrisOlson wrote: »Another consideration could be your inverter having a calibration error?
Perhaps. I'll need to find out. However, I don't think so. This is my second XW and they're both exactly the same.ChrisOlson wrote: »Because ours is usually dead on:
It depends on charging current. It's very pronounced at 100A. At about 50A, VA and W are about the same. Without charging, VA goes slightly higher than W as it should be.
You probably won't be able to summon enough charging current and still have some loads to be able to see that.
But someone with grid or a big generator could. Anyone reading this wants to try? -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.Negative is below 0. 1.65 is above 1. It is opposite of negative.
Not necessarily. A negative PF can happen when the current leads or lags the voltage by 90 electrical degrees or more, and will give readings exactly like you have in your photo. Usually caused by a reactive load acting as a generator, returning power to the source. But can also be caused, like Bill said, by AC Mains power being present, or some other power source feeding the batteries causing a backfeed.You probably won't be able to summon enough charging current and still have some loads to be able to see that.
True. On a cool day, and if I buy some good gas without ethanol for it, I can get 60 amps out of the charger with our little Honda. I need to put a Roots blower on it and boost it to sea level manifold pressure, and put an intercooler on it. Then I'll get 100 amps out of that charger. I've been known to do things like that too 8)
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »Not necessarily. A negative PF can happen when the current leads or lags the voltage by 90 electrical degrees or more, and will give readings exactly like you have in your photo. Usually caused by a reactive load acting as a generator, returning power to the source. But can also be caused, like Bill said, by AC Mains power being present, or some other power source feeding the batteries causing a backfeed.
If you have a pure sine wave, then 90 degree shift will give you PF equal to zero. On average power doesn't move at all.
The 180 degree shift will give you PF equal to -1, which means that the power is moving backwards. This requires a power source on the load side.
If you move it further, it'll start moving towards zero again, and will be zero at 270 degree shift.
After that, it'll be positive again until it becomes 1 at 360 degrees.
It cannot be above 1. This is a mathematical impossibility. PF of one means that every ounce of available voltage has been 100% utilized. PF is one when the current exactly follows the voltage with purely resistive load. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »True. On a cool day, and if I buy some good gas without ethanol for it, I can get 60 amps out of the charger with our little Honda. I need to put a Roots blower on it and boost it to sea level manifold pressure, and put an intercooler on it. Then I'll get 100 amps out of that charger. I've been known to do things like that too
Actually, you probably can see this effect if you produce enough charging current by combining RE charging with generator charging. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.Actually, you probably can see this effect if you produce enough charging current by combining RE charging with generator charging.
It might. But on a good day I can produce 180 amps of RE charging, so the generator never runs for that. The only time it ever gets used for battery charging is if I start it manually, or we get a week of bad weather.
I finally got my gen disabling switch wired in on wires 5 & 6, and my manual on/off switch on wires 7, 8 & 11 on the AGS harness today. I had never gotten that done and got tired of having to manually disable the generator at the SCP just to check oil in it. Of course, again it's not like the old SW Plus GSM. Once you hit that disabling switch the SCP flashes a fault and says:
Attachment not found.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.
I took a half-dozen different measurements on our XW at various loads and arrived at the conclusion that the VA reading, as I had known from before, appears to be obtained from actual measurements the XW makes on the split-phase RMS voltage and amps. I cannot find an instance where it is off by more than rounding error and reasonable meter reaction time and accuracy (+/- about 3-5%), even at virtually no load on our system with just parasitic loads going in the house - electronics in stuff mostly:
Attachment not found.
However, the watts reading appears to have nothing to do with Power Factor, or the VA reading, as I had assumed. It appears to be actual measured DC consumption from the battery bank, at least on the Load meter. I don't know what it does on the AC2 Input meter. Over several different readings with my Snap-on AVR on the DC supply to the inverter, multiplying the real-time DC amps consumed by the inverter by voltage measured at the inverter power studs yields what it shows for watts within rounding error, and the reasonable reaction time and accuracy of the meters.
