Skip to content →

ITPower Posts

Solar Cart Guide

This post is a work in progress!

The Cart


Meet the ITP Energy cart! This is your new best friend if you’re doing a solar-powered project, or any energy-related project at all. However, the cart must be understood and respected. This blog post is an introduction to the many things it has to offer.

What’s In The Cart?



This is what we all came here to see. Designed as a grow lamp, presumably for indoor gardening enthusiasts, this lamp simulates the light of the sun, which makes it work great for solar projects!

How Powerful Is It?

Depends on the settings, and where you put your project! The lamp has two switches: “Growth” and “Bloom”, which offer different red/blue balances for different stages of plant growth. Since we’re not growing plants, however, we’re mostly concerned with how much power these different settings offer. Let’s break out the handy AMPROBE Solar meter, located right next to the panel on the cart!

This device can measure both Watts per meter squared and BTUs (if you need that for some strange reason). If you’re using the AMPROBE to test solar power, make sure it’s set to W/m2 and you’ve hit the “R” button to reset the range to measure ranges above 200 w/m2. If you see a decimal point between the numbers, you’re in the wrong range.

Some Context

The standard measurement for direct sunlight is 1000 watts per meter squared. Often, people will try to test their solar projects next to a window. Let’s see how that measures out.


This was on a cloudy day, but… not great.

Now let’s try right outside the window.


Still not great. If you’re not in direct sunlight, your watts per meter squared will decrease veeeeeery rapidly.

Growth Setting


Right next to the LEDS — 530 Watts per meter squared, or about 50% of direct sunlight.

Base level.

At this setting, if we place our meter right up to the LED panel, we get about 530 W/m2 — around half the wattage of direct sunlight. However, if we move our meter even a foot away, that measurement drops to about 198.

Bloom setting


Not great, Bob.

The bloom setting gives us waaaay lower light. So what do we do to get close to the sun?

Both switches on


90% of direct sunlight! That’s bright!

About 30% of direct sunlight at the lowest point. That’s less good.

Now we’re getting somewhere useful. Still, if we move our meter even a foot away from the panel, we lose well over half of that wattage. This could be useful for testing low-light scenarios, but if we want to test for direct sunlight, we’re going to need to be as close to the light as possible. Luckily, it’s adjustable for just that reason.

Setting Up Your Project


First, measure the height of your project to determine how close the light needs to be to it.


Then, very carefully, with ONE OTHER PERSON HOLDING THE LIGHT, take turns adjusting the chains on each side to lower the light. Just unhook the chains from the S-hook attachment and rehook them at a lower or higher height.

You can now put your project under the light. But DO NOT TURN IT ON YET.



Next to the light, there’s a big green fan. Make sure this is on and blowing air before you turn on the light.



Congrats! Your project is ready to go.

Equipment (in the bin underneath)

Drok USB intelligent electronic load — great for testing projects with a variable load, or checking if a power source can power something before you buy it.

SunTransfer Solar Lamp/Mobile Phone Charger x5 (at least 2 broken) — I don’t know how old these are, but their website doesn’t work anymore, so it’s anyone’s guess how well they work. Could be fun to take apart!

Adafruit 6V/6W solar kit x4 (kit 1 currently missing DC/DC converter) – These are probably going to meet most of your solar project needs in an ITP context.

SunTransfer 9V/2W panel

Watts Up watt meter This is an incredible device that you should use in any project where you’re measuring energy consumption. Some basic instructions for setup are below.

Tempest Security Battery — this is a big-ass battery, if you need one of those.

350 farad capacitors & individual solar cells in cardboard box

KillAWatt PSx — electricity usage monitor, good for monitoring project consumption over a longer period of time or just solving your curiosity about how much power your macbook consumes.

Open Circuit Voltage With A Solar Panel


On a multimeter, connect the black wire to the COM socket and Red to the standard voltage socket, usually found on the right side. Set the multimeter to measure DC voltage.


Using your handy alligator clips, which you definitely remembered to buy, connect your voltage to the red wire on solar panel and ground to the white wire.


Congrats! You can now measure your solar panel’s output. If it’s working correctly, it should give you a decent output. Using one of the 6V Adafruit solar panels, you’ll get about ~5V if it’s close to an indoor lamp.

Short Circuit Current With A Solar Panel


Keeping your multimeter attached to the solar panel, switch the red cable to the 10A socket, usually found on the left. Set it to the 10A measure setting.

This may not work out well, depending on the current your power source is putting out. However, there’s an easy fix for that!

