Test Equipment Anonymous (TEA) group therapy thread

[Zoli]

Who would have thought putting a PC and windows in a spectrum analyzer was a bad idea ?

Sometimes ago I found myself wanting some newer gears. Maybe running some version of windows, with nice UI and everything. Well I got some of those now and I'm currently living in PC hell.

Is it booting ? no of course not.
...

So I'm trying to upgrade the harddrive to something better than those old mechanical one prone to failure. Got a adapter, because everything is SATA now and not IDE. Which one of those is actually going to work ?
...

The drive is detected correctly.
...

Going back to Acronis True Image, now this time actually pay attention to what the software is doing. The original drive is quite simple, just 2 partitions that bootup with MBR.
...

Now when I copy the image on another drive Acronis Is actually converting the drive to GPT (different way of representing the master boot record), wait what 
...

Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

Clonezilla to the rescue; practically, is a "nice" interface around dd, as capt bullshot recommended; as bonus, is bootable.

[mansaxel]

mnem
*huggles his 189 close*

Bah.

I'm set on this one. 

That is a very wise decision. I've only got the Metrahit 25s, but even that one has higher resolution (more steps) than the 87V (87V has 19999, the 25s has 31000).  I do sense some english-language-centric fanboy-ism in the 87 cult, and I'm not unaffected: I've got as many Flukes as every other handheld meter brand together ( 7 Flukes, 3 Avometers, 2 Gossen, one -hp-, one H&B)

All mentioned meters can do enough for most situations, I believe..

[BU508A]

Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

Capt. Bullshot is right. Use Linux / dd and it should work out in your sense.

https://serverfault.com/questions/4906/using-dd-for-disk-cloning

I recommend to use the Linux tool box for such cases: grml.

https://grml.org/

[BU508A]

That is a very wise decision. I've only got the Metrahit 25s, but even that one has higher resolution (more steps) than the 87V (87V has 19999, the 25s has 31000).  I do sense some english-language-centric fanboy-ism in the 87 cult, and I'm not unaffected: I've got as many Flukes as every other handheld meter brand together ( 7 Flukes, 3 Avometers, 2 Gossen, one -hp-, one H&B)

All mentioned meters can do enough for most situations, I believe..

<whisper>
I don't own a 87V ...
</whisper>

[mansaxel]

<whisper>
I don't own a 87V ...
</whisper>

<comfortingly>
Me neither. That is OK. You can always buy it after you've got yourself a 27CE. With Probulator Kit.
</comfortingly>

[dl6lr]

We know, courtesy of the session with the insulation tester last week, that the extension cord is good, so I put it to use.



It didn’t occur to me to test the hedge trimmers with it until I got them out, but it turns out that wouldn’t have been possible in much of a meaningful way since they’re double insulated with no ground.

It's not level !   

Clearly missing a caliper 

[dl6lr]

Now when I copy the image on another drive Acronis Is actually converting the drive to GPT (different way of representing the master boot record), wait what 


Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

I am too stupid to use those fancy windows software, so I am still copying my partitions with simple linux dd command. Usuually just copying the whole disk is fine, afterwards run fdisk once to adapt the partition table  to match the drive size. Just converting the Prema MC8047 rotating rust to solid state. Here I even downsized the disk size, as I replace the dd with a DOM. Even 64MB is barely filled with the small DOS and dmm software.

[mansaxel]

Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

Capt. Bullshot is right. Use Linux / dd and it should work out in your sense.

https://serverfault.com/questions/4906/using-dd-for-disk-cloning

I recommend to use the Linux tool box for such cases: grml.

https://grml.org/

I second the general direction of advice too, even if I'm inclined to use a BSD flavour instead. I find that OS X has a lot of good tools for manipulating for instance ISO images on and off reflective and flash media.  When you request to get the results of a MRI or CT scan out, you unvariably get them on a DVD-R or CD-R. Long-time storage and access of those is not something I'm fond of, so I tend to clone them to spinning rust with cloud backup. This also makes things like shipping a copy to another doctor quite easy, you just "hdiutil burn <path to iso file>" directly off the networked storage, and OS X DTRT.

