This is a placeholder topic for “Makerverse Nano Power Timer” comments.
Periodically cycles power to your project, greatly extending battery life.
Read more
1 Like
This is a great device if you want something to wake up, do something, then go back to sleep.
I wanted something to run until an event occurs, then shutdown for an hour or so. You cannot do that with this device because of how Texas Instruments designed the timing chip.
The timing cycle runs all the time and at the end is a 50ms gap where power is removed, causing the micro to reset, not good if the micro is doing something and not ready to reset.
Maybe I should have researched it more, anyway, annoying.
Jim
3 Likes
This is a possible solution to my previous post.
It uses a mosfet to switch ground to the circuit.
A 22uF capacitor holds the mosfet on while the Nano Timer switches off for 50ms.
The mosfet allows for a much higher current drain.
Its not ideal because it switches ground, but it works for my purpose.
Cheers
Jim
EDIT: This does NOT work correctly, see further posts.
2 Likes
Well I thought I was getting somewhere, but alas not.
To this point I had not connected the DONE signal to the Pico. Just tapping the pin to VDD to trigger a power off cycle.
With the Pico GPIO connected the Nano timer continually cycles.
Pressing reset on the Nano fixes it, but only until the Pico switches it into power off again.
When it comes out of power off it continually cycles again.
I think the problem is due to the capacitor on the gate of the mosfet charging and delaying the start up of the Pico. It’s a catch 22, cap needs to be large enough to hold the gate on, but small enough to charge quick enough. Maybe a different value resistor …
At this point I have had enough and have abandoned using the Nano Timer.
While the Makerverse Nano Timer is a good design by Core Electronics, it is let down by the way Texas Instruments designed the TPL5110 chip.
Regards
Jim
EDIT: btw the 50ms power off at the end of the timing cycle, every cycle regardless of the DONE pin state, has been noted in other forums. Also using a high active DONE can be problematic, TI should have designed low active or allowed the option to use either.
3 Likes
Did some further testing, setup similar to the Tutorial using various timing intervals.
As long as the on time of the micro is less than the power off time it works nicely.
The 50ms drop in power at the end of the cycle is still the problem.
btw if you have a micro run time of 10 seconds and a cycle time of 15 seconds. It will be off for only 5 seconds.
The continual cycling I was experiencing is due to my solution to try and keep power on over the 50ms gap. So not really a solution, needs some work.
I would like a device that keeps power on until I tell it to go off and consumes almost no power like the Nano Timer. I think the real problem is the power cycle runs all the time. I have briefly looked at the one shot mode, but without testing it I don’t know if it would provide a solution. I don’t really understand it from the datasheet. Unfortunately the Makerverse board has the one shot pin set to VDD. I would need to modify the board. It would be nice if this was a jumper on the board.
Had a close look at the board, not easy to modify and easy to damage other stuff. Oh well …
Regards
Jim
1 Like
Hi James
I don’t know anything about the details of your requirement but if you want the Mosfet to stay on until it is told to switch off just remove the 10kΩ resistor from gate to ground. The 22µF cap is not needed and will only slow down the switch on time as you have already found out. The internal gate capacitor of the Mosfet will keep it on until you turn it off by taking the gate to ground. This will actually stay on even when power is removed. Unless that gate cap is discharged it will still be “ON” when next powered up.
Cheers Bob
PS: An interesting experiment for you. Grab a bare Mosfet, connect a LED and current limit resistor between a voltage (5V and logic level mosfet) and drain, negative supply to source.
Touch gate to positive V momentarily to turn LED on.
Remove gate connection, the LED should stay on.
Remove power, LED goes out.
Restore power, do nothing to gate. The LED should come back ON. ie; the mosfet is still ON.
While doing this make sure there is no discharge resistor between gate and source.
2 Likes
Hi Jim,
Thanks for sharing your knowledge and experiences so that others can see if the Nano Power Timer will work for them. Perhaps there is a need for a more advanced version if we can find another timer chip that doesn’t have the same limitations.
2 Likes
Thanks for your reply Bob.
The problem is the Nano Timer runs all the time at whatever time interval is set by the resistors. At the end of each interval the drive pin is taken low for 50ms (TI must have a good reason for doing this).
This resets the Pico when I don’t want it reset. The 22uF holds the gate of the mosfet long enough to bridge this gap. That part works nicely. Without the cap the Pico resets, too low a value and it resets too.
The Nano Timer has a DONE pin which switches it off for the duration of the timing interval. This is driven by a micro pin, HIGH to activate.
I thought the cap charging might be the problem. After some thought I now think it is because I am switching ground. I am working on a solution to switch VDD rather than ground.
cheers
Jim
1 Like
Hi James
I have not gone into that timer to know how it works or what it does so can’t comment further really.
That cute little trait of a Mosfet can be handy to remember sometimes. Can be useful.
