For the USB data logger lock up problem
Message from Davis:
We are no longer swapping dataloggers for this issue after discovering the
source of the problem and a couple of solutions.
Apparently, the issue is noise and quite a problem for any SUB device that is
constantly connected and always on.
We have found that if the user simply unplugs their AC adapter from the
console and run on the console’s C batteries alone (they’ll last a year and
warn you when they’re low) the problem goes away.
Some customers cannot unplug their AC adapter. Sometimes they have cabled
systems and must use the AC adapter or they may exercise the datalogger
enough (5 minute internet uploads, bulletin 24/7, etc.) that the C batteries
run less than a year and makes replacement inconvenient. These users should
then use a torroid placed on the AC line, USB cable or both to filter out the
The torroids we used were of the type detailed at this link.
From the other thread about lock ups, I had done both of the things that Davis asked me to do, batteries only and the purchase of the cores from radio shack (at $12 each I think) Still the same problem. Maybe thats why they replaced mine.
It also brings into question the quality of the usb cables, apparently in some cases CAT5E shielded has better performance over a longer distance than oem USB cables. Anyone tried an alternative cable?
More to the point, if Davis know this is a regular problem, what are they doing to sort it?
They say they have located the source of the lockouts and come up with some solutions but will they be building these solutions into their equipment so that customers get a reliable unit when it’s used with the USB datalogger? :-k
Good points. It’s interesting that users with a separate USB-serial adaptor don’t seem to suffer from this issue so I’m not sure I accept that it’s “a problem for any USB device that is constantly connected and always on”. I’ve not seen similar complaints on the Labjack forum, that’s a USB device that’s used even more intensively - but it does use a different interface.
Would also be nice if they gave away the toroids instead of referring to a vendor with a minimum order policy.
It is actually a problem that’s encountered even with 8434 serial-to-USB adapters occasionally but certainly the incidence of the issue is lower than with the CP210x chipset of 6510USB. So it’s more a question of degree than an absolute distinction. I believe that Davis might have been looking at alternatives to that USB chipset a year or two back when the noise problem was first recognised, but that every option they looked at still had some degree of vulnerability to the same problem. The USBXpress option does seem to be somewhat more resistant to the problem than the VCP (Virtual COM Port) approach and I guess it was felt that USBX was just as good an option as completely re-engineering the logger.
There’s no doubt that this is a complicated problem, since noise/RFI can take many forms - through-space (ie spurious wireless transmissions) as well as through the mains supply. Adding extra ferrite cores is probably the single simplest measure that’s likely to have any benefit but won’t resolve every situation. Using a good UPS on the mains supply may help. But often trying to identify and remove the source of the interference can be quite effective too.
as above, I do not have that problem so far. I think one of the issues on the console is that the AC power cord is so darned close to the data cable, it is thin enough not to have adequate shielding as doesthe brief space between the data logger usb connection and the data cable connection. Definitely both need adequate shielding. I am not sure why Davis doesn’t connect the long cable right to the datalogger rather than have a connector as it does.
I did some looking around about the random crashes with Vista … it is something adherent to Vista’s architecture that Microsoft needs to sort out. I see an error that something was trying to shut down and the system froze. Not much more information in the data logs what was because a hard reboot is needed and that stuff doesn’t get saved.
the problem with davis hinting that usbexpress is a fix is that only weatherlink works with that
(and the only help given by davis for how to work with usbexpress was a url to a sdk from the manufacturer…but no examples…basicly flogged off)
The first thing I did was swapped the USB cable with a “premium” one when I had a problem right after installation. Now it’s been running for over two years without a problem. However this doesn’t work for everyone.
Brian, I really wouldn’t take it too personally - the programmers at Davis always seem to be hard to contact - by anyone. Davis isn’t Microsoft - they only have fairly limited programming resource and AIUI they’re under a lot of pressure at present with several projects and deadlines all fighting for attention (just like any other smaller business really). Work is tough for everyone nowadays and maybe time spent on goodwill, but which doesn’t help towards meeting corporate objectives, is being squeezed more now.
And USBX is only one approach to mitigating the USB dropout issue. As has often been said previously and as you’ve highlighted again, attention to eliminating or suppressing RFI whether mains-borne or through-space seemingly remains the most effective answer.
I have often felt that their programmers have to do their homework, get their braces on their teeth adjusted and in bed by 8 pm on a school night. They probably home that Sally Sue will take them to the prom.
In contrast, Windy, has a lot to do with his life, his kids, his job, and avoiding having to weed the garden, yet he turns out a lot in a little time frame. Davis just makes a nice instrument,their software hasn’t changed much over time. I do like their direct update for the VP2 when there needs to be something changed on the console.
However… I do use both and have sent them suggestions regarding Weatherlink and have always been told that they pass it on to the programmer. I think that person or persons are not a full time and hence the high school geekie kid type.
I am just setting up my new Davis and had trouble of disconnection emailed them and got the reply that is at the top of this page
Tried to buy a couple of clamps but unable to, but a couple of PC shops gave me some clamps from some cables that they where throwing away seems to have sorted the problem .Changing batteries would have been a hassle as I am using the cabled Davis
you can uncross the fingers
I recommend people to get a serial data logger and then a good quality USB to serial convertor if not serial port (I have always recommended that) (or a PCI serial card even)
OK, here is a cheap solution which works fine for me and I don’t have any crashes anymore.
You see the USB connection close to the station. This is the weak point. Davis shouldn’t split the cable but should have one cable going straight from the WS to the PC. I guess their designers were asleep at Uni while the subject was “Interferences”.
All you have to do is wrap the connection with Aluminium foil.
Also pull the power cable apart from the USB cable and don’t let them run too close side by side from the WS to the PC. Don’t wrap the power cable around the little pins inside the station as Davis recommends it.
PS: Davis, if you need a hand with your stations and need some more tips and design solutions, you can contact me (and in return send my money back for the station I bought).
PPS: An additional component (eg. a Daytime Fan Aspirated Radiation Shield Kit or an UV-sensor) would do too.
I recently bought a Davis VP & have experienced the same lock up problems. I have followed all the advice on this forum. Ferrites on the premium USB cable & the AC adaptor. I have also installed the latest drivers for the UART from Silicon Labs. I did notice that my cheap & nasty computer speakers locked the data logger within minutes so they have now gone in the bin. Everything is a lot better now but it still locks up after about 36 hours. The computer I am running is a two year old Dell Dimension with windows XP. Has anybody got any more ideas?
Rumour has it that Brian is going to come out with a version of WD that supports Davis USBXpress. Is this true & if so when?