But, this branch aside, yes, Linux is your friend, and it might be so that for some things an ancient Windows machine might be required. Regardless, the cloning part is best done using "dd".

[Cubdriver]

I got in on that evilbay sale of the Fluke 27/AN meters; listed at $150, offered $130 and it was accepted, so wound up getting it for about $150 after shipping and paying the .gov's cut.  It arrived today.  Pretty sure it's brand spankin' new old stock - don't think it was ever out of the box.  It includes the 6 kV and 40 kV probes, the RF probe, an operator's manual and a service manual complete with schematics and parts list.




A few more pics at: https://pmanning.smugmug.com/Electronics/Fluke-27AN-Meter

(Of course I don't have any 9V batteries handy; need to get one tomorrow.)

According to the listing, there are still more than ten left.  https://photos.smugmug.com/Electronics/Fluke-27AN-Meter/i-6zzCL8P/0/9350235a/X2/Fluke%2027_AN%20Kit%20-%20Cpl-X2.jpg
NAWTS, but if you'd like a like new 27/AN, go for it.

-Pat

[med6753]

It finally showed up yesterday. Thanks to BU508A. 

[Specmaster]

mnem
*huggles his 189 close*

Bah.

I'm set on this one. 

That is a very wise decision. I've only got the Metrahit 25s, but even that one has higher resolution (more steps) than the 87V (87V has 19999, the 25s has 31000).  I do sense some english-language-centric fanboy-ism in the 87 cult, and I'm not unaffected: I've got as many Flukes as every other handheld meter brand together ( 7 Flukes, 3 Avometers, 2 Gossen, one -hp-, one H&B)

All mentioned meters can do enough for most situations, I believe..

Pah, the Brymen 867 & 869 can beat that with 50000 as standard and special 500000 mode, but I don't huggle that close, as a rule its useless, I find a tape measure is way better for that job 

[Specmaster]

That is a very wise decision. I've only got the Metrahit 25s, but even that one has higher resolution (more steps) than the 87V (87V has 19999, the 25s has 31000).  I do sense some english-language-centric fanboy-ism in the 87 cult, and I'm not unaffected: I've got as many Flukes as every other handheld meter brand together ( 7 Flukes, 3 Avometers, 2 Gossen, one -hp-, one H&B)

All mentioned meters can do enough for most situations, I believe..

<whisper>
I don't own a 87V ...
</whisper>

[Robert763]

Who would have thought putting a PC and windows in a spectrum analyzer was a bad idea ?

Sometimes ago I found myself wanting some newer gears. Maybe running some version of windows, with nice UI and everything. Well I got some of those now and I'm currently living in PC hell.

Is it booting ? no of course not.

So I'm trying to upgrade the harddrive to something better than those old mechanical one prone to failure. Got a adapter, because everything is SATA now and not IDE. Which one of those is actually going to work ?


The drive is detected correctly.


Going back to Acronis True Image, now this time actually pay attention to what the software is doing. The original drive is quite simple, just 2 partitions that bootup with MBR.

Now when I copy the image on another drive Acronis Is actually converting the drive to GPT (different way of representing the master boot record), wait what 


Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

You think that's bad, try doing an install of an older sysem on a R&S CMU200 / CRTU-RU. I did this to put CMU200 software on a CRTU-RU. This gives a CMU200 for a fraction of the price. Of course the CRTU hardware is slightly different so you can't just clone a CMU200 disk. The install program (specific versions) for the CMU200 will run on a CRTU giving you all the basic functions e.g AM and FM on the signal generator.
The problem is the installer will ONLY run on a HDD that was booted from MS DOS 6.22. Not  Win 98 "DOS", FREEDOS or anything else. The low level format and system files must be DOS 6.22 (6.2 might work but I didn't try it). The install puts an odd dual boot system on it with the basic functionality programs in one and Windows in another. If you don't do cellphones you don't use Windows.
For hardware I used an "industrial" CF card in an IDE converter. Having an old PC with IDE an d floppy support was essential.

[BU508A]

It finally showed up yesterday. Thanks to BU508A. 

For the next piece of art for you I'll add an Odysseus to it. 

