Author Topic: Swapping in a 6312 console for an Envoy  (Read 5015 times)

0 Members and 1 Guest are viewing this topic.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Swapping in a 6312 console for an Envoy
« on: January 30, 2017, 06:56:14 PM »
My current home configuration has one Vantage Pro2 ISS with three receivers.  Two receivers are consoles at different locations in the house where my wife and I wish to view live data, while one is a 6316 Weather Envoy connected to a PC from which I host (using Cumulus) a web site and provide input to Weather Underground.

From a few months after initial setup reception reliability on the Envoy has been a problem.  At one stage I diagnosed that a new habit of recharging my cell phone a few feet from the Envoy correlated with much worsened outage rates.  That I easily resolved by banishing that charger to the other end of the house.  But I have a continuing distracting rate roughly 25% of all days having a 5-minute dropout in which successful reception drops from the usual 100% to somewhere between 0 and 10% (i.e. many, many consecutive misses).  That has little consequence, as I only publish data at 5 minute sampling rate for anything save the detail Cumulus graphs.  But less frequently one of those outages stretches to two or more hours of total dropout.  That has consequence in published data and my own retained data, even to records.

I've noticed that my 6312 consoles, on checking their diagnostics for a day on which the 6316 had a multi-hour outage, still show the customary 99% reception since midnight.  So it is not a question of the ISS completely failing to transmit.

Since more than half the 5-minute dropouts start within a few minutes of 4:05 a.m. in the morning, it seems likely there is an intermittent interfering signal which not only stops the 6316 from getting correct data, but sometimes gets it into a state from which it has trouble recovering even after the interfering source is gone.

I recently experimented with temporarily locating a 6312 console very close to the Envoy.  I've already had a multi-hour Envoy outage with better than 99% console reception.

I'm guessing there is either a systematic design difference with makes the Envoy more sensitive to my problem than either of my consoles, or else my Envoy is to some degree a marginal sample.

I'm strongly considering buying a new console (from Scientific Sales, most likely) and putting it in the kitchen, while moving the kitchen console down to my study, extracting the weatherlink dongle from the Envoy, and trying to get my WeatherLink/Cumulus software running with the revised configuration.

Just to make things more fun, the current PC host is months overdue for retirement, so I also face moving the WeatherLink/Cumulus stuff to a new PC.

Steve has posted hots of detailed advice on the Cumulus move part of this.  That does not scare me.

I seek expertise and advice here on two points:
1. how likely do you think I am to see a sustained improvement in reception success by substituting a 6312 for my 6316 (which may be a defective sample of that breed?)
2. do you have any advice and a good order of actions in doing the reconfiguration on the existing PC host?

P.S. I plan to steer clear of the whole "green dot" compatibility issue by using the console that came in the same box with my ISS on the PC, hoping it is maximally likely to be compatible with my ISS and my concurrently purchase WeatherLink dongle).  I'd put the new console up in the kitchen, running fully standalone just for display.
P.S.S. If the reconfiguration while still on the same host to just swap 6316 for 6312 is simple and reversible, I intend to do a trial run for proof of improvement without purchasing the new console.

Offline Garth Bock

  • Table Rock Lake Weather
  • Forecaster
  • *****
  • Posts: 2745
Re: Swapping in a 6312 console for an Envoy
« Reply #1 on: January 30, 2017, 08:27:50 PM »
My Envoy is in the computer room (3 computers and monitors and wireless router) and doesn't have any drop outs like what you are describing. Have you called Davis about getting a replacement ? You can always downgrade the firmware to 1.9 to match your current logger. You won't be sending in the logger...just swapping the Envoy itself.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #2 on: January 30, 2017, 09:09:57 PM »
Have you called Davis about getting a replacement ?
Only recently did I realize that I could use contrast with the consoles to explore the question of bad Envoy performance.  By now the unit is about 5.5 years old and I've just assumed without checking that it is far out of warranty, if it is even true that it is defective.

I assume that my 4:05 in the morning highly repeatable trouble onset is not internal to the Envoy, but some form of external interference (more likely at the IF frequency than at the signal frequency), to which my Envoy seems more vulnerable than are my consoles.

I failed to mention that the study where my Envoy lives is fairly far from the ISS (almost opposite corners of a long house, so multiple walls and some distance).  Signal strength display by a console I moved around the study as a test device typically showed mid-30s signal strength , though there were some readings in the 20s mixed in.  I also failed to mention that when I have happened to notice a multi-hour outage before it ended, that twisting the antenna on the Envoy some often got things to start up within the next few minutes.  I've tried moving the Envoy about in the room a little, hoping to find a sweet spot, but have not noticed material effect from that.

