I used the logging function for real for the first time last week. It works well, although the software has now started to crash occasionally under XP.
The crash is not a big deal, because it only occurs when you begin logging. In other words, if it starts logging, it's fine. If not, you know before you lose any data.
This would be a total pill if you were waiting for an event before you started logging, because it might crash as soon as you press "Start". But you could always start logging ahead of the event.
The only other problem is that the software does run out of buffer at about 45 hours, max. If you want to do weeks of logging - like I did - you have to save and reset the buffer every 2 days tops or lose data.
Oddly enough, what little I have found on the MS8226T meter indicates it does not have this problem. It still has the RS-232/USB interface - same style as the MS8218. In fact, aside from the 50,000 count display, I'm hard-pressed to find any real difference between them.
The 8226 does have a temperature function, and even comes with a thermocouple. This is good, as I was really looking for a temperature meter with a peak hold, because I'm doing a lot of temperature monitoring that I can't always keep a close eye on right now. I can get the VC99 for about $34, or the 8226 for about $60.
Unfortunately, the adapter looks like it's permanently attached to the thermocouple, so it may only accept special thermocouples. I would have to get one to see if the adapters I have gotten from other meters could be used with an 8226. The VC99 accepts a standard bladed thermocouple native.
(I know that Dave over at EEVblog did not like the VC99. Everything he said is spot-on correct, I'm sure. I love Dave, he's the best damn reviewer out there. They probably are cheap, somewhat poorly built, and have sucky battery life.
But for what I want, it seems the best option. Not many cheap meters have temperature, peak hold, and the ability to take a standard K-type thermocouple with no accessory adapter. Dave covers nearly everything, but not the presence or absence of min/max/peak, and didn't give the VC99 any credit for being the only one to take a standard thermocouple.)
It's also possible the "new"? software from an 8226 would operate the 8218. In which case I could "upgrade" my 8218 to seconds/minutes/hours logging from the current seconds-only software. I may give it a go.
I could always buy both the VC99s and the 8218s. I may just do that. If I was on the road all the time, in all kinds of weather conditions, I'd want 1 excellent meter. If the meter is getting frozen or heated, or bashed around, and I can't get another one if it breaks, by all means.
But in my lab, I'd rather have 14 cheap meters and one good meter than one excellent meter.
Why? 2 DMMs are tied up right now with battery tests. 3 are presently doing temperature tests for me. I could use one more for ambient temperature monitoring, so that's 4 just for that test, or 6 in total. That's nearly all I have on hand right now.
If I had more loggers, I could be doing more battery tests, or do more data collection per test than I can do now.
Plus I do not mind having meters for different reasons. Three VC99s would be nice for temperature logging, as they accept standard thermocouples, but can also do other things. Two 8226s for data logging would probably cover logging, as well as filling in for temperature and other things. One 50000 count 8218 for current logging or whatever is probably fine.
Plus they're so cheap - why not? As I argued in my previous post, they are plenty accurate. Especially when you consider they are never used outside of room temperature, or at least not very much.
For temperature accuracy, +/-1*C is not going to change anything at all. Voltage and resistance are dead easy. Current is hardly any more difficult until you get to the microampere range, in which case having one pretty good meter is covering that. And so on.
If I do invest in the 8226 or VC99, I'll try to remember to post what I find.
Saturday, April 28, 2012
Wednesday, April 18, 2012
Adobe Acrobat "These characters cannot be used in a file name"
You type in a path and file name into Acrobat, you get this:
Apparently, this is because Acrobat is too "stupid" to append the ".pdf" to your information.
For example, if you type "C:\Temp\MyPDF", you get the error.
If you type "C:\Temp\MyPDF.pdf" instead, no error.
Am I the only one who thinks this is stupid?
Apparently, this is because Acrobat is too "stupid" to append the ".pdf" to your information.
For example, if you type "C:\Temp\MyPDF", you get the error.
If you type "C:\Temp\MyPDF.pdf" instead, no error.
Am I the only one who thinks this is stupid?
Saturday, April 14, 2012
Unexpected and sudden "Android System shutting down"
Recently my Xperia X10i has started "crashing" and shutting down when attempting to take a picture or video.