For low-power current sources, simply switch the red cable back to the standard output, and try measuring microAmps and milliAmps. If you’re getting any current at all, it should show up. For example, on one of the AdaFruit 6V solar panels, I get about 5 mA under a standard LED lamp.


Still confused? Check out this helpful video: Solar panel: how to measure open circuit voltage & short-circuit current.

Measuring A Battery While It’s Being Charged/Measuring Energy Consumption Over Time/Measuring Pretty Much Anything With The Watt’s Up


Always good to test with an LED first.

A 9V AA powering an Elegoo — all is well here.

Meet the Watt’s Up watt meter. If you’re doing an energy project, you’ll want to be best friends with this guy.

This meter measures current, voltage, and time, and calculates peak current, peak power, minimum voltage, power (W), energy (Wh)and charge (Ah) values on any circuit you attach to it.


Simply attach your power source to the “source” wires on the left of the screen and your load to the “load” wires on the right.

Testing low-power batteries (anything below 4V)

If you’re measuring a power source below 4V (which is going to be a lot of the LiPo batteries used in ITP projects and any smaller batteries like a coin cell), you’ll need to connect the auxiliary power on the Watt’s Up to get accurate measurements — the three wires on the left side by the “source” wires. These should be marked, but since things at ITP are rarely the way they should be:

The blue wire furthest from the source input is ground.

The green wire in the middle is the voltage input.

The yellow wire closest to the source input resets the Watt’s Up.


These can easily be connected to a wall power source or a 9V battery — anything above 4 volts should do.

In the extremely confusing picture below, I’ve:

  1. attached a 3.7V LiPo battery to the “SOURCE” input.
  2. attached an Elegoo to the the “LOAD”.
  3. attached the wires from a wall charger to the Watt’s Up auxiliary power source.

Look at those sweet, sweet measurements.

This way I can get accurate measurements from my low-power battery.

Still confused? The Watt’s Up manual (TODO: find permanet file storage for PDF) is extremely detailed and helpful if you run into any issues.

Measuring Variable DC Load With The Drok Intelligent Electronic Load


This is a crazy little machine! It mimics an electronic load on a power source so you can check how it will perform in different power consumption scenarios. To set it up, connect the VOLT connectors to a reliable power source (a wall charger should work fine) to power the electronic load itself, then attach your battery or other power source to test its performance.


As you can see here, I’ve attached it to a 3.7V LiPo battery, which should offer 2 amp hours when fully charged (so it should last two hours with an output current of 1 amp).


Rotate the dial to adjust current.

Press down on the dial to get different measurements.

By rotating the dial on the front, I can increase the current draw of the electronic load, and by pressing down on it, I can check the wattage, amp hours, and hours left of charge at the current draw on my power source. In this example, it’s only measuring 00.05Ah on my battery — chances are the battery either needs charging or isn’t very good.

Helpful resources:

The Killawatt PSX


This thing is a lot of fun and a great way to measure long-term power consumption on larger projects. Simply plug it into a wall socket and plug your project in, and it’ll give you tons of information about its energy consumption and efficiency: voltage, amperage, power factor (the percentage of available energy being consumed), kilowatt hours of power consumed since it’s been reset, and leakage (amount of power in the circuit returned to ground).


If you cycle the button for each, it’ll also give you the minimum and maximum values recorded. If you have a project you want to leave plugged in for an extended period of time, this is a great way to track its energy consumption.

You can check out the manual here

Leave a Comment

Current Sensing 4/5/19

Using a 60w lightbulb, a current sensor, a precision rectifier circuit (pictured below) and the ReadAnalogVoltage program from the Arduino Examples (thanks, Tom!) we got readings of 3.1-3.4 volts

rectifier circuit diagram

If you’re curious to see this stuff IRL, here’s the setup:

The rectifier circuit and Arduino
rectifier circuit from above. Hopefully the wiring is clearer?

Many credits to Sid for figuring out how to use the rectifier circuit! 