[mansaxel]

I got in on that evilbay sale of the Fluke 27/AN meters; listed at $150, offered $130 and it was accepted, so wound up getting it for about $150 after shipping and paying the .gov's cut.  It arrived today.  Pretty sure it's brand spankin' new old stock - don't think it was ever out of the box.  It includes the 6 kV and 40 kV probes, the RF probe, an operator's manual and a service manual complete with schematics and parts list.

Interesting! Some notes:

• You, as I, seem to miss the standard test leads.
• But the slip-on croc clips for TL71 (or similar) leads are in the 85RF bag. Just as for me.
• I did not get any service manual.
• Mine is full of USMC calibration stickers and has a tag on the case handle also indicating a calibration made back in 2011. This -- from your pics -- seems to be missing for you.
  

Based on these differences and similarities I have two guesses:

• Yours was not opened while in Govt. possession
• Someone went through all the meter cases and removed the test leads to sell separately.

[mansaxel]

Pah, the Brymen 867 & 869 can beat that with 50000 as standard and special 500000 mode, but I don't huggle that close, as a rule its useless, I find a tape measure is way better for that job 

I'm a gear slut enough to be affected by nostalgia, and Brymen, while certainly good enough and definitely so for the price, are not an Old Brand. There has to be a legacy. A Bakelite past. Brymen is like a Japanese sports car. It is impressive, even objectively better, but it is not right.

The Brymen leaves my cockles ice cold, while -hp-, Fluke and Gossen pass that very subjective test.


Yeah, I'm silly. But so is the world, mostly.

PS:
The Miyata is excused, because it is very much in tune with say the desires and impressions of a small MG or Triumph.

[capt bullshot]

A short one from the Fun with TEA department:

Testing the internal HS-Oscillator of the 53131A vs. the internal OCXO of the FA-2:



Both of them have their 10MHz output connected to this little setup:



One of the CTI OCXOs in a simpe test jig, used as the reference input of my dual XOR mixer frequency difference measurement board. 10MHz outputs of the counters are connected to the comparison inputs. Here, it's running in its MCU based mode (not the offset reference DMTD mode), reference and input frequency inputs don't have an offset, frequency difference is measured by analog to digital conversion of I and Q output of the XOR gates (two per input channel).

As this setup is quite sensitive to ambient temperature, it's normally covered with a highly sophisticated thermal shield.



For a first simple result, one can see the short term stability of the 53131A (pink) internal oscillator is by one decade better than the FA-2 (blue). For long term measurements, I'd either have to let the CTI OCXO stabilize for one or two more days, or use the GPSDO as reference. The True Position GPSDO isn't as short term stable as the undisciplined OCXO, so probably I'll wait and see if the OCXO stabilizes good enough.

[mnementh]

I think I'm gonna go look ad DIY vids on replacing a window screen... try and work smarter, not harder.

mnem
no, no, no, no, no... not Home Improvement!!! bad dwagon!!!

Ah, that job I did yesterday. The tiny white fluffy puppy knocked the screen out of a patio door.  It took all of 3 minutes to put it back.

The key point is to know what size of slot is in the door/window.  Then know the thickness of the screen (I use a very heavy "animal" mesh here for good reason). After that, choose the appropriate thickness of spline; thick enough to hold tight, but not to the point where it is too thick to install. There are generally 4 sizes of spline in the GWN.

EDIT:  I had no idea of Uncle Bumblebutt's shenanigans when I posted that.... nice spline!

I got lucky and was able to just pick out and re-roll the spline on the one side that came out; took all of 3 minutes (and before my first cuppa, no kidding!  ) for me too, and most of that was fighting the storm window pane off and back on the door. There is one area that doesn't feel as taut as the rest; however it all looks nice & flat, so I'm gonna live widditt.

Right now I'm shopping a 32" towel bar to attach right at butt-height to prevent recidivism; if I don't find something "just right" at a sane price, I'll sacrifice a cheap broom handle from Dollarama and print some brackets.

mnem

[mnementh]

Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

Don't use "smart" tools to copy the HDD. Use an as dumb as possible tool to copy the old hard drive content to the new one sector by sector. This will completely preserve the layout etc. Best results are achieved using "dd" on a Linux box.