The phone was working fine, and I don't recall doing any significant updates. It's always possible that an application updated in the background, though.
Still, the phone was working fine for quite a while, and for at least a month since my last application was installed.
First, the phone started to shut down whenever I tried to take a video using the built-in camera application. I chalked this up to some kind of bug between the camera app and CameraZoomFX, which I had tried but no longer used.
Next, it started to crash and shut down when taking a photo.
One time it began "crashing" and shutting down over and over. I began to restart it, over and over, and it would shut down over and over before I had a chance to see what was going on.
After four or five attempts, I noticed the battery indicator had dropped from 40% to less than 10% in those few restarts. I assumed the "crash" had left something active that should not have been.
After a bit I got a little suspicious of the crash theory. I uninstalled CameraZoomFX, but it did not help. Also, the unexpected shutdowns always seemed to occur when the battery was less than 50%.
Just recently the phone shut itself down without me trying to take a picture. The battery indicator was at 32%.
I realized then what was really going on - my battery is worn out. Plugging the unit in, I could not get it to shut down taking a picture or a video. The low-battery indicator was giving me bogus information.
I've got a new battery on order, we'll see if that fixes the problems.
So - if you start getting sudden and unexpected Android system shutdowns of your device, be sure to consider the battery. Given how these devices are used, 2 years for the battery will probably be a good lifetime.
The phone was working fine, and I don't recall doing any significant updates. It's always possible that an application updated in the background, though.
Still, the phone was working fine for quite a while, and for at least a month since my last application was installed.
First, the phone started to shut down whenever I tried to take a video using the built-in camera application. I chalked this up to some kind of bug between the camera app and CameraZoomFX, which I had tried but no longer used.
Next, it started to crash and shut down when taking a photo.
One time it began "crashing" and shutting down over and over. I began to restart it, over and over, and it would shut down over and over before I had a chance to see what was going on.
After four or five attempts, I noticed the battery indicator had dropped from 40% to less than 10% in those few restarts. I assumed the "crash" had left something active that should not have been.
After a bit I got a little suspicious of the crash theory. I uninstalled CameraZoomFX, but it did not help. Also, the unexpected shutdowns always seemed to occur when the battery was less than 50%.
Just recently the phone shut itself down without me trying to take a picture. The battery indicator was at 32%.
I realized then what was really going on - my battery is worn out. Plugging the unit in, I could not get it to shut down taking a picture or a video. The low-battery indicator was giving me bogus information.
I've got a new battery on order, we'll see if that fixes the problems.
So - if you start getting sudden and unexpected Android system shutdowns of your device, be sure to consider the battery. Given how these devices are used, 2 years for the battery will probably be a good lifetime.
Tuesday, April 3, 2012
Sudden and unfixable "Connection was reset" error message when accessing 3ware 3DM2
If 3DM2 suddenly stops working in your browser it is probably because your browser has recently been updated.
LSI (the company that bought 3ware) has information on this in this rather cryptically titled knowledge base article.
It seems from my experience that Firefox 11 and IE 9 are now both unable to access the 3DM2 web GUI.
[Update]: This is a fundamental problem with the SSL and TLS protocols. The browser updates fix the flaw, but the clients (i.e. 3DM) must be updated to be compatible.
There may be a fix described here in the Mozilla forums, but I have not investigated it yet. It seems LSI has updated the 3DM executable to be compatible, but you have to install it manually, which should hardly be a big deal.
[Update #2 (2015)]: LSI KB article describing fix.
I haven't done it yet, but it >seems< you do the following:
1. Stop the 3dm2 service (Windows) or daemon (Linux/FreeBSD).
Tip: /etc/init.d/tdm2 stop (to stop the 3DM2 daemon)
2. Copy the attached binary and overwrite the current binary.
Tip: 3dm2 lives in /usr/sbin
3. Restart the service/daemon (or reboot the system).
Tip: Command line: 3dm2
or /etc/init.d/tdm2 start (to start the 3DM2 daemon)
I haven't done it yet! (I'm too busy.) Use at your own risk!
LSI (the company that bought 3ware) has information on this in this rather cryptically titled knowledge base article.