Here’s what we need to do next:

Leave a Comment

Energy Monitoring/Current Sensing Update

Summary of the 3/28/19 Meeting
By Luming, Rushali and Bev (ed. Jenny)
– Existing setup was to plug three current sensors into one 6-pin clippy-plug thing (is this an official name for this?). 
– we had trouble sticking wires leads in and seeing anything interesting on a multimeter. so we looked up the datasheet and looked at voltages on a multimeter and three oscilloscopes while we had the current sensor wrapped around a (SINGLE WIRE) of an ac powered lamp (see pic below). the datasheet was incredibly lacking in data despite being a sheet of data but from this exercise, we determined that the current sensors are the kind that go from A_Lot_of_current -> Small_But_Proportional_Measurable_Range_of_Current not A_Lot_of_current -> Small_but_Measurable_Range_Of_Voltage. current sensor wrapped around a single wire of an AC-powered lamp
1. We worked with two bulbs primarily of 130 and 240 mA (if Rushali recalls correctly)
2. Our signals were super noisy but that’s probably cause we were looking at voltage
3. This website is SUPER helpful: and It seems like we need to add a “burden resistor” to convert the current output to a voltage output. 
Here are the charts from the data sheets we looked up that (Bev thinks) shows what our expected current output should be based on the current input:
input-to-output chart
It looks like this would be the chart from the data sheet to look at to pick a resistance, so we can get the appropriate output voltage:
chart of resistors and expected output voltage
Goal for next meeting:
– use Elizabeth’s current or similarly set-up op amp to convert this output current to voltage
– observe the range of voltage from the op-amp in relation to the current of the lamp (as measured by some commercial device i suppose)
here is the datasheet for the current sensor we’re using
Expand below for some pictures!
Leave a Comment

The Latest on Energy Monitoring at ITP

The Latest

A few of us got together today to test the Precision Rectifier Circuit on a breadboard with a dimmable lamp.

We debugged and saw lots of signals on the oscilloscope we were not expecting. Hurray! Something to work off of.

The oscilloscope not showing us we want

The highlights are…

  • We need  different Current Transformer Sensors than the ones we have.
  • We might need to use two different sensors: first, the high current sensors we have, for very high energy usage appliances – probably only the air filter connected to the laser cutters. Second, possibly lower current sensors for the rest of the circuits. We also debated whether to stick with the High Precision Rectifier circuit we have or not. The current circuit has two stages of op amps, one for current-to-voltage conversion and another for amplification. We might be able to replace this with a simple a Differential Circuit/Amplifier.
  • We can use an op-amp calculator like this one to predict the components we need to use for various circuits.

Read on for more.

Current Transformer Sensors

 The current transformer sensors we have are “too large” for what we are monitoring on the floor. They are designed to convert a 0 – 100 Amp AC current input to a 0 – 1 Volt output, with a 1% margin of error. This means that one amp of current in the 120V AC circuit translates to 0.01 volts in the output of the sensor. Since the margin of error is 1%, any reading that’s 0.01V or less could be an error. However, many appliances we want to measure consume less than 1 amp of current. For example, a 65-watt Apple adapter for a laptop consumes 0.54 amps at 120 volts. This means that it, and devices in the same range, will be giving us readings that are less than one volt. Understanding that these sensors that can’t reliably read less than one amp (i.e. output 0.01 volts) is helpful to us.

Now we will look for sensors that give us the resolution we need for our devices.

We likely want sensors rated to  take an input of about 0 – 15 A at 120V AC and output about 0 – 1 (or up to 0-3, ideally) volts.

Two Different Circuits

We also wondered if we can simplify our lives a little and use a different circuit for our applications. We need our system to turn a reading of amps into volts at one point or another, and our current Transformer Sensor already does this for us (for example, giving us a reading of 0 – 1 volts for anything 0 – 15 volts). Therefore, we don’t need the amp to voltage conversion of the first half of the circuit we’ve been using. The sensor and the LM358 appear to be performing the same work twice.

Instead, we might be able to use a differential amplifer circuit, which only amplifies a reading as needed. Theoretically, we can rely on the Current Transformer Sensor to give us a voltage value.

Mission Reminder

As a reminder, generally we want to get energy readings of appliances on the floor so that we can analyze, act on, and play with these numbers. Whenever we choose an appliance to monitor, we need to do two things to prepare the signal to be readable by a microcontroller: reduce its voltage down to a range of 0 – 3.3 or 5 volts, and convert or “rectify” any AC signals to DC. If an appliance’s voltage is not already within 0 – 3.3 or 5 volts (the allowable range for microcontrollers ), then we would rather reduce the voltage to an even smaller amount and amplify it back up again. To reduce voltage down, rectify it, and amplify it back up, we are using the Precision Rectifier Circuit and possibly the Differential Circuit / Amplifier. 

Visualizing It All

Hopefully this infographic helps new folks see how it all fits together.

