Hi @ddabrowa,

I get issues with WiFi too.
Even if the device is turned on since days (its hanging on a charger) and the settings menu shows that it is sending packages, the website uradmonitor.com shows that my Model D is offline. I have to reconnect to WiFi through the debug menu.
Unfortunately, the measured data is lost?

The data is permanently stored in the CSV files. As discussed with Scott, I’ll look into this, but it is a complex code, so I will need to get a time frame for it. Hopefully soon.

please keep the reports coming. It’s the only way to improve what we have so far.

hi @Koplin, please see my added replies to Scott, https://www.uradmonitor.com/topic/model-d-feedback/#post-4651 .

I have noticed that the delay in input is annoying and so far I don’t know the pattern however I did notice if I turn off wifi, gps and activate mute, that the input lag reduces a bit. I do have issue with my USB port cables don’t firmly seat due to the PCB sliding in the case. Likewise for the power button is a tad deep but I actually like this because it prevent accidental activation. Finally the SD card is actually unreachable without tools. I just noticed Radu’s tips on adding something to keep the PCB from floating in the case and will give that a go, might fix the USB cable issue as well.

Stopping some of the modules might help as this frees some of the CPU and there is more time to react to taps. I detailed this in my reply to Scott, but one thing I need to highlight again is that you should be careful with the LCD, not to stress it. The LCD itself is sensitive and touches are recorded, but the CPU can’t process them in real time. So best thing to do when operating the model D, at least in the current firmware, is to click and hold until the action gets registered.
A newer firmware might implement a more clever mechanism to get along this limitation.

Critically my last 3 trips have not shown up online. and I made sure to sync my data and it says “deleting file” etc… so not sure where the data is but I took a trip around my town, cruise ships running their equipment (PM 2.5 samples) around my hospital (Radiation etc) then I took it to my office and hit it with the X-Ray head. I love that the alarm was immediate, activating before the x-ray system finished its cycle (sub 1 second). Regardless of display the device alerted to the high dose (tested several times). These trips/data do not appear online for some reason. Had GPS on and locked until I went inside and my phone tethered the entire time with typical cell tower handoff and such so I am sure there was coverage dropouts at time.

All data is logged in the CSV files, separately, but it’s strange that the sync failed. Please try this repeatedly and let me know if you can see a pattern. Would love to see your x-ray tests myself! Send me the CSVs if you like.

While driving around occasionally behind a bus or big truck I would get what I suspect is an air quality alert (likely PM 2.5) however I couldn’t tell and I was hoping to see the samples online but again some data didn’t appear to make it to the site for some reason. Having a way to “tag” or mark an interesting moment would be neat.

Excellent suggestion! This is a great idea, we will have it for sure.

Overall I like the device and hope that with this sort of feedback that should there be a Rev 2 or firmware changes in the future that we can fix some of these issues. Sounds like firmware will improve as time goes on. I would really like the option of keeping my screen on. I used my portable battery to power the thing and would love to use it with the screen on more. Maybe have the geiger click a separate item you can mute or make louder etc. Our beaches got a lot of debris from the tsunami that caused Fukushima so having a portable device is really nice, not sure what I might find. The Model D didn’t like my BBQ when I lit it 🙂 gave me a nice warning from several feet away. One more addition would be localization options, timezone, MPH vs KPH, pressure, temp C vs F etc. Being in the states I am used to my backward ways.

This kind of feedback is what brought us here, initially the model A got some attention and the community slowly pushed things to more complicated features. So this is gold for me, thank you for taking so much of your time to enter this input.
Thanks for these additional suggestions, they are on the list for the future firmware changes.

Overall I love the device and its rough edges. I am very grateful you kept working at it rather than rushing something out. Even the way it is now I don’t see it as nonfunctional in any way just sub optimal in a few areas and personally I am willing to accept this especially after reading about the challenges you had to overcome to bring this to market and the great feedback/support you provide such as the tape/toothpicks.

Thank you, having your support and encouragements like this certainly made things better.

Keep up the great work

A few important things you should all now:
1.Many reported the PCB moves inside the enclosure. This should have not happened, but the manufacturer left out some self adhesive tape that should have been installed on the inside, at the end with the LND712 tube (that requires extra care). Picture with the foam attached below. Please install such stoppers yourself, the board is 1mm shorter than the enclosure. Alternatively, you can force the tips of wood toothpicks between the PCB and the aluminium rails, cutting the extra.

