I actually did an experiment last week and used a twisted pair of CAT6 305meters straight off the drum(box) ... worked absolutely no issues. I think the input for the inverter CT is differential rather than one leg connected to "ground" and voltage measured from this point. Anyway, customer installed 110meters of CAT6 and the system is working perfectly!
You can have two fairly different situations using the same CAT6 cable. I am speculating here somewhat, but if you should use two pairs as the one conductor and the other pairs as the other conductor, then you will get a higher series inductance than if you use one conductor of each twisted pair as the one conductor and the other conductor of each twisted pair as the other conductor. If you keep the series inductance low and therefore the burden impedance of the CT, I can not reason why this cable could not run more than 15 meters. But of course it remains to be verified.
We use CT cable over long runs and it works fine. However we split the pairs and join the 4 white and 4 coloured ones together. So it remains twisted and protects against interference
A CT is a current source, and the inverter is looking for current. You are talking about voltage anomalies, noise, volt drop etc are not an issue. Your problem is using all pairs of a multi twisted pair to transmit the same signal. This is like not using a twisted pair at all!!. Get the proper size SINGLE twisted pair cable.
The Eastron energy meters is not that expensive, that is if you buy them in quantities. But of course it is just way easier to install the CT than an energy meter. BUT, we have one issue in South Africa that I think disqualifies the Modbus energy meters and that is that our prepaid meters allows very little reverse flow of energy. The secondary CT current that is fed to the SunSynk reacts immediately to any change in load (obviously) and therefore the inverter should be able to adjust quicker to sudden load variations, while rate of feedback from the energy meter will be much slower and therefore inverter reaction as well. I would therefore expect more nuisance trips when making use of an energy meter than using a CT. What do you think?
@@pieterkruger4134 I have no idea what purpose that would serve. You need to measure the property's input current so putting a CT in the inverter is totally bizarre.
@@greenpro-greenproconsultin7420 I've noticed another solar installer do that and he has installed hundreds of Sunsynks and he does this sometimes when the db is far and everything is on the inverter. So I was wondering. As I was wondering about this very topic and couldn't get a clear answer from my suppliers until I found this post
I would be very cautious of the above method. The reason is as follows. Modern households no longer have resistive type loads. This results in current drawn that may have harmonics into the kHz. Most new electronic meters either trip on reverse feed ( seen as a tapering fault) or you are charged for the reverse feed. (Meter cannot tell the direction of energy flow.) If you extend a cable from a CT clamp to the device that measures the CT output, you may encounter several problems: 1.Signal attenuation: The signal from the CT clamp may weaken as it travels through the cable, which can result in a lower signal amplitude at the measurement device. This can reduce the accuracy and reliability of the measurement. 2.Noise interference: The cable may pick up external noise sources, such as electromagnetic interference (EMI) or radio frequency interference (RFI), which can add unwanted signals to the CT output. This can distort the measurement and make it difficult to extract the true signal from the CT clamp. 3.Impedance mismatch: If the cable has a different impedance than the CT clamp or the measurement device, this can cause reflections and standing waves in the cable. This can result in signal distortion, phase shifts, and other measurement errors. 4.Grounding issues: If the CT clamp and the measurement device are not grounded properly, this can lead to ground loops and other interference issues. This can cause unwanted signals to be introduced into the CT output, which can affect the accuracy of the measurement. Although the above method tries to address item 1, and possible 2 (with twisted pairs) I don't think it addresses item 3. This may result in reverse feed at the incorrect time. (That may not be immediately apparent) In a nut shell without knowing the circuitry in the inverter that is used to measure the current, and how the inverter makes a decision to reverse feed it difficult to tell. For me it would be best to get hold of SunSynk (Keith I think) for an opinion.
Two CTs in series, if anything provides more harmonic filtering that the small CT. The intent of the 2 CTs is to have more current in the long wire, so that losses are minimised. There is a voltage loss but no current loss and as long as the VA consumed by the 2nd CT plus the cable resistance power wasted stays within spec of the 1st CT there are no losses that matter. Anyway, this solution costs R500 and it works. These fancy meters etc. solutions costs an order of magnitude more. Also the accuracy is not that important if you are controlling around zero.
