Motion Sensor numitron clock

More
8 months 2 days ago #11119 by Ty_Eeberfest
Yeah, this one isn't your fault.

Have you measured Vin at the sensor? You said you're getting "Motion detected 0 sec ago" which makes me think voltage at the sensor is low which (at least in the case of my microwave sensor) causes the sensor to assert "motion detected" constantly.

I can make it work, but it's not a practical solution. One setup that works is feeding the sensor from its own separate power supply. Another setup that works is to power the whole works from a lab supply that has a sense input: feed the power into the clock as usual but connect the sense input to Vin at the sensor. Thus the power supply regulates to maintain a constant voltage at the sensor rather than at its own output terminals.

Both of the above take care of the low sensor Vin issue but neither one fully solves the "no boot" issue.

Also, just as an FYI, sticking big ass electrolytic capacitors (went as high as 3300uF before giving up) across Vin & ground at the sensor seems like it should help but it did not.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
8 months 2 days ago - 8 months 2 days ago #11120 by Galesloot
Hello to all.
My name is Wim and I am new here!
The same happened with my very recent built 4 digit wemos clock. (first one)
Many false triggers. After I fed the RCWL-0516 from a battery it worked fine.

My solution was to feed the sensor with an MT3608 stepup converter. In 7.5 volts ( same as to the clock ) and out of 12 volts. ( to the sensor )
I applied a 10k pulldown resistor to the output from the sensor to the clock.
It has been working smoothly for a few days now.
The dimmed hours can be seen nicely on the web page of the clock.
BTW. I am very happy with the clock and the sensor!

Hope it will help, Wim.

Attachments:
Last edit: 8 months 2 days ago by Galesloot. Reason: add schematic
The following user(s) said Thank You: Ty_Eeberfest, Ian, Hartmut E.

Please Log in or Create an account to join the conversation.

More
8 months 2 days ago - 8 months 2 days ago #11121 by Ian
Replied by Ian on topic Motion Sensor numitron clock
I had this at the beginning and somehow I resolved it. I don't remember how. I have a feeling that I might have used the regulated 3V3 line to drive it in the end, but I'll have a look at one of my clocks at home: It runs absolutely reliably.

I can backup what Ty was saying that the Microwave sensor is really, really sensitive to supply voltage fluctuations. It appears that microwave magic wants special conditions.

Ty: Are you using the "bang" date transition? I can imagine that this causes a bit of a glitch in the supply voltage.

Another possibility might be to use a high quality power supply. That might nail the fluctuation a bit.

I'll have a look at putting a 4V21 Zener on the board for the stabilization if all else fails.
Last edit: 8 months 2 days ago by Ian.

Please Log in or Create an account to join the conversation.

More
8 months 2 days ago - 8 months 2 days ago #11122 by Ty_Eeberfest
I've tried both the transitions. Both cause a spike up in voltage during the transition but "bang" causes a sharper spike than "wipe" or, IOW, dV is equal for both transitions but dV/dt is greater for "bang".

Just now I tried powering the sensor with 3.3V from the Wemos's 3.3 pin. Tried feeding it into the sensor's Vin pin, then into its 3.3 pin, then into both. No good. The sensor powers up, output initially pulls low for a couple seconds, then swings high and stays high forever as if always detecting motion. This behavior is consistent with what I've seen in other low Vin situations, Vin seems to need to be >= 4.75ish volts and very steady if the thing is to work properly.

I've been thru various power supplies today. The only setup with single power supply that worked was using the lab supply with "sense" inputs. I feed the power into the +5 and Gnd points on the clock board, Taking "sense" from at the sensor board works well, taking "sense" from the +5 and Gnd points on the clock does not work well. The implication here seems to be that well regulated power into the clock != well regulated power to the sensor.

Look into it later when the dust is clearing off the crater.
Last edit: 8 months 2 days ago by Ty_Eeberfest.

Please Log in or Create an account to join the conversation.

More
8 months 2 days ago #11123 by judge
Replied by judge on topic Motion Sensor numitron clock
The voltage fluctuation thing is interesting. I have two radar sensors (same module). One is working fine, the other definitely false triggers. The clock for the working one is powered from a 3A, 5.1V power supply intended for use with Rasp Pi's. The one that keeps triggering is powered from some USB adapter I had lying around. Guess I'll check the bad one. When the todo list gets a little smaller.

Please Log in or Create an account to join the conversation.

More
8 months 2 hours ago - 8 months 2 hours ago #11124 by Torsten Lang
Hi all,
I would be highly interested to see some photos (top and bottom side) of these sensors.

I use them myself for a short time and connect them to the USB power supply of my clocks. I had no trouble so far even when connecting the clock to a USB port of my PC which is not exactly a prefect example of a stable and noise free supply...

Regards,
Torsten
Last edit: 8 months 2 hours ago by Torsten Lang.

Please Log in or Create an account to join the conversation.

Time to create page: 0.189 seconds

Search

Tube Suppliers

Go to top
JSN Boot template designed by JoomlaShine.com