Author Topic: bad AC power = bad wind readings??? (Read 10743 times)

coyotefred · « **on:** February 13, 2009, 01:09:25 PM »

* Although this is a continuation of an earlier discussion (http://www.wxforum.net/index.php?topic=1600.0), I decided to start a new thread with a more specific Subject line...

To summarize: I'm running a Lacrosse 2315 hooked up to an 5 year-old emachines laptop running XP Home. Most recently I've been running Cumulus software, but I've tried several other programs (WD, WU, VWS, HW) with identical results. The "errors" I’m trying to get rid of are erroneous wind readings that show up every 2-7 minutes. The most common erroneous readings are 14.1, 28.4, 71.3 and 85.1 mph. The readings show up both on the 2315 console and the software display, but only occur when the console is hooked up to the laptop.

I originally thought the problem was in the cabling between the sensors and the console, so I replaced all of that cabling with CAT-5. I’ve used several different serial-USB adapters, and most recently bought and tried out a PCMCIA RS232 serial card. Finally, I bought some ferrite chokes and installed these on the cables thinking RF interference might be the culprit. But none of these attempted “fixes” eliminated the erroneous readings.

But last night I had a "breakthrough" of sorts. I decided to run the laptop from battery power only and the erroneous readings stopped (or at least for 10 minutes before the battery died since it is old and I never run the laptop from battery power). When the AC adapter cord was plugged in, the erroneous readings reappeared with 2-3 minutes. So I think it’s safe to say that the problem involves the AC adapter cord—or at least something involving the AC powering of the laptop.

Incidentally, I've tried all sorts of combinations of plugging in both the 2315 console and the laptop to different outlets around my house, both with and without basic power strips. It *seemed* that plugging the laptop into an outlet farther away from the outlet into which the 2315 was plugged reduced the frequency of the readings, but it did not eliminate them… I live in a rural area in a late 1940s house with some but not all of the wiring updated, and our rural power supplier has been known to have issues now and then with minor “spikes,” etc. (although I’ve never had any other problems with the rest of my electronic equipment).

So…. any ideas on the best next step in light of this “discovery”? Should I be exploring some kind of “cleaner” or regulated power supply, or maybe some kind of more sophisticated power strip with “filtering” capabilities? The AC adapter that came with the laptop looks pretty typical, with a rectangular “block” in the middle and a detachable power cord out of one side and a cord with a small choke on it that runs to the laptop plug. The specs are 18.5V 1.5A output (Lite-ON model PA-1900-05).

Any advice on what might solve the problem? Thanks for your time-

ironton · « **Reply #1 on:** February 13, 2009, 01:44:21 PM »

That is a good piece of investigation, coyotefred.

I have a 2310 that exhibited the same spikes in wind at the same multiples (cat5 solved it for me). Mine was connected to a desktop so I did not have the option of running on battery power.

From what you describe, the first thought I had was something to do with the ground - perhaps a loop between the Laptop and console that results in some kind of spike. It could also be related to spikes in the line voltage due to a motor start (like in fridge) or static build up in the weather station lines. Does the console have a battery that could run for a while to see if you get similar success?

Another thought, you mention the house wiring is 1940's. Do you have 3 prong outlets? and do the ground wires all return to the main panel? It would not be the first time that someone put in grounded outlets but did not run all ground wires back to a ground.
It will be interesting to read other ideas.

coyotefred · « **Reply #2 on:** February 13, 2009, 01:55:42 PM »

I'm wondering the same thing. Most of my outlets (and the ones I've mainly using for the console and the laptop) are 3-prong, but I haven't investigated whether there is true grounding back to the box. The breaker box is relatively new, but I'm not sure whether or how much of the original wiring was redone when that was put in...

wuhu_software · « **Reply #3 on:** February 13, 2009, 06:05:32 PM »

La Crosse using a scheme where they are creating a "virtual ground" by using the DTR and RTS lines of the serial port.

There are only 4 wires run to the console, RX, TX, RTS, DTR.

