SparkFun Forums

[fll-freak]

I have a few questions on the SEN-10341 30A current sensor. http://www.sparkfun.com/products/10341

I have received two units and started to experiment with them. My overall goal is to instrument my home breaker panel to monitor energy usage on each branch. Similar to thread viewtopic.php?f=14&t=28775&p=129410

Q1) My units did not come with a 3.5mm stereo male phone plug but rather 3.5mm mono. Not a big issue, but did other purchasers have the same issue? Just a little bit worried about counterfeit parts. My electrical inspector demands UL listed parts in the breaker box. I do not want my house to burn down (go figure...)

Q2) The comments on the product page indicate that the transfer function is 1V out at 30Amps. From experience I do not think this is the case. A 60W light bulb at 120Vrms should draw 0.5Amps RMS or 0.71 amps peak to peak. In this condition, the CT generates about 36mv making the transfer function 50mv per amp. Can anyone confirm?

Observations:
It has been a fascinating experience looking at the CT output for various items I have monitored. A 60W bulb is a nice clean sinusoid (if you filter the fuzz of the signal). My LeCroy scope pulls current in a very strange way. It goes sinusoidal for about 1/3 of the half wave and them jumps up in a spike. My hot air gun on low only consumes current on 1/2 the cycle. On high it pulls current for the whole wave. And I will not even talk about my bench grinder!

My original hope to simply rectify and low pass filter the output to get a good idea of power consumptions has been dashed. If all my appliances are as ill behaved as my scope and hot air gun, I will have to sample at a high rate and compute the RMS value. Grrr. Need to pull out my copy of "IC OP AMP Cookbook" and look for an RMS circuit.

[jremington]

The "data sheet" published for the SEN-10341 current transformer on the Spark Fun website is a disgrace and nearly useless -- the manufacturer couldn't even be bothered to state whether the voltage output is rms or peak! Obviously one has to make measurements to know how the CT will respond to different current levels. The output is likely to be nonlinear, so you will probably need to investigate the response over the full stated current range. Note that if the core of the CT saturates at overcurrent, the rms/peak output voltage will actually drop from its maximum value!

I did that for my power monitoring system by making a plug adapter so that I could try out different resistive loads, such as a free standing room heater, light bulbs of various wattages, etc. The results for the Coilcraft CS series is a pretty decent calibration curve -- a straight line, in fact, over the entire measurable range. If you could make those measurements for the SFE CT and post them here, other potential customers would be very appreciative.

[fll-freak]

I have hacked a power strip and instrumented it for the exact purpose of running some tests. At this point I have one data point (a resistive load of 60W). Since the outlet is on a 15 Amp branch, not sure how I am going to calibrate up to 30 Amps! But then again, I really do not need to go to 30 as each branch is protected by at most a 20Amp circuit breaker.

One good thing so far is the the two samples behave the exact same way. The bad news is that I am beginning to get an uncomfortable feeling putting them in my power panel. If the data sheet is so bad and the connectors do not match, how does the rest of the quality stack up? Do I even trust the UL listing. I may look for more trustworthy parts even if I do need to pay more for them.

[Mee_n_Mac]

My original hope to simply rectify and low pass filter the output to get a good idea of power consumptions has been dashed. If all my appliances are as ill behaved as my scope and hot air gun, I will have to sample at a high rate and compute the RMS value. Grrr. Need to pull out my copy of "IC OP AMP Cookbook" and look for an RMS circuit.

In retrospect it seems "obvious" why those appliances do what they do ... but I'd not have guessed that up front. While I suppose you could do some form of square law true RMS circuit with op-amps, it might just be worth the $8+ to buy a true RMS IC like the AD737. Digikey has them.

I wonder how off your power estimate would really be w/o such a circuit though. If you use an electric range/oven or clothes dryer or water heater ... those should ( ) be all nice resistive loads. I'm less sure about the refigerator though, even if it's just a motor. TV and PCs and microwaves would probably be akin your o-scope. I wonder what a CFL looks like (current-wise) ?

[esklar81]

Do I even trust the UL listing. I may look for more trustworthy parts even if I do need to pay more for them.

Skye,
As a general rule, I don't trust claims of being listed, unless I'm familiar with the manufacturer. UL (partly as a service to those of us who buy and use components and partly as a means of defending the use of their marks) makes it pretty easy to check a claim of listing.

My first question, though, is to what "UL listing" are you referring? I didn't see any such claim on the datasheet and I don't see anything that looks like a UL listing mark in the photos SparkFun put on its catalog page.

