To usb idea

We also posted a tutorial on calibrating the NTC thermistor on those loggers with a method that's easy to scale for batch processing of multiple loggers: thecavepearlproject.org/2024/04/20/a-practical-method-for-calibrating-ntc-thermistors/

We also posted a tutorial on calibrating the NTC thermistor on those loggers with a method that's easy to scale for batch processing of multiple loggers: thecavepearlproject.org/2024/04/20/a-practical-method-for-calibrating-ntc-thermistors/

We also posted a tutorial on calibrating the NTC thermistor on those loggers with a method that's easy to scale for batch processing of multiple loggers: thecavepearlproject.org/2024/04/20/a-practical-method-for-calibrating-ntc-thermistors/

We also posted a tutorial on calibrating the NTC thermistor on those loggers with a method that's easy to scale for batch processing of multiple loggers: thecavepearlproject.org/2024/04/20/a-practical-method-for-calibrating-ntc-thermistors/

😄your project make me alive again

you can conect this to type c cable?

Hello Edward , Thank You for making this video, Thank You for Spending your time to Share with us from your knowledge and Thank You for smart choice using simple words to explain - for us , those no native English speakers ... Now, I am struggling for more than 2 month now, in my attempt to make an Arduino Nano simple clock, some temperature from DS3231 and a TM1637 LED display ... I have a "major" problem with DS3231 RTC module - I can upload the Code, I can see the time in Serial Monitor, even on TM1637 module, but when I Disconnect the Power of the Arduino from computer and plug the 5vDC PS the time has "left behind" at the moment I was upload the Sketch ... Thos Mod You did to the RTC module will Solve this Issue or is a Program/Code's "fault" ? Thank You and Bless You and your family ...

Also. In the past you cut Vcc pin (pin 2) and connected to ground. Now you left it floating. Are both ways acceptable?

Hello! Great job!! I can see you have modified the way you feed DS3231. Some time ago you used 2 1n4148 diodes to feed it from the coin batt and from Vcc. Now, it seems you had joined both sources. Could you please give me a link to the current system diagram to see how are you really feeding the RTC? Thank you in advance!

You dont have to put that on top of the original regulator location 😂 just let it on the side

Wonderful, I like how you have solve many issues of enviromental protection of these loggers with commodity parts!

Thank your for making your work public, Ed. It's very helpful.

Thanks for the video, this is a great idea. Unfortunetly these PEX swivel connectors are not available in the UK. I understand as you said in the video it can still be done with the 1" cap but its not ideal as its a bit tricky to stop the wires twisting in the housing due to it being fixed / no independant spinning nut.

In one of your loggers you used bme280, how much pressure they can handle beyond 1100hpa as it says in doc?

Recently won an auction ($35) and got a "box" of those DS3231 modules with 24c32. Turned out the box had 132 of them with battery in each! I am [now] searching for every way I can use them. Your information is great!

Hi, I have an idea for you. You can use hall-effect sensors and magnets as switches. In other words, if you wanted to have like a button to download information--instead of drilling a hole through the housing, you could attach a hall sensor, such as a Honeywell bipolar sensor inside the housing--with no hole drilled. Then it can be activated by a magnet. Also, with some experimenting linear or ratiometric sensor can be used for with springs to measure pressure or flow, or perhaps with a lever to show water level. Cool, inspiring projects.

I love the banter between the two of you! I'm a little surprised you're not using syringes of flux and solder paste, and the hot air gun, to solder the replacement EEPROMs. That method is a lot easier than using an iron and fluxing by wire. Although this is for beginners you expect them to have a hot air gun already. Flux and solder paste syringes are so convenient I find myself using them all the time now, with the iron as well.

Thanks for your proyect! is very inspirational for me! Diving into the CPP website is like falling in a Tolkien world of open hardware, electronics and environmental monitoring. I was wondering if the process of adding the epprom will be the same for At24c256 dip...

Part 3 sensors and code?

