diy solar

diy solar

No AC loads yet Quattro reports a 244W AC draw while charging?? Pic attached.

Time to resurrect this thread and try to figure out this problem. I've contacted my Victron dealer who in turn contacted Victron.. still waiting on a response.

I now have test loads averaging 165W consisting of:
Starlink ~ 45W
Two freezers ~45W
Incandescent light bulb ~75W

The following graph shows a huge spike in Output Power (AC loads on L1). This happens consistently EVERY time the charger kicks in. I put an amp clamp on L1 out while charging and have observed roughly one amp concurring with my test loads. The issue seems data related.

Mysterious AC Loads.jpg

Here is a screenshot of VRM Dashboard from the same time the above example was captured:
52717-hi-ac-draw-dashboard.jpg



PV input has no effect on AC Output Power (AC Loads on L1):
1683203166101.jpeg

It seems like the efficiency of the charger is to blame here and that the AC Loads has become a dump for losses BUT if this were true then why does the graph @sunshine_eggo posted above not show the same jump in AC Loads?

@Will Prowse ?
 

Attachments

  • 1683196676044.jpeg
    1683196676044.jpeg
    109.7 KB · Views: 1
The inverter has a current sensor on grid and generator ports, and current sensor on actual inverter AC path.

The AC output port current is not directly measured, it is a calculation of subtracting measured AC input current and measured inverter AC current.

CT transformers are not super accurate and are not linear versus current level. The output current reading is likely just CT measurement error at the high current.

They usually do calibration at low AC current because it is more critical to prevent inverter back feeding to AC input. They also hold an error margin on AC input current which prevents AC input max current limit for inverter load shaving from being set lower than about 5 amps to prevent any possible back feed current due to inaccuracy in CT current sense reading and inverter output adjustment response time lag to changing AC output loads.
 
Last edited:
RCinFLA said:
The inverter has a current sensor on grid and generator ports, and current sensor on actual inverter AC path.

The AC output port current is not directly measured, it is a calculation of subtracting measured AC input current and measured inverter AC current.

CT transformers are not super accurate and are not linear versus current level. The output current reading is likely just CT measurement error at the high current.

They usually do calibration at low AC current because it is more critical to prevent inverter back feeding to AC input. They also hold an error margin on AC input current which prevents AC input max current limit for inverter load shaving from being set lower than about 5 amps to prevent any possible back feed current due to inaccuracy in CT current sense reading and inverter output adjustment response time lag to changing AC output loads.
Click to expand...

I'm going to try and run my generator this weekend and see if I see the same thing on my identical Quattro. I just don't recall seeing this happen on my system on the rare occasions that I need to run the generator.
 
RCinFLA said:
The inverter has a current sensor on grid and generator ports, and current sensor on actual inverter AC path.

The AC output port current is not directly measured, it is a calculation of subtracting measured AC input current and measured inverter AC current.

CT transformers are not super accurate and are not linear versus current level. The output current reading is likely just CT measurement error at the high current.

They usually do calibration at low AC current because it is more critical to prevent inverter back feeding to AC input. They also hold an error margin on AC input current which prevents AC input max current limit for inverter load shaving from being set lower than about 5 amps to prevent any possible back feed current due to inaccuracy in CT current sense reading and inverter output adjustment response time lag to changing AC output loads.
Click to expand...
Sounds like I've got some research to catch up on.. I'm sure I will have questions later :) Thank you.
 
Because only the grid OR generator input is allowed to be active at any given point in time, there is no reason to put CT sensors on both AC inputs. A single CT current sensor is put after the pass-through relays to measure AC input current, covering both AC input use cases.

Generator or grid input current should have the same CT accuracy, for the same AC input current level, since it is the same CT sensor.

CT sensors usually have greater error on current reported the greater the absolute current level. This is due to the CT sensor magnetic core 'S' curve non-linearity.

You may get a slightly different result if there is distortion from generator output. This would cause some variation on current demanded from inverter during distorted, sinewave waveform matching between generator and inverter.

(ignore the current numbers in diagram, it was for a different inverter. I believe the 5kW Quattro has 100 amp pass-thru relays)

inverter functional diagram.png
 
Last edited:
Totally fair, but I always run the carb bowl dry by shutting the fuel off until it dies. Used fuel stabilizer as well.

I ran a Coleman 3000 completely dry before mothballing it, and it started on the 2nd pull after.... NINE YEARS... :)

I'm confident there should be no issues. :)
 
RCinFLA said:
The inverter has a current sensor on grid and generator ports, and current sensor on actual inverter AC path.
Click to expand...
So basically an amp clamp on AC in and another on the inverter right?
RCinFLA said:
The AC output port current is not directly measured, it is a calculation of subtracting measured AC input current and measured inverter AC current.
Click to expand...
Am I correct to assume generator amps and charging amps displayed in dashboard are reported from the internal AC input CT and inverter CT sensors?
1683249033193.png
RCinFLA said:
CT transformers are not super accurate and are not linear versus current level. The output current reading is likely just CT measurement error at the high current.

