Work over, I've had a play with the Yokogwa TA520 Time Interval Analyser.
Well it certainly works. Hooked it up to a GPStar GPSDO on Ch.A and my cased up HP 10811D OCXO on Ch.B. In Time Interval mode you can clearly see the 1G turnover. With the OXCO box flat fine trim for no movement, Place on left side and indication moves left. Flat again it stops. On right size indicaton moves right.
Now I just have to learn how to use it properly and make sense of the results. And find somewhere to put it 
I'm so considerably tempted. Perhaps work should buy it. There's 1PPS and 10MHz i/o on nearly every network switch we buy.
I was thinking along the exactly same lines!
If either of you want the one I bought, drop me a PM. I'd be OK passing it on at cost plus shipping. I'm not sure I'll be able to do it justice.
My name is @nfmax and I am a TEA addict. I confess it was me that bought the other TA520. Just recently, I've been using it to help investigate some quirks of the HP/Agilent 33120A function generator:
https://www.eevblog.com/forum/testgear/hp-33120a-frequency-instability-~100-hz/I'm still getting the hang of it, but it seems to work OK. Though it has a fan that shakes the whole bench while it is running
4x 2TB WD Gold Enterprise class spinning hard drives in Linux Software RAID.
I wonder if it is better than ZFS...
4x 2TB WD Gold Enterprise class spinning hard drives in Linux Software RAID.
I wonder if it is better than ZFS...
Something I need to research before setting up a new PACS next month. I don't know much about ZFS. I'm planning to use Ubuntu 22.04, which is on kernel 5.15.
Linux RAID is in the kernel. Last time I checked, ZFS was not. I don't want to use a file system that is not well supported.
For the production PACS, reliability and uptime are far more important than speed.
$ cat /proc/mdstat
Personalities : [raid10] [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4]
md0 : active raid10 sdb[1] sdd[3] sda[0] sdc[2]
3906766848 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
bitmap: 0/30 pages [0KB], 65536KB chunk
unused devices: <none>
I've had linux RAID member disks fail in my "Sandbox" Mac Pros. Swapping in a spare drive and rebuilding the array is painless.
I have stack of old Mac Pros I use instead of a VM. The disks in them are 10+ years old and occasionally fail. RAID 1 is free. New drives are not.
For the production PACS, reliability and uptime are far more important than speed.
Amen Brother,
Praise God the Almighty!
ZFS is used often in production, according to my limited experience, is exactly what you want/need for production.
Next.
A friend spotted a seller on Leboncoin.fr that had lots of cheap stuff. He was interested in a few items he could shift for a quick profit, but seller would not ship... but turns out he lives a stone throw from me, I mean almost literally : less than 5 kms... 5 minute drive, the next village, neighbouring mine. I traverse this village at least twice a day to get anywhere. Looking at all the ads from this seller, IO spotted a little something I was interested in... a small lot of random components, for 10 Euros. In the pic it was hard to tell what was what exactly, but I for sure saw a few bags full of brand new LEDs of various colours. Was worth the Euros alone, and I have no LEDs in sotck other the ones I salvaged 30 years ago, so short legs, tired/dim, and not two identical ones... it sucks. This new LEDs would solve that problem 
So I went to pick-up that lot. As for the 4 items my friend wnated, one had already been sold this morning, too late (a little Geiger counter...), 2 more items had suddenly seen an impromptu price increase : " Oh it'(s my daughter who put the ads, but she got teh price wrong ! ". So my friend decided to give up on these two items. So in the end he got only one single item out of his shopping list... but a cool item !
A modern GPS DO... for 10 Euros ! 
Those Chinesium ones. He said the cost about 90/100 Euros. Cheap I thought.. but 10 Euros is even better ! I need one of course but well, he saw it first !
Been running the thing for hours now. Frequency is very stable since the beginning really... 9,999,986 Hz +1/-0 Hz. just stays there. I don't know the accuracy of that GPSDO but I would hope it's much better than 13 Hz in 10MHz ! So I guess the frequency I see is that of the item base in my old Metrix Nixie counter ! Looks like the oscillator needs to be trimmed !
I don't know if the GPS is working oir not. The status LED on the front panel, since cold, keeps displaying a solid yellow colour, but don't know what that means.
In my world, green means all good and red means all bad... so yellow is in-between ! It's not all good... yet not all bad... there is some good, and some bad... I guess the bad is that it fails to synchronize, and the good is that it can talk to the GPS chip locally ?! Who knows ! RTFM yuo say ! I can"t ! There is no model number nor even a manufacturer name no this thing, ZERO ! It's just a black box !!! 
The Power ON LED is troubling too... it lights up green, which is good but... it's blinking all the time !
