-
GPIB FUN
Oh, RESULT !!!
Look at that !
The TEKTOOL program looks like it's working now !!!
Now, it's about getting the ISA GPIB card working... luckily since Win95 is still on the drive along W2K, I can access the MS-DOS config / startup files to see what I had to put in there 5 years ago to get the driver to load properly.... hopefully that driver will work on W2K as well !!
Stay tuned...
-
Ok, we can accept some rounding. And the latter should be easily fulfilled by aquisition of some quality crimp tools!20% of any Bernstein, Erem or Lindstrom catalogue content.
25% of any Wiha, Wera, Knipex, Xcelite, PB or OK/Jonard catalogue content
as well as 10 hand tools not included in the above which represent a new value of 100$ or € each.
I don't know about the exact percentages, but I definitely have Lindstrom, Wiha, Wera, Knipex, and Xcelite stuff in my toolbox(es). Not sure I have anything that retails for more than $100 each though. Which is probably a Good Thing.
I have a Greenlee crimper now, that ought to count for something!
-
Call to my neighbor report
First of all, he meant to say 1.6 GHz. He said the "investigators" were using a spectrum analyzer in the field. One of them was a "physicist". They were looking for any strange signals near this Skinwalker ranch because that's just one aspect of their investigation. They noticed "something" at 1.6 GHz that looked "noisy". But they also had seen a radio station signal at lower frequencies that looked like that signal. They got hold of the station and asked them to stop transmitting music. Who can ask radio stations to stop transmitting? Then they saw the "noise" go away around the signal (carrier was left). I think we would call that side bands.
He said they had recorded audio of the "noise" that the 1.6 GHz signal made and they asked the radio station to play their recording so they could see if the signal looked like the one they saw on 1.6 GHz.
Don't ask me I didn't watch the show!
Supposedly when the 1.6 GHz signal was present they observed "mysterious objects" in the air. Yea uh huh. Probably birds. The noise could be produced simply by the 1.6 GHz signal being a data link with modulation of several different types used for data. Duh. IMHO I wouldn't let these people near anything more complicated than a battery tester.
There isn't anything to it because there is a military radio link in the area on that freq. So ... much ado about nothing. Guess they produced an hour long show out of it and generated advertising revenue.
Just google 1.6 GHz Skinwalker ranch ... it's all BS.
What did he mean by me creating a sound that can be seen on a scope? He wants to observe a signal like they did that is transmitting sound, so he can see it for himself. I said no problem, come over tomorrow and we'll set my spectrum analyzer to the AM or FM broadcast band. Knock yourself out LOL. -
The little heated wire anemometer arrived and I took a look at it, as I would need it soon. I also contacted Testo, which took two days to tell me that they do not have a manual of this instrument anymore. They enclosed the very same picture of the rear side (with short instructions) which was included in the auction as 'the only help they could give me'!
As one could guess from the photo, there was something wrong with the pointer. No matter whether I shook or tilted it, the pointer did not bodge. Ok, a nearly complete disassembly along some rather un-intuitive lines did show that inner side of the little rotating plastic part doing the zero adjustment was crumbled and did jam the movement. After removal, the pointer behaves normal and can be adjusted to zero.
Partially reassembled, keeping functional testing for the weekend. No other visible damages found. There is also something blocking the telescopic action of the probe, which I need to look at.
-
Call to my neighbor report
And if they were using just one SA to observe a 1.6GHz signal, how would they know it wasn't just a spur.....if you look deep enough all spectrum analyzers have them.
First of all, he meant to say 1.6 GHz. He said the "investigators" were using a spectrum analyzer in the field. One of them was a "physicist". They were looking for any strange signals near this Skinwalker ranch because that's just one aspect of their investigation. They noticed "something" at 1.6 GHz that looked "noisy". But they also had seen a radio station signal at lower frequencies that looked like that signal. They got hold of the station and asked them to stop transmitting music. Who can ask radio stations to stop transmitting? Then they saw the "noise" go away around the signal (carrier was left). I think we would call that side bands.
He said they had recorded audio of the "noise" that the 1.6 GHz signal made and they asked the radio station to play their recording so they could see if the signal looked like the one they saw on 1.6 GHz.
Don't ask me I didn't watch the show!
