Thanks for the detailed project review of the LTC2945. I had been looking for a way to monitor higher dc voltages other than using voltage dividers. Your project paved the way for using the LTC2945 with a ESP8266-12E on my 24 vdc solar collector. Now I can remotely monitor the collector voltage on my favorite IoT site.
The connector could add more than 0.02 ohm resistance between the main board and the board with resistor. Hopefully the contacts in the connectors are gold plated, but I doubt it. You should have placed just two resistors in series on the pcb spaced enough apart and with two half circle pads to allow user to connect two traces together with a blob of solder - if user wants to enable that resistor, just add a blob of solder over the two pads to create the connection, and unsolder the other blob of solder to disconnect the other resistor. This way the pcb copper layer would also behave like a tiny heatsink for the current sense resistors and reduce the drift with temperature.
Hi, I am using your code and your calculated shunt resistors in one of my projects and my measurements only accurate with the 10 Ohm resistor. If the resistor is smaller than 10 Ohm for example: 1 or 0,1 or 0,01 the mesaurments are not accurate. The higher the current higher the devitation. The devitation always higher. Thank you in advance for your help!
The TI INA230 is just above $1.00 on digikey and TI has 16bit A/D. The LTC2945 is well over $4.00 and it has only a 12bit A/D. Both are pretty similar otherwise with TI still holding an edge on other small things. except for the 80volt upper voltage limit which reduces resolution anyway, I see no reason to even consider this part for anything. And I am not a big TI fan either. Sorry LT, just another overpriced part.
I've tried a couple other current measurement techniques (built-in sensing on a motor controller chip, and a separate current sensing chip inline with the main power to the motor controller, but they didn't have such resolution! Since I would like to be able to measure to 6-7 A (for motor-stall conditions), could I use a 1/2-w sense resistor (wound wire? How would I get that low a resistance... another puzzle!) to be able to sense more than 5A? That's a versatile board design. Thank you for sharing the design!
Just confirming, since the current is calculated from the input voltage and the sense resistors, this would not work for a non-ohmic device such as a diode?
I would like to be able to measure the power being used in my Sous Vide controller using a SSR using PWM from a Chinese temp controller. I want to track the temp vs power curve to see what is happening during operation. It is 115VAC slow cooker for now ( 350 watts) and later a bigger heating unit of 1000 watts. From another question I cannot use your device (80VAC MAX). What can I do to get the readings in the arduino? I eventually want to make my own controller on the esp8266 or ESP32).
hello I have more than one question about an old video it was about (How: Heart Beat Monitoring! Arduino/Processing) Ido it for my college project please if you can help me how can I contact with you
Wouldn't you want the sense resistor as close to the pins as possible? Wouldn't the traces introduce added resistance, especially at lower ranges? Or is the resolution not enough to significantly skew the results?
Look at the schematic or board layout. He is using a Kelvin connection ("Kelvin" not "Kevin" ;) ) as he should. Trace resistance is irrelevant in that case.
Very interesting. Sadly, I need something that will go up to 10A at least. Well, maybe LTC has a chip that can in that range or I will have to run it through a divider or alike to make sure the Amps aren't too high.
nope, actually bought 50 of them to hand out to the Patreons... yep not cheap. I'm definitely going to add the TI part as an alternate in the description of the video
Please suggest me to measure th DC voltage upto 500V and Current of 20A I want to make power meter for my solar panals whcih are capable of giving volage upto 300V and current upto 5A please suggest me the sensors
That depends on your requirements. I like the ACS712 because it's completely isolated, can measure positive and negative curents and has an internal resistance of only 1.2mOhm (so almost no voltage drop). But it will have higher noise than a dedicated chip that uses a shunt resistor to measure the current. Since you need "only" 50Hz you can use a simple RC-low pass filter on its analog output and should be able to reduce that considerably (the ACS712 on it's own has a 80kHz bandwith, so it's definitely not your limiting factor). You can always go other current sensing chips, that use a sense resistor but have analoge output. So you provide the ADC yourself (the µC internal one for example). I know there are alot of these chips, but can't really give you any part number, because I have never used one of them.