I'm going to be the one guy who says otherwise, and that you should use a smart clone utility (preferably the one offered/recommended by the SSD MFR)... and that's because I've been doing SSDs since before Windoze "did it all for you".

dd and *NIX utilities don't do one critical thing for you: align the SSD for optimal block size. While you probably aren't going to get TRIM with the prehistoric (I'm guessing slimlined embedded version) OS that thing runs, just copying a disk image from spinning rust to a SATA SSD is going to perform like poop. In many cases, no differently than the spinning rust it replaces. Maybe even worse.

Windoze now does the alignment and enables TRIM and a number of other things automagically on any clean install; this is one of the main reasons every DIY you'll see on upgrading to a SSD recommends a clean install. Well, that and not bringing your existing "Windoze Rot" to the new drive. 

While you should be able to get by with dd or (my preferred utility) Image For Linux (which has the handy side effect of automatically making a archivable disk image as part of the process) and set the block size manually to 4K with your favorite partition manager, the smart cloning utilities will check your SSD/controller and make sure it works and that 4K is actually the optimal alignment for the combination, and some of them will actually check for and ask to install the latest firmware for SSD and HDD controller.

Also, check your BIOS for AHCI in the SATA controller before you do the image and at least try to enable it. Otherwise you could wind up running the OS at USB 1.0 data transfer rates.

Good luck,

mnem
 

[capt bullshot]

Kind of explain why it's not booting. Now I need to find another image software or an older version of Acronis or maybe something that can convert a GPT partition back to MBR ?

Arrggg stupid PC and windows!!! 

Don't use "smart" tools to copy the HDD. Use an as dumb as possible tool to copy the old hard drive content to the new one sector by sector. This will completely preserve the layout etc. Best results are achieved using "dd" on a Linux box.

I'm going to be the one guy who says otherwise, and that you should use a smart clone utility... and that's because I've been doing SSDs since before Windoze "did it all for you".

dd and *NIX utilities don't do one critical thing for you: align the SSD for optimal block size. While you probably aren't going to get TRIM with the prehistoric (I'm guessing slimlined embedded version) OS that thing runs, just copying a disk image from spinning rust to a SATA SSD is going to perform like poop. In many cases, no differently than the spinning rust it replaces.

And I'm the one who says using any "smart" tool is not the right way here, because I've been doing HDDs in TE from when they were 20MB (yes MegaBytes) in size and SSD didn't exist at all ...
All the modern SSD related stuff mentioned is neither needed nor understood by that old TE. Any SSD will perform way much better here than the orignal HDD, regardless of any kind of "optimization". In contrary, one might run into issues if the original partition layout isn't exactly replicated. At least this BIOS appears fresh enough to "understand" the SSD at all and at its native size. Older BIOSes might even fail if the SSD reports e.g. a size of larger than 32GB (IIRC).

[BU508A]

There is one discipline, where dd is really shining; exact 1:1 copies of harddisks.

This is also the reason, why dd is one of the forensic tools being used.

https://forensicswiki.xyz/wiki/index.php?title=Dd

I'd stay away from any smart tools in this respect. And when it comes to old harddisks / filesystems etc. I'd highly advise to avoid any kind of optimization.

Just my 2 cents.

[Cerebus]

Ah, that job I did yesterday.
The tiny white fluffy puppy knocked the screen out of a patio door.  It took all of 3 minutes to put it back.

Better that than go through the wall leaving the screen door intact.

[mnementh]

There is one discipline, where dd is really shining; exact 1:1 copies of harddisks.

This is also the reason, why dd is one of the forensic tools being used.

https://forensicswiki.xyz/wiki/index.php?title=Dd

I'd stay away from any smart tools in this respect. And when it comes to old harddisks / filesystems etc. I'd highly advise to avoid any kind of optimization.

Just my 2 cents.

Yes, I've done my FileSystems Forensics coursework; I have used dd to make forensic copies of a hard drive which are legally admissible as evidence. That is not what I'm getting at here. A 1:1 copy from one drive to a different drive is not always the best thing for a live operating system.

