by

|   May  9, 2009 at 11:25 AM

Actually, you shouldn't need a giant ground lead as shown. It's only being used for a little current for sensor biasing. so a much lighter gauge wire should suffice. This would also allow the ground pad in a revised board to both shrink and move away from the power pads, reducing short circuit possibilities considerably.

by

|   August 25, 2009 at 11:52 AM

I couldn't use this on an AC line, without modification, right? I managed to melt an extension cord connector pulling too much power through it, so yeah, I need to add some tools to watch the power draw.

by

|   August 30, 2009 at 10:27 PM

Whoa, careful! Read the 2nd sentence of the description above: "DC current is determined by...". Also the description in the attached spec sheet says "Compact DC Voltage and Current Sense PCB with Analog Output".

Sounds like you tried to use this on AC? Also, it should be really clear from spec sheet and "Features" section above that voltage max is 51.8 volts. Definately not 110 AC.

Disclaimer: I am the creater of this sensor re-sold through SFE.

by

|   August 30, 2009 at 10:30 PM

SOISentinel - This sensor was made for RC airplane application (UAV actually) where 14 and 12 gauge wire is typical, and when I ship this with an autopilot the 12 gauge wires come pre-soldered. All the end-user has to do is solder on Deans Ultra or Power Pole connectors. You are right in that as far as the sensor is concerned, only a tiny common ground need to be connected back to the power sourse and load.

Also, I am now out of stock of my original 500 PCB batch, and this is being re-designed to be larger in some ways, specifically more space between both the "+" and the GND and with notches in the PCB so that Deans connectors can be soldered direct to the PCB.

by

|   September  2, 2009 at 11:50 AM

This won't handle 24VDC up to 120A, would it? I saw your note that you're scaling up the board... would love to have something exactly like this for 6S1P lipo or larger.

by

|   September  2, 2009 at 11:48 PM

6s (22.2 volt nominal) is not the problem as this unit goes up to 51.8 volts if you are using a 3.3V ADC, the issue is the drawn 120 Amps.

I meant by comments to scale it up that it would have side slots to allow direct solder of Deans Ultra which are ubiquitous in the small unmanned aerial systems that use the AttoPilot autopilot system.

I also plan to make a 180 Amp version.

by

|   October 18, 2009 at 4:24 AM

I'll ask, although i prob know the answer. I'd like to use this for monitoring charge to a 12v battery, can this do 'negative' current reading? (ie read charge as well as discharge rates?)

I think it can't but want to be sure! If it can't would it just read as 0v if current is going the other way?

by

|   October 25, 2009 at 9:31 PM

Hi smartroad. The INA169 chip from Texas Instruments is the main holdup to 'reverse' usage of this PCB. It expects one side of the shunts to be at a higher ptoential than the other side, in other words is measures the voltage drop across a precision shunt resistor, and it assumes that one side is higher potential. I am not sure what sensed current output would appear on the "I" analog pin in this case. I have never flowed current through this board the other direction. The voltage sense portion is a lot simpler in that it's only a resistor divider. The way the circuit is layed out the resistor dividor measures the high side of the shunt relative to ground, so that any losses (however trivial) across the shunt is not degrading the voltage read directly from the DC source. If you flow current backwards the reported voltage would be just a very slight bit lower than in the forward direction (assuming something catastrophic doesn't happen to the INA169 IC and smoke it and thus mess up the entire PCB). The 'slight lower' voltage depends on how much current is flowing... at the 90 amp upper limit the V drop across the shunts (2 X 0.001 Ohm in parallel, so the combo is 0.0005 Ohms) would be only 0.045 volts. Pretty trivial voltage drop especially at lower currents.

Dean G

by

|   October 25, 2009 at 9:33 PM

I recommend if you want to track discharge after a charge of a battery as you described, then it would be best to re-connect the battery to the other side so that current still flows the way the circuit expects.

Dean G

by

|   December 11, 2009 at 5:29 PM

Hi I just got mine in the mail. I plan to use these to monitor solar power generation for a project.
As soon as I opened the box I noticed that the in, ground and out traces appear to all be combined. There were no instructions. Also the board is slightly different from what is shown here even though it's the same part number (SEN-09028). Each of the inputs appear to be two small holes. The three sets of two holes are separated by large holes. Do I use a Dremel tool or similar to cut the trace where it is thinnest?

Thanks and best regards.
Will

by

|   December 11, 2009 at 10:21 PM

After looking at the AttoPilot some more I realized that they just rearranged the input traces slightly to make the ground take up one end of the board. There are two notches in my board instead of the holes. The in+ and out+ wire is supposed to solder into that while the ground solders into the end of the board much like the instructions indicate.

by

|   December 14, 2009 at 1:58 PM

Hi - I am Dean the creator / manuf of this board. I owe Sparkfun an updated spec sheet and instructions. The old sheet is posted to this page, but all the technical details are the same as the new board.

The notches are to take Deans Ultra connectors directly with the GND overlapping onto the large tinned area then lots of solder flooded on the entire GND area to solidly connect both leads. The notches also allow heavy gauge insulated wire to "plug in" to the sockets: stripped wire goes in small inner notch right up against the shunt leads with solder flooding it well top and bottom of PCB, and the insulation of the wire near the end can go in the larger portion of the notch.

I will provide a new spec sheet in the next 1 week.

Dean

by

|   January  4, 2010 at 8:13 PM

I want to use this device but at the 5V level, any chance someone has done the calculation for what replacement resistor would be needed?

by

|   January  7, 2010 at 6:54 PM

"I want to use this device but at the 5V level, any chance someone has done the calculation for what replacement resistor would be needed?"

Can you clarify what you mean? The spec sheet talks about usage with 5V referenced ADC, and no modifications are required. You lose a small amount of resolution using 5V ADC, but the resolution is already small at 22mA / bit (12 bit ADC with 3.3Vref) and 13mV per bit. Going to 5V ADC at 12 bits, the resolution is still small at 37mA and 21mV per bit.

I'm confused about why someone would be thinking about changing resistor values on the PCB though (no offense, I am just trying to figure out motive). Changing the 73.2 kOhm resistor would change the scale factor by the INA-169 shunt monitor IC, and changing the 1k/10k/4.7k resistors would change the Vsense scale factor away from 15.7:1. In other words, you wouldn't necessarily do a specific resistor swap to make the PCB 5V specific.

Simply going to a 5V ADC allows the Vmax sensed to increase from 51.8V to > 75V, however this approach (having access to higher range) doesn't apply to the current sense because top end (~91 amps) is defined by max wattage of the two shunts (4 watts together, which happens at around 90 Amps).

If you have further questions, you can e-mail me directly at dmgoedde@gmail.com,

Dean