Author Topic: Repeated signal dropouts  (Read 22309 times)

0 Members and 1 Guest are viewing this topic.

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #200 on: January 24, 2018, 10:37:12 AM »
I'm a bit skeptical about the role of QC/QA - I would not expect the individual products to be tested at the temperature limits.

I was thinking strictly of the components themselves and/or the assembly process. Usually there is a control process to weed out substandard units.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #201 on: January 24, 2018, 02:01:32 PM »
This has been an interesting topic.

I'm a bit skeptical about the role of QC/QA - I would not expect the individual products to be tested at the temperature limits.

I was thinking strictly of the components themselves and/or the assembly process. Usually there is a control process to weed out substandard units.

So, what's your estimate of the number of reports that Davis has received, regarding cold-weather signal dropouts?

And how many VP2 systems do you suppose that Davis has manufactured?

Offline SnowHiker

  • Forecaster
  • *****
  • Posts: 352
Re: Repeated signal dropouts
« Reply #202 on: January 24, 2018, 02:08:10 PM »

miraculon, who lives in northern Michigan, has 6 transmitters and 2 VP2 consoles he says do NOT experience dropouts. He would have to confirm how many of those transmitters each console receives, but I think it's pretty safe to assume at least one of those consoles has more than 2 IDs selected.

I would think even if each console isn't set up to receive all the transmitters, the extra signals would provide a certain amount of interference, thus befuddling* the console to some extent.

*Sorry for my lack of technical jargon.

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #203 on: January 24, 2018, 03:03:48 PM »
So, what's your estimate of the number of reports that Davis has received, regarding cold-weather signal dropouts?

And how many VP2 systems do you suppose that Davis has manufactured?

The total units produced question is probably better directed toward johnd. But my guess would be in the 10-25 thousand range. That wouldn't include supplemental consoles or transmitters, just base systems.

Given the qualifiers necessary to experience the problem, I'd say dozens have reached out for support to Davis. But for every one that is willing to do so, there's got to be at least 10, 50, maybe even 100 that don't, especially since the problem may not happen during the warranty period. I live just a few blocks from Lake Superior and didn't really notice the problem until this year. 2 warm winters previous to that didn't consistently expose it.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #204 on: January 24, 2018, 03:10:27 PM »
Maybe I should revise the total VP2 stations produced to 50-100k given we're talking worldwide. But, I'm just spit balling here.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #205 on: January 24, 2018, 05:09:10 PM »
I'd say dozens have reached out for support to Davis.

I'd guess fewer than that, but - if we say TWO dozen, then...

That's roughly an average of two reports per year.

If we assume 100,000 units, then that's  0.024%

In a non-life-safety product..........

Online mcrossley

  • Forecaster
  • *****
  • Posts: 1132
    • Wilmslow Astro
Re: Repeated signal dropouts
« Reply #206 on: January 24, 2018, 06:21:20 PM »
Has anybody with the problem actually checked on their VP2 console to see if the frequency correction being applied is getting larger in cold weather? And  if it is showing a larger value than usual, warming a transmitter to see if it comes back to a smaller value?
Mark

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #207 on: January 24, 2018, 07:07:18 PM »
....to see if the frequency correction being applied is getting larger in cold weather?

Note that it could actually be getting smaller.  It's the change that's a factor here.

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #208 on: January 25, 2018, 12:26:48 AM »
I'd say dozens have reached out for support to Davis.

I'd guess fewer than that, but - if we say TWO dozen, then...

That's roughly an average of two reports per year.

If we assume 100,000 units, then that's  0.024%

In a non-life-safety product..........

Well, if we're talking 100k units, then I'd up my report estimate to the low hundreds.

In your initial calculations, you forgot to multiply the reported cases. I'm guessing you don't believe that the quantity of support interactions == the quantity of VP2 owners affected. Let's assume that a "high friction" issue such as this one would be 1 report per 50 customers affected (1:50 ratio) which is a middle range estimate between 1:10 and 1:100. So, at 200 support touches (again middle of the range "low hundreds"), that would be 10,000 people potentially affected or 10% of the customer base. That's somewhere in the ballpark I would think. Hard to say.