They usually do calibration at low AC current because it is more critical to prevent inverter back feeding to AC input. They also hold an error margin on AC input current which prevents AC input max current limit for inverter load shaving from being set lower than about 5 amps to prevent any possible back feed current due to inaccuracy in CT current sense reading and inverter output adjustment response time lag to changing AC output loads.
Click to expand...
There must be others around here with Victron or similar inverters that report somewhat erroneous AC out. Perhaps mine is a just a little more noticeable.
 

Attachments

  • 1683248817827.jpeg
    1683248817827.jpeg
    88.9 KB · Views: 1
sunshine_eggo said:
Totally fair, but I always run the carb bowl dry by shutting the fuel off until it dies. Used fuel stabilizer as well.

I ran a Coleman 3000 completely dry before mothballing it, and it started on the 2nd pull after.... NINE YEARS... :)
Click to expand...
Nice!
sunshine_eggo said:
I'm confident there should be no issues. :)
Click to expand...
I struggle with small engines every year.. often because they are not used on a regular basis. Running them dry is definitely a good thing to do!
 
NorthTown2022 said:
So basically an amp clamp on AC in and another on the inverter right?

Am I correct to assume generator amps and charging amps displayed in dashboard are reported from the internal AC input CT and inverter CT sensors?
View attachment 147846
Click to expand...

Since DC current shunts are expensive and take up space, the battery current is often calculated from battery voltage, inverter AC voltage and AC current, with a derating for inverter efficiency based on absolute power level. I believe Quattro offers an optionally purchased battery monitor addition that does actually measure battery current.

You can usually tell it is calculated from AC power when charging AC input current remains relatively constant but charging bulk DC battery current reading drops a bit as battery voltage rises during charging. Inverter AC input current is actually doing the charging DC bulk current regulation.

In the picture, 2061 watts is calculated net charging power after inverter AC CT sensor reading x AC input voltage x inverter efficiency.

The 2061 watts charging is with efficiency derating, for that power level it is likely about 90-92%. That would mean actual input AC current to inverter is about 2290 watts / 122.2v =18.74 amp rms.

The VRM Dashboard has an issue as to what it should present. If it reports efficiency corrected number for charging current, which it should, there is going to be a discrepancy in the AC input, AC output, and charging power summation when inverter efficiency is taken out. They likely don't want to clearly expose there is over 200 watts being pissed away in efficiency loss.

In this case, the efficiency loss, about 229 watts, is missing from the power summations. At least that would take the unaccounted for 414 watts AC output power down to 414w - 229w = 185 watts. That would be 185w/122.2vac = 1.5 amps AC of measurement error.

That is a believable error for the 22.8 A absolute AC input CT amperage measurement, minus the 18.7 amps absolute inverter AC input CT amperage measurement.

Some of the approximate 1.5 amp error is due to AC input CT sensor reading, some of 1.5 amp error is due to inverter AC input CT sensor reading. There is also a little contribution due to AC voltage reading accuracy.

There is still some power unaccounted for to run the controller electronics, display, and pass-through relay coil. That summation is probably in the ballpark of 5 watts. Also, although pretty good, the charging AC power factor is about 0.90-0.92 which may add some error in power calculations.

The charging AC current profile does follow a sinewave shape fairly well due to PWM switching, but the charging AC sinewave current does drop out for a small time period gap around zero crossings of sinewave. This causes some error in CT sensor rms current reading. CT sensor is not to be blamed for this AC current shape distortion.
 
Last edited:
RCinFLA said:
Since DC current shunts are expensive and take up space, the battery current is often calculated from battery voltage, inverter AC voltage and AC current, with a derating for inverter efficiency based on absolute power level. I believe Quattro offers an optionally purchased battery monitor addition that does actually measure battery current.
Click to expand...
Victron offers a BMV-712 battery monitor which I believe offers more features and better Bluetooth range over Victron Smart Shunt which I currently have installed.
RCinFLA said:
You can usually tell it is calculated from AC power when charging AC input current remains relatively constant but charging bulk DC battery current reading drops a bit as battery voltage rises during charging. Inverter AC input current is actually doing the charging DC bulk current regulation.
Click to expand...
I'll keep an eye on this.
RCinFLA said:
In the picture, 2061 watts is calculated net charging power after inverter AC CT sensor reading x AC input voltage x inverter efficiency.

The 2061 watts charging is with efficiency derating, for that power level it is likely about 90-92%. That would mean actual input AC current to inverter is about 2290 watts / 122.2v =18.74 amp rms.