Offline Garth Bock

  • Table Rock Lake Weather
  • Forecaster
  • *****
  • Posts: 2745
Re: Swapping in a 6312 console for an Envoy
« Reply #3 on: January 30, 2017, 10:38:55 PM »
Makes you wonder what kicks on at 4am that would interfere with your Envoy. One thing to try.....some people that have wifi problems with their router employ some tricks that might help. One trick is with an aluminum can (ie. Cantenna) and one is with a vegetable strainer.... http://www.instructables.com/id/Wifi-Receiver-Extender-WRE/
http://www.instructables.com/id/My-24p-WiFi-Booster/.

Just a thought.....

Offline johnd

  • Forecaster
  • *****
  • Posts: 4827
    • www.weatherstations.co.uk
Re: Swapping in a 6312 console for an Envoy
« Reply #4 on: January 31, 2017, 03:30:08 AM »
AFAIK the main receiver boards in Envoy and 6312 are literally identical. Any differences in reception will be down to trivial differences in the plastic case and the fact that the 6312 antenna will sit a few inches higher.

So I really doubt that there's any measurable difference in general between the two console types, but that's obviously not to say that there might not be slight differences between two individual examples due to eg manufacturing tolerances.

TBH it sounds like you've still got some issues of occasional background interference, maybe wireless, maybe through the AC mains (unless you're on battery-only operation with AC adapter disconnected). Or maybe that particular Envoy that you're using is now performing sub-par.
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #5 on: January 31, 2017, 08:56:17 AM »
AFAIK the main receiver boards in Envoy and 6312 are literally identical.
I suspected that might be true--it would be the simplest development approach.

Quote
TBH it sounds like you've still got some issues of occasional background interference
The part that seems really spooky to me is the way it occasionally reacts to what otherwise seems to be a temporary interference condition by going into a multi-hour state of failing to reacquire the signal when, I suspect, the actual interference is long gone.

I repeat the actual question I asked, to which no one has yet offered a reply: can anyone give configuration advice on an effective way to do a temporary swap in which I extract the Weatherlink dongle from my Envoy and plug it into a console and presumably need to do a bit of configuration adjustment to WeatherLink and Cumulus to accommodate the hardware change?

Offline johnd

  • Forecaster
  • *****
  • Posts: 4827
    • www.weatherstations.co.uk
Re: Swapping in a 6312 console for an Envoy
« Reply #6 on: January 31, 2017, 09:17:02 AM »
I repeat the actual question I asked, to which no one has yet offered a reply: can anyone give configuration advice on an effective way to do a temporary swap in which I extract the Weatherlink dongle from my Envoy and plug it into a console and presumably need to do a bit of configuration adjustment to WeatherLink and Cumulus to accommodate the hardware change?

Well, there is only one way: power off both Envoy and console and physically move the logger before repowering. (Assuming that both consoles are compatible with the logger, ie if one is a relatively recent unit then the logger will need to be green dot of course.)

I suspect that the logger will just restart OK and no reconfiguration needed at all. It's just possible that you might need to clear the logger via WL, but unlikely.

I think you're making this more complicated than it actually is.
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Swapping in a 6312 console for an Envoy
« Reply #7 on: January 31, 2017, 10:21:51 AM »
Have you considered changing the channel on all of the devices, to see if that changes the symptom?

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #8 on: January 31, 2017, 10:58:08 AM »
Have you considered changing the channel on all of the devices, to see if that changes the symptom?
When the unit was new, it received a neighbor's unit (I had left it on the default channel.  That caused me much confusion for a day or so, but I've left things alone on the current channel for years.

To answer your question, I've not considered changing it.  I assume the interference is more likely entering the receiver at the IF stage than actually coming in on the intended frequency.  I also guessed that the spread spectrum coding scheme in use makes the channel distinctions without actually using a different frequency by channel.  In other words I thought it unlikely to help, but I've not tried it.

Can someone here familiar with the Davis transmission protocol comment on how the different channels actually differ?

Offline johnd

  • Forecaster
  • *****
  • Posts: 4827
    • www.weatherstations.co.uk
Re: Swapping in a 6312 console for an Envoy
« Reply #9 on: January 31, 2017, 11:02:55 AM »
Can someone here familiar with the Davis transmission protocol comment on how the different channels actually differ?