Here's a follow up question for you: what percentage of owners would contact Davis if the solder joints connecting the battery terminals to the board in every VP2 console came loose within 2 years?

I'd bet 5% (tops!) for an issue that affects 100% of consoles eventually. And just like in the case of cold weather drop outs, there are going to be some owners who never notice for one reason or another. Regardless, that's a 1:20 ratio for a higher profile, easier to comprehend, lower friction issue.

For something immediately obvious, like a cracked display out of the box, I'd expect closer to 1:10 or less. Most people would just return it to the dealer though.

So, no, I don't think support requests potentially representing 10% of the customer base, give or take, is a vanishingly small issue.

Full disclosure: I've spent the last 20 years in the IT support field, including consulting for businesses.
« Last Edit: January 25, 2018, 09:19:34 AM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #209 on: January 25, 2018, 10:25:18 AM »
Full disclosure: I've spent the last 20 years in the IT support field, including consulting for businesses.

We have somewhat similar backgrounds.

Of course, I was talking about reported instances, and you are talking about affected installations.  Quite different topics.

And reports of IT system problems are almost always easily reproducible, while the problem under discussion here is not so easily reproduced - particularly based on a few user reports that are short on details. 

I know you understand that, if users experience a problem but nobody reports it, it doesn't get fixed.  And that problem-investigation-and-fix resources get allocated based on the number and severity of reports. 

What's your evaluation of the severity of this signal dropout issue?  [Compared to solder joints connecting the battery terminals to the board in the same number of VP2 consoles.]

And did you significantly increase the number and details of signal-dropout support calls to Davis?

Yes, this is an interesting problem, and (perhaps) indicates that the VP2 system does not completely meet the advertised specifications under the entire range of conditions and combinations of accessories.

If you were hired as a consultant to Davis, what would you advise them to do?

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #210 on: January 25, 2018, 12:47:15 PM »
We have somewhat similar backgrounds.

Of course, I was talking about reported instances, and you are talking about affected installations.  Quite different topics.

And reports of IT system problems are almost always easily reproducible, while the problem under discussion here is not so easily reproduced - particularly based on a few user reports that are short on details. 

I know you understand that, if users experience a problem but nobody reports it, it doesn't get fixed.  And that problem-investigation-and-fix resources get allocated based on the number and severity of reports. 

What's your evaluation of the severity of this signal dropout issue?  [Compared to solder joints connecting the battery terminals to the board in the same number of VP2 consoles.]

And did you significantly increase the number and details of signal-dropout support calls to Davis?

Yes, this is an interesting problem, and (perhaps) indicates that the VP2 system does not completely meet the advertised specifications under the entire range of conditions and combinations of accessories.

If you were hired as a consultant to Davis, what would you advise them to do?

Posing speculative questions like "What's your evaluation of the severity of this signal dropout issue?" or "how many VP2 systems do you suppose that Davis has manufactured?" only to cleverly attempt to falsify any answer (which is non-falsifiable by nature since the facts are clearly unknown or the answer can only be subjective) isn't helping to clarify matters.

And, no, IT support issues are not "almost always" reproducible any more than car troubles are almost always reproducible. There's a fair number that are all around vexing.

Let's simplify this. In the end, it doesn't matter how many users are aware or have reported it or even whether Davis is aware. If a product owner cannot operate their device normally according to the instructions and stated limitations in the manual without repeatable failures, the product is defective. Period.

Quote
I agree with your speculation about component tolerances and hardware/firmware at the limits, and would add individual siting factors, local RF environment, and peopleware

Let's examine your list of possible confounding factors (reasons why the operator or environment could be at fault).

First up, the idea that there may be an unfavorable "RF environment". Not only has no evidence has been presented to support this contention, but, in fact, the available evidence goes the other way. We know that signal strength is acceptable even under the conditions in which failure is likely. Background noise has also been shown to be low. The Vue console's 0 resync count compared to an adjacent VP2's console's positive resync count is further evidence that the issue is not external.