I took a quick look on the UL site. I found several listings for things with "sct" in their names, but nothing that looked like a match. I found nothing when I tried "yhdc" as the manufacturer's name.

The only mention I saw on the datasheet of UL is the claim "in accordance with UL94-V0". UL94 is a standard for flammability of polymeric parts of electrical devices. On the bright side, "V0" is an appropriate classification of polymer to use for this sort of device. On the dark side, "in accordance with" usually means that the manufacturer thinks it meets a standard, but that the product has not been tested by an appropriate agency.

Good Luck,
Eric

[fll-freak]

Eric,

Your post are always informative.

Somehow I seem to have gotten my wires crossed (figuratively). I could have sworn I remember seeing a UL listing on the device at some point. When I placed my order, I got back ordered for a long time. Now I go back and the UL listing I remembered was not there. Perhaps the data sheet has changed since, or perhaps it was one of the other devices that I was looking at that was listed.

When I called my town's electrical inspector to tell him what I wanted to do, he wanted to see UL listed devices. The sensor I was originally doing to use was designed for satellites. Even a mil standard was not going to cut it at town hall. But now that you bring up the fact that there are many UL standards, I must go back and ask in more detail what they want.

The two guys I purchased will be great for HW and SW development. For implementation I must look elsewhere. What I liked about the SF units were the size. They were snap on and yet I figured I could get a dozen inside my packed breaker box.

Mee_n_mac,

I have seen numerous hall effect type sensors from various vendors. To me they appear to be much safer than a traditional CT. If the load resistor fails, you can get some pretty interesting failures including smoke and flames. In my judgement "Flames are Bad". But again, Town Hall was not keen on my cable tying chips to the branch circuits.


So I think I will keep my hunt up for small (preferably snap on) CTs that have a "real" UL listing that matches what Town Hall wants. We will have to see how that works out!

[fll-freak]

So I collected a few more data points on the sensor. I used incandescent lights as the load trusting to the wattage printed on them. I may cross check against a Kill-A-Watt, but from what I have been reading, they may not be any more accurate (especially for reactive devices).

Code: Select all

Watts  Amps  mv RMS  mv P2P  mv Peak
60       0.5    15.7       47.6     23.8
150      1.3    41.0     114.0     57.0
200      1.7    55.1     154.0     77.0
260      2.2    71.2     198.0     99.0
350      2.9    96.5     264.0    132.0
410      3.4   113.0    310.0     155.0

Attached is a screen shot from the 200 watt tests. It shows that P2P (peak to peak) is twice my column labeled Peak. This is fine for a resistive load. A little bit of Excel work gives the equations:

Amps = mv (instantaneous) * 0.022

Have NO clue how to read that from the data sheet!

For the sake of looking at an ugly reactive/non resistive load, I included my hot air gun at low setting.


For extra credit what wattage lights bulbs did I have on hand?

[jremington]

Interesting! Rearranging your fit to the data gives 30A/0.022 = 1364 mV.

This is close (less than 4% error) to 1.414V, which would be the peak voltage corresponding to 1.000VAC rms for a 30 amp rms resistive load. So now we know what the data sheet means.

[n1ist]

You had a 60W, 2 75W, and 2 100W bulbs...
/mike

[fll-freak]

You had a 60W, 2 75W, and 2 100W bulbs...
/mike

Sorry, I only had three bulbs.

[Mee_n_Mac]

You had a 60W, 2 75W, and 2 100W bulbs...
/mike

Sorry, I only had three bulbs.

Where did you get a 200W bulb ?


(60, 150, 200)

[fll-freak]

Photo flood light.

[fll-freak]

jremington,

Thanks for the insight. Of course that small error could be in the sensor or in the fact my bulbs may not be at exact wattage.

I am also looking only at one tenth the scale of the sensor. At 410 watts, I was only pulling 3.4 amps when the sensor is good to 30. I might be able to find a few more lamps and sockets to push this up to perhaps 1000 watts or about 8 amps, but I do not think I would be comfortable pushing 30 amps through my house wiring and a hacked outlet strip. If I continue with this sensor (doubtful) then I will just have to trust that calibration curve stay linear.

[waltr]

I would be comfortable pushing 30 amps through my house wiring

Hopefully the house circuit you are using has a breaker rated for either 15 or 20Amps.

[fll-freak]

Very true. I can't push more than the cb will allow. But I am not even sure I want to push the whole cb rating through one branch and a power strip. Visions of wisps of smoke start flashing through my head like sugar plums in winter.