I used the ACS712 in the past to measure the current draw of two servo's (Dynamixel RX-24F). I need 50Hz (or higher) to be able to see the spikes of current consumption used to drive the servo's. I didn't use any RC-filtering because I thought that would average or smooth my signal to much. I'm looking for other options because over time the voltage on the output (with an input of 0A) was drifting. I only need to measure positive currents ranging from 0A to 6A with a resolution of a 1-100mA. I had a quick look at the datasheet of the ina219 of TI and this chip seems faster.
Hey for a mechanical relay for arduino which comes with octocouplers as the ones sold on ebay, would you worry that much for noise that could possible damage the relay or not really? Thanks
and of course if you need more than the 10a you could use a hefty big current shunt off ebay and with some maths on the arduino correct ofr the shunt resistance i may add one to the cars charging system to show charge....
hello Kelvin, how small can the shunt resistor be, once changed is it just a matter of code changes. looking at getting a .000200 ohm 7W shunt resistor. au.rs-online.com/web/p/surface-mount-fixed-resistors/1249340/
Hello there! Nice try. Your presentation is confusing. There is nothing "EASY CURRENT & VOLTAGE MEASUREMENT WITH ARDUINO" as your video title says.. A lot of improvement still has to be done. At any, thanks for your effort.
Thanks for the detailed project review of the LTC2945. I had been looking for a way to monitor higher dc voltages other than using voltage dividers. Your project paved the way for using the LTC2945 with a ESP8266-12E on my 24 vdc solar collector. Now I can remotely monitor the collector voltage on my favorite IoT site.
cheers kevin....this is exactly what I needed for my project.
many thanks and virtual beers to you man ;)
The connector could add more than 0.02 ohm resistance between the main board and the board with resistor. Hopefully the contacts in the connectors are gold plated, but I doubt it.
You should have placed just two resistors in series on the pcb spaced enough apart and with two half circle pads to allow user to connect two traces together with a blob of solder - if user wants to enable that resistor, just add a blob of solder over the two pads to create the connection, and unsolder the other blob of solder to disconnect the other resistor.
This way the pcb copper layer would also behave like a tiny heatsink for the current sense resistors and reduce the drift with temperature.
Hi, I am using your code and your calculated shunt resistors in one of my projects and my measurements only accurate with the 10 Ohm resistor. If the resistor is smaller than 10 Ohm for example: 1 or 0,1 or 0,01 the mesaurments are not accurate. The higher the current higher the devitation. The devitation always higher. Thank you in advance for your help!
The TI INA230 is just above $1.00 on digikey and TI has 16bit A/D. The LTC2945 is well over $4.00 and it has only a 12bit A/D. Both are pretty similar otherwise with TI still holding an edge on other small things. except for the 80volt upper voltage limit which reduces resolution anyway, I see no reason to even consider this part for anything. And I am not a big TI fan either. Sorry LT, just another overpriced part.
wow, that is a cool part. I did need the higher input voltage though, so that's why I chose LT. Might have to do a part 2 with the one you found
I've tried a couple other current measurement techniques (built-in sensing on a motor controller chip, and a separate current sensing chip inline with the main power to the motor controller, but they didn't have such resolution!
Since I would like to be able to measure to 6-7 A (for motor-stall conditions), could I use a 1/2-w sense resistor (wound wire? How would I get that low a resistance... another puzzle!) to be able to sense more than 5A?
That's a versatile board design. Thank you for sharing the design!
Just confirming, since the current is calculated from the input voltage and the sense resistors, this would not work for a non-ohmic device such as a diode?
I would like to be able to measure the power being used in my Sous Vide controller using a SSR using PWM from a Chinese temp controller. I want to track the temp vs power curve to see what is happening during operation. It is 115VAC slow cooker for now ( 350 watts) and later a bigger heating unit of 1000 watts. From another question I cannot use your device (80VAC MAX). What can I do to get the readings in the arduino? I eventually want to make my own controller on the esp8266 or ESP32).
Nice work (: Do you disable the internal pull-up resistors for the I2C bus, when you are using external 10k pull-up resistors? (:
Hi Kevin,
I am interested in this board, but cannot find it via the link you provided. Is is out of order?
hello
I have more than one question about an old video it was about (How: Heart Beat Monitoring! Arduino/Processing) Ido it for my college project please if you can help me how can I contact with you
is there room on that board for banana jack ring footprints, standard spacing ? I looked at these chips before, but $$$
Can i have to measure 50A is it possible???? R=0.001ohm
Power=2w
Wouldn't you want the sense resistor as close to the pins as possible? Wouldn't the traces introduce added resistance, especially at lower ranges? Or is the resolution not enough to significantly skew the results?