You mentioned siting. Is placing an ISS 60' away from the console through normal residential wood-frame construction considered poor siting?  This would seem especially dubious given the reception statistics mentioned above.

Up next we have "peopleware" (a vague, elastic term). This "peopleware" is active when the same person fires up their new Vue console and it operates normally under identical conditions causing failure in the VP2 console. Curious.

No reasonable or supportable alternative hypothesis has been forwarded to date as to why certain VP2 consoles fail in cold weather. We can speculate on the exact nature of the error(s) or exacerbating factors, but we have sufficient first-hand accounts, statistical data and expertise to conclude the responsibility primarily rests with the manufacturer.





« Last Edit: January 25, 2018, 12:55:39 PM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #211 on: January 25, 2018, 01:16:29 PM »
we have sufficient first-hand accounts, statistical data and expertise to conclude the responsibility primarily rests with the manufacturer.

I agree.

Full disclosure: I've spent the last 20 years in the IT support field, including consulting for businesses.

If you were hired as a consultant to Davis, what would you advise them to do?

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #212 on: January 25, 2018, 02:41:39 PM »
If you were hired as a consultant to Davis, what would you advise them to do?

Your question deliberately confuses roles. I'm a consumer of a defective product. Why should I offer the manufacturer free consultation, especially when they've shown not the slightest interest in understanding this issue (or other, less consequential issues)? Or was your thinly-disguised, patronizing point that I should take a moment to consider Davis' perspective? I assure you I've already carefully reviewed Davis' position on this matter.

If johnd, who's livelihood is connected to selling Davis equipment, can remain fair and impartial, why can't others who presumably lack a conflict of interest refrain from reflexively minimizing and blame shifting?
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #213 on: January 25, 2018, 02:48:19 PM »
If you were hired as a consultant to Davis, what would you advise them to do?

... I'm a consumer of a defective product....
 
.... I assure you I've already carefully reviewed Davis' position on this matter....


I believe I understand.

I'm happy to have answered some questions.

Offline Old Tele man

  • Singing in the rain...
  • Forecaster
  • *****
  • Posts: 1365
Re: Repeated signal dropouts
« Reply #214 on: January 25, 2018, 03:56:19 PM »
DEVILS ADVOCATE:  Opinions do not equal facts.
• SYS: Davis VP2 Vue/WL-IP & Envoy8X/WL-USB;
• DBX2 & DBX1 Precision Digital Barographs
• CWOP: DW6988 - 2 miles NNE of Cortaro, AZ
• WU - KAZTUCSO202, Countryside

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #215 on: January 25, 2018, 04:07:58 PM »
DEVILS ADVOCATE:  Opinions do not equal facts.

I'm not sure what that means. You'll need to clarify.

I'll point out, though, there are sufficient facts in this thread to make a prima facie case for design failure. Of course, there are plenty of opinions too!
« Last Edit: January 25, 2018, 04:12:12 PM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline johnd

  • Forecaster
  • *****
  • Posts: 4827
    • www.weatherstations.co.uk
Re: Repeated signal dropouts
« Reply #216 on: January 25, 2018, 04:53:13 PM »
OK.  The following is based on work by dekay and several others:  And the actual Davis algorithm might be slightly different (we have no way of knowing), but generally has to work this way:

OK, many thanks for the explanation. That's broadly the picture I had, but it's at the next level of detail that I start getting more lost.

So, just starting off thinking about a single transmitter on #1: AIUI, data is transmitted in eg 20-byte packets with 8-10-byte payload & some of those bytes used for comms overhead, ie start/stop bytes etc. Also AIUI/BICBW there's a trade-off of baud rate and sensitivity which, within limits, is up to the developer to decide on. But if eg 38.4K baud was being used and assuming in this case that bps = baud (not sure whether that's true or not, but if...) then each packet takes ~4ms to transmit.

Then do we think that each complete packet is transmitted on a single frequency, which I'm guessing would simplify the design? And then the transmitter sleeps for a relatively long period, which I thought was how it manages to use so little battery/supercap power. At what point will it hop to the next frequency? Maybe after 62.5ms? So the next packet arriving would arrive on the next frequency in the sequence? And so on? If this isn't the correct picture then how often do the frequencies change?