Found you by accident. I like your dual talking combination. Nice work. I have also been mounting my capacitors incorrectly. I thought the bar meant negative. Seems the standard has changed for SMD. Cheers

That is a cool upgrade. Thanks for the inspiration and your willingness to share the details.

Edward Mallon, an inspiration to me. Great Job

Have you looked at using the Macetech Chronodot V2? No need for ugly surgery. The V3 now has eeprom as well.

] 6 п5

I imagine in the four years since this video was posted you have already determined this, but that tap marked "8x1.25" is a metric tap for an M8 hole with 1.25mm thread pitch.

Hello there. We work manufacturing robots used to underwater inspection for the last 8 years. I'll share the most valuable secret: rubber mounted under pressure. In our case (and I can see it would be very useful in the equipment showed in the video with proper application) o'rings do the job pretty well. We tried multiple kinds of glue to the connectors... Epoxy isn't good to keep water vapor out, silicone glue is a little better, but the contact with water causes it to become more rigid and losing mechanical properties over time. Our actual solution to cables is using good quality cable glands, filled with silicone glue as an extra barrier. To mechanics we use rubber mounted under pressure, well "hydrated" with good quality industrial vaseline grease. Since we changed to this approach, our robots do a good job for years, and when they come back for servicing they're dry and well preserved inside.

Have you tried using a msp430 they will run on a coin cell down to 1.8v for a long time if you use sleep mode. Certainly not as easy to program as an arduino, but its worth the hassle if battery life is critical.

Does the EEPROM consume any power? Don't really need it.

I have time series data in uneven intervals, does that break this tool? Or how do I adjust?

You are awesome! Working on some DIY projects and this has got to be the best helpful tutorial that's more than just a quick fix or "it'll do." This also looks better than some of the other approaches out there.

32:30 is that a NPT nipple? (Sorry I didn't read your blog first...) If it's NPT why don't you put some Teflon tape on the threads?

Anyone watching this older build video might be interested in our new 2-Part logger that NO LONGER NEEDS THE SD CARDS! It stores sensor readings in EEprom and the data download is handled through Arduino IDE's serial monitor. Larger EEproms are easy to add via the I2C bus. thecavepearlproject.org/2023/12/01/the-e360-a-classroom-data-logger-for-science/

FYI: we build our own loggers to generate the data for these tutorial videos. Our latest 'mini' logger uses only two parts: A Pro Mini & an RTC module. The entire assembly can be done in about 15 minutes, and it runs for a year on a just the coin cell. thecavepearlproject.org/2022/03/09/powering-a-promini-logger-for-one-year-on-a-coin-cell/

The latest 2022 iteration of our 'mini' logger can be made using only two parts: A Pro Mini & an RTC module. Assembly can be done in about 15 minutes, and it runs for a year on a Cr2032 coin cell. thecavepearlproject.org/2022/03/09/powering-a-promini-logger-for-one-year-on-a-coin-cell/

Anyone watching this would probably like to know about our latest 'mini' logger that can be made using only two parts: A Pro Mini & an RTC module. Assembly can be done in about 15 minutes, and it runs for a year on a just the coin cell: thecavepearlproject.org/2022/03/09/powering-a-promini-logger-for-one-year-on-a-coin-cell/

Anyone watching this would probably like to know about our latest 'mini' logger that can be made using only two parts: A Pro Mini & an RTC module. Assembly can be done in about 15 minutes, and it runs for a year on a just the coin cell: thecavepearlproject.org/2022/03/09/powering-a-promini-logger-for-one-year-on-a-coin-cell/

Folks watching this video might find our latest post interesting, as we now power the entire logger from just the coin cell on the RTC module. The two-part combination of a cheap ProMini clone with this RTC sleeps below 5uA, so should easily run for more than a year on the Cr2032: thecavepearlproject.org/2022/03/09/powering-a-promini-logger-for-one-year-on-a-coin-cell/

I have broken several MCP1700 with heat when soldering. The broken ones still can regulate the voltage correctly, but the quiescent current is way too large, like 1 ~ 10 mA.

