Hey mate actually making it smaller! Hah. I have already 3D printed some cases that they just neatly snap-into too, which are semi transparent so you can see the status LED. Working on v0.2 now, which should be out in a few weeks. Just umming and arring about whether I move down to an Attiny88 w/ OpenPicoRTOS, or move to an STM32L011... which is likely going to be the lowest power option that's widely supported. Ah choices!
@@PlatimaTinkers I want to incorporate it into a single enclosed that will hold my sensor array and this power controller. So I don’t want it in its own enclosure. :-)
@@ScottCLemon Aaah okay I see. Valid use case and point! I've added notes for my v0.2 prototype (aiming to order a batch next week) about this. It'll be pushing it with space, but would 2x M2 holes suit? I'm scrapping the SPI headers, and changing to a smaller MCU because of that, so should have some room!
@@ScottCLemon Yeah I'd usually leave it with 1mm of track around it so the head isn't biting straight into the FR4, and it can be treated as a grounding for any ESD discharge
I can see an issue here with ground loops when power is cut.... If anything on the SBC side of this is measuring/sensing and has a ground or grounded shielded cable, and the power ground is cut, the power will be maintained by the grounds on sensors etc. also while the voltage drop across the mosfet is going to be very very small, this could read to odd readings on sensors once again due to the ground loop issue. two good reasons to /never/ design something that does anything with the grounds. You could switch the + rail with either a P channel or an opto coupled N channel fet just as easily.
Yeah I was thinking about that a fair bit during the design, but a) cutting GND removes the need for the added complexity of a P-channel MOSFET requiring a driver, and b) this is for low powered setups - any sensors are going to be powered by the SBC, as if there was external power, DeepSleepr wouldn't be required. So there is absolutely the potential for a ground loop, however, if there was, then this product wouldn't be needed. Not strictly, but close to, mutually exclusive scenarios? Thank you for your input either way my friend!
@@PlatimaTinkers yes agreed, just a small possibility, it's been a curse of electronics and electrics since the dawn of time. I have a dashcam I wanted to keep the internal battery fully charged so tried one of those tp405x modules only to find the module charged by switched positive and the dashcam cut off the charge by switched negative so while the battery wasn't overcharged the charging never actual ceased. Amplifier's are the worst for loops with PSU noise and odd windy sounds.
@@lezbriddon Yeah you know what, this stuck with me overnight, and in the middle of the night I realised that if there is any mounting/enclosure, if it's metal, that might be acting as a common ground. V0.2 will have a P-channel MOSFET 😅 Thank you for your support/encouragement!
Looks like I'll be squirting one out in about two weeks, as the test code I roughed up yesterday worked! Just optimising it for lower power and accuracy now, then will 3D print some cases!
Yeah hence I ordered them assembled without the USB type-C ports so I could experiment with it like that. Might leave some THT pin holes under the USB pads. Added to my notes 👍
Had not actually been measured/rated yet, but the Attiny414 draws 0.69µA in standby mode w/ external xtal. The MOSFET would likely be the biggest power sink with 4µA for zero gate voltage. Still, anything under 150µA should give ... what, 2 years on a 2600mAh LiPo? Minus time the SBC is awake for operations of course.
Hey mate thanks for the supportive words! I actually had 20 made for this prototype batch, and as of last night, they work! The code is rudimentary, and could be lower power, but does the job. 2 I am using for development, 2 I keep for historic purposes, 1 I destroyed in thermal testing, but that leaves 15 I could sell if you're actually interested in buying some?
love that first sentence! awesome intro.
Haha very welcome
Phil's Lab is a golden gem here on UA-cam.
Absolutely 🤘
This is great! One suggestion: Make the PCB slightly larger so you can add some holes for screws to mount it within a 3D printed enclosure!
Hey mate actually making it smaller! Hah. I have already 3D printed some cases that they just neatly snap-into too, which are semi transparent so you can see the status LED.