My Fluke is +/- 2% and I doubt the internal hall effect sensors in the XW could be much better than +/- 5% on accuracy.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »However, the watts reading appears to have nothing to do with Power Factor, or the VA reading, as I had assumed. It appears to be actual measured DC consumption from the battery bank, at least on the Load meter. I don't know what it does on the AC2 Input meter. Over several different readings with my Snap-on AVR on the DC supply to the inverter, multiplying the real-time DC amps consumed by the inverter by voltage measured at the inverter power studs yields what it shows for watts within rounding error, and the reasonable reaction time and accuracy of the meters.
So, they just don't display the correct number for AC load watts. Probably the best way is to just disregard it. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.So, they just don't display the correct number for AC load watts
The thing is, it's not easy to calculate AC watts. It usually requires a regular wattmeter because the voltage and current waveforms have to be measured over a precise period of time, the measurements must be simultaneous, and the average must be calculated over the measurement time period. VA is easy to calculate and measure, and why it is the standard in AC circuits for measurement for load, breaker and wire sizing.
So if think about it, it is unrealistic to expect an instantaneous and accurate watt measurement on a digital meter because of the time involved for the measurement, and the reasonable reaction time of digital meters which are rapidly changing values that your eye cannot discern. The good old analog watt meter (and ammeter) is still hard to beat because it shows information that registers instantly on the human brain by needle position.
It is as easy on a DC circuit to get an instantaneous watt reading as it is on AC to get an instantaneous VA reading. And that's why I had a sneaking suspicion that the watts shown on the SCP has nothing to do with the VA reading, or if it does, at best could be considered a long term average. Watts, being easy to measure on DC circuits, is why I compared the two with my AVR on the inverter input. And I found a direct relationship there.
So it entirely possible to see a higher watt reading than VA, especially with leg imbalances where the inverter is less electrically efficient.
Again, I do not know where the watt measurement is coming from in the inverter (especially with AC input). But I suspect it is coming from some hall effect sensors in the DC section of the inverter someplace. I've never had the cover off the right side of the inverter but I might take that off and investigate a bit and see if I can figure it out.
Edit:
Ha! Just verified by Schneider. The inverter only has two measurement sections on power - DC in and AC out. The watts is DC in, the VA is AC out. The values on AC1 and AC2 are derived values based on AC out to loads and DC to batteries, and are not actually measured on the AC inputs. Due to the limitations in the measurement circuitry, just like changing scales on a high end meter, the values shown on the SCP are least accurate at light loads because they're on the high scale all the time.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »Again, I do not know where the watt measurement is coming from in the inverter (especially with AC input). But I suspect it is coming from some hall effect sensors in the DC section of the inverter someplace. I've never had the cover off the right side of the inverter but I might take that off and investigate a bit and see if I can figure it out.
I've never opened the right side too. If you do, please post the pictures.
There are four current transformers on the left side. That should be enough to measure AC amps on both legs on both input and output. If they do high frequency measurements (even 3-5kHz would do) they can calculate watts with good precision, but I'm afraid you're right and they don't do high frequency, but rather use some sort of algorithm to estimate AC watts from DC side measurements. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.but I'm afraid you're right and they don't do high frequency, but rather use some sort of algorithm to estimate AC watts from DC side measurements.
I don't think so. I'm pretty sure there's no shunts on the DC side - probably hall effect sensors in there. And hall effect sensors can be accurate at a certain range, but can be highly inaccurate outside that range too, or affected by magnetic fields from nearby stuff.
I think it's important to remember that the SCP gives you a reasonably accurate representation of loads and battery current for your system. But it is not designed to be as accurate as precision measuring equipment. That's why they make precision measuring equipment. Even the speedometer in your car is probably off by 5%. If you want to measure it precisely, then use radar. Otherwise it's "good enough" to know how fast you're going.
I found that the actual and what it showed on the SW Plus was sometimes off too. And on different days it would be different sometimes. So when you make a setting on these inverters, then end up with an overload or surge problem on a generator, it pays to pull out the precision measuring equipment to diagnose the problem. It might be the difference between what the radar says vs what your car speedometer says, that determines whether or not you're going to get the ticket, and the precision radar usually wins - although even it can be inaccurate sometimes.
--
Chris -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »Here's that video that I snapped this morning showing how the transformer enhances generator performance on a split-phase power system with a leg imbalance. Without the transformer, our little generator is not allowed to charge batteries much because it's being amp-limited based on the high leg amps. With the transformer it increased the output of the generator by about 1 kVA and that power can then go to charging batteries.