Supposedly when the 1.6 GHz signal was present they observed "mysterious objects" in the air. Yea uh huh. Probably birds. The noise could be produced simply by the 1.6 GHz signal being a data link with modulation of several different types used for data. Duh. IMHO I wouldn't let these people near anything more complicated than a battery tester.
There isn't anything to it because there is a military radio link in the area on that freq. So ... much ado about nothing. Guess they produced an hour long show out of it and generated advertising revenue.
Just google 1.6 GHz Skinwalker ranch ... it's all BS.
What did he mean by me creating a sound that can be seen on a scope? He wants to observe a signal like they did that is transmitting sound, so he can see it for himself. I said no problem, come over tomorrow and we'll set my spectrum analyzer to the AM or FM broadcast band. Knock yourself out LOL.
-
GPIB FUN
OK, spent some time looking at the GPIB ZIP file san dPDF documentation I had gathered back then. Evnetually figured out again what I had to do to get it to work.
In short : does not work. All the utiliy programs do load and execute fine, they can run, but they can't find the GPIB card.
Triple checked all the jumpers and dip switches against the NI PDF documentation for that board.. it's basically set PC2 not PC2A (still can't figure out what's the difference between the two modes... other than PC2A allowing you to use multiple adpaters in the same computer....). and all settings are at their default basically : 2B8, IRQ7 and DMA 1.
H/W confiug utility can't detect the board using these parameters.
Most likely cause is that the driver doesn't work.
It's loaded in CONFIG.SYS :
" device = \GPIB-PC\GPIB.COM "
... but there is NO config.sys in W2K.... so I created one and added that required line to load the driver. No joy.
I don't even know if W2K can even load MS-DOS drivers and how to do it, to begin with !
Maybe its MS-DOS compatibility is limited to running programs that don't need any drivers.... I don't know.
But if you DO know, make sure to speak up, because unless I know how to use that driver on W2K, I have now hit a road block, it's game over.
Sorry Terra... I did my best
-
And if they were using just one SA to observe a 1.6GHz signal, how would they know it wasn't just a spur.....if you look deep enough all spectrum analyzers have them.
How would they know? Because they had a physicist!
-
Hmmm... DigiKey sez it's a part #CR-2354/GUN; a whopping $2.11 per. Dammit, now I need to make a minimum order for free shipping...
mnem
Ughhhhh... DigiKey isme again... my account is locked to the Canadian site... both accounts. Why the fuck did they force me to make a new account for the Canadian site if they were going to do that...?
... I'm gonna go look in my crap for a for a CR2032 holder. If I can make it fit, it's going in there.
mnem
one step forward, two steps back...
-
GPIB FUN
OK I googled a bit and bad news, as I feared it's a dead end. IT will not work.
1) Found a document describing the W2K Professional architecture, a document for nerds. It said that W2K, for security reasons, will not allow MSDOS drivers to run, so they can't get hold directly of the H/W and potentially do nasty things. Only exceptions it's willing to make is common drivers like Keyboard and mice. Anything else is meant to not work.
2) found this post on NI's own support forum, where a guy asked just what I am asking... has an old ISA NI car that he wants to use on a W2K machine.... the guy from NI replied to him : " Sorry it just won't work, unsupported. W2K is 32 bits, won't handle old 16 bits drivers like this old car uses. Have you considered our newest products instead ?!.... "
https://forums.ni.com/t5/Instrument-Control-GPIB-Serial/pcii-gpib-card-with-windows2000/td-p/253178
So... sorry Terra. All that messing around for nothing.
Maybe if I had a newer NI card, that uses PCI instead of ISA, maybe those cards can use a more modern 32 bits driver that could run on W2K. More expense... would rather wait to be able to afford the USB NI adapter and use that instead, since W2K managed to get USB working on that machine...
Anyway, it looks totally hopeless for my old card and machine, so I guess it's time to clean the bench of all that mess ?!
It's sad.. all that work for nothing, kills me.
Well at least I learned interesting stuff for my future local network which I want to implement when I computerize the lab. So it's was not all wasted after all....
Oh yes, I need to try Mansaxel's instruction on getting FTP to work, to share files directly between the two computers... want to see if I can get that to work, that would be really cool...
-
Oh no, mansaxel said Win95 has everything that's needed out of the box... how do I do then ?