They're just different time offsets from a reference timing AIUI.
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline Garth Bock

  • Table Rock Lake Weather
  • Forecaster
  • *****
  • Posts: 2745
Re: Swapping in a 6312 console for an Envoy
« Reply #10 on: January 31, 2017, 11:35:01 AM »
I did some reading and the procedure would be to unplug the power to the Envoy, remove it's batteries. Remove the batteries in the console. Pull the data logger and insert it in the Console. Put the batteries back in and connect a power brick to the console. You need a power brick to power the console with datalogger. No other configuration is necessary. You will lose some data points while everything is disconnected (unless you put the ISS in test mode). Since the console and envoy receive the same signal and data there should be no difference. If the dataloger and console are of the same vintage (firmware) then it should pick up where everything left off. You may need to reset/clear the Datalogger using WeatherLink. One thing I missed was what Dalecoy said....try a different channel to see if it changes the situation. You may have a new source of interference that has appeared recently.

I found this on the net......http://athena.trixology.com/index.php?topic=1424.0

Offline johnd

  • Forecaster
  • *****
  • Posts: 4827
    • www.weatherstations.co.uk
Re: Swapping in a 6312 console for an Envoy
« Reply #11 on: January 31, 2017, 12:01:12 PM »
You need a power brick to power the console with datalogger.

Only true if you're using the IP logger. The standard USB/serial logger is perfectly happy with batteries alone.
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Swapping in a 6312 console for an Envoy
« Reply #12 on: January 31, 2017, 12:04:57 PM »
Have you considered changing the channel on all of the devices, to see if that changes the symptom?
  In other words I thought it unlikely to help, but I've not tried it.

I agree it's unlikely to help - but the symptom that you are experiencing is also unlikely. 

It will be interesting to see whether swapping consoles fixes it.

Offline Garth Bock

  • Table Rock Lake Weather
  • Forecaster
  • *****
  • Posts: 2745
Re: Swapping in a 6312 console for an Envoy
« Reply #13 on: January 31, 2017, 12:22:46 PM »
You need a power brick to power the console with datalogger.

Only true if you're using the IP logger. The standard USB/serial logger is perfectly happy with batteries alone.

That is true. I misstated it. You need a power brick to power the console with datalogger...IF you want uninterrupted data. The batteries serve as backup. With no power brick you run the risk of dead batteries overnight and no data or during the time it takes to change them. Unless you stock up on C batteries....this can be a hassle.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #14 on: January 31, 2017, 12:31:54 PM »
My concern about software configuration arose from a garbled memory that both WeatherLink and Cumulus have station type fields--but that is for the ISS type, which is not changing.

The only other configuration item I've spotted in skimming the manuals so far is for baud rate on the console after initial turnon.  That may not even be applicable.

Regarding power being off during the datalogger connection/disconnect, one of the manuals is very strong in requiring this, threatening damage if not.

So my current plan is to review the manuals later today, and make myself a step-by-step procedure before I actually do it.  I also plan to review many months of data logs to get a better idea of how long I need to see good behavior to suppose something has improved.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Swap plan for tomorrow
« Reply #15 on: January 31, 2017, 09:41:19 PM »
I spent some time with the manuals for Envoy, Console, and WeatherLink.  As of now, my detailed plan is:

1. Prep console:
   a. place console in setup mode
   b. disconnect wall wart
   c. remove battery cover and batteries
2. assure both WeatherLink and Cumulus are shut down on PC
3. extract logger from Envoy
   a. disconnect USB cable and power cable from Envoy
   b. open the Envoy battery port and remove three AA batteries
   c. open Envoy (three screws on back), then pivot along long side opposite antenna and extract logger
4. insert logger in console
5. connect USB cable to logger
6. insert batteries in console
7. check baud rate on console
   enter Setup mode by holding down console DONE key, then pressing down arrow
   Use BAR and DONE to scroll to the Baud Rate settings screen
   Use the up (+) and down (-) arrows to change the baud rate setting, if needed, to 19200
   Press DONE when the correct baud rate is displayed
   When finished press and hold DONE to return to current weather screen
8. start WeatherLink and see whether it is in contact with the logger
9. shut down WeatherLink and Start Cumulus
10. check that Cumulus is receiving data.
11. check diagnostic screens for current signal strength.  Consider moving if below 30.
12. log console startup time for comparison to error graphs later.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Observed problem days per month
« Reply #16 on: January 31, 2017, 10:00:36 PM »
I did a fair bit of review of WeatherLink stored data to put some numbers on the rates at which I've seen some symptoms.  Also I looked back in time to see how far back I could find the 4:05 a.m. syndrome.