Cheers Bob
1 Like
This is NOW my solution, a bit messy having to add another Mosfet but switching VDD works like a charm. Pico runs for as long as it wants, sets DONE and turns off for the time set in the Nano Timer. Starts again just like it should. Happy with this.
The output from the Nano Timer is 4.85V or open circuit (P-channel mosfet on board).
The 100 ohms limits the inrush current to the cap to about 50mA, RC time constant 2.2ms.
The 22uF cap discharges through the 3k3 resistor holding the gate active for about 75ms covering the 50ms gap from the Nano Timer.
The 3k3 to VDD allows the FQP27P06 to switch off when the 2N7000 switches off by pulling the gate high.
Cheers
Jim
3 Likes
Hi James
Looks like you have it sorted with a high side switch
I was going to suggest you have a read through this thread which I had some input to.
Pi power switch using car ignition logic
for some possible ideas for a latching switch . Last post was 2 days ago so it is not far down the list.
I haven’t figured out how to link another post yet
Cheers Bob
2 Likes
Read briefly through the ignition post.
All too often someone puts something out there (internet) without really testing it.
I gather the original circuit was slightly flawed. Maybe I need to re read it, I might have that wrong.
Your last post is very interesting; your points I have seen already and good advice to be aware of the intentions of persons who post for other than hobbyist reasons.
Cheers
Jim.
1 Like
Hi James
The original circuit may be switching a 12V device or did not use the buck converter so probably worked fine. I think the problem here could be the 220µF input cap in the inverter back feeding voltage preventing the switch from unlatching properly. That is my theory anyway and I don’t have the time to build it and prove right or wrong.
So it could be the situation where a circuit is being used for a task it was not originally designed for. Seems to happen quite often.
Cheers Bob
2 Likes
I did the experiment as you suggested and the LED stays on for a long time, even through multiple connects and disconnects of power. It would make an interesting classroom experiment with discussion as to why it is happening. It is a practical example of how the mosfet is constructed.
My rule of thumb has been to include a high value resistor between gate and source of a N-Channel to discharge the gate capacitance to ensure the mosfet turns off. In the case of the Makerverse Nano Timer the output is 4.85V or open circuit. The mosfet would not turn off for a long time hence I used the 10k to ground.
In the classroom this could be shown as the reason you design a circuit a certain way, to make sure the device does what you want it to do, each and every time.
Thanks for your advice.
Cheers
Jim
EDIT: The LED has been on for about 5 hours now, power was removed for about 30 minutes and it still came on again. Something I was not expecting, I thought the gate capacitance charge would eventually dissipate. Very interesting.
Thought I’d put a couple of pics of what I am doing with the Makerverse Nano Timer.
The cart will randomly wander the house trying not to bump into anything.
If it finds a dark place it will power off for a time, the Nano Timer bringing it out of that state.
This means at night it will not move.
When it is active it will randomly play a tune or noise at random time intervals.
If something moves in front of it, it will stop and play a tune or something.
There is a PIR sensor hidden behind the IR sensor.
With just the Pico running it consumes about 40mA, in power off mode 0.5mA.
This is the quiescent current of the power supply.
The project was also an attempt to develop a steering system. Originally the wheel was attached directly to the servo but this placed to much stress on the servo movement and it did not align properly. You can see the marks on the board where it originally turned. Using the lever proved better.
Cheers
Jim
2 Likes
Hi James
The input resistance of most Mosfets disregarding the capacitor is in the region of several TΩ (lots) and would take a LOOOOOOOONG time to discharge the cap.
Cheers Bob
2 Likes
Hi Bob,
I think you’ll get a lot of use out of linking previous posts so I’ve taken some screenshots that show the process.
Start by locating the post you want to link to (probably in another browser tab/window) then click the chain icon below the post. The dialog box that appears will have a link to that specific post in the thread. Clicking the two-pages icon will copy that link to your clipboard so you can post it in your next forum reply.
1 Like
Hi Bob,
The mosfet is still activated after almost 2 days. Something I will have to keep in mind for future projects. Was not expecting it to stay on for so long. Thanks for your suggestion, it has enhanced my knowledge.
Cheers
Jim
Thanks Trent. Will copy that for future reference.
Cheers Bob
Hi James
I think that gate capacitor is pretty close to the perfect cap. That is it will never discharge due to its own leakage.
Cheers Bob
By the way, a capacitor is the only thing that will “store” “electricity” or electrical energy.
Batteries are a chemical action, the chemistry changes, they don’t “store” electrical energy in the true sense of the word. It is only convenient to refer the charging chemistry as “storing” energy.
An inductor might be thought to “store” energy but as soon as the magnetic field collapses the energy is returned to the outside world. Could be likened to having a lend of it for a short time then giving it back.
Cheers Bob
1 Like