Thanks for a very useful video! My guess as to what consumes the 100uA would be the Vref inside the "power control" circuit (datasheet page 10). When you're powering directly from Vbat this power control and thus Vref is likely mostly off. Now, my problem is that apparently the DS3231 on my board is a scrapped version that consumes about 350 uA *more* than specified. I guess cheap ebay stuff is cheap for a reason...

Thank you for your real world advice, 95% of KZbin tutorials are just bench / theoretical stuff. I made a timer that turns on and off a water feature following the real rising and setting of the sun, using the ds 3231 but it kept going off at unexpected times i narrowed it down to the ds resetting itself to 1/1/2000 8 am for some reason, I remembered seeing this video and looked it up again. Did the capacitor mod and so far so good. I guess turning on the relay gives such a jolt to the electrical system that the ds resets itself just like you said. So this was excellent. many thanks.

One thing to watch out for when retrofitting a new regulator onto 'some' cheep promini clones - they have a diode that may prevent some regulators from stabilizing because they cant 'see' the 10uF input capacitor on the promini board: "You can see there are two components (labelled “h” and “S4”) between the RAW pin and the regulator and C19 capacitor that aren’t on the schematic. It took a fair bit of Internet searching to find out that S4 is a diode and the thing marked “h” is a fuse. Apparently these add extra protection in the case of someone mistakenly powering the board via the header and RAW at the same time." diyelectromusic.wordpress.com/2020/08/01/mini-usb-midi-to-midi/ So attach the raw input pin on the replacement regulators to the far side of that diode (instead of the pin shown in this video) - this will assure that the titanium capacitor is there to stabilize the supply regulator, OR you could just jumper out the diode? The MCP1700's I tried work either way, but the HT7333's I tried really need an input capacitor for stable operation.

Re: drill hole size. Your LED looks like a 5mm. 3mm "small" and 8mm "large", are other common sizes.

Just want to say, I use colourless nail-varnish to protect circuits (e.g. ESP8266's) from water and t works well. May even be a viable substitute for your PVC glue.

Great to know. Thanks for sharing!

Much appreciated! Was driving me nuts scratching my head :)

Hey Ed. I just started reading some of your documentation - you have a ton of stuff! I've just scratched the surface. I need to make a low power data logger, (entirely powered by CR2032) and I will try to re-use what you have already created. One quick question: Regarding this power saving mod. Adding the 1N4148 diode won't do anything, since VBAT is already connected to the cathode of the existing diode. The new diode will be shorted. Am I missing something?

BTW another handy thing to do with Excel's Fourier tool is to identify when 50 or 60 Hz mains noise is bleeding through into your sensor data ( if they are deployed indoors)

Great mod! If you used a lower voltage drop diode like a schottky type, seems like the RTC can run even longer as the battery drains and VBAT drops. Vmin recommended is 2.3V for the DS3231.

One minor nit -- the total time in cell G5 on your FFT sheets should be the number of samples times the sample interval, not the time interval between first and last samples. The deltas you calculated are the sample interval times 4095; the value you need is your value times 4096/4095 (exactly 42 2/3 days, for a 15 minute sample interval). If you do a lot of analyses of stuff largely driven by the tidal cycle, you might also consider whether it would be worthwhile changing the sample interval to get an even power-of-2 samples per tidal lunar day (~ 24 hours 50.4 minutes). That should slightly narrow any peaks driven by the moon. What surprises me about your result is why the tilt data looks like it does. In the first couple minutes of the video, the lower plot appears to be of the tilt vs time data, with most of the noise removed. But the tilt is unipolar -- it looks like all negative values have been clipped. I'd have expected either raw tilt data looking like a full-wave rectified signal (with 2 positive peaks per tidal cycle, corresponding to max inflow and max outflow), or tilt data that had been corrected for flow direction to give both positive and negative peaks (one direction corresponding to max inflow; the other to max outflow). So why does the tilt data look like all negative values have been clipped?