Status LED for the OCXO seems good : lit up red when cold, then turned off after a few minutes when warmed up. I do get a 10MHz stable output so I guess at least the OCXO itself is good.. bit the GPS part, not so sure... I did bring the antenna right to the window n the living room, no joy, LED still yellow. A friend told me these crappy antennae are not very sensitive and you need to take them outside in the garden, hoping they would catch the satellites up there... might try that tomorrow to see if that makes a difference...
Also tomorrow, will feed the thing to my Ferisol counter to compare it my Metrix counter, see if it's also as far off. Ferisol has a OCXO at close to 10-9, but the Metrix I am not sure what it's got for a time base... I am not sure it's an OCXO, might just be a regular/bare crystal. Need to check...
Piccies....
-group-therapy-thread/?action=dlattach;attach=1437727;image)
Vince,
The GPSDO looks like one of these from AliExpress:
https://www.aliexpress.com/store/group/GPSDO-Signal/4951008_515594303.html?spm=a2g0o.store_pc_home.pcShopHead_38580915.1_4Tony Albus has a couple of videos, including one on how to configure GNSS, etc.
Thanks....not my "problem" anymore though, I have already packaged it and shipped to my friend, it's intented owner. Played with it for only a day.
My old friend told me about a nice yet affordable GPSDO, not chinese "youd 'ont know what you get" toy, but rather a gnetleman in the UK who made a business designing and selling these things.
So closer to me and more credible. At least he shows graphs of the testing of his design. Once it' warmed up / 15 minutes, it gets to 8 or 9x10exp(-13).
So say about 10exp(-12), so good for a 11 digit counter hey ? ... if I can ever afford a thing probably not, and doubt I will have a need for that anyway.
It's plennnnty good enough for my 8 digits NIxe counters, and good enough for a more modern counter once I can afford one.
165 quid / 200 Euros or so... good value for me compared to that Chinesium one, it's a no brainer.
https://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=107&products_id=234Such a pain to have to endure the import bullshit to get it, though...
And now for something a little different, slightly off-topic, although I suspect most people will have at least one of these items in their labs.
The other day I wanted to do a colour printout, but my printer refused to play ball, leaving the image more than lacklustre, looked like the rats had been at it. After doing many head cleaning cycles and test pages, it was still the same, yellow had only 1 or 2 jets on its head working, cyan and magenta was marginally better. I was forced to give up and do the print on the laser printer instead, which is far more costly to do and also requires manual intervention to do duplex printing.
I searched around and found a company called Printheadhospital that claimed to be able to revive most ink jet printer heads, they sell their kits via Amazon and operate both in the UK and North America. I saw on Amazon that they offer upto 80% refund on their kit if their product does not cure your printhead problems. I ordered a kit last night, took delivery today and tried to clean my printhead and got 100% success with it and now my inkjet works just as good today as it did when I first purchased it a few years ago
Their website is
http://www.printheadhospital.com/ Video is here
https://printheadhospital.com/brother-print-head-cleaning-kit/So if you have problems with your inkjet printer, then this product will may very well help you to cure your problems. NAWTS.
For the production PACS, reliability and uptime are far more important than speed.
Amen Brother,
Praise God the Almighty!
ZFS is used often in production, according to my limited experience, is exactly what you want/need for production.
I have not lost a single bit to ZFS problems, and I've run it on suitable and unsuitable platforms for 15+ years. I run ZFS on bare metal FreeBSD, VMWare hosted Freebsd, and I've got extensive experience on Solaris (but won't run that anymore, because Larry is a right despicable cunt.). ZFS is the file system all others oughta be. The only thing coming close is jfs on lvm in AIX. The rest, including every native Linux file system I've ever seen, is just a joke in comparison. Of course I'd rather not run it on Linux, because I'd rather not run
anything on Linux.
Had a typical interaction with Centos and it's network manager UI today. It is mostly incapable of creating a LACP paired connection. To smooth over this, the vendor who ships the application running in a bunch of bad copies of FreeBSD jails on top of the OS, ships a Python script that fails to interact with the native config UI and its backend.
Also, the LACP config CLI says to type a string in its help output, and when I type that exact string, complains that there are too many or too few parameters.
I want to hurt people.
TE update: 5316b has shipping label created. Investigating oscillator options. Perhaps should buy a 10-pack of those CTI oscillators..
And now for something a little different, slightly off-topic, although I suspect most people will have at least one of these items in their labs.
The other day I wanted to do a colour printout, but my printer refused to play ball, leaving the image more than lacklustre, looked like the rats had been at it. After doing many head cleaning cycles and test pages, it was still the same, yellow had only 1 or 2 jets on its head working, cyan and magenta was marginally better. I was forced to give up and do the print on the laser printer instead, which is far more costly to do and also requires manual intervention to do duplex printing.