Look at the schematic or board layout. He is using a Kelvin connection ("Kelvin" not "Kevin" ;) ) as he should. Trace resistance is irrelevant in that case.
yep it's a Kelvin connection, but ideally you want to have this close to the device... I really wanted the resistor to interchangeable
Very interesting. Sadly, I need something that will go up to 10A at least. Well, maybe LTC has a chip that can in that range or I will have to run it through a divider or alike to make sure the Amps aren't too high.
His board will take you up to 10A - there is a jumper for that.
In short, can it measure AC Voltage and AC Current? I need to embed in ESP8266..
Kevin if you made your own multimeter I would buy the shit out of it!
for higher voltage levels ... can I use a voltage divider, as long as the chip input impedance is pretty high?
NO, the maximum input voltage will be exceeded when measuring current.
Did Linear Tech give you this part for free?What other reason to use it?
nope, actually bought 50 of them to hand out to the Patreons... yep not cheap. I'm definitely going to add the TI part as an alternate in the description of the video
Please suggest me to measure th DC voltage upto 500V and Current of 20A I want to make power meter for my solar panals whcih are capable of giving volage upto 300V and current upto 5A please suggest me the sensors
Use a HV depletion NMOS from Infineon and float the LTC2945, this is shown in app ckt. Use optos to send the digital data to the operating ground
What happens if the current flows in the opposite direction? I want to measure the current in two directions.
Use LTC2992, essentially LTC2945 x2 and only priced slightly higher
Great video.. Thankyou
Hey Kevin how does it behave when the current is bidirectional?
Reads code 0 for anything
What is the maximum polling frequency achievable via the I2C bus for this chip?
The value will be in the datasheet, which there's a link for in the description. Won't get easier than that.
The value I found is 7.5 Hz. If I want a polling frequency of 50Hz I'm better off with the ACS712 series for a current draw of 6A max?
That depends on your requirements. I like the ACS712 because it's completely isolated, can measure positive and negative curents and has an internal resistance of only 1.2mOhm (so almost no voltage drop). But it will have higher noise than a dedicated chip that uses a shunt resistor to measure the current. Since you need "only" 50Hz you can use a simple RC-low pass filter on its analog output and should be able to reduce that considerably (the ACS712 on it's own has a 80kHz bandwith, so it's definitely not your limiting factor).
You can always go other current sensing chips, that use a sense resistor but have analoge output. So you provide the ADC yourself (the µC internal one for example). I know there are alot of these chips, but can't really give you any part number, because I have never used one of them.
I used the ACS712 in the past to measure the current draw of two servo's (Dynamixel RX-24F). I need 50Hz (or higher) to be able to see the spikes of current consumption used to drive the servo's. I didn't use any RC-filtering because I thought that would average or smooth my signal to much. I'm looking for other options because over time the voltage on the output (with an input of 0A) was drifting. I only need to measure positive currents ranging from 0A to 6A with a resolution of a 1-100mA. I had a quick look at the datasheet of the ina219 of TI and this chip seems faster.
Hey for a mechanical relay for arduino which comes with octocouplers as the ones sold on ebay, would you worry that much for noise that could possible damage the relay or not really? Thanks
and of course if you need more than the 10a you could use a hefty big current shunt off ebay and with some maths on the arduino correct ofr the shunt resistance
i may add one to the cars charging system to show charge....
Why have the i2c resistors external ?
Because if they were internal to the board and you put it into an existing system you would have too much pullup on the bus.
I see
ina219 is another good option
Orkhan AmirAslan but not up to 80V
only for 26V (safely)
hello Kelvin, how small can the shunt resistor be, once changed is it just a matter of code changes. looking at getting a .000200 ohm 7W shunt resistor. au.rs-online.com/web/p/surface-mount-fixed-resistors/1249340/
unfortunately it costs more than a few micro controllers.
yep, not a cheap part
Not only can't the small resistor take more than 5A, the pins aren't made for 10A either (even if there are three in parallel).
yea good point!
1:25
Hello there! Nice try. Your presentation is confusing. There is nothing "EASY CURRENT & VOLTAGE MEASUREMENT WITH ARDUINO" as your video title says.. A lot of improvement still has to be done. At any, thanks for your effort.