Device sends %CO2% instead of the actual value

[admin]

hmmm it consumes at maximum 300ma in very short bursts
and maybe if the other PS cannot deliver so quickly the device restarts..
will have to play with some power supplies

[René Arts]

I have encoutered these issues as well, it definitely needs a stable supply. On my laptop it did not have any problems. On one USB charger (quite cheap one) it doen't even work. On another one it runs mostly ok, but it's one which also powers a RasPi so might not be able to deliver all the ESP+sensor needs.

[Ed Davies]

I've been trying it on various supplies too and getting mixed results.

Off my Toshiba laptop: works fine - no problems noted. I think the same with my Asus netbook but I'll re-test that at some point.

Off a quite old Staples four-port powered hub on its own mains supply: sometimes works with intermittent %CO2% as per my previous post on page 1 of this thread.

Off a 5000 mAh USB “power bank” bought in Lidl yesterday: doesn't work at all, I assume because the power bank shuts down if you draw less than 80 mA.

Off the Lidl power bank via the Staples hub: stops after a while. E.g., tried to record CO₂ in the bedroom last night and it just stopped at 00:37 after running for about an hour or so. Power bank still had plenty of power left and was still on with a red LED on the hub lit. Occasional %CO2%, etc.

Off a USB wall-wart originally for a Tesco Hudl Android tablet: seems to work fine with just one %CO2% reading in a few hours testing though also with a rather odd drop in CO₂ over two readings:

2016-10-04T19:36:36 /?temp=23.18&hum=53.65&co2=1047.00
2016-10-04T19:37:38 /?temp=23.22&hum=53.64&co2=535.00
2016-10-04T19:38:40 /?temp=23.26&hum=53.60&co2=533.50
2016-10-04T19:39:43 /?temp=23.31&hum=53.56&co2=1065.50

Might have been a waft of fresh air as I went out to make dinner or something but rather strange.

I did update the firmware this morning so I'll retest with the Staples hub with the various power sources later. Still, it does seem to be sensitive to the supply in use in some way.

[admin]

Thanks for the comprehensive testing!
I just got back from vacation and now i will also give it a try with the different power supplies.

for the readings 1047 / 533, i will have a look again in the firwmare. In general there is a part that averages the last two readings.
And in case some error during reading appears there is a -1 internal value. It should not get into the averaging part, but maybe there is some way it happens

[admin]

I played a bit more.. it turned out that while the USB spec says - maximum Voltage is 5.25V, if the USB Power supply outputs more - then the Sensor (also rated at maximum 5.25V does not work any more).
So, when the device fails to communicate with the sensor - it does not populate the %CO2% value - i will fix this by reusing the last stored value

there is yet an unknown reason why the output sometimes fluctates +/- 100 ppm each 10-20 seconds.. i hope to find it soon

[admin]

After playing around with the CM1102 and CDM7160 sensors i found out that apparently such fluctations are normal, but the CM1102 solves this by having a large averaging buffer built in, while for the CDM7160 it has to be in the software. Up to now i was averaging the last 2 values, now i switched to 6 which made it quite stable
The sensor is read each 20 seconds, so this means that within 2 minutes any change would be reflected
You can try it out by updating to the latest firmware with "otah"

[admin]

(10-04-2016, 07:38 PM)Ed Davies Wrote: Off a 5000 mAh USB “power bank” bought in Lidl yesterday: doesn't work at all, I assume because the power bank shuts down if you draw less than 80 mA.

aah.. this was quite annoying indeed i also noticed it with a power bank. I the next version of the FW i changed in a way that the device wakes up each 30 seconds for a couple of seconds, this way the power bank does not die.

just that the issue with powerbanks is that some of them put out a bit more than 5.25 v and the sensor does not like it. one option is to patch the power supply to the sensor it with some Schottky Diode so that also those power supplies are accepted. I will check if this works indeed

[baz123]

It would be good if the power draw was configurable. An option for normal, low power draw, and power bank keeep alive draw with the sleep time configurable?

[baz123]

Just to add, that on startup mine sends out the %CO2% rather than a value. It would be better if it checked to see if there was a valid value to send.

[admin]

yeah.. the intial %CO2% is still on my task list... the problem is that there is one part in the FW that fills in those replacement values, and another that simply does replacements.. this is what makes it a bit harder

as for the configurable power draw - i can add this feature easily i guess