One of the lines (DTR or RTS) is held low (I forget which) and is tied to the ground plane of the console. It is also tied to the - terminal of the power pack.

If you are using a surge suppressing power bar, try one without surge suppression (just a regular multi-outlet cord), that may make a difference. I am not sure if there is a difference if the power pack is oriented (hot/neutral) as the PC cord, that may or may not make a difference.

You might also try attaching the console to a regular PC serial port (versus the laptop serial port). Some laptop using a +/- 5vdc scheme rather than the +/- 12vdc used by RS-232. That voltage range might make a difference.

As I recall, I tested the IO Gear USB lines, I believe they were +/- 12vdc. Not all USB to serial adapters are alike.

coyotefred · « **Reply #4 on:** February 13, 2009, 07:37:22 PM »

Hmmmmm....interesting. I appreciate your detailed thoughts here. I will try to experiment a little bit with this over the weekend...

DanS · « **Reply #5 on:** February 13, 2009, 07:55:09 PM »

I had similar erratic wind readings on my WS-2310 and discovered my desktop P.C. I was using had a floating ground. I measured approx. 90 volts AC between the P.C. and ground. By grounding the P.C. the problem with the erratic readings went away. I'm wondering what would be the result if you ran a ground from somewhere on the laptop

coyotefred · « **Reply #6 on:** February 13, 2009, 08:57:00 PM »

Good question...how did you go about grounding your PC?

Quote from: DanS on February 13, 2009, 07:55:09 PM

I had similar erratic wind readings on my WS-2310 and discovered my desktop P.C. I was using had a floating ground. I measured approx. 90 volts AC between the P.C. and ground. By grounding the P.C. the problem with the erratic readings went away. I'm wondering what would be the result if you ran a ground from somewhere on the laptop

DanS · « **Reply #7 on:** February 13, 2009, 09:21:50 PM »

You could use one of those little 3-prong AC plug adapters with the ground tab and connect a wire from there up to one of the laptop's comm ports (serial / parallel) connector screws? Might not be pretty but would check to see if that's the problem or not anyway.

coyotefred · « **Reply #8 on:** February 14, 2009, 04:36:13 PM »

That's a good idea. I'm going to give this a try tomorrow and try to check on my house wiring/grounding situation tomorrow and I'll report back what I find...

DanS · « **Reply #9 on:** February 18, 2009, 07:04:59 AM »

Quote from: coyotefred on February 14, 2009, 04:36:13 PM

That's a good idea. I'm going to give this a try tomorrow and try to check on my house wiring/grounding situation tomorrow and I'll report back what I find...

Not trying to be nosey but wondering if you nailed down your erroneous readings problem and if you have what the cure was. Just good to know for future reference.

Thanks,
Dan

coyotefred · « **Reply #10 on:** February 18, 2009, 12:56:20 PM »

Well mostly good news....the grounding issue seems to be the main culprit, but I haven't totally solved the problem

I discovered that none of my (1940s) house wiring was grounded, despite the fact that there were plenty of 3-prong outlets around. I therefore created a ground by running a grounding rod into the ground outside the house and running a grounding cable in from that inside the house through the cable run where the sensor and other outdoor cabling comes in. I then created some "grounded" outlets by using outlet adapters where you can attach an external ground. Finally I plugged both the laptop and console into these new "grounded" outlets.

I've observed this setup for about 2 days now and the number of erroneous readings has definitely been reduced. In fact initially I gloated prematurely as I saw none of the erroneous readings for an hour two. But as I spent a little more time observing last night and reviewing the software data log, I could see that I was still getting some erroneous readings from time to time--most frequently the 14.1mph but also a 28.4 once or twice.

So I will continue to "tinker" and see if any of that helps eliminate the readings entirely. I may run that ground wire to the laptop itself (rather than just relying on the grounding through the AC cord). And I've just been "testing" with the PCMCIA card rather than the USB-serial adapter, so I may try that as well. But of course I'd welcome any suggestions from the group here on other things to try. I feel like I'm getting close...!