It seems from my experience that Firefox 11 and IE 9 are now both unable to access the 3DM2 web GUI.
[Update]: This is a fundamental problem with the SSL and TLS protocols. The browser updates fix the flaw, but the clients (i.e. 3DM) must be updated to be compatible.
There may be a fix described here in the Mozilla forums, but I have not investigated it yet. It seems LSI has updated the 3DM executable to be compatible, but you have to install it manually, which should hardly be a big deal.
[Update #2 (2015)]: LSI KB article describing fix.
I haven't done it yet, but it >seems< you do the following:
1. Stop the 3dm2 service (Windows) or daemon (Linux/FreeBSD).
Tip: /etc/init.d/tdm2 stop (to stop the 3DM2 daemon)
2. Copy the attached binary and overwrite the current binary.
Tip: 3dm2 lives in /usr/sbin
3. Restart the service/daemon (or reboot the system).
Tip: Command line: 3dm2
or /etc/init.d/tdm2 start (to start the 3DM2 daemon)
I haven't done it yet! (I'm too busy.) Use at your own risk!
Ergotron LX Triple Monitor stand fits three Acer v223W 22" widescreens
As the title says. They WILL fit, with an inch or so per side to spare.
OBVIOUSLY, you will not have any clearance to rotate the monitors from portrait to landscape. I can't imagine this will deter too many people.
With the Acer's going for about $135 each, this is a damn good package with tons of LCD real estate.
OBVIOUSLY, you will not have any clearance to rotate the monitors from portrait to landscape. I can't imagine this will deter too many people.
With the Acer's going for about $135 each, this is a damn good package with tons of LCD real estate.
Monday, April 2, 2012
Setting up HTTP and FTP servers behind an Asus RT-N56U wireless router
I've been using the same router in my system since - gasp - 2004! It was a Xincom, and it's big claim to fame back then was it supported dual WAN inputs.
This was a big deal when internet bandwidth was hard to come by. The easiest way of getting more bandwidth was to sign up with more than one ISP. I did just that, and bought hardware to match.
However, I dropped the dual internet connections a long time ago, and pretty much ignored the poor thing until it died last night. It's been a fantastic piece of kit, having outlasted nearly every other piece of hardware I've ever bought. Including the second Linksys WRT54G that I had for both wireless and backup purposes, which also turned out to be busted.
The failure was weird, too - it got Alzheimers. Stopped remembering it's own password, as well as all of the other settings, and usually (but not always) forgot to answer the door when you came round to knock. I suspect the EEPROM/Flash memory finally wore out.
Router benchmarks are hard to come by. Thank God for the SmallNetBuilder Router Charts. I would certainly have picked the 'wrong' router if not for these little babies.
("Wrong" being a relative term, of course. Once the Xincom finally kicked it, I quickly realized that it was probably bottlenecking my whole internet experience. It did not do LAN routing - I have dedicated switches for that - but I can't deny my internet connection should be faster.)
Anyway, ended up buying the RT-N56U. Nice unit. But it has some irritating quirks in the setup.
First is the web server. It has it's own built-in web server, theoretically to allow you to remotely configure it over the internet.
But all you get in the configuration screen is a poorly worded warning about potential conflicts between an external server and the internal server:
"When your network's firewall is disabled and you set 80 as the HTTP server's port range for your WAN setup, then your http server/web server would be in conflict with RT-N56U's web user interface."
Well, ain't that helpful.
The FTP service has the same issue:
"When you set 20:21 as your FTP server's port range for your WAN setup, then your FTP server would be in conflict with RT-N56U's native FTP server."
Googling this problem only showed a few forum posts that look like they were written in Punjabi, run through Babelfish to Korean, then translated again into English. Barely comprehensible and damn near useless.
The solution to these problems is nothing like what you expect from the warning messages, and a damn sight easier than it first appears. But it did take me a while to figure it out.
First, go to Firewall / General and set "Enable Web Access from WAN?" to NO. This should disable the internal service that attempts to hijack port 80 from you.
Next, go to WAN / Port Forwarding and set up forwarding for port 80:
Your web server should now be 'live', provided your DNS or dynamic DNS service is working properly. If not, fix up your dynamic DNS first.