A system diagram of the energy monitoring system


Leave a Comment

Precision Rectifier Circuit

Elizabeth has been doing some really great work on the Precision Rectifier Circuit. You can see her presentation here. While we keep working on it, here is some recommended reading from Practical Electronics for Inventors

  • 2.21 – AC and resistors, RMS voltage and current
  • 3.6.9 – Capacitor applications 
    • Coupling and DC + power supply decoupling
  • 4.2.5 – Diode/Rectifier Applications
  • 7.4 – Oscilloscopes 
Leave a Comment

Summary of the Year (so far)

The main tasks of the 2017-18 academic year were:
1. To restore the function of the solar charging station on the 12th floor of Tisch, and
2. To investigate what it would take to replicate the current sensing system put in place by Enertiv with a system designed in-house.

1. Solar

solar panel on roof

The solar charging station  was put in place as a demonstration project for students to get direct experience with the conversion of solar energy into usable electrical energy. An 80W solar panel (seen above, on roof of 721 Broadway) is fed to a charge controller on the 12th floor, which both regulates the electrical output of the panel for safe charging of a battery pack (below right), and provides a data feed (below left) of the changes in current and voltage coming from the solar panel.

Left, readings from solar db. Right, charge controller, pi and battery.

The data feed had been inactive for a couple of years prior to 2017 since the original computer attached to it had reached the end of its life. This year, we restored the data feed by connecting it to a Raspberry Pi embedded linux board, which connects to a database on the ITP server via the NYU-legacy wifi network.

2. Energy Management System (EMS)


The Enertiv system provides circuit-level current and voltage monitoring on the floor with an enterprise-level web API. Above, you can see one of Enertiv’s circuit monitoring boxes installed on the floor.

While Enertiv is useful as a tool for students to monitor the floor’s energy use data, it does not provide transparency as to how that data is collected. Since ITP values an understanding of the workings of electrical and electronic sensing and communication systems as a part of the curriculum, we felt it would be useful for students to gain first-hand knowledge of the physical layer of the data collection system. Enertiv’s early hardware was based on the open-source BeagleBone platform, providing that transparency, but the electrical details of its current systems are no longer published. So this year, we decided to build  and document an equivalent voltage and current sensing system from the physical layer through  to the database.

Pictured below is one of the circuit boxes on the floor. The black boxes with numbers on them are Enertiv’s current sensors. The blue clamp on the left side is one of our current sensors.

Panel opened

Below we summarize the tasks performed, along with the point people for those tasks.

Fall 2017:

Replaced batteries in Malia (yes, the battery is named after Malia Obama), so we had one working solar battery! – Jenny

Wrote code for pi and set up solar db, which now runs on NYU server. The pi sends daily emails to Jenny, Mathura, Steph, Jeff and Tom, and hourly updates to the db. The data can be accessed via Heroku. – Jenny, Mathura

Peter and Vishal from Bug Labs introduce their API platform, signalpattern, that makes Enertiv easier to integrate with other API platforms. Signalpattern had nice qualities but it felt like it was another layer of dependency, further removing us from the actual data itself. – Jenny, Mathura, Steph

Researched making our own energy management system, inspired by Enertiv. – Jasmine, Jenny, Mathura, Steph, Yeseul (YAY)

Spring 2018:

Replaced batteries in Jenna (why yes, the battery is named after Jenna Bush); we now have two working solar batteries.  – Jenny, Andrew

Tried and tested multiple rectifying circuits for EMS. We’ve settled on THIS ONE and will make more after May 1. – Rushali, Yeseul

Tested MQTT as the protocol for sharing current sensor values in our homemade EMS. We’re not sure whether this will be used moving forward. – Mathura, Steph, Yeseul

First year workshop: hello Andrew (’19), Beverly (’19), Caleb (’19), Hayley (’18), Lola (’18), Sam (’19), Yue (’19), and Jeff Feddersen (’02). – Jenny, Yeseul

Made video with Jeff summarizing the good times.

To-do, starting May 2:

Move solar data management from current pi to new microcontroller that permits 5GHz connection. This way we can move from NYU-legacy to the regular NYU wifi network. – Jenny, Steph, Yeseul

Figure out the future of the solar panel. – Jenny

Homemade EMS: make more circuits, figure out the database system. – Andrew, Mathura, Sam, Steph

Suggestions for year 2018-2019:

Homemade EMS: Complete one circuit box. Make system ready for Brooklyn (and test there). Create user-friendly dashboard for our energy usage data.

…and who knows?

One Comment

Battery Maintenance

Today, Andrew and I replaced the batteries in Jenna. With Malia, this now makes two working xPower 1500 portable batteries for our solar panel!