The farthest back I saw the 4:05 problem was on May 11, 2014, though my review before that date was not at all thorough.

For trouble calibration I defined three symptoms, as to which any given day I could classify as displaying any, all, or none.  The parameter I was looking at is called "ISS reception" in my version of WeatherLink, and is given as a precentage between 0 and 100.  I set the graph vertical axis limits to 0 and 100 for consistency, and reviewed data on the "Day" time scale.  Each data point in the WeatherLink graph appears to summarize 5 minutes of reception

1. I regarded a date as possessing the 4:05 problem if the 5-minute sample starting at 4:05 a.m., or either adjacent sample, scored below 90%.
2. I regarded a date as possessing a significant reception problem if at least one 5-minute sample, any time in the day, scored below 70%
3. I regarded a date as possessing the extended outage problem if at least 3 consecutive 5-minute samples showed 0%.  (most of these in fact went on for between one and three hours)

The numbers in the table are a count of the days showing that specific symptom during the month in question. 

Code: [Select]
Month Year 4:05  Sub 70 >15 minute
June 2014 13 10 0
January 2015 14 18 1
January 2016 16 20 3
January 2017 14 24 4

This is a bit tedious, and I only scored 4 months this way:

If my Envoy to console swap works, and if I see no 4:05 errors or sub-70 spikes in a week, I think I'll be clear that for some reason my console receives better than my Envoy in the face of my problem (whatever it is).  But the real thing I want to improve is the extended outage problem, and that is irregular enough in appearance that I'd need a couple of months just to begin to be reassured.

[edited the next day to add one to the January 2017 count for extended outages.  I had a two hour one in the evening--after I had made the data reviews posted here.]
« Last Edit: February 01, 2017, 10:31:27 AM by archae86 »

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #17 on: February 01, 2017, 10:51:40 AM »
The swapout where I moved the logger from Envoy to console was done along the steps I mentioned.

Just in case some neophyte comes along, here were the sticky points along the way (but most are perfectly normal, just a little disconcerting):

1. getting the batteries out of the Envoy was bad--had to use a screwdriver just to get the first end out.
2. the console "saw" my ISS fairly soon in the scan (number 3 showed up on the lower screen receiving row).  But seeing is not receiving.  I saw an R in the lower right corner for some minutes, which meant it was trying to resume ISS reception, but had not yet succeeded.  Only when the regular X display resumed was it actually getting live data from the ISS.  That took perhaps five minutes or so.
3. on initially starting up the WeatherLink program on my PC, the indications were ambiguous (and still are an hour later).  On the one hand, pressing the "test" button on the communication configuration page gets the response that it found a Vantage Pro (actually Pro2 in my case) station, but on the other hand for quite a while after connection attempts to download the station got a "no new data response".  And when it finally did, the first screen looked anomalous.
4. While I am used to Cumulus warning me that shut down may take several minutes, it normally takes very few seconds, so I was startled that both the data download on startup of Cumulus, and the "wait while I shut down" steps took far longer than normal.  Ball park 2 minutes, instead of a few seconds.

However at this point the Cumulus data output to my web site looks good.  You can see a small discontinuity in the internal temperature graph at the changeover, as the room temperature sensed the the thermometer inside the console was not a perfect match to that sensed by the  Envoy (using the add-on temperature probe).

Trying for an initial fair comparison, I've placed the console in the place I had the Envoy.  A signal strength quick check just now (second diagnostic screen) shows values of 38 to 41.  The bad packet count so far (since the batteries went back into the console a bit over an hour ago), is just 1, so I'm seeing the customary 99% "percentage of good packets received".  This does not mean things have improved, just that they seem to be working properly.

I'm a bit concerned over whether I need to do something to re-introduce the station to WeatherLink, as I am dependent on that for my detailed reception error data and historic comparisons.  As always, I don't run Cumulus and WeatherLink simultaneously, so I plan just to build up some hours in standard Cumulus operation, then try starting up Weatherlink and seeing whether a fresh download and data display work OK.  I probably should not fiddle with it until another day, at least.  If it still seems strange tomorrow, I think I'll try initiating the "New Station" procedure.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #18 on: February 01, 2017, 07:09:22 PM »
I'm a bit concerned over whether I need to do something to re-introduce the station to WeatherLink, as I am dependent on that for my detailed reception error data and historic comparisons.  As always, I don't run Cumulus and WeatherLink simultaneously, so I plan just to build up some hours in standard Cumulus operation, then try starting up Weatherlink and seeing whether a fresh download and data display work OK.  I probably should not fiddle with it until another day, at least.  If it still seems strange tomorrow, I think I'll try initiating the "New Station" procedure.