Working on v0.2 now, which should be out in a few weeks. Just umming and arring about whether I move down to an Attiny88 w/ OpenPicoRTOS, or move to an STM32L011... which is likely going to be the lowest power option that's widely supported.
Ah choices!
@@PlatimaTinkers I want to incorporate it into a single enclosed that will hold my sensor array and this power controller. So I don’t want it in its own enclosure. :-)
@@ScottCLemon Aaah okay I see. Valid use case and point! I've added notes for my v0.2 prototype (aiming to order a batch next week) about this.
It'll be pushing it with space, but would 2x M2 holes suit? I'm scrapping the SPI headers, and changing to a smaller MCU because of that, so should have some room!
@@PlatimaTinkers that sounds great! Make sure to leave clearance for the screw heads. :-)
@@ScottCLemon Yeah I'd usually leave it with 1mm of track around it so the head isn't biting straight into the FR4, and it can be treated as a grounding for any ESD discharge
great idea, will follow-up on next versions
Cheers mate should only be a few weeks!
I can see an issue here with ground loops when power is cut.... If anything on the SBC side of this is measuring/sensing and has a ground or grounded shielded cable, and the power ground is cut, the power will be maintained by the grounds on sensors etc. also while the voltage drop across the mosfet is going to be very very small, this could read to odd readings on sensors once again due to the ground loop issue.
two good reasons to /never/ design something that does anything with the grounds. You could switch the + rail with either a P channel or an opto coupled N channel fet just as easily.
Yeah I was thinking about that a fair bit during the design, but a) cutting GND removes the need for the added complexity of a P-channel MOSFET requiring a driver, and b) this is for low powered setups - any sensors are going to be powered by the SBC, as if there was external power, DeepSleepr wouldn't be required.
So there is absolutely the potential for a ground loop, however, if there was, then this product wouldn't be needed. Not strictly, but close to, mutually exclusive scenarios?
Thank you for your input either way my friend!
@@PlatimaTinkers yes agreed, just a small possibility, it's been a curse of electronics and electrics since the dawn of time. I have a dashcam I wanted to keep the internal battery fully charged so tried one of those tp405x modules only to find the module charged by switched positive and the dashcam cut off the charge by switched negative so while the battery wasn't overcharged the charging never actual ceased. Amplifier's are the worst for loops with PSU noise and odd windy sounds.
@@lezbriddon Yeah you know what, this stuck with me overnight, and in the middle of the night I realised that if there is any mounting/enclosure, if it's metal, that might be acting as a common ground.
V0.2 will have a P-channel MOSFET 😅 Thank you for your support/encouragement!
Can we get a demo video?
Looks like I'll be squirting one out in about two weeks, as the test code I roughed up yesterday worked! Just optimising it for lower power and accuracy now, then will 3D print some cases!
the sleeper must awaken
It will when it wants to :)
maybe also a version without usb connectors, just through hole solder pads, so the device powered by it would control the shutdown signal ?
Yeah hence I ordered them assembled without the USB type-C ports so I could experiment with it like that. Might leave some THT pin holes under the USB pads. Added to my notes 👍
Nice 👍
Cheers mate!
Sorry if I missed it - what's the quiescent draw of the DeepSleepr?
Had not actually been measured/rated yet, but the Attiny414 draws 0.69µA in standby mode w/ external xtal. The MOSFET would likely be the biggest power sink with 4µA for zero gate voltage.
Still, anything under 150µA should give ... what, 2 years on a 2600mAh LiPo? Minus time the SBC is awake for operations of course.
I want 10 …
Hey mate thanks for the supportive words! I actually had 20 made for this prototype batch, and as of last night, they work! The code is rudimentary, and could be lower power, but does the job.
2 I am using for development, 2 I keep for historic purposes, 1 I destroyed in thermal testing, but that leaves 15 I could sell if you're actually interested in buying some?