Great! Thanks! _______________ Short: DE-NOISE. Simplest: By capacitor. De-ring by resistor. (In theory: make "integration" of noisy signals better than deye" ;)) Trim current values: Too high: place (adjustable trimmer) Resistor at terminals of deye Too low: Use more windings and divide back (Resistor (adjustable) in parallel to terminals at Deye). More windings: 2x through the same coil or get a CT with more windings. Don't forget to roughly double the terminating resistance of the Deye PCB then adjust to test current constantly flowing (heating) and measured by eg (smart or dumb meter, one can also count seconds of 1 rotation on disc based meters :) (As a bit power can be dissipated at Resistor: I do not know if a small multi-windings trimmer pots is overloaded, one could use a sturdy normal resistor as base value and add serially the trimmer to it.) ___________ De-Noise: it could be we have to filter out "mantle waves", there are standard circuits for this, eg on any terminal PCB of PC power supply symmetric to shielding/protection ground.This also can be done by de-coupling, see bottom. So we clamp shield of TP only on one side, not to introduce vagabonding shield currents into the cable, and use only ONE twisted pair per CT. Then there remains the issue of high-frequency harmonics, the electronics of the deye is not sampling anyways. I would recommend to dive into signal shaping a bit with a few RC components to filter down to 10Hz ca., but the Deye should do this anyways... As you correctly stated, a CT is a nice thing as it is a differential current loop transmission only suffering distortion for high frequencies. I have to re-read my measurement electronics book to look up the limits of current loop transmission. As the Deye is sampling I think 5Hz: This is next to DC, so you can just patch any noisy situation by Capacitors you de-couple a bit by series resistors (yes, it is a contradiction making the wire even lossier;) so no ringing to the coil occurs. R: (decoupling anti-ringing, perhaps small L as series HF blocker) R Deye CT_n a ____._______===_____(TP pair n, a cable)____. = C(filt) L(CT Deye CT_n b ____._______===______(TP pair n, b cable)___. We should ask someone actively developing RF circuits, such a pro would not have a problem shooting out values that probably work out of the cold air. Also, some capacitor or other filter at the beginning of the CT could be beneficial. In doubt, I would shield the setting of CT and 1 phase by mu metal or coke tin. Be careful not to touch hot wires there, isolate by self-attaching foil. I find WAGO current transformers for hat rail mounting quite nice. Everything pre-shielded and robust. As the CT coils have high resistance internally, which you can measure quite easily, by Ohm-meter to the coil, you should not worry about 2Ω - 200Ω ;)) This should also be the ca. size of the terminating resistance on PCB board. Mine have 2000 windings and some kOhm. Rather, you should mindfully do ablation of the insulation of the TP cable, so you can see which cables are twisted together and pair them by drilling them tighter a bit ; so you do not cross-link pairs. Then, some "noise filter" setup for standard slow-speed serial current loop protocols can be copied here, as we nearly have DC from a cable's point of view. Some kHz is still DC when a TP is transferring several hundred MegaHerz normally. We do not have impedance mismatch here - we just have to de-noise it back to "quasi-DC". And block or get rid of parallel parasitic imprinted waves. If you like, I can attach some similar DC CT current coil and measure spectra or time domain with a scope. What I observed is that they have a) protection diodes that sometimes function as a nonlinear component in very noisy environments, so we could place them on terminals of Deye after some decoupling series resistors. b) do not work without terminating resistance that "burns" the current transferred. It could possibly be that if the terminating resistance is some 2kΩ and cable has a few hundreds ohms, we would do well to adjust the terminating resistor. This could happen if the Clamp voltage without load is lower than the voltage needed to drive the resistors. Does this ever happen? As we can't do this, we could use a CT with a few more windings and some R current dividing network to calibrate back, so in the easiest form place a parallel shorting resistor (10rounds calibrating resistor for testing) parallel to the CT clamping terminals of the Deye. I watch the deye discussion for some while, and have yet to buy one, so please forgive me I can only measure for you with a "surrogate" ;)) Hope it still helps and am willing to do measurements on long winding CAT cables by oscilloscope if you like. What REALLY can help is DC decoupling the long wire from the deye ,as the greatest source of strange signals around is the inverter itself. This can be done by some 1:1 transformer. But I think that mantle wave filter will do fine. ___
Marius, I have only one issue with your explanation and that is that the original CT has a 2000:1 ratio, but your two CT combination has a total ratio of 40x100:1, that equals 4000:1. Care to explain? (It seems that going from 63A to 3.2A requires a 20:1 CT ratio and not a 40:1 CT ratio.)