I searched around and found a company called Printheadhospital that claimed to be able to revive most ink jet printer heads, they sell their kits via Amazon and operate both in the UK and North America. I saw on Amazon that they offer upto 80% refund on their kit if their product does not cure your printhead problems. I ordered a kit last night, took delivery today and tried to clean my printhead and got 100% success with it and now my inkjet works just as good today as it did when I first purchased it a few years ago
Their website is http://www.printheadhospital.com/
Video is here https://printheadhospital.com/brother-print-head-cleaning-kit/
So if you have problems with your inkjet printer, then this product will may very well help you to cure your problems. NAWTS.
Successfully recovered dried out inkjet print heads with just leaving them overnight on wet paper towels that dissolves dried out ink and gets them going again.
Inkjets don't like sitting around unused and power cycling them occasionally so their head clean utility runs keeps them in running order.
And now for something a little different, slightly off-topic, although I suspect most people will have at least one of these items in their labs.
The other day I wanted to do a colour printout, but my printer refused to play ball, leaving the image more than lacklustre, looked like the rats had been at it. After doing many head cleaning cycles and test pages, it was still the same, yellow had only 1 or 2 jets on its head working, cyan and magenta was marginally better. I was forced to give up and do the print on the laser printer instead, which is far more costly to do and also requires manual intervention to do duplex printing.
I searched around and found a company called Printheadhospital that claimed to be able to revive most ink jet printer heads, they sell their kits via Amazon and operate both in the UK and North America. I saw on Amazon that they offer upto 80% refund on their kit if their product does not cure your printhead problems. I ordered a kit last night, took delivery today and tried to clean my printhead and got 100% success with it and now my inkjet works just as good today as it did when I first purchased it a few years ago
Their website is http://www.printheadhospital.com/
Video is here https://printheadhospital.com/brother-print-head-cleaning-kit/
So if you have problems with your inkjet printer, then this product will may very well help you to cure your problems. NAWTS.
Successfully recovered dried out inkjet print heads with just leaving them overnight on wet paper towels that dissolves dried out ink and gets them going again.
Inkjets don't like sitting around unused and power cycling them occasionally so their head clean utility runs keeps them in running order.
I have done that in the past on printers that the heads as part of the replaceable cartridges, but most new ones tend to cartridges that have separate heads on the end of long ink tubes and the heads are none replaceable, such as multifunction units like mine where that is not an option.
@Vince, here are some current clamps that Robert didn't buy yet...
... with an AUDIO JACK for a connector ?!
@Vince, here are some current clamps that Robert didn't buy yet...
... with an AUDIO JACK for a connector ?!
Don't lose sight of the fact it is only carrying a few mA at mV normally.
Just received a 100x scope probe, NOS, reasonable money.
It's an Elditest GE 3121, 150MHz, 2kV, though I don't want to ever be within a mile of anything oscillating at that frequency and magnitude! Fabrique en Allemagne, so decent quality I should think.
What's the voltage derating curve
2kV at 100MHz aross even 1pF is a whole lotta current!
Just received a 100x scope probe, NOS, reasonable money.
It's an Elditest GE 3121, 150MHz, 2kV, though I don't want to ever be within a mile of anything oscillating at that frequency and magnitude! Fabrique en Allemagne, so decent quality I should think.[/color=teal]
Worthless unless you have the frequency derating chart.
@Vince, here are some current clamps that Robert didn't buy yet...
... with an AUDIO JACK for a connector ?!
Technically a plug...
It does have an extended insulation that fits in the meter socket iirc (this type is usually used with a 3ph+N meter to help with load balancing, which is why you quite often see them in fours), and in fairness there's no direct connection to the test subject, so even if you go all Mr Bean and stick the plug in your mouth while the clamp is on a current carrying cable, you're not going to get more than a mild tingle at most, if that, as I doubt FSD is more an a few volts.
Oh, looks like I was misunderstood.. my bad !
I didn't at all choke on the safety aspect of it, it did not even occur to me... I was merely surprised to see a jack plug for an electronics piece of TE, that is all.
I would have expected either a BNC plug if meant to plug on a scope, or 4mm plugs if meant to plug in... well anything other than a scope.
I didn't know it was common / normal for current clamps to use a jack connector, that was all !
I have never seen such a plug on a current clamp before, hence my surprise..
(...)
TE update: 5316b has shipping label created. Investigating oscillator options. Perhaps should buy a 10-pack of those CTI oscillators..
There are also China surplus "VECTRON C4550A1", some of them even have the V-Ref output option (they sadly all look the same from outside).
Mine have arrived and were bought when they were only slightly more expensive then the CTI, which seems to have changed recently
A quick check indicates that 3 of 10 seem to posess the option. You'll have to decide for yourself if that sample size is representative enough.
This guy made nice A-Dev Graphs for them and gave some more details.
Could be upwards of 6 inches or more and later the wind will kick up and blow it into drifts. Lots of fun.