2.The clicking sounds are for the Geiger Counter. They indicate a radiation pulse has been recorded. You can mute them in Settings

3. The LCD is “unstable” or “works intermittently”. The screen can either work or not work at all, but for the clicks on screen, please keep what you want to click, pressed a little longer. No need to force the screen, it is very sensitive and fragile.
The problem is the CPU is a bit low with all the many features added to the design. So sometimes , it is unable to process clicks as fast as we’d like. Once you connect to Wifi it will run smoother. Failing to connect burns CPU cycles.
Also the LCD uses a resistive touchscreen, not the capacitive one you’re used with from your mobile phones. The resistive ones require precise focused taps (using a stylus or the nail).

4. The reset button. It can be located using the picture attached. Use it if there are any issues with the unit.

5. Screen turning off. There is a 30seconds timeout to the LCD screen to save power. Tap the screen to turn it back on.

6. Battery normal voltage. The units have an internal Lithium Ion battery, and while the nominal voltage is 3.7V, the normal discharge path goes from 4.2V (more or less when fully charged), to about 3.2V. The uRADMonitor unit will turn off automatically when the voltage is below 3.4V, because there is a real time clock that needs the extra power.

7. VOC readings. The VOC numbers are not comparable from one sensor to another. There is no UM for VOC, and the manufacturers have different approaches on what they consider VOC. Normally the technology implies a sensor with an embedded burning filament (300*C) that consumes various gases from the air (both Oxidising and Reducing reactions), and as a result, the electric resistance changes.
On your D units, you see this electric resistance, in Ohms, fluctuating to air quality. Bosch did a great job with the BME680, as stability is a big issue on this approach and not many sensors are up for the job (tested many). A clean air is when the resistance is higher: no burned products contributting to conductivity across the sensor crystal. On the other hand, for polluted air we’ll see lower ohmic value. The quantities of detectable gases is listed on: http://www.uradmonitor.com/uradmonitor-model-d/
Now the good news is Bosch provided us a post processing BME680 software library that we are working to integrate on the server. Then, accessing your unit’s data from the server via the API (see http://www.uradmonitor.com/dashboard) , will give you the processed VOC readings as an air quality score index.
This library also does auto-calibrations to temperature and humidity data, that can be affected by the enclosure and the extra heating from the other components (battery, charger, wifi module, microcontroller).

(I’ll add more when needed)

Hi Scott, thanks for the input , I’ll get back to it in detail, also do the formatting, but tomorrow, as it was a very long day.
Try to format the sdcard in your PC. Also use the RESET button when / if needed. See my post above, there’s a picture with the button location.
Your flight log is very interesting, this is a first I think. There was a bug with the colour representation, and I just fixed it, see pictures attached.

LE: I promised an answer.
Hardware

1. The PCB in mine seems a little loose and rattles if softly shaken lengthways. I suspect it might need a thin piece of medium- or high-density foam on the inside of the top and/or bottom plates
2. The power switch moves the entire PCB lengthways, before operating (see previous point)

There should have been a self adhesive foam installed inside the enclosure on the lid facing the PM2.5 sensor. The manufacturer left this out. Please get a piece of foam, they are affordable and easy to source, and secure the PCB. PCB movement is definitely not
ok and can impact the LCD connector.

3. How should we orient the device if sitting at desk (for example) – it seems as though there are sensor or antenna holes on all surfaces, except the curved sides, and the USB cable also complicates things

I envisioned it placed with the USB connector down, in vertical position. This position is not for charging. You might want to see this: https://www.uradmonitor.com/topic/uradmonitor-model-d-desktop-stand/

4. The lag on the touchscreen can be a few seconds, at times – I’ve noticed as much as 8-second jumps, when on the Measures screen

This is important as not knowing what happens can lead to LCD damage. The new features added as extended goals impacted the processing performance and the microcontroller was updated to partially compensate for that. But not entirely. What happens is that sometimes, while you press the screen, the microcontroller is busy reading sensors, communicating with the server, decoding GPS or writing to the SDCARD. What you need to do is tap and hold, gently, as the LCD is very sensitive and doesn’t need to be abused. It’s not an issue of touchscreen contact, but lack of processing power. This might be addressed in software too, perhaps in a future firmware I’ll be able to do the touchscreen driver code better and react to the touches with higher priority then the other subcomponents. Bottom line, please don’t force the LCD, instead click and wait a little if the touch is not recognised the first time.

5. What is the black antenna/sensor on the front, above the screen? It’s not described in the specifications on the website.

