Wondering if Midnite has considered using current sensors similar to what a clamp-on meter uses so that there is no insertion loss or isolation added to the circuit.
The Whizbang is what is used because it can mount on and monitor via a precision shunt that is permanently installed in a system. This is a must if being used for control. There is only a small mv drop on the shunt so not any appreciable loss.
I don't see the use of clamp on amp monitoring when it is already built into the system .
If someone wanted to just read current on any particular cable it is easy to do that with a $30 clamp on digital multi meter.
Larry
I believe that inside the Classic, there IS a Hall-Effect current sensor for the CC output.
The calibration for Hall Effect sensors do drift, and usually need to be re-zeroed periodically.
Some of these sensors may need to be degaussed from time-to-time, as well.
As Larry mentioned, 50 mV, 500 A Shunts are really the industry-standard for DC current measurement, and are shipped, installed, in the DC portion of many e-panels and DC conduit boxes, and are useable by many existing solar subsystems. Disregarding connection resistance of the large cables to the Shunt, 100 A of current through the Shunt, yields 5 mV of voltage drop on a 50 mV 500 A Shunt.
FWIW, Vic
I remember another discussion on these forums where someone said that a supercap is a good idea when putting one or more shunts in series with an inverter due to an isolating effect the .005 ohm resistance that a 50mv shunt would have.
But this clamp-on sensor is not in addition to already existing shunts this is before any other shunts are installed or wired in.
Mostly thinking about the ease of installation and ease of adding more shunts at any time.
Instead of buying terminals, cutting wires and securing them to terminals then mounting the terminals to the shunt,
all that would need to be done is slide a sensor unit over the wire and leave it.
It's going to take over an hour preparing connectors and terminals for ,say, 5 shunts and these connections should probably be checked from time to time for voltage drop and / or re torquing. When it could take less than a minute to slip 5 sensors over wires and secure the ends and eliminate several mechanical connections in places where there is somewhat high current.
Just a thought.
Hi Eric,
OK, BUT, a 50 mV 500 A Shunt has a resistance of 0.0001 Ohms, really infinitesimal.
The traditional Clamp-On DC current sensors are quite unstable, and need frequent zeroing. Only really acceptable for their intended purpose.
A better all-Effect sensor (not a clamp-on, but a donut type device), would probably still need to be re-zeroed, on occasion. This can remove much of the utility as a general-purpose device, but would be fine for a purpose-built subsystem.
IMO, Vic
Vic, Hall Effect sensors are exactly what I am thinking.
Some of the Fluke clamp-on meters use Hall Effect IC's as sensors.
They might not as stable but I don't know, they are solid state though. I'm thinking of exact donut sizes made for the wire size as needed. This would help with repeatable results.
Also mechanically assembled shunts are not free of drift either.
Quote from: Vic on March 13, 2017, 07:37:09 PM
I believe that inside the Classic, there IS a Hall-Effect current sensor for the CC output.
A couple of years ago I asked about the current sensors in the Classic, I thought they were likely hall-effect devices, but was informed they were in fact shunt based.
dgd
EWZ,
All types of clamp-on current sensors tend to be notoriously inaccurate with readings drifting (for reasons as Vic noted above).
If you want to monitor currents at different places in your system then the easiest way is to use in circuit H-E devices such as the AC758, you would have to use a cpu
(PC,Arduino, rPi, BBB etc), and write some code to read and display the data.
http://www.ebay.com/itm/ACS758LCB-050B-PFF-T-Hall-Current-Sensor-Current-Module-MF-/172379257623?hash=item28229abb17:g:xQ8AAOSwzaJYBiCV (http://www.ebay.com/itm/ACS758LCB-050B-PFF-T-Hall-Current-Sensor-Current-Module-MF-/172379257623?hash=item28229abb17:g:xQ8AAOSwzaJYBiCV)
If you want MN to do this then you may have a long wait as it may not be a commercially viable considering development costs vs potential sales.
dgd
vic, "....a 50 mV 500 A Shunt has a resistance of 0.0001 Ohms,really infinitesimal....."
That was the exact point I made but it seemed to be dismissed in that particular discussion about the need for a supercap.
Well, like I said, this was only a thought. If a Hall IC attached to a current lead is not practical because the sensors are not accurate then it isn't going to work.
Hello again Eric (I believe),
If you were referring to me being dismissive, then I apologize ... perhaps that part of a discussion was on another Forum.
Hall devices can be very accurate. Historically they have needed to be recalibrated, momentarily, when in use ... I have not followed the progress in this area. Quite possibly, the cost of devices that are more stable has come down ... dunno.