(Note: the RT-N56U has "dyndns.org" and not "dnydns.com". .org is correct, and should work.)
Next up is the FTP server. Apparently this one is another internal service that is supposed to let you share a big hard drive connected directly to the RT-N56U USB port, without any other fuss. Unfortunately, the internal service grabs port 21 away from you as well.
To fix, go to USB Application / FTP Share. You will see three buttons near the top of the screen.
Annoyingly, they will initially appear as grayed out and inactive, but will suddenly come to life once you wave your mouse pointer over them. This can be really confusing.
Press the left-most button that says "Disable FTP". The notification area to the right should say "Share via the FTP is disabled". This should free up port 21.
Again go to WAN / Port Forwarding and set up forwarding for port 21:
Note that you do not have to touch the router firewall on/off setting for either of these. The firewall can stay on.
Of course, you still need an external connection to be able to properly test these. You can at least test the port forwarding using Dyn's Open Port Tool or some other similar service. It's really quite handy for nailing down router mis-configurations, such as putting in the wrong LAN address or something.
Personally, I use my Android phone to do final tests of my internal systems using the cellular 3G data network. It works fine and is easily fast enough for a final check or routine test.
I've only had the RT-N56U for a few hours, so far so good. The benchmarks put it into the top 3 for every test, so I'm hoping it's going to live up to its promise. It smokes the oft-quoted WNDR3700, which is now falling behind, and should beat the newer WNDR4200 as well.
Who knows if the wireless performance is really better. But the wired performance is really why I have it, and in wired is wipes the floor with everything else out there.
The only thing I wish is that it was metal. I do not like these cheapo plastic-wrapped consumer devices any more. In my experience, they tend to run too hot.
This is only really a problem for me - my office is small, not well ventilated, has a lot of sun exposure, and is packed to the gills with several PCs and monitors that can get the interior temperatures up into the 30s. This has killed many a gadget in here.
I have upgraded to industrial-grade redundant network switches and even an industrial grade USB hub by Moxa - pricy, but so worth it. I have cooling fans on the switches, and my PCs are all packed with as many vents and fans as I can get in them. But the monitors and older PCs do pump out some joules, and my monitors, PSUs, UPSs and other equipment all end up dying sooner or later.
From this, I would feel much better if the RT-N56U had a metal back or case. The metal would dissipate the heat better, and be easier to cool if I needed to. I can't help but wonder if the Xincom lasted so long because it has a metal case, while the plastic WRT54G died twice in the same amount of time.
Who knows, I may strap a fan onto the back of the Asus to help cool things off a bit. Even just sitting there it gets warm. There are vents on the top (good) but none on the front, back, sides or bottom (stupid).
Still, for $170 or so, when it dies I'll just get another one. Not everything is going to last as long as that Xincom.
This was a big deal when internet bandwidth was hard to come by. The easiest way of getting more bandwidth was to sign up with more than one ISP. I did just that, and bought hardware to match.
However, I dropped the dual internet connections a long time ago, and pretty much ignored the poor thing until it died last night. It's been a fantastic piece of kit, having outlasted nearly every other piece of hardware I've ever bought. Including the second Linksys WRT54G that I had for both wireless and backup purposes, which also turned out to be busted.
The failure was weird, too - it got Alzheimers. Stopped remembering it's own password, as well as all of the other settings, and usually (but not always) forgot to answer the door when you came round to knock. I suspect the EEPROM/Flash memory finally wore out.
Router benchmarks are hard to come by. Thank God for the SmallNetBuilder Router Charts. I would certainly have picked the 'wrong' router if not for these little babies.
("Wrong" being a relative term, of course. Once the Xincom finally kicked it, I quickly realized that it was probably bottlenecking my whole internet experience. It did not do LAN routing - I have dedicated switches for that - but I can't deny my internet connection should be faster.)
Anyway, ended up buying the RT-N56U. Nice unit. But it has some irritating quirks in the setup.
First is the web server. It has it's own built-in web server, theoretically to allow you to remotely configure it over the internet.