The VRM Dashboard has an issue as to what it should present. If it reports efficiency corrected number for charging current, which it should, there is going to be a discrepancy in the AC input, AC output, and charging power summation when inverter efficiency is taken out. They likely don't want to clearly expose there is over 200 watts being pissed away in efficiency loss.
Click to expand...
As unattractive the calculation may be it would sure be nice to see it accurately displayed, or at least my AC/DC out watts! Can't blame them I guess :rolleyes:
RCinFLA said:
In this case, the efficiency loss, about 229 watts, is missing from the power summations. At least that would take the unaccounted for 414 watts AC output power down to 414w - 229w = 185 watts. That would be 185w/122.2vac = 1.5 amps AC of measurement error.

That is a believable error for the 22.8 A absolute AC input CT amperage measurement, minus the 18.7 amps absolute inverter AC input CT amperage measurement.

Some of the approximate 1.5 amp error is due to AC input CT sensor reading, some of 1.5 amp error is due to inverter AC input CT sensor reading. There is also a little contribution due to AC voltage reading accuracy.

There is still some power unaccounted for to run the controller electronics, display, and pass-through relay coil. That summation is probably in the ballpark of 5 watts. Also, although pretty good, the charging AC power factor is about 0.90-0.92 which may add some error in power calculations.

The charging AC current profile does follow a sinewave shape fairly well due to PWM switching, but the charging AC sinewave current does drop out for a small time period gap around zero crossings of sinewave. This causes some error in CT sensor rms current reading. CT sensor is not to be blamed for this AC current shape distortion.
Click to expand...
Thank you for the detailed response this is starting to make more sense.

One thing I'm still a bit confused about though is weather or not the internal AC input and inverter CT sensors are providing amperage readings for the above calculations since I have a smart shunt installed?
 
If you have the optional battery shunt, the battery DC current displayed should be a direct reading from shunt. With comm link it can also get info from an external Smartshunt or 712 monitor.

I am not sure if it uses direct DC shunt current reading for all entangled calculations. For example, it may indirectly modify the calibration of AC sourced charging bulk current regulation by inverter AC input by 'tweaking' based on actual battery shunt DC current reported. If remote battery shunt DC current does not reasonably agree with the raw AC current, AC voltage, and battery DC voltage calculations, it will ignore the battery DC shunt reported current in bulk charging current regulation.

The VRM screen still lacks displaying inverter losses so the 'goes into', and 'goes out of', are not going to add up, even if the actual inputs and outputs reported are accurate. Then there is AC power factor and crest factor further mucking up the numbers. (Victron is one of the few inverters that actually specs maximum crest factor load allowed)
 
RCinFLA said:
If you have the optional battery shunt, the battery DC current displayed should be a direct reading from shunt. With comm link it can also get info from an external Smartshunt or 712 monitor.

I am not sure if it uses direct DC shunt current reading for all entangled calculations. For example, it may indirectly modify the calibration of AC sourced charging bulk current regulation by inverter AC input by 'tweaking' based on actual battery shunt DC current reported. If remote battery shunt DC current does not reasonably agree with the raw AC current, AC voltage, and battery DC voltage calculations, it will ignore the battery DC shunt reported current in bulk charging current regulation.

The VRM screen still lacks displaying inverter losses so the 'goes into', and 'goes out of', are not going to add up, even if the actual inputs and outputs reported are accurate. Then there is AC power factor and crest factor further mucking up the numbers. (Victron is one of the few inverters that actually specs maximum crest factor load allowed)
Click to expand...
Sounds like adding a second Smart Shunt for DC loads may be a partial solution.

After a bit of googling and read over your explanation several times. I've got a better understanding now. Thank you.
 
Whelp, I ran the genny last night, and I'm completely confused. I see similar... but opposite behavior.

My system is split phase, but I only have the generator connected to the L1 inverter. I have plotted AC in L1, AC out L1 and AC out L2:

1683468654126.png

My system is under BMS control, and I have two shunts - the Batrium shunt and a BMV-702 in series. They always read the same current value within 0.1A, and they are both measuring only the net current flow to/from battery.

L2 is clearly unaffected, but I'm seeing the exact opposite behavior on L1 with a DROP instead of a bump.
 
sunshine_eggo said:
Whelp, I ran the genny last night, and I'm completely confused. I see similar... but opposite behavior.

My system is split phase, but I only have the generator connected to the L1 inverter. I have plotted AC in L1, AC out L1 and AC out L2:

View attachment 148167

My system is under BMS control, and I have two shunts - the Batrium shunt and a BMV-702 in series. They always read the same current value within 0.1A, and they are both measuring only the net current flow to/from battery.

L2 is clearly unaffected, but I'm seeing the exact opposite behavior on L1 with a DROP instead of a bump.
Click to expand...
Similar but opposite is right. Not really the result I expected to see. As for the drop I wonder if this has something to do with pass-through.
 
You must log in or register to reply here.

Similar threads

L
Refit electrical system 35ft/11m catamaran with LiFePo4. System recommendations anyone?
Replies
5
Views
640
5teve
5
Back
Top