Some resources on the internet point out that, while Davis advises putting the console in setup mode before powering it down to assure a data transfer to the WeatherLink module is not interrupted in the middle with possible corruption, there is no provision to put the Envoy in a similarly safe state.  That person advocates clearing the logger archive memory after making such a change, just in case.  I did so later, but still that did not make things look right.  Finally I realized that somehow in the course of the swap the archiving interval got bumped up from the 5 minutes I have used for years to 30 minutes.  I put it back, which cleared archive memory again.  Possibly the unintended change from 5 to 30 came without clearing, which may have posed a problem to WeatherLink reading things properly.  Cumulus seemed unbothered. 

Anyway, I figure the real experiment started about three hours ago.  Zero errors logged in that time, but that is the expected value for so short an interval.

Offline PaulMy

  • Forecaster
  • *****
  • Posts: 5509
    • KomokaWeather
Re: Swapping in a 6312 console for an Envoy
« Reply #19 on: February 01, 2017, 11:11:22 PM »
Quote
While I am used to Cumulus warning me that shut down may take several minutes, it normally takes very few seconds, so I was startled that both the data download on startup of Cumulus, and the "wait while I shut down" steps took far longer than normal.  Ball park 2 minutes, instead of a few seconds.

Quote
Finally I realized that somehow in the course of the swap the archiving interval got bumped up from the 5 minutes I have used for years to 30 minutes.  I put it back, which cleared archive memory again.  Possibly the unintended change from 5 to 30 came without clearing, which may have posed a problem to WeatherLink reading things properly.  Cumulus seemed unbothered. 

This from the Cumulus install readme.txt likely explains the longer than usual time to read the logger data.
Quote
It is strongly recommended that you set the archive interval in your station and the Cumulus logging interval to the same value. This is particularly important if you have a Davis station as otherwise you may activate the 'feature' in the Davis stations where they send the entire contents of the data logger when Cumulus asks for the data since it was last running. You will need to use the Davis Weatherlink software to set the station's logger interval, and note that changing the station's logger interval also deletes any existing data in the logger.

Enjoy,
Paul


Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
faulty logger?
« Reply #20 on: February 02, 2017, 12:02:27 PM »
Now things are getting really strange.  I'm forced to wonder whether a defective or marginal logger might be my primary problem.

Details:
Last evening I stopped Cumulus to make a quick check of the transmission error log as downloaded and displayed by WeatherLink software.  But I neglected to restart Cumulus when I stopped Weatherlink, so no active software program was talking to the logger overnight (though quite possibly the driver was still somewhat active).

On finding neither program running this morning, I first used WeatherLink.  I got my first surprise in seeing what looked like my classic 4:05 a.m. 5-minute transmission gap, but somehow it was at 3:50 a.m.!  It was a nasty one, fully reaching zero percent success in the center 5 minute sample, with some loss in the two adjacent samples. 

A bit curious and dismayed, I shut down WeatherLink, restarted Cumulus, and went on about my morning.  A couple of hours later I glanced at the console display, and noticed it looked very abnormal.  Local temperature and humidity were displayed, with dashes where received data from the ISS ought to be.  I imagined I was in the course of the dreaded extended outage, and quickly went to the diagnostics display screens, and got  NOTHING USEFUL.  One or two of the fields had abnormal characters.  Many simply had zero.  I did not even get a signal strength display.

So far I had been running the console on batteries, which I have done successfully (without a logger) on this specific sample for 5 years.  Further I had put in freshly recharged (I use NiMH) ones just before this exercise.  Still, perhaps the console pulls spikes with this logger installed that are a problem with batteries, so I now have a wall wart plugged in.  And for some minutes it has run OK.  The Cumulus and the console graph suggest the console stopped receiving data very shortly after I restarted Cumulus (but not immediately, as Cumulus displays the logger archived data it successfully downloaded when first restarted).

In sum: I've seen at least some of the disturbing behavior I detected when this logger was plugged in my Envoy occur now that it is plugged into this particular console.  This same console did not have these symptoms when it ran in my kitchen for years, and ran in my study (within feet to inches of the long-term Envoy) location, for days.

On a conventional parts-swapping debugging doctrine, any PC tech would declare the logger proven bad, and seek to replace it.  But I'm really shy of imagining what fault in the logger would have my array of symptoms.

