Page 1 of 1

rapidfire doesn, t work properly

Posted: Mon 28 Jul 2014 6:40 pm
by Jorginho
Here is the problem:
I use a davis vp2 station in the countryside. It sends the data to a notebook with xp on it sp3. In it is huawei 173 gsm modem or dongle.
It sends the data to the net.

I have set standby out also usb standby com3 is not allowed.
now the huawei monitors the output. The only thing running on it is cumulus vrrsion of dec 2012.

what i see in the data is three 5kb/s upload. These correspond with updates on wunderground. Then nothing for 10, 20 s up to 1 or 2 minutes. Than another spike and wunderground is updated again.

Now there is also some download data so is it cumulus not sendingits data to wundergr. Or is wunderground not responding bavk. Idid not find an answer on the net and wondrrnwhat is going on. Anyone? Thx in advance!

Re: rapidfire doesn, t work properly

Posted: Mon 28 Jul 2014 7:22 pm
by steve
As far as I am aware, a large number of Cumulus users are successfully uploading to WU using the rapid fire setting.

If you turn on the debug log, the uploads to WU and any error messages from WU are logged to the debug.log file.

I'm sorry but I cannot support such an old version of Cumulus as the one you are using. The first thing to try if you have a problem with an old version of Cumulus is to install the current version. But I suspect the problem is something to do with your setup, e.g. the GSM modem.

Re: rapidfire doesn, t work properly

Posted: Mon 28 Jul 2014 8:55 pm
by Jorginho
Hi, thx for the fast response.

I read you no longer support old version, but thought may be you (or someone else) knew the answer out of the top of your hat.

I too think it has got something to do with the GSM module. I have to sya the first month or so (in 2013, it is that long that I have problems) it ran with zero problems with the exact same modem. In 1,5 year so many things have happened and gone wrong it is nex tot incredible. It really is. LAst three weks have been crazy to, cost me lots of time but also made me more determined. How difficult can it be to get a PC onto internet? Apparantly very difficult.

Anyway: with a Davis VP2, arguably the most convenient system out there but not the most accurate, it is not that bad. I have calibrated datalogger in there too and that is the data I will use. Davis gives way to low rainrates and the tempsensor is really soso. Today 40 mm measured rainfall the traditional way became 15,4 mm from Davis. And its indicator in the pluvio says it is 100% horizontal....Temp: within 0.7 C of the real temp and generally a bit better it is really good.

So for this, I have found the program AmpWinOFF. I have been looking for a way to schedule reboots. Now of the minor problems with the system is that the EEEPC is second hand and Windows scheduler in the ends wants a password. Which I don't have....Anyways: AMpWinOff is great. If people have programs that can crash or connections that can get lost it is very good. Youc an chose to reboot every day. or every X hours etc. But you can also tell the PC to reboot whenever a certain speed of data traffic (either combined, up or download) has not been reached for a time. I saw the packages by Cumulsu alone are 4-5 Kb/s upload max. So you can say: if for 10 minutes 3 Kb/s is not reached just reboot. You can also combine this with a x-hourly or daily scheduled reboot.

In short: I now let the PC restart every 1,5 hours and if this works really well will go up to minimise the amount of startups. The Envoy logger will have the data anyways, so a connection that is lost (which is my main problem still btw) is not a real problem. Cumulus and Tmobile internetmanager software are set in WIndows so they will open automatically after reboot. I'm fine.

I am reluctant to install the latest veriosn though. If I do so, will it automatically retain the old settings and will it also reboot in windows? Thx again and thanks for a great software package once again!

Re: rapidfire doesn, t work properly

Posted: Tue 29 Jul 2014 6:54 am
by steve
Jorginho wrote:I am reluctant to install the latest veriosn though. If I do so, will it automatically retain the old settings and will it also reboot in windows?
Yes, of course. But make sure you read all of the release notes (in the announcement threads) for the versions since the one you are running.

But if you don't want to install the latest version and just want someone to guess at what the problem might be, that's fine. The debug log is probably going to be helpful, though.