Now if there are multiple transmitters in play: My previous picture, which I know may well be wrong was that when a new transmitter starts up, it first listens on the primary frequency in the sequence and if it hears another active Tx then synchronises with that signal. Otherwise surely the console has to be listening on 50-odd (US) frequencies simultaneously, which sounds like a clever trick.

Sorry, I know that this is rather OT for this thread, but it would simply be interesting to learn in more detail just how the data comms actually work.
« Last Edit: January 25, 2018, 04:57:09 PM by johnd »
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.

Online mcrossley

  • Forecaster
  • *****
  • Posts: 1132
    • Wilmslow Astro
Re: Repeated signal dropouts
« Reply #217 on: January 25, 2018, 05:07:27 PM »
Each transmitter just round robins the set frequencies for the geography.

Each packet from a transmitter is sent on a different channel.

There is no link between transmitters, the id set determining the gap between transmissions. Because you have no control over when each transmitter starts it's sequence the gap variation between channels assures that the transmission collisions are minimised.
Mark

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #218 on: January 25, 2018, 05:14:03 PM »
The baud rate is 19.2
The packet is transmitted on a single frequency.
The transmitter then sleeps for 2.5 seconds (if it's on ID #1) - and then transmits the new packet on the next frequency in order.  [The frequency order is the same for all transmitters]

Quote
when a new transmitter starts up, it first listens on the primary frequency in the sequence

No, the transmitter doesn't listen.  And there is no "primary frequency". 

It transmits a packet on a frequency, sleeps for 2.5(+) seconds, and transmits a new packet on the next frequency in order.

That greatly simplifies the job of the transmitter - and complicates the job of the receiver.

Let me know what's confusing here.


Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Re: Repeated signal dropouts
« Reply #219 on: January 25, 2018, 05:18:00 PM »
Sorry, I know that this is rather OT for this thread, but it would simply be interesting to learn in more detail just how the data comms actually work.

Nope, not OT at all. I find it illuminating as well and applicable to this issue.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Online mcrossley

  • Forecaster
  • *****
  • Posts: 1132
    • Wilmslow Astro
Re: Repeated signal dropouts
« Reply #220 on: January 25, 2018, 05:56:28 PM »
Now if there are multiple transmitters in play: My previous picture, which I know may well be wrong was that when a new transmitter starts up, it first listens on the primary frequency in the sequence and if it hears another active Tx then synchronises with that signal. Otherwise surely the console has to be listening on 50-odd (US) frequencies simultaneously, which sounds like a clever trick.

The way it is implemented on my transceiver, is that it:

1. Listens on a frequency for the transmission id you expect to hear.
2. If nothing is heard after a period of time, switch to the next frequency and listen again (there may be interference on the first frequency you pick).
3 Repeat 2. until you receive the transmission you want.
4. When you get that first packet, store the expected time for that channels next transmission, and..
5. Just before the next transmission is due, set to the expected frequency and switch on the receiver.
6. If the packet is received, repeat 4 & 5. If it is missed, then skip that channel on to the next frequency and expected time anyway.

You set the time for the next packet from the received time of the last, however if you miss one then you have to set the time to the last 'expected' time plus the interval - so you can see a timing error can accumulate from a missed or sequence of missed packets.

Of course the above is duplicated for each channel you want to receive. I set a timer interrupt for the next expected transmission from any channel, as soon as you receive a packet the receiver is switched off and an interrupt set for the next expected packet - that way you can receive packets on different frequencies quite close together in time.
Mark

Offline dalecoy

  • Forecaster
  • *****
  • Posts: 6447
    • Lee's Summit, MO
Re: Repeated signal dropouts
« Reply #221 on: January 25, 2018, 07:36:46 PM »
Thanks, Mark.  As I said earlier:
.... the actual Davis algorithm might be slightly different (we have no way of knowing), but generally has to work this way:

There's certainly a lot of ways for increased "smarts" in the firmware, with corresponding better performance in normal and off-normal conditions.  Limited by the amount of firmware memory space available.  The Vue console seems to be smarter than the VP2.