As I wrote earlier, yes, that is true. But you have to put on your 90s head to do it.
On the Linux box, install or enable a FTP server. (google "enable ftp server ubuntu" or whatever distribution you have) Make a note ("ifconfig") about the IP address of that box.
On the windows box, open a DOS prompt, and start going back in time:
c:\> cd c:/vince/gpib/dumps/ <- where you put the file
c:\> ftp 192.0.2.44 <- swap in IP address of your Linux box (that address is a RFC 5737 example address)
Login with your user and password of the Linux machine.
This will go in CLEAR TEXT over the LAN. That's why I was adamant on it being almost OK on only that LAN and nowhere else.
For classic ftp daemons, as long as you're approved. you'll now be in your home directory on the target machine. Check with "ls":
ftp> ls
229 Extended Passive Mode OK (|||45379|)
150 Opening ASCII mode data connection for '/bin/ls'.
....
You should see the same as "ls" on your linux box shows, in that folder.
On the target, you might want to do a new folder:
ftp> mkdir gpib
257 "gpib" directory created.
ftp> cd gpib
250 CWD command successful.
Now you're in that new folder. Time for file transfer.
The memory dump is binary, right? Not a text file?
Then, put the file transfer in binary mode.
ftp> bin
200 Type set to I.
Transfer the file (I'm assuming it's called vincescp.dmp):
ftp> put vincescp.dmp
local: vincescp.dmp remote: vincescp.dmp
229 Extended Passive Mode OK (|||38311|)
150 Opening BINARY mode data connection for 'vincescp.dmp'.
100% |********************************************************************************************************************************************| 1172 KiB 40.36 MiB/s 00:00 ETA
226 Transfer complete.
1200989 bytes sent in 00:00 (11.53 MiB/s)
ftp>
Done! Don't forget to be polite:
ftp> bye
221 Goodbye.
And you have the file on the usable computer. Easy as Tarte Tatin!
Thanks !
I just tried that, works just fine !
Was indeed not that difficult, thanks to your instructions !
I googled a bit, looked like " vsftpd " was a popular FTP server on Linux, so I installed that.
I failed to connect to the old Windows machine though... maybe because W2K doesn't enable the FTP server by default, maybe I need to click somewhere to enable it...
However the client was already installed and I could "ftp" to my Linux from the command line just fine !
I could see my Home directory, I created a new directory in there, then copied a file into that directory. So easy indeed !
I like this stuff a lot, really cool !
So my future local network might not be that hard to make then, after all !
EDIT : found this web site that explains how to enable the FTP server on W2K..
https://www.techrepublic.com/article/get-it-done-setting-up-an-ftp-server-with-windows-2000/
it was indeed not enabled by default. Followed the instructions to create an FTP "site" as W2K calls it... appears to work, locally at least, if I type " ftp://localhost " in the file explorer, W2K can see itself. It also shows there, that the server is set to "anonymous".
So then I moved to my Linux box and try to connect to W2K. See screenshot... the Linux box, using the IP address of the W2K box, can "connect", I mean it can identify the machine, because it prints its name, and knows it's a Microsoft FTP server... so that's good. I know they can talk to each other. However despite being set to anonymous, I am still being asked for a password... I type it anyway ( "admin" ...), but it fails to login with that, crap.... I tried giving no password at all, since it's supposed to be anonymous, but still doesn't work.
Oh well..... I will call the customer support !
-
Ahhhh, the good old days...
When real men were SysOps and our password was the ass word...
mnem
yes, you may touch it... -
The Facom R1Pico very compact ratchet set has won the heart and desire of everyone who's laid hands on it. If you have an addiction problem, don't look at it. If you have a sensible interest in high-quality tools, by all means.
(The fact that Stanley has bought Facom makes me nervous. I think that Stanley should stick to wood planes and breaker bars, and not be buying up fine tool companies and brand-rape them. Like Crap-On with Bahco. (And Bacho with Sandvik and Belzer, someone from Germany is going to complain, pretty soon... )
Facom is one brand that I currently own none of. Along with PB Swiss and Ko-ken. Great, more opportunities to feed my HTAS (Hand Tool Acquisition Syndrome). Then again, I also own nothing made by Mac, Matco, or Proto either.