But all you get in the configuration screen is a poorly worded warning about potential conflicts between an external server and the internal server:
"When your network's firewall is disabled and you set 80 as the HTTP server's port range for your WAN setup, then your http server/web server would be in conflict with RT-N56U's web user interface."
Well, ain't that helpful.
The FTP service has the same issue:
"When you set 20:21 as your FTP server's port range for your WAN setup, then your FTP server would be in conflict with RT-N56U's native FTP server."
Googling this problem only showed a few forum posts that look like they were written in Punjabi, run through Babelfish to Korean, then translated again into English. Barely comprehensible and damn near useless.
The solution to these problems is nothing like what you expect from the warning messages, and a damn sight easier than it first appears. But it did take me a while to figure it out.
First, go to Firewall / General and set "Enable Web Access from WAN?" to NO. This should disable the internal service that attempts to hijack port 80 from you.
Next, go to WAN / Port Forwarding and set up forwarding for port 80:
- Service Name = anything you like.
- Port Range = 80
- Local IP = the static IP of your web server machine
- Local Port = 80
- Protocol = TCP
- Protocol No. = blank / no entry
Your web server should now be 'live', provided your DNS or dynamic DNS service is working properly. If not, fix up your dynamic DNS first.
(Note: the RT-N56U has "dyndns.org" and not "dnydns.com". .org is correct, and should work.)
Next up is the FTP server. Apparently this one is another internal service that is supposed to let you share a big hard drive connected directly to the RT-N56U USB port, without any other fuss. Unfortunately, the internal service grabs port 21 away from you as well.
To fix, go to USB Application / FTP Share. You will see three buttons near the top of the screen.
Annoyingly, they will initially appear as grayed out and inactive, but will suddenly come to life once you wave your mouse pointer over them. This can be really confusing.
Press the left-most button that says "Disable FTP". The notification area to the right should say "Share via the FTP is disabled". This should free up port 21.
Again go to WAN / Port Forwarding and set up forwarding for port 21:
- Service Name = anything you like.
- Port Range =21
- Local IP = the static IP of your FTP server
- Local Port = 21
- Protocol = TCP
- Protocol No. = blank / no entry
Note that you do not have to touch the router firewall on/off setting for either of these. The firewall can stay on.
Of course, you still need an external connection to be able to properly test these. You can at least test the port forwarding using Dyn's Open Port Tool or some other similar service. It's really quite handy for nailing down router mis-configurations, such as putting in the wrong LAN address or something.
Personally, I use my Android phone to do final tests of my internal systems using the cellular 3G data network. It works fine and is easily fast enough for a final check or routine test.
I've only had the RT-N56U for a few hours, so far so good. The benchmarks put it into the top 3 for every test, so I'm hoping it's going to live up to its promise. It smokes the oft-quoted WNDR3700, which is now falling behind, and should beat the newer WNDR4200 as well.
Who knows if the wireless performance is really better. But the wired performance is really why I have it, and in wired is wipes the floor with everything else out there.
The only thing I wish is that it was metal. I do not like these cheapo plastic-wrapped consumer devices any more. In my experience, they tend to run too hot.
This is only really a problem for me - my office is small, not well ventilated, has a lot of sun exposure, and is packed to the gills with several PCs and monitors that can get the interior temperatures up into the 30s. This has killed many a gadget in here.
I have upgraded to industrial-grade redundant network switches and even an industrial grade USB hub by Moxa - pricy, but so worth it. I have cooling fans on the switches, and my PCs are all packed with as many vents and fans as I can get in them. But the monitors and older PCs do pump out some joules, and my monitors, PSUs, UPSs and other equipment all end up dying sooner or later.
From this, I would feel much better if the RT-N56U had a metal back or case. The metal would dissipate the heat better, and be easier to cool if I needed to. I can't help but wonder if the Xincom lasted so long because it has a metal case, while the plastic WRT54G died twice in the same amount of time.
Who knows, I may strap a fan onto the back of the Asus to help cool things off a bit. Even just sitting there it gets warm. There are vents on the top (good) but none on the front, back, sides or bottom (stupid).
Still, for $170 or so, when it dies I'll just get another one. Not everything is going to last as long as that Xincom.
Subscribe to:
Posts (Atom)