Two working batteries

The wiring of the xPower is fairly straightforward. Follow the picture below:

Inside the battery

The only thing of annoyance is that the xPower was designed with this kind of replacement battery in mind, which has M5 Insert Terminals built in. Unfortunately, when we ordered our replacement batteries we ended up with this kind of terminal:

Battery terminal: requires nut

As opposed to the M5 Terminal, this one requires a nut. The replacement batteries actually came with nuts and screws, but it was too late – I’d gone to buy some. On the upside, there are now spare nuts and screws in the lid compartment of Jenna. And should it come up again, we bought 5mm nuts with 0.80 Pitch.

Leave a Comment

ITPower Workshop!

On April 6th, we hosted the ITPower workshop to share what we’ve been doing and what we’ve learned from the projects. We started the workshop by introducing how the ITPower started and what our vision has been. After talking about the solar panel project and DIY energy monitoring system that we’ve been building, we delivered a brief interactive (!) lecture on two topics: MQTT, a lightweight IoT protocol we’re using on the projects and Precision Rectifier Circuit which is an essential part of the energy monitoring system. Here is the link to the slide.

We invited Lola to talk about Sustainable ITP. She gave us a great presentation about the group.


We invited the workshop participants to solder precision rectifier circuits in case they want to find a peace of mind, learn more about the circuit, or contribute to ITPower. Many of them joined us :)


We setup a circuit testing corner in the room so people can test their circuits after making.


Here are the soldered circuits made before/during the workshop. Some of them still need debugging and testing.


Thank you everyone who came to the workshop, Tom, Jeff, Lola, and Power Rangers!


Leave a Comment

Precision Rectifier Circuit on Perf Board

On Saturday, we successfully soldered three precision rectified circuits on perf boards.

Parts we used:


Circuit diagram:

rectifier circuit diagramIt’s ready to be used:

rectifier circuit

The circuit is doing the right thing!
yellow: reading of the signal after the first op amp
blue: reading of the final signal

rectifier circuit with reading

Leave a Comment

Simplified Rectifier Circuit on Perfboard (FAILURE!!)

March 13

Today, Rushali and I built and tested a simplified rectifier circuit.

The simplified rectifier circuit is a modified version of the precision rectifier circuit that Jasmin and Aaron shared with us. What the original precision rectifier circuit does is 1) rectify AC signal (the first op amp) and 2) amplify the signal (the second op amp). When Jenny, Mathura and I built the precision rectifier circuit last time, we figured that the second op amp which is used for amplifying the rectified signal is not necessary for our purpose.

Below is the circuit diagram of the simplified rectifier circuit.

circuit diagram

We tested on a breadboard first.
correction: pin2 – 680ohm – ground(-)  –>  pin2 – 680ohm – power(+)

simplified rectifier circuit

It gave us a good result (at least, we thought). Blue is AC input signal (60Hz) from the function generator and yellow is rectified and smoothened signal that we need.


We made the same circuit on perf board. There are two sets of pins: input on the output, and output on the left. See the image below to see the details. Rushali and I each made one so we now have two working circuits.

rectifier circuit on perfboard

Wondering what was our working station like today?

working station!

The next step is to test this circuit with the current sensor.

After doing all of those above, we took a closer look into the oscilloscope reading and realized that the waves from the precision rectifier circuit and the modified version that we built today are a bit different–as seen below. Our observation is that in the modified circuit, the peak of AC and rectified signal doesn’t exactly match. It seems like the rectified wave is a bit off to the right–does it mean that it’s being delayed or what? We would like to examine this more.



March 24

It turned out that the simplified circuit we made here had not been powered properly and the oscilloscope reading that we were looking at meant (almost) 0 voltage… Make sure to read the voltage from the right 0: the 0 line is not the bold dotted lines in the middle, but the tiny arrow with channel number (1/2) on the very left.

When we powered correct, the reading looked like this.

simplified-with capacitor

When we replaced the 47 uF capacitor with a wire, the reading looked like this (voltage is 0 because we’re basically measuring a point that is directly connected to the ground).

simplified-without capacitor

We learned that for the capacitor to be able to hold and release energy, using resistors in series or the second op amp (pin 5, 6, 7 of LM358N) between the output from the first op amp (pin 1 of LM358N) and the capacitor is necessary. Eric recommended us to keep the second op amp because it’s simpler than using several resistors. The IC we’re using (LM358N) have two op amps in one part.

There might be a more clever way, however, we decided to keep the first and second op amps in the circuit for now.

Thanks to Tom, Mathura, and Eric for advising us.

Leave a Comment