would I be safe using one of the current sensors?
- Thread starter bbsux
- Start date
CAI_Support
Senior Member
That is basically a transformer. Do you plan to measure its output to determine the current? Analog input is over voltage protected, so that even the analog input voltage is higher, it is still okay. Digital input does not have that kind of protection.
bbsux
Active Member
yes to measure the current. I'm probably not going to go this way but for people that may...
won't I have to pull the voltage down to levels the webcontrol can handle using a burden resistor? Like here?
http://openenergymonitor.org/emon/buildingblocks/ct-sensors-interface
just need to adjust the values accordingly.... Or am I wrong?
won't I have to pull the voltage down to levels the webcontrol can handle using a burden resistor? Like here?
http://openenergymonitor.org/emon/buildingblocks/ct-sensors-interface
just need to adjust the values accordingly.... Or am I wrong?
CAI_Support
Senior Member
BBSUX,
[font="Arial""]I think the need for burden resistor is for more accurate measurement. Because AC voltage through transformer normally are different when reading RMS, AVG, and PEAK, the burden resistor is helping to reduce the differences. Properly adjusted burden resistor will produce more accurate measurement. CT sensor is mostly for large AC signal measurement.[/font]
[font="Arial""]However, the most accurate way to measure AC is not through CT sensor. Have you ever looked Kill-A-Watt inside? It uses a LM2902 op amp directly amplifying the AC voltage drop on a piece of big copper wire. That will do much better on the small signal AC measurement, because op amp is very linear on small AC signal than diode or transformer. LM2902 is a low cost, very easy to wire op amp on a prototype board. The only bad thing in this approach is to put whole WebControl board on the hot wire voltage.[/font]
[font="Arial""]The WebMux analog input is over-voltage protected, so that it will not cause damage to WebControl.[/font][font="Arial""][/font]
[font="Arial""]I think the need for burden resistor is for more accurate measurement. Because AC voltage through transformer normally are different when reading RMS, AVG, and PEAK, the burden resistor is helping to reduce the differences. Properly adjusted burden resistor will produce more accurate measurement. CT sensor is mostly for large AC signal measurement.[/font]
[font="Arial""]However, the most accurate way to measure AC is not through CT sensor. Have you ever looked Kill-A-Watt inside? It uses a LM2902 op amp directly amplifying the AC voltage drop on a piece of big copper wire. That will do much better on the small signal AC measurement, because op amp is very linear on small AC signal than diode or transformer. LM2902 is a low cost, very easy to wire op amp on a prototype board. The only bad thing in this approach is to put whole WebControl board on the hot wire voltage.[/font]
[font="Arial""]The WebMux analog input is over-voltage protected, so that it will not cause damage to WebControl.[/font][font="Arial""][/font]
I went to great lenghts to get that CT sensor to work for me. I tried an several methods including the op amp idea. None of them worked very well. My goal was to know when my sump pumps were running. I really didn't care about how much power the pumps were using so an acurate reading was not a priority. What was a priority was getting the CT sensor to put out the voltage to the board.
What I did was put a resistor in parrell with two leads, then another resistor in series. I measured the current that was being produced (Very important) and using Ohms law I figured out what resistors I needed to get the voltage I wanted. I found out the hard way that this method creates half the sign wave, so the board was seeing rapid on/off states. I had a counter setup and each run was getting about 300 hits. Putting an electrolitic capacistor in the mix, leveled out the current and fixed that issue. I haven't had a problem since.
If you know what current you want to measure, you can calculate the right resistors.
What I did was put a resistor in parrell with two leads, then another resistor in series. I measured the current that was being produced (Very important) and using Ohms law I figured out what resistors I needed to get the voltage I wanted. I found out the hard way that this method creates half the sign wave, so the board was seeing rapid on/off states. I had a counter setup and each run was getting about 300 hits. Putting an electrolitic capacistor in the mix, leveled out the current and fixed that issue. I haven't had a problem since.
If you know what current you want to measure, you can calculate the right resistors.
bbsux
Active Member
accuracy is not a concern of mine, I just need to know when my washer/dryer are running. I'll probabaly stick to the hall sensor version I have
http://www.ebay.com/itm/230684998854?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649
http://www.ebay.com/itm/230684998854?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649
CAI_Support
Senior Member
The problem with using current sensor to detect if washer running is that the main motor may pause during the washing cycle. If your washer has a finish LED, it is best by wiring to that LED.
Here's an idea if they have just knobs that rotate to select cycle and time. Take the knob and put a solid contact point just above the selection point with a wire going to the digital input. Then on the plate put a spring lever contract point on the finish point on the face plate with a wire going to the +5v. When the washer/dryer are done with there cycle, the contact will connect and trip the board. You could even put multiple finish points on the face plate for different cycles. I would definitely put in some safety switches in the PLC code to prevent early emails when selecting a cycle and to prevent multiple emails being sent when the cycle is done. If you don't care about a warrenty, you could tap some holes and make it look really nice without wires showing like mine did. My wife didn't care what it looked like because they were in the basement.
Just an idea for an older style washer/dryer. Mine had led lights, so I used a photoresistor to let me know when they were done. I've since moved and my washer/dryer are on the first floor where I can hear them. Although I still use the photoresistor setup for other device on/off status.
Just an idea for an older style washer/dryer. Mine had led lights, so I used a photoresistor to let me know when they were done. I've since moved and my washer/dryer are on the first floor where I can hear them. Although I still use the photoresistor setup for other device on/off status.
CAI_Support
Senior Member
This article described how to replace the old mechanical control with a circuit board. It uses two relays to control the washing machine motor and total washing time. You could borrow this idea to have WebControl and two relays to handle the whole washing cycles, add three more relays to control hot water valve, cold water valve, and water discharge valve. At the end, to send email notice.
http://www.electronicsforu.com/EFYLinux/circuit/sep2003/cir-5-washing.pdf
http://www.electronicsforu.com/EFYLinux/circuit/sep2003/cir-5-washing.pdf
Good Lord! The phrase "reinventing the wheel" comes to mind. You might as well just build the washing machine from scratch. I don't see the point in rebuilding something that is working fine just to get an email. Don't get me wrong, if your into that kind of thing more power to you. Personally I got the board to make my life easier and less complicated. Again, just my opition. Good luck with that though.
bbsux
Active Member
I agree. Now if I can find a pinout or maybe I can even talk to an old school Maytag tech -- maybe I can find a place on the existing control boards I can tap onto (even if its 110vac and I need to run a wall wart to get down to 5vdc and run a relay, heck I'd much rather it be digital input than analog anyway.). I read one guy that actually probed the round selector knob and found a 110vac place to tap into, that he used for a notifier.
Yes, that's where I got the idea. So that when the dial rotates, it doesn't catch and stop the dial. The contact point on the dial would be fixed and the ones on the face plate would we on springs. Just setup code to prevent early emails. I think that would be a lot easier than monitoring power.