Coyotefred

PS One thing that is odd....if I look at the most common erroneous readings I'm getting there does appear to be a basic "pattern" at play: 14.1, 28.4, 55.7, 71.3, 85.1--all of these are pretty close to multiples of 14-ish...perhaps someone with more mathematical/computer knowledge than me can surmise what this might mean as far as diagnosing the problem...

Scalphunter · « **Reply #11 on:** February 18, 2009, 03:15:45 PM »

They are what we call harmonics. It is use in radio to get frequency with out going thru trouble of build a complete rig to get desired freq. Those of you who use CB" and wonder why your Channel 2 on the TV would got awry. 27 MHZ raido 54 MHZ channel on TV. But that why you are seeing the multiple there 14, 28, ETC. The above practice was use in Hamming to get some of the VHF/UHF freq. when there was not store bougth gear to be had.

John
KL7IFP

DanS · « **Reply #12 on:** February 18, 2009, 04:07:37 PM »

Sounds like you're narrowing it down.

Another thing you might want to try is "shielding" your cable from the console to the computer. I used foil tape (or metal tape). I laid the cable out on the floor and applied the tape length-wise to wrap it all. The end of the tape stops just before the console end so no contact at the console. The end that goes to the P.C. should make contact to the P.C. chassis anywhere there's a chassis ground available (connector screw, etc.). This connection gives the shield a ground/drain. This would eliminate the possibility of RF interference on this cable anyway. The anemometer cable maybe experiencing this as well. Make sense? It's not always easy to convey a picture or thoughts with a keyboard.
Good work finding the AC power ground (or lack of) problem. That will definately cause problems.

Albuquerque Weather · « **Reply #13 on:** February 18, 2009, 05:13:09 PM »

Hi
T had same problem here with no ground into the room where pc and wx is located. Fixed that same way - grounding rod.

Also did this mod
I didn't realize it then but the rain gauge was also affected--10-12 inches accumulating every hour or so, in the middle of a drought! Having seen somewhere in this forum that this was the same problem as the 57mph wind spikes: interference displayed instead as inches of rain, I decided to deal with it first.

Since the rain sensor is easily accessible, I went directly to circuit board on it and replaced the unshielded telco wire with CAT5e twisted pair. It was a complete success: no more problems with crazy rainfall readings.

I have done the same to the wind sensors, opting for the "hard" method shown by Kenneth Lavrsen's on his site http://www.lavrsen.dk/sources/weather/windmod.htm and it worked great also.

coyotefred · « **Reply #14 on:** February 18, 2009, 11:16:10 PM »

Hi there,

Those mods make sense, but I'm not sure those are the problems in my case. Unless there is something in those exposed cables in the rain and wind sensors that would cause an issue only when my computer is running off AC (remember I only get the errors when the lacrosse console is connected to the computer and the computer (laptop) is running off AC power and not a battery), I have to think my remaining problem has to do with something "closer to home" with the computer and the console. As DanS recommended, the next thing I'm going to try is shielding (and grounding on the computer end) the cable between the 2315 console and the laptop...

Bushman · « **Reply #15 on:** February 18, 2009, 11:44:42 PM »

Do you have a ferrite core on the power to laptop cable? Nearest to PC.

DanS · « **Reply #16 on:** February 18, 2009, 11:52:13 PM »

... also I believe you mentioned this already but I'll remind again. The power being the way it is, it probably wouldn't hurt to have the laptop's power cord plugged into the same power source as the console uses. With the laptop having it's ground provided and the shielded cable "referencing" the console to the same ground you should be getting close to nailing this issue.

coyotefred · « **Reply #17 on:** February 19, 2009, 10:36:23 AM »

Yep. The power cable had one already, but I put a couple more on there earlier on in my testing process to see if that would make a difference (it didn't)

Quote from: Bushman on February 18, 2009, 11:44:42 PM

Do you have a ferrite core on the power to laptop cable? Nearest to PC.

tomcj2 · « **Reply #18 on:** February 19, 2009, 01:17:00 PM »