Your point about the connections to a large Shunt potentially having greater resistance than the Shunt is generally correct, and this is one factor that can work against using the simple Shunt.
The Hall devices would need an interface, power, etc. This is a bit of a disadvantage. Resistive Shunts are the industry Standard, and many monitoring/control functions readily interface to these Shunts, so there would appear to be a wide application for them, using existing monitoring/control devices ...
Those were the points that I was trying to make, perhaps, too clumsily.
73 Have Fun, Vic . .
Vic, to be honest I don't remember who was making the claim that a supercap would be needed if a shunt was installed. I wasn't trying to get personal about it.
Just saying that you supported my conclusion where I stated that a 50 millivolt shunt wouldn't create any more resistance than a normal connection.
In dealing with this topic I see that some people are using the resistance in wire as shunts. This might be a great idea. By calculating the resistance of a particular wire over a defined length one should be able to come up with a very accurate ohmic value.
For instance about 7' of #2 would have about a .001 ohm value.
We use a shunt (500A 50mV) because this is what our industry has been using for many many years now.
It is also very low on power being 0.0001 Ohms. At say, 25 amps draw, this is 25 * 25 * 0.0001 = 62.5 milliwatts.
0.0625 watts Which is really low.
But the BIG advantage to this, as others have pointed out is that it way more accurate than a Hall effect device.
Hall effect devices also draw power to operate and possibly more than a shunt at low power levels.
Hall effect devices and magnetic current sensors in general are also sensitive to external magnetic fields which can contribute to inaccurate readings.
The Whizbang Junior calibrates for zero drift in-between every sample it takes which is about 10 times per second.
It's really accurate.
boB
Quote from: boB on March 21, 2017, 02:09:01 PM
The Whizbang Junior calibrates for zero drift in-between every sample it takes which is about 10 times per second.
It's really accurate.
boB
boB, I see anywhere from 0.5-1.0 Amp of jitter on the WBjr data dumped by the LA at 2 sec/reading resolution. Is that normal? What is the typical jitter on the WBjr data stream.
Hi John,
I see about a total variation of about 1.0, or perhaps 1.2 A in WBjr Jitter.
This is not huge -- about 0.1% of battery capacity, but I do wish that it were lower.
Had assumed that this was due to a small voltage jitter, which results in a larger current jitter.
Had also guessed, that due to the relatively high String voltage on the main systems here (106 V at the STC Vmp of the PVs), that this might well increase the Jitter, compared to using a lower String voltage - guessing.
If the WBjr Jitter could be reduced, that would be great.
AND, on a slightly different subject, sure DO wish that there could be a way for the Classic could have a battery voltage Sense device (perhaps the WBSr, WBPlus ... WBDeluxe, or similar that cold measure battery V at the battery terminals, and communicate it to the Classic for use as THE battery voltage. Realize that this could be a part of the B17 inverters, but we may not need replacement Inverters for some time ... knock, knock.
FWIW, Vic
Quote from: Resthome on March 23, 2017, 05:50:49 PM
boB, I see anywhere from 0.5-1.0 Amp of jitter on the WBjr data dumped by the LA at 2 sec/reading resolution. Is that normal? What is the typical jitter on the WBjr data stream.
Now this intrigued me as my arduino is reading the WBjr modbus registers every second or just slightly over. The values for current are sent directly to a web page current gauge that is refreshing at the register read rate of near every second.
The gauge is sensitive enough that a small jitter or even fifth amp would make the needle visably jitter.
I have not investigated more deeply but unless the modbus register is amps averaging and there is another 'raw' amps register then I cannot see this jitter in data from the WBjr...
dgd
Vic, dgd
Thanks for the replies. So Vic is seeing close to what I see. I seem to recall this was the topic once and boB did some kind of averaging to try to smooth it out. At least this is what I recall. Not sure what Modbus reg the LA is reading for it. And not sure where the jitter is coming from or if it is normal. It would make End Amps easier to hit if it was smoother and more steady in my case. I did have to extend the purple wire a few feet to get to the Classic, but don't think that effected it. Unless it is running near something else that is causing it.
Agree with Vic on the remote voltage sense, that would help with the variations in voltage with current and no current. The tweets offset doesn't help this.
Will try to post some data tomorrow of what I see.