I do own exactly one Snap-On tool though. A Snap-On F50CSA driver handle. It's a 3/8" drive driver handle with the socket cap in the head so it can be used as an extension. I bought one off the Snap-On truck back in like 1994 or so and it's been my favorite hand tool ever since. Sadly they don't make this model anymore, so if I ever lose mine then I'm stuffed. Well, maybe you could find a used one on Ebay or whatever, but it would probably cost out the yazoo.- you have at least 3 different cross-slot screwdriver types (Philips, Pozidrive, JIS, R&P)
This actually got me thinking (and you know that's a dangerous thing) - I do not currently own a Reed & Prince screwdriver. So I thought I'd look and see if there are actually any still being made, and how hard it would be to acquire one. And... wow. I figured there'd be at least a few on Ebay and probably plenty of R&P hex-bits on even Amazon or whatever.
No.
I found one R&P screwdriver - an ancient, crusty looking Craftsman model - and that was it for actual screwdrivers. R&P hex bits appear to be readily available, but not to the extent of being listed on Amazon. You can find them from McMaster Carr and some specialty tool sites.
I'm maybe a tiny bit surprised by that. Hmm... I think I heard once that R&P was particularly popular in the marine / nautical world at one time, so maybe some specialty shops in that space would have R&P screwdrivers still?
Anyway.. not that I actually need one or anything. Just found it interesting to look into that, because... well, why not?
EDIT:
Looks like Snap On also sell a R&P hex bit but not a traditional screwdriver in R&P.
-
No worries Vince, a fun adventure was had and now we know how not to do it.
Win NT can be a right bitch for drivers and such that work on a 'bare metal' level.. I have an old parallel port EPROM programmer that won't work on any OS beyond Win98SE due to Windows blocking low level access to the port (because we all know how dangerous it is to let hackers have access to that port....
I'll have to see if I can figure out a way to pull the firmware without the tektools application. Once I get the PSU for my PC working......
Speaking of which, anyone got a spare TSM110CN or TSM111CN 'Triple Voltage and Current Supervisor) chip laying around?
https://www.st.com/resource/en/datasheet/cd00001698.pdf
I've tested almost everything else, and it seems the Current Mode Controller chip (uC3845B) that switches the main switching transistor isn't getting a signal from the TSM110CN, but the optocouplers seem to test ok.
It looks like Digikey have the TSM110CN chips in stock, but I need to make a $50 order to avoid the $20 shipping charge.......
No-go in Akihabara too, the chip is too esoteric, unless I find a cheap PSU to salvage parts from.. -
@Vince
I do have some not so high endian audio stuff, too.
And I cannot go past a decent amp that someone is throwing out for a tenner.
I am in the fortunate position that I did get my high end stuff mostly in trade for PC stuff that I no longer need which is in turn used for astronomical image processing and thus needs to be reasonably powerful.
Which in turn means that I got my Threshold overthetop power amp as trade for a PC. The Threshold will not depreciate as fast as PCs do. And I am getting far more joy out of it.
And the difference between a *really* good stereo system and one average system is *huge*.
I can do the blind fold test any time. Also I want to have what I like, and I build my stuff accordingly.
Which in turn means that stuff that I don't like that much has got to go.
I do find the price of those Audyn copper caps excessive, though.
My friend who does all these speaker developments uses them in his high end designs. I never considered those, as they are way beyond what I would be willing to pay, however listening to his builds is "impressive".
https://www.hifi-spectrum.de/cms/produkte.php?kategorie_id=2&kat_id_list=_0_2 -
let's play a game.
Plug J137 back in and bingo
Spot the gorilla, umm, idiot, umm failure ...umm, no. look at the jumper wire and the trace that is burned away. a new power board ist 350€, a pinball repair company charged them 2500 for that jumper wire.
The could at least have repaired the trace properly
Not to mention J128, 129, 131, and 132. It's not that unusual to have unused connectors, these boards will be common to multiple different machines. That "repair" is UGLY. Why the heck didn't they just solder a wire bridge to the track?
I can see the value in a big loop during troubleshooting, to get a current clamp in there, if other means of measuring it aren't practical, but you don't leave it like that once you're done.
Also J133 may need re-seating. Bloody Molex IDC's...
Aaaaaand... the screw is behind the keyhole slot in the board at the top right, meaning the board is flapping in the breeze, risking stress fractures at the other mounting points.