Have a look at something like the Eastron SDM230 or the Chint DDSU66. They measure the power flowing through themselves and transmit the signal digitally to the inverter over RS485 so you dont have the same losses
@@SasVDWesthuizenBecause it prevents stray fields NOT inducing noise currents, this only works by TWISTING, not by doing loops. OK, it is still in outer shield, but way less effective shielded. Catch the TWISTED pair by mindfully ablating and then twisting it tighter to see by eye or place some shrink tube over the pairs. I think you can run 3 CT in one TP cable no problem and even transfer another signal. Your audio from hifi ;) or infrared pulses or other stuff like some serial link.
Hello, a little different topic. My problem is that approximately 20seconds after switching inverter on it trips RCD it is connected to. It doesn’t have an earth bond except normal earthing. Can’t figure out why this happens, fault with inverter??
What the heck happened to proper instrument cable freely available at Communica ? All CAT cables were designed for high speed digital communications surely. Stop being lazy, use the right stuff !
But Guys, you can install the CT coil, right at the inverter, on the AC output that goes to the DB. Why would length be an issue? If you purchase a Blue nova CPS5000, witch is a pre-assembled Sunsynk 5KV and Blue nova 5.2 KWh, it comes wired like that, and it is fully certified.
This would not allow you to feed anything to the non essential loads. You want the CT coil downstream of the non essential loads so it only measures what cones and goes from the utility supply, not the house loads.
You need the CT to be between the Grid-Tie common point and the grid. This would need to be in the main DB, but the actual inverter could be 50m away in the garage at the bottom of the garden.
I actually did an experiment last week and used a twisted pair of CAT6 305meters straight off the drum(box) ... worked absolutely no issues. I think the input for the inverter CT is differential rather than one leg connected to "ground" and voltage measured from this point.
Anyway, customer installed 110meters of CAT6 and the system is working perfectly!
Thanks Carl. I would like to discuss if I have such a problem again. Please send email to energy @ greenpro . co . za for your contact.
You can have two fairly different situations using the same CAT6 cable. I am speculating here somewhat, but if you should use two pairs as the one conductor and the other pairs as the other conductor, then you will get a higher series inductance than if you use one conductor of each twisted pair as the one conductor and the other conductor of each twisted pair as the other conductor. If you keep the series inductance low and therefore the burden impedance of the CT, I can not reason why this cable could not run more than 15 meters. But of course it remains to be verified.
Interesting discussion !
I use the DDSU666 Chint Energy Smart Meter with Cat 6 cable using the RS485.
Hello, Can I ask you for DDSU666 Chint Energy Smart Meter? Is it can use other brand for DDSU666 like huawei DDSU666 Chint
We use CT cable over long runs and it works fine. However we split the pairs and join the 4 white and 4 coloured ones together. So it remains twisted and protects against interference
A CT is a current source, and the inverter is looking for current. You are talking about voltage anomalies, noise, volt drop etc are not an issue. Your problem is using all pairs of a multi twisted pair to transmit the same signal. This is like not using a twisted pair at all!!. Get the proper size SINGLE twisted pair cable.
Great idea, but would it not have been easier to change to a modbus over RS485 meter and run over cat5/6/whatever?
2 CT's at R 200 each and you simply wind wire through them. Must be less than 10% of any other option.
The Eastron energy meters is not that expensive, that is if you buy them in quantities. But of course it is just way easier to install the CT than an energy meter. BUT, we have one issue in South Africa that I think disqualifies the Modbus energy meters and that is that our prepaid meters allows very little reverse flow of energy. The secondary CT current that is fed to the SunSynk reacts immediately to any change in load (obviously) and therefore the inverter should be able to adjust quicker to sudden load variations, while rate of feedback from the energy meter will be much slower and therefore inverter reaction as well. I would therefore expect more nuisance trips when making use of an energy meter than using a CT. What do you think?
Why not just put the cts in the inverter where you connect your input power and output power?
@@pieterkruger4134 I have no idea what purpose that would serve. You need to measure the property's input current so putting a CT in the inverter is totally bizarre.
@@greenpro-greenproconsultin7420 I've noticed another solar installer do that and he has installed hundreds of Sunsynks and he does this sometimes when the db is far and everything is on the inverter. So I was wondering. As I was wondering about this very topic and couldn't get a clear answer from my suppliers until I found this post
I would be very cautious of the above method. The reason is as follows. Modern households no longer have resistive type loads. This results in current drawn that may have harmonics into the kHz. Most new electronic meters either trip on reverse feed ( seen as a tapering fault) or you are charged for the reverse feed. (Meter cannot tell the direction of energy flow.)