Found this little snap on current sensor with 2% accuracy.
http://cdn.sparkfun.com/datasheets/Sensors/Current/ECS1030-L72-SPEC.pdf
Hi Eric,
It seems to me, that these are AC current transformers:
"Rated Primary Current(Amp.) 50/60Hz"
Perhaps you need an AC Current transformer, but had thought that you had been referring to DC Current Sensors.
FWIW, Vic
Quote from: Vic on April 07, 2017, 04:54:03 PM
Hi Eric,
It seems to me, that these are AC current transformers:
"Rated Primary Current(Amp.) 50/60Hz"
Perhaps you need an AC Current transformer, but had thought that you had been referring to DC Current Sensors.
FWIW, Vic
Vic;
Probably why they are called current
transformers. You need an active component to read DC. with the AC Current Transformer you can read volts produced from the sine wave through the transformer with a known amps to millivolts ratio. With steady DC you can't drive a transformer once the flow is steady. Talking about non contact methods here, of course (no shunt).
You probably know that but there it is in case others wonder about it.
Tom
Quote from: TomW on April 07, 2017, 05:03:29 PM
With steady DC you can't drive a transformer once the flow is steady. Talking about non contact methods here, of course (no shunt).
You probably know that but there it is in case others wonder about it.
Hey Tom, I hate to beg to differ... and it's not exactly the same thing, but....
When I was a young'n, my father had a NON-CONTACT, DC CURRENT METER.
It was used to measure battery cranking amps, headlamp current, and looking for other residual current draws like running lights etc.
It was basically about a 2" diameter centre-zero meter that showed amps charge and discharge. One simply held it against the wire in question (it had a little guide on the back to make sure it sat exactly along the wire the right way).
It worked great, and had absolutely no electronics in it whatsoever.
So there you go, it CAN be done!
So was it some sort of hall effect sensor with a sensitive microvolt meter? I can't imagine it being very or even nearly accurate.
dgd
Quote from: dgd on April 08, 2017, 07:43:37 PM
So was it some sort of hall effect sensor with a sensitive microvolt meter? I can't imagine it being very or even nearly accurate.
dgd
I believe that this type DC Ammeter just used the large magnetic field surrounding the cable with a LARGE current flowing in it to disturb the magnetic field from a fixed magnet in the meter, IIRC.
So a direct reading of the field, and calibrated in tens of amps, or so.
Vic
Quote from: Vic on April 08, 2017, 08:00:23 PM
Quote from: dgd on April 08, 2017, 07:43:37 PM
So was it some sort of hall effect sensor with a sensitive microvolt meter? I can't imagine it being very or even nearly accurate.
dgd
I believe that this type DC Ammeter just used the large magnetic field surrounding the cable with a LARGE current flowing in it to disturb the magnetic field from a fixed magnet in the meter, IIRC.
So a direct reading of the field, and calibrated in tens of amps, or so.
Near as I recall (I haven't seen it in decades), it was pretty much a bog-standard moving-iron type meter movement, but instead of having a coil and passing current through the insides of the meter, the wire was held directly against the back of the meter movement.
It was certainly more sensitive than merely "tens of amps" - it could quite easily indicate the modest current of the interior light coming on and off, so in the order of an amp or so.
I think it must have been a logarithmic meter, cranking amps would take it out to pretty much full scale one way or the other, and generator (or alternator) charging current of 40A would show about 1/2 scale.
I've had a quick search of google images and haven't found a picture of it or anything quite like it. I best ask Dad next time I see him if he still has it, and take some pictures for historical reference!
OK, this isn't the same but it's very similar.
(http://tools.rossw.net/mt-110-current-indicator/front.jpg)
(http://tools.rossw.net/mt-110-current-indicator/back.jpg)
Quote from: RossW on April 08, 2017, 06:24:07 PM
So there you go, it CAN be done!
I actually have one of those in the toolbox. Not sure why I ignored that fact.
Wonder what kind of voltage it puts into that meter from a steady flow of current?
You seem to be my own personal fact checker. ;D
Now on to your regularly scheduled programming.
Tom
Quote from: TomW on April 09, 2017, 09:03:27 AM
I actually have one of those in the toolbox. Not sure why I ignored that fact.
Wonder what kind of voltage it puts into that meter from a steady flow of current?
None. We already know a static field won't actually
generate a voltage in any useful form, but a magnetic field - even static - is a different thing. Think back to your compass - earths magnetic field is relatively static yet can deflect the compass needle in a very useful way. These hold-over-the-wire type indicators are pretty much the same thing and work in pretty much the same way.
Quote
You seem to be my own personal fact checker. ;D
Well, we all miss you in IRC, so I have to come here to catch up!