Re: rapidfire doesn, t work properly

Posted: Tue 29 Jul 2014 5:18 pm
by duke
not the most accurate, it is not that bad. I have calibrated datalogger in there too and that is the data I will use. Davis gives way to low rainrates and the tempsensor is really soso. Today 40 mm measured rainfall the traditional way became 15,4 mm from Davis. And its indicator in the pluvio says it is 100% horizontal....Temp: within 0.7 C of the real temp and generally a bit better it is really good.
If you are certain that your other equipment is 'more accurate' than the VP2, both of these instruments can be calibrated to match. Have you compared other local stations for similar measurements to your VP2? How do they differ?

The temp can be calibrated from the console and the rain gauge manually, various 'how to's' on the net for this. Are you sure you have the correct adapter in the rain gauge?

Re: rapidfire doesn, t work properly

Posted: Tue 29 Jul 2014 9:19 pm
by Jorginho
Hi Duke,

Thx for your response and input. Dabis uses a sensirion sensor and compared to a calbirated fast reponse datalogger it has the difference sit has. I use a Vaisala radiation screen (KNMI) for that sensor. The Davis is next to it in its own radiation shield. It is already "calibrated: agfaisnt it. In the past I have had Weather Wizard 3 and a WM2 in a Stevenson screen together with two calibrated min/max mercuty thermometers and much the same differencces were measured. Others have noted the same thing. Moreover Davis themselve guarantee a 0,5 K deviation from the real temp between -20 and +20 (range I am not entirely sure). Their radiation shield is pretty good, but it is not black inside which on very radiant days will add a little to the error. You cannot expect pro performance from a setup tjhat is only 600 euro and that is why I am okey.

The raingaugs are known to be highly inaacurate. BTW: I got mine cleaned just the day before and checked the airbubble in the shield that indicates the wether it is level or not. It is dead in the middle.

I found the Peet Ultimeter two things:
- The tempsensor (but I did not test the temohygor sensor!) is incredible. it is dead on. Just out of the box dead on. It is dead on the minima and dead on in the maxima. I mean within 0,2 K from a calibrated logger in the same shield. The fast response means this IS easy to calibrate. With davis last year I had for instance -13,3 with the datalogger and -12,8 on the Davis. But when it comes to the max. it registers lower. That is difficult to correct for. Mind you out of the box the deviations was more on sided, so the minima especially were way too high.
- Peet registers every 0,5 or even 0,25 s. For wind this is according to WMO standards. It also starts to spin at 0,1 m/s.
- Peets small raingauge has a system that is non tipping bucket. It was dead one. It standardizes the siz eof dropplets coming through. Everytime one such drop means a 0,25 mm or 0,1 mm (yes it registers that!) precipitation. You need larger funnel to register the latter. Which costed me 12 euro or so. The advantage here is that in high downpoors the tipping bucket simply splashes water away. Also if the pluviometer is a bit of balance it does not influence it as much as the tipping bucket which now will not tip at equal amounts. But the latter simply means you need to be accurate.

I am in a commission that writes the guidelines on equipment for the Dutch Society for Weather and Climate and my task is to set the standard for accuracy of the insturments used and the siting and how a station should be sited. It means I probably know a bit more than the average weatherbuff on equipment and how to measure according to some standards/ My reports on the Davis are not unique, they are pretty common.

The Davis to my mind is however ideal for home use and for the money it is incredible value. Also it is so reliable even wireless that it makes Peets claim that wires are so much more reliable a bit laughable. Their station especially lacks usable software like Cumulus. They appear not to be very helpfull towards the maker of Cumulus. You can use Weather Display which I do not like nearly as much as Cumulus and WUHU. Wuhu was okeyish...

And that is a bit sad I think. I plan to use a minipc (like E3350 minisystems) with this Peet, connect it via WifI to a MIFI router and let it run on solar power. A fun project to do. But it will be timeconsuming I think. Well: I fear...;-)

Re: rapidfire doesn, t work properly

Posted: Tue 29 Jul 2014 9:28 pm
by steve
Jorginho wrote:They appear not to be very helpfull towards the maker of Cumulus.
Peet do at least publish their protocol spec. So it's possible that one day, Cumulus might work with one of their stations.