If you extend a cable from a CT clamp to the device that measures the CT output, you may encounter several problems:
1.Signal attenuation: The signal from the CT clamp may weaken as it travels through the cable, which can result in a lower signal amplitude at the measurement device. This can reduce the accuracy and reliability of the measurement.
2.Noise interference: The cable may pick up external noise sources, such as electromagnetic interference (EMI) or radio frequency interference (RFI), which can add unwanted signals to the CT output. This can distort the measurement and make it difficult to extract the true signal from the CT clamp.
3.Impedance mismatch: If the cable has a different impedance than the CT clamp or the measurement device, this can cause reflections and standing waves in the cable. This can result in signal distortion, phase shifts, and other measurement errors.
4.Grounding issues: If the CT clamp and the measurement device are not grounded properly, this can lead to ground loops and other interference issues. This can cause unwanted signals to be introduced into the CT output, which can affect the accuracy of the measurement.
Although the above method tries to address item 1, and possible 2 (with twisted pairs) I don't think it addresses item 3. This may result in reverse feed at the incorrect time. (That may not be immediately apparent)
In a nut shell without knowing the circuitry in the inverter that is used to measure the current, and how the inverter makes a decision to reverse feed it difficult to tell. For me it would be best to get hold of SunSynk (Keith I think) for an opinion.
Two CTs in series, if anything provides more harmonic filtering that the small CT. The intent of the 2 CTs is to have more current in the long wire, so that losses are minimised. There is a voltage loss but no current loss and as long as the VA consumed by the 2nd CT plus the cable resistance power wasted stays within spec of the 1st CT there are no losses that matter. Anyway, this solution costs R500 and it works. These fancy meters etc. solutions costs an order of magnitude more. Also the accuracy is not that important if you are controlling around zero.
Great! Thanks!
_______________
Short: DE-NOISE.
Simplest: By capacitor. De-ring by resistor.
(In theory: make "integration" of noisy signals better than deye" ;))
Trim current values:
Too high: place (adjustable trimmer) Resistor at terminals of deye
Too low:
Use more windings and divide back (Resistor (adjustable) in parallel to terminals at Deye).
More windings: 2x through the same coil or get a CT with more windings. Don't forget to roughly double the terminating resistance of the Deye PCB then adjust to test current constantly flowing (heating) and measured by eg (smart or dumb meter, one can also count seconds of 1 rotation on disc based meters :)
(As a bit power can be dissipated at Resistor: I do not know if a small multi-windings trimmer pots is overloaded, one could use a sturdy normal resistor as base value and add serially the trimmer to it.)
___________
De-Noise:
it could be we have to filter out "mantle waves", there are standard circuits for this, eg on any terminal PCB of PC power supply symmetric to shielding/protection ground.This also can be done by de-coupling, see bottom.
So we clamp shield of TP only on one side, not to introduce vagabonding shield currents into the cable, and use only ONE twisted pair per CT.
Then there remains the issue of high-frequency harmonics, the electronics of the deye is not sampling anyways.
I would recommend to dive into signal shaping a bit with a few RC components to filter down to 10Hz ca., but the Deye should do this anyways...
As you correctly stated, a CT is a nice thing as it is a differential current loop transmission only suffering distortion for high frequencies.
I have to re-read my measurement electronics book to look up the limits of current loop transmission.
As the Deye is sampling I think 5Hz:
This is next to DC, so you can just patch any noisy situation by Capacitors you de-couple a bit by series resistors (yes, it is a contradiction making the wire even lossier;) so no ringing to the coil occurs.
R: (decoupling anti-ringing, perhaps small L as series HF blocker)
R
Deye CT_n a ____._______===_____(TP pair n, a cable)____.
= C(filt) L(CT
Deye CT_n b ____._______===______(TP pair n, b cable)___.
We should ask someone actively developing RF circuits, such a pro would not have a problem shooting out values that probably work out of the cold air.
Also, some capacitor or other filter at the beginning of the CT could be beneficial.
In doubt, I would shield the setting of CT and 1 phase by mu metal or coke tin.
Be careful not to touch hot wires there, isolate by self-attaching foil.