Basically it makes the most of your gen run time by not allowing the generator to run partially loaded. Putting the transformer between the gen and the inverter does not work. It balances the generator, but the inverter still amp-limits the generator based on its high leg amps. Putting the transformer on the output of the inverter balances both so the inverter "sees" a lower load on the higher loaded leg.
[video=youtube_share;IdMYUUh5j2s]http://youtu.be/IdMYUUh5j2s[/video]
People are always reluctant to buy these transformers because they $400-500 and they think they don't need it. So they only get bought, primarily, for stepping up 120V to split phase to run well pumps. But they got more uses than that.
--
Chris
Chris, I'm unsure if I can install a balancing transformer on my system and still maintain capacity. With stacked Sw5548s will 2 transformers be required? Can you point me to a wiring diagram with that kind of a installation? Thanks......John -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.ChrisOlson wrote: »I think it's important to remember that the SCP gives you a reasonably accurate representation of loads and battery current for your system. But it is not designed to be as accurate as precision measuring equipment. That's why they make precision measuring equipment. Even the speedometer in your car is probably off by 5%. If you want to measure it precisely, then use radar. Otherwise it's "good enough" to know how fast you're going.
Sure precision is not that important for SCP. However, these very measuremets (or calculations) are used to regulate the load on generator. It's quite possible that the oscillations in charging current are caused by the biased measurements. -
Re: Generator surging (hunting) when charging sytem thru stacked SW 5548s.Sure precision is not that important for SCP. However, these very measuremets (or calculations) are used to regulate the load on generator. It's quite possible that the oscillations in charging current are caused by the biased measurements.
It's possible I suppose. However, the XW is considerably better than our SW Plus units were in this regard. And as shown in my video our Honda generator has little or no problem with it at all. I also believe our auto-transformer is helping tremendously to get the most from the generator (and inverter). Due to the nature of our loads, it's virtually impossible to balance it by swapping circuits around and having the legs out of balance by only 3 or 4 amps compared to 17-20 amps makes a big difference, especially with a smaller generator.
So these are the things I've learned over time and why we got what we got today. We bought the stacked inverters originally because we were told it was the cat's meow. It wasn't. Running series stacked dual inverters for normal loads is horribly inefficient. It took minimum 2.5 -3 kW load for them to reach their peak efficiency. At light loads below 1 kW (500 watts per inverter) they were ~50-60% efficient. A bigger single inverter with a transformer was better. But still suffered from too many amps coming from one unit on that one wire. The XW is heads and shoulders above both of them and does a much better job of managing the generator on auto-start/stop.
We use our generators in two different ways. You depend on your generator for battery charging (as most do, and they use it as a last resort). Experience has shown me that using a generator for battery charging is not good, no matter which way you look at it. So we use our generator to manage power consumption so the RE can keep up and do the battery charging. There is some fundamental differences in how your system is set up for the two different uses of gen power, so we might be comparing apples and oranges.
I find it hard to do any nitpicking on the XW because I've had way worse inverters
--
Chris
Categories
- All Categories
- 243 Forum & Website
- 150 Solar Forum News and Announcements
- 1.4K Solar News, Reviews, & Product Announcements
- 191 Solar Information links & sources, event announcements
- 919 Solar Product Reviews & Opinions
- 256 Solar Skeptics, Hype, & Scams Corner
- 22.5K Solar Electric Power, Wind Power & Balance of System
- 3.5K General Solar Power Topics
- 6.7K Solar Beginners Corner
- 1K PV Installers Forum - NEC, Wiring, Installation
- 2.1K Advanced Solar Electric Technical Forum
- 5.5K Off Grid Solar & Battery Systems
- 438 Caravan, Recreational Vehicle, and Marine Power Systems
- 1.1K Grid Tie and Grid Interactive Systems
- 652 Solar Water Pumping
- 829 Wind Power Generation
- 626 Energy Use & Conservation
- 626 Discussion Forums/Café
- 303 In the Weeds--Member's Choice
- 91 Construction
- 124 New Battery Technologies
- 108 Old Battery Tech Discussions
- 3.8K Solar News - Automatic Feed
- 3.8K Solar Energy News RSS Feed