For the moment I intend to continue on wall-wart power with the logger installed in my "known-good" console.

Just for extra joy, of course my logger is old enough for the green-dot issue to apply.  So on my already stated intention of picking up a new console, it now appears I am on course to pick up a new WeatherLink product as well.  The only plus it that I won't have to try to find the old WeatherLink optical disk in order to install on my new daily driver computer.

I am all ears to any evidence, experience, or logic as to how my logger could be, or could not possibly be, at fault in all this.

Does anyone know whether the Envoy has the green-dot compatibility issue?  Or is it possible I could buy a new logger and plug it into the (5-year old) Envoy?

Offline Garth Bock

  • Table Rock Lake Weather
  • Forecaster
  • *****
  • Posts: 2745
Re: Swapping in a 6312 console for an Envoy
« Reply #21 on: February 02, 2017, 12:13:30 PM »
You can get a non-green dot logger on eBay. They are still out there. OOOORRRR.....you can get a Belfryboy logger for half the cost. One thing....you are sure that it is not the ISS doing it ? Just don't want it to get left out of the diagnostics loop.....you know supercap problem....battery...that sort of thing.

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #22 on: February 02, 2017, 12:40:26 PM »
you are sure that it is not the ISS doing it ?
I'm not sure of anything.  But the ISS would have to be pretty creative to misbehave in a way which caused multi-hour dropouts to whatever receiver had the logger installed while one or two other receivers listening to the same ISS had no trouble at all in the same time period.  After typing this but before hitting "post", I checked the diagnostic screen for the console at the other end of the house.   The first diagnostic screen listed 60 bad data packets and 14088 good ones, which it rounded up to 100% success (I usually see 99%).  I believe those numbers are since midnight, so include the couple of hours in which my console with the logger installed was completely out of ISS contact.

I've noticed an additional oddity:  Cumulus displays intact indoor data (temperature and humidity) for the dropout period today.  So either live communication between PC and logger (transmitting current console local data) continued through the ISS reception outage, or data was archived into the logger and successfully downloaded later.

As to what possible logger fault might  contribute to some of this, I'll propose a state-dependent excess power consumption defect.  As an old semiconductor guy (long-term employer of ye olde microprocessor werkes, in design, reliability, and manufacturing data analysis jobs), I know such defects exist, are hard to test for, and some folks don't want to discard the faulty parts even if testing finds them.

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Swapping in a 6312 console for an Envoy
« Reply #23 on: February 02, 2017, 02:10:07 PM »
As to what possible logger fault might  contribute to some of this, I'll propose a state-dependent excess power consumption defect. 

...which only happens early in the morning, around 3:50 or 4:05 AM.......

On the other hand, you have two different receivers that, with this logger, in the same location, same connection cable, to the same PC, same software programs, and with the same channel, exhibit similar symptoms.

(While other receivers, in other locations, do not have the problem.......)

I'm betting it's not the logger - but of course one way to find out is to buy another logger.

[Oh, yeah - re: 3:50 event - is the console time the same as the PC time?]

Offline archae86

  • Senior Contributor
  • ****
  • Posts: 184
    • LynxStoll weather
Re: Swapping in a 6312 console for an Envoy
« Reply #24 on: February 02, 2017, 02:44:50 PM »

...which only happens early in the morning, around 3:50 or 4:05 AM.......

The issue of concern is not the 4:05 matter, whatever it is, but rather the extended outages.

Quote
[Oh, yeah - re: 3:50 event - is the console time the same as the PC time?]
It is now.  I don't know what opinion the Envoy has of time, though I suppose it must have some degree of elapsed time awareness in order to support the prescribed archiving interval. 

Before your query, I separately realized that I don't know just how the time stamps get onto the history data in either the Cumulus software or the WeatherLink software.  In my configuration the usual case is that Cumulus acquired the data "live" and one would suppose the time stamp would be the host PC's idea of time (which by the usual Windows methods is kept very close to correct time by use of Internet resources).  However after host PC outages, Cumulus shutdown to run WeatherLink, and so forth, data clearly gets to the Cumulus history that was obtained from the logger archive.  Now the console does have an opinion on the time, and one needs to nudge it now and again if one prefers it not to drift too far.  If the Envoy has a full-up internal clock, I am sure I have not updated it myself in years.

So, to be long-winded, perhaps the 4:05 vs. 3:50 difference was the difference between the Envoy internal clock as of yesterday, and this console as of early this morning.

 

anything