Lastly, that braided earthing strap really should be either tinned or crimped/sleeved at the end to prevent it fraying, but I'm probably nit-picking at this point. -
The thin jumper wire will not be able to handle the load and will cause additional issues. . I will rework this board accordingly and put in a new board for the time being.
-
The thin jumper wire will not be able to handle the load and will cause additional issues. . I will rework this board accordingly and put in a new board for the time being.
It looks to be the same as the wires in the IDC the trace goes to
-
@Vince,
I do have a PCI NI GPIB card........
Or you could try WIN98 SP2
-
Crate update.
It's a mill
Wedding anniversary gift from SWMBO.
It's a SIEG SX1LP. Small but with long X travel and a high (for non CNC) spndle speed, 5000 RPM.
It's optimised for light alloy and plastics. Apparently the supplier called SWMBO back because she had ordered an "electonics" type mill and they wanted to check what sort of engineer I was. Very good service in this day and age.
I was fearing that she might be making slow preparations to turn you between centres, or buying a very big vise to put some delicate part of your anatomy in.
Let me know how you get on with it. I have toyed with the idea of getting one because it's about all the mill that would fit into my space poor residence.
It should be fine. ARC euro trade have a good reputation. The basic design of the mill is common to many makes and models. ARC has specfic features and improvements. It's a brushless DC motor with belt drive rather than the plastic used by many. I've Used a Clarke (Machine Mart) CMD10 and it will be much better than that, just the lack of gear noise will be great. It also has more table travel than even the larger Clarke CMD300. This is important if making front panels. The SX1LP will cover most of the usable area of a 4U 10" panel without repositioning. The non-tilting pillar is a positive advantage as it is much more rigid. and of course the 5000 RPM spinle speed is much better when using small cutters and drills in non ferrous metals and plastics. The <2000 RPM loaded speed of the CMD 10 was a big pain.
Now I've just got to make a bench and dig out my tooling etc.
Not even turned it on yet and I'm already a fan. -
PSU constant current response timing shoot out
Having collected some PSUs with CC capability I’d like to know how quick they are responding to an overcurrent condition.
I put the results into a seperate thread but dare to post a link here, because the topic is triggered by a serious seizure of TE collection syndrom.
https://www.eevblog.com/forum/testgear/psu-constant-current-response-timing-comparison/msg4329532/#msg4329532
Test setting is like this:
PSU is configured to deliver 25V/1A.
PSU is connected to a 10 Ohms resistor.
This load circuit is switched on/off by a MOSFET-Module, controlled by a function generator.
off: No load, PSU idles at 25V/0A
on: PSU would deliver I=U/R=2,5A but limits current to 1A by lowering the voltage to 10V. The question was: How quick does this happen?
The results show some surprises.
The mighty TDK Lambda 30A switcher is by far the slowest one, and a cheap consumer lab supply is by far the fastest. -
Suddenly the good old days don't seem to be as good as we remembered them to be, eh?
Oh no, mansaxel said Win95 has everything that's needed out of the box... how do I do then ?
As I wrote earlier, yes, that is true. But you have to put on your 90s head to do it.
On the Linux box, install or enable a FTP server. (google "enable ftp server ubuntu" or whatever distribution you have) Make a note ("ifconfig") about the IP address of that box.
On the windows box, open a DOS prompt, and start going back in time:
c:\> cd c:/vince/gpib/dumps/ <- where you put the file
c:\> ftp 192.0.2.44 <- swap in IP address of your Linux box (that address is a RFC 5737 example address)
Login with your user and password of the Linux machine.
This will go in CLEAR TEXT over the LAN. That's why I was adamant on it being almost OK on only that LAN and nowhere else.
For classic ftp daemons, as long as you're approved. you'll now be in your home directory on the target machine. Check with "ls":
ftp> ls
229 Extended Passive Mode OK (|||45379|)
150 Opening ASCII mode data connection for '/bin/ls'.
....
You should see the same as "ls" on your linux box shows, in that folder.
On the target, you might want to do a new folder:
ftp> mkdir gpib
257 "gpib" directory created.
ftp> cd gpib
250 CWD command successful.
Now you're in that new folder. Time for file transfer.
The memory dump is binary, right? Not a text file?