Quote from: coyotefred on February 18, 2009, 12:56:20 PM

I discovered that none of my (1940s) house wiring was grounded,

There will be a ground at your service disconnect (which is most likely your main panel). The ground may be to one or more ground rods, but, considering the age of the building, is probably only connected to a the metal plumbing system of your home. I have come across many homes that a section of water pipe has been changed to plastic, not good! If that ground is not "good" you will have problems, the most common would be substantially different voltages between one side of the main buss to ground compared to the other buss to ground. Both should be about 120V . The new ground rod that you just installed should be tied to the the existing grounding conductor with #6 copper wire to ensure there is not a potential difference (voltage) between the two.

coyotefred · « **Reply #19 on:** February 20, 2009, 11:37:05 AM »

UPDATE

So far so good with DanS' suggestion of shielding the data cable from the console to the laptop and then grounding that shielding to the laptop

However, winds are real high today so I'm not able to get a good "read" on the common 14.1 and 28.4 erroneous wind readings. I'll monitor over the weekend and play around a bit more and report back...

coyotefred · « **Reply #20 on:** February 24, 2009, 09:45:03 PM »

UPDATE:

I appear to (finally!) have normal readings with no erroneous wind errors. The final "fix" was grounding the laptop by running a small grounding wire from the ground I had created outside (see earlier message) and attaching to the one of the screws on the laptop used for attaching a serial cable. I'm not sure what one would do for a newer computer; my newer one has no obvious attachment-points for grounding the laptop. But this final step made the difference. 'Just out of curiosity I backtracked a bit and plugged both the 2315 console and laptop into one of my other (*non*-grounded) outlets. I still had no errors. So it appears that the decisive "fix" for me was grounding the laptop itself; grounding the outlets into which the laptop or console is plugged wasn't relevant. Nor was shielding/grounding the data cable from the console to the laptop.

So to summarize for folks of the future who may be skimming to the end of this thread...

1. I've been struggling with erroneous wind readings when I connected the Lacrosse 2315 weather station console to my (4-5 year old) emachines laptop via the included data cable and a USB-serial cable adapter. The most common erroneous wind readings I would receive were 14.1 (mph), 28.0, 57.0, 85.0, 71.3 and readings very close to those specific ones. These readings would only show up when the computer was connected; they would stop completely when the computer was disconnected.

2. I initially tried "fixing" these readings by replacing the factory-included cable with CAT-5 shielded cable. This seemed to reduce the frequency of the readings, but they were still quite common (every 7-10 minutes).

3. I later tried fixes including attaching ferrite cores to all wiring, using a PCMCIA card to connect the data cable to laptop rather than a serial-USB connector, and various combinations of plugging the console and the laptop into the same, different, etc. outlets.

4. In the end, it appears the decisive "fix" for my particular situation was grounding the laptop, which I did by running a wire from an outdoor grounding rod to the laptop itself as described in earlier posts. I had to do this because my older home did not have grounded outlets.

So for folks suffering from similar erroneous readings in the future, I would definitely start my problem-solving by looking at the grounding situation of your computer...

Thanks everyone for all of your comments and suggestions for help...
Coyotefred

DanS · « **Reply #21 on:** February 24, 2009, 10:03:21 PM »

Glad to hear you nailed it down. Strange things can happen with floating grounds. I frequently water the ground around my grounding rod during dry spells to help keep the connection.
Good work!

Dan

riverratmike · « **Reply #22 on:** June 06, 2009, 09:38:22 PM »

I also have this wind spike problem with my ws-2310. I heard somewhere a couple of years ago there was a step by step fix for the anometer. Just today I got a wind spike of 68.0 mph.

Albuquerque Weather · « **Reply #23 on:** June 06, 2009, 10:01:12 PM »

Riverratmike

I use a 2308 and had same problem till i did this mod

http://www.lavrsen.dk/sources/weather/windmod.htm

riverratmike · « **Reply #24 on:** June 06, 2009, 10:43:58 PM »

thanks, I'll check it out.

News:

Author Topic: bad AC power = bad wind readings??? (Read 10743 times)