It’s a plastic cover for the GPS ceramic antenna

Battery

6. It won’t power up if connected to charger but it will charge after being powered up; so, if the battery is empty, you need to press the On button and then immediately connect power, otherwise it will power-down again

Please press and hold the on button. If there are any issues starting the unit , you will need to hit the reset button before attempting to press the ON button again.

7.It seems to power down if connected to a PC but then appears as USB-serial port; this is fine if you have AC USB adapters everywhere but means that you can’t use a PC/laptop to charge it and have it uploading at the same time

This is a glitch that will be fixed.

8. Are you planning to release firmware updates for the Model D? Is that what the USB-serial functionality is for?

Absolutely. All the other models received tens of firmware updates. Your feedback is what generally leads to these updates. I try to make it as good as possible.

9. The battery indicator didn’t reach full, from empty, after being connected to a 5V/1A USB charger for 8.5 hours uninterrupted, despite only drawing 0.5A; it does with a 5V/2A charger. This may have been location specific – at my office it can’t receive a GPS signal and the wireless is busy, so that may use more power

The charging rate is adjusted in hardware. It can be changed using a resistor.
The GPS uses a custom antenna. It needs a strong signal to get a lock (and also to keep it). Wireless can do a lot of interference on the GPS receiver.

10. Battery life seems to be about two-four hours

Battery life is dependant on WIFI and GPS use, but also the number of times the LCD is turned on. Disable WIFI when you don’t need it.

Software
General

I implemented this feature already, I’ll double check.
Wireless

13. Wireless config – if you were viewing/setting the Wi-Fi key and exit back to the Settings menu, when entering Wi-Fi settings again, the SSID button is selected but it displays the key

I’ll need to check this.

14. Wireless config – the keyboard shows each key as being depressed for each character used in the SSID or key
It only seems to connect to the configured SSID if its broadcast – is that correct?

The key refresh was left like that as refreshing each key would have consumed additional resources, which unfortunately are not available.

15.A single SSID is a moderate limitation, particularly given the method of configuration. A method of saving a couple of profiles would be handy, even if the profile has to be manually selected

Excellent suggestion, will keep this in mind for the next firmware update.
Offline

a.Could you add an option to automatically upload offline data when reconnected?

Yes.

b.The offline samples counter resets after power-cycle – I initially, wrongly, assumed that it was losing offline data after reboot (battery was drained)

The offline counter shows indications per session. Data is not lost.

c.Whilst uploading offline data, it doesn’t seem to be also uploading live data or adding it to the offline file, to be uploaded after the sync

That is correct. If it becomes possible, I will add this feature. But the idea is you will have a way to measure important data, eg. during a trip, then when you get home you sync all data. The model D is not intended as a monitor, instead a portable unit with multiple functions.

d.Whilst uploading, the battery indicator is hidden but the screen stays on – so it drains the battery quickly but you have no idea how much battery is left

Good suggestion, thank you.

e.If the battery runs out whilst uploading offline data, it starts again when another sync is initiated after powering back up

f.Estimated time for upload of 912 updates is 912 seconds however, for me, each update took ~5 seconds to upload, so it took ~95 minutes. This number could adapt over time, with a value saved to NVRAM or SD (wherever you save Wi-Fi settings), or at least adapt during the current sync session. Using my phone as a hotspot, it was taking ~7-8 seconds per sample.

Excellent, we definitely need these added to the firmware.

Measures

a.Barometric units – can this be changed to hPa or mbar (which are equivalent)? 101000 Pa seems silly for pressure, when weather services use 1010.0 for the same pressure

Yes

b.Switching the labels and values between sides on the Air Q screen would enable you to change the units from Pa to hPa without losing precision, although I doubt anyone would mind losing hundredths of hPa

Thank you, good suggestion.

“Alarm” on the notification bar doesn’t tell you which parameter is triggering the alarm

Indeed, this will be updated too.

All in one, this is very constructive feedback, and I can tell already that most of this will become implemented in the next firmware upgrade. Thanks for the contribution, Scott.

Chuck probably wouldn’t have posted this, he sent me a nice email instead checking what to do.
After the long wait with the model D production, I was happy to ship the first units, including his the 64000003, but the outcome is terrible, see the picture attached below.

The unit was fine prior to shipping: https://www.uradmonitor.com/?open=64000003&sensor=cpm . I’m posting this to show how things don’t always go as planned. Of course, 64000003 will be repaired asap.