I've been doing SSDs on embedded versions of Windows for over a decade... and it really isn't that hard to have both a working duplicate of an existing custom build of Windows and a SSD that is fully optimized. I can't count the number of information kiosks I converted from 2.5" spinning rust to SSD.

First, you use the utility offered/recommended by the SSD manufacturer to prep the drive in the machine it's going to be used in. Most of the time, that's all you need to do. Then clone the drive in the target machine.

Then, if that fails, you use the same utility to prep and/or clone on a newer machine, and try again. Sometimes that works.

Then, if none of that works, a direct 1:1 image on the target machine. But usually, that's the worst performing of the options, and in some cases can result in a machine which is so slow it is unusable.

If you take the time to prep and get the right utilities to begin with, it's easy painless work; a couple-three hours even if you have to do all 3 options.

Cheers,

mnem
 

[Cerebus]

What you're missing out here Mnem is that this isn't stock hardware running stock Windows. This is a bit of embedded gear. Having watched EEs who believe themselves to be genius* programmers in action it is entirely possible that the muppet in question may have relied on knowing the exact physical disk layout in use to have done something like hide the enabled options in a specific disk sector, called out by a "magic number" hardcoded into the code, which he them proceeds to read directly from the disk with a BIOS call. (If done, the options sector will of course be cleverly encrypted by XORing each byte with the number '42'.**) Or may have created a backing buffer for memory at a known, again hard coded with magic numbers, location on disk that is read and written directly, bypassing the operating system entirely.

Faced with the possibility of this level of "coding brilliance" being present it's wise for embedded gear to produce a 1:1, sector by sector clone.

The aim here isn't to get a perfectly optimised, or even partially optimised, disk system. It's to get a working disk system that will be more reliable than the ageing drive that's currently there and, with a little bit of luck, may even be faster. The later is "nice to have" but isn't part of the objective and is easily lived without.


*The word "genius" here is to be pronounced exactly as it was by Phillip Pope playing "Leonardo Acropolis" in the Blackadder II episode "Money" as he declaimed "I am genius".

**True story, I've seen this done as 'encryption' in an option encoding scheme.

[mnementh]

What you're missing out here Mnem is that this isn't stock hardware running stock Windows. This is a bit of embedded gear. Having watched EEs who believe themselves to be genius* programmers in action it is entirely possible that the muppet in question may have relied on knowing the exact physical disk layout in use to have done something like hide the enabled options in a specific disk sector, called out by a "magic number" hardcoded into the code, which he them proceeds to read directly from the disk with a BIOS call. (If done, the options sector will of course be cleverly encrypted by XORing each byte with the number '42'.**) Or may have created a backing buffer for memory at a known, again hard coded with magic numbers, location on disk that is read and written directly, bypassing the operating system entirely.

Faced with the possibility of this level of "coding brilliance" being present it's wise for embedded gear to produce a 1:1, sector by sector clone.

The aim here isn't to get a perfectly optimised, or even partially optimised, disk system. It's to get a working disk system that will be more reliable than the ageing drive that's currently there and, with a little bit of luck, may even be faster. The later is "nice to have" but isn't part of the objective and is easily lived without.


*The word "genius" here is to be pronounced exactly as it was by Phillip Pope playing "Leonardo Acropolis" in the Blackadder II episode "Money" as he declaimed "I am genius".

**True story, I've seen this done as 'encryption' in an option encoding scheme.

No, I do understand. I do grok what's being said in here, quite fully thank you.

I'm not saying that isn't the case... just that it isn't always the case. It costs nothing but a little time to try and set up your SSD correctly first.

I know exactly what a dick embedded Windoze can be; and some of the kiosks I've worked on required DOD clearance to even get onsite or login, and have a lot of customized hardware that is not consumer-grade.

It is always worth the effort to try, and in some cases, your machine simply will not work correctly if you don't.

Jeezus... why is it always the same old abitrary binary choice "Do this, never do this!" BS  in here all the time?

Just because machine logic is binary doesn't mean we have to be.

There's a reason we still keep the meatsacks in charge; machines are terrible at making "judgement calls".

mnem