I find WAGO current transformers for hat rail mounting quite nice.
Everything pre-shielded and robust.
As the CT coils have high resistance internally, which you can measure quite easily, by Ohm-meter to the coil, you should not worry about 2Ω - 200Ω ;))
This should also be the ca. size of the terminating resistance on PCB board.
Mine have 2000 windings and some kOhm.
Rather, you should mindfully do ablation of the insulation of the TP cable, so you can see which cables are twisted together and pair them by drilling them tighter a bit ; so you do not cross-link pairs.
Then, some "noise filter" setup for standard slow-speed serial current loop protocols can be copied here, as we nearly have DC from a cable's point of view.
Some kHz is still DC when a TP is transferring several hundred MegaHerz normally.
We do not have impedance mismatch here - we just have to de-noise it back to "quasi-DC".
And block or get rid of parallel parasitic imprinted waves.
If you like, I can attach some similar DC CT current coil and measure spectra or time domain with a scope.
What I observed is that they have
a) protection diodes that sometimes function as a nonlinear component in very noisy environments, so we could place them on terminals of Deye after some decoupling series resistors.
b) do not work without terminating resistance that "burns" the current transferred.
It could possibly be that if the terminating resistance is some 2kΩ and cable has a few hundreds ohms, we would do well to adjust the terminating resistor.
This could happen if the Clamp voltage without load is lower than the voltage needed to drive the resistors. Does this ever happen?
As we can't do this, we could use a CT with a few more windings and some R current dividing network to calibrate back, so in the easiest form place a parallel shorting resistor (10rounds calibrating resistor for testing) parallel to the CT clamping terminals of the Deye.
I watch the deye discussion for some while, and have yet to buy one, so please forgive me I can only measure for you with a "surrogate" ;))
Hope it still helps and am willing to do measurements on long winding CAT cables by oscilloscope if you like.
What REALLY can help is DC decoupling the long wire from the deye ,as the greatest source of strange signals around is the inverter itself.
This can be done by some 1:1 transformer.
But I think that mantle wave filter will do fine.
___
Marius, I have only one issue with your explanation and that is that the original CT has a 2000:1 ratio, but your two CT combination has a total ratio of 40x100:1, that equals 4000:1. Care to explain? (It seems that going from 63A to 3.2A requires a 20:1 CT ratio and not a 40:1 CT ratio.)
I wind the secondary of the 1st CT thru the 2nd CT to form 2 windings.
Tx Marius. My installer said the 5m cable could not be extended. Please please do a basic video on wiring of CT
I need to run the CT back 100m because the solar is mounted in the garage, any suggestions?
Have a look at something like the Eastron SDM230 or the Chint DDSU66. They measure the power flowing through themselves and transmit the signal digitally to the inverter over RS485 so you dont have the same losses
1:51 "... we used all 4 pairs of a CAT6 cable "
This is how NOT TO USE paired-wire.
Providing no other explanation for why not is of no use as well...
@@SasVDWesthuizenBecause it prevents stray fields NOT inducing noise currents, this only works by TWISTING, not by doing loops. OK, it is still in outer shield, but way less effective shielded. Catch the TWISTED pair by mindfully ablating and then twisting it tighter to see by eye or place some shrink tube over the pairs.
I think you can run 3 CT in one TP cable no problem and even transfer another signal. Your audio from hifi ;) or infrared pulses or other stuff like some serial link.
so awesome....
Hello, a little different topic. My problem is that approximately 20seconds after switching inverter on it trips RCD it is connected to. It doesn’t have an earth bond except normal earthing. Can’t figure out why this happens, fault with inverter??
What the heck happened to proper instrument cable freely available at Communica ? All CAT cables were designed for high speed digital communications surely. Stop being lazy, use the right stuff !
But Guys, you can install the CT coil, right at the inverter, on the AC output that goes to the DB. Why would length be an issue? If you purchase a Blue nova CPS5000, witch is a pre-assembled Sunsynk 5KV and Blue nova 5.2 KWh, it comes wired like that, and it is fully certified.
This would not allow you to feed anything to the non essential loads. You want the CT coil downstream of the non essential loads so it only measures what cones and goes from the utility supply, not the house loads.
You need the CT to be between the Grid-Tie common point and the grid. This would need to be in the main DB, but the actual inverter could be 50m away in the garage at the bottom of the garden.