Then, put the file transfer in binary mode.
ftp> bin
200 Type set to I.
Transfer the file (I'm assuming it's called vincescp.dmp):
ftp> put vincescp.dmp
local: vincescp.dmp remote: vincescp.dmp
229 Extended Passive Mode OK (|||38311|)
150 Opening BINARY mode data connection for 'vincescp.dmp'.
100% |********************************************************************************************************************************************| 1172 KiB 40.36 MiB/s 00:00 ETA
226 Transfer complete.
1200989 bytes sent in 00:00 (11.53 MiB/s)
ftp>
Done! Don't forget to be polite:
ftp> bye
221 Goodbye.
And you have the file on the usable computer. Easy as Tarte Tatin!
Thanks !
I just tried that, works just fine !
Was indeed not that difficult, thanks to your instructions !
I googled a bit, looked like " vsftpd " was a popular FTP server on Linux, so I installed that.
I failed to connect to the old Windows machine though... maybe because W2K doesn't enable the FTP server by default, maybe I need to click somewhere to enable it...
However the client was already installed and I could "ftp" to my Linux from the command line just fine !
I could see my Home directory, I created a new directory in there, then copied a file into that directory. So easy indeed !
I like this stuff a lot, really cool !
So my future local network might not be that hard to make then, after all !
EDIT : found this web site that explains how to enable the FTP server on W2K..
https://www.techrepublic.com/article/get-it-done-setting-up-an-ftp-server-with-windows-2000/
it was indeed not enabled by default. Followed the instructions to create an FTP "site" as W2K calls it... appears to work, locally at least, if I type " ftp://localhost " in the file explorer, W2K can see itself. It also shows there, that the server is set to "anonymous".
So then I moved to my Linux box and try to connect to W2K. See screenshot... the Linux box, using the IP address of the W2K box, can "connect", I mean it can identify the machine, because it prints its name, and knows it's a Microsoft FTP server... so that's good. I know they can talk to each other. However despite being set to anonymous, I am still being asked for a password... I type it anyway ( "admin" ...), but it fails to login with that, crap.... I tried giving no password at all, since it's supposed to be anonymous, but still doesn't work.
Oh well..... I will call the customer support ! -
@Vince
You know, sometimes I just wonder if that difference is as huge as you think it is. The only true way to find out, is to go somewhere else, where you have never been before, so you are not aware of the acoustics of the room. Then be blindfolded, while the room is totally devoid of any Hi-Fi, and then the equipment (again nothing you're used to), is brought in and set up. Then play the identical piece of music, again, something you are not familiar with and then ask you after all the systems have been used, to rank them in order of quality and cost based purely on your perception and then visibly reveal the systems used in the order they were used in. I reckon that the results may well astound you.
I do have some not so high endian audio stuff, too.
And I cannot go past a decent amp that someone is throwing out for a tenner.
I am in the fortunate position that I did get my high end stuff mostly in trade for PC stuff that I no longer need which is in turn used for astronomical image processing and thus needs to be reasonably powerful.
Which in turn means that I got my Threshold overthetop power amp as trade for a PC. The Threshold will not depreciate as fast as PCs do. And I am getting far more joy out of it.
And the difference between a *really* good stereo system and one average system is *huge*.
I can do the blind fold test any time. Also I want to have what I like, and I build my stuff accordingly.
Which in turn means that stuff that I don't like that much has got to go.
I do find the price of those Audyn copper caps excessive, though.
My friend who does all these speaker developments uses them in his high end designs. I never considered those, as they are way beyond what I would be willing to pay, however listening to his builds is "impressive".
https://www.hifi-spectrum.de/cms/produkte.php?kategorie_id=2&kat_id_list=_0_2 -
Hence why I use AirPods for music
Plus they have a feature to silence other dickheads in the room
-
This is a bit painful to watch...
-
Actual TE content. I'm on a roll.
Fixed the three remaining issues with the 465B:
1. I put the channel A knob on the wrong position. Have fixed it.
2. The trigger BNCs were a little gunky and not making contact. Cleaned and fine now.
2. The intensity pot was a little janky. This seems to have cleaned up with some exercise and is nice and smooth now.
I let it burn in for 2 hours and nothing exploded. It's working nicely now and perhaps surprisingly the cal checks out fine.
This is going on "bench 2" which will be set up in around a month from now. Woohoo!