Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

Yeah, I looked it up yesterday over lunch and found a similar article; obviously NOT SHINE as that is a single momentary event which will not cause more than a momentary disruption, or loss of a few packets... Unless you run your backbone copper completely commando without any form of surge suppression or proper line filtering; then I suppose it could cause a power cycle/reboot that would take things down for 3-20 minutes depending on how long it takes for the switches/routers in question to reboot. Which would obvi generate a fault log a mile long in multiple affected devices, and should have alarm bells ringing on half a dozen maintenance consoles. 

More likely just another technical term misused by fuckwidgets who don't know a SMPS is, or what to call ordinary everyday switching noise. 

mnem

[Cerebus]

Erm, not 'backbone' as these are last mile services, literally as far away from the backbone as it's possible to get. Pedantry about terminology I'll grant you, but as you're being pedantic about their terminology I declare this "fair game". 

[Specmaster]

I, to, looked up the https://kitz.co.uk/adsl/rein.htm site and discovered about the AM radio tuned to 612kHz can be used to detect the emissions from a LCD monitor or TV, and even the DSL signal of a router even. So the question is not really about the persons TV set but surely about why, if this was already known to be a potential problem, why was it that those frequencies were used for the broadband in the first place, a few kHz higher or lower in the spectrum might have prevented any problem from occurring.

Bear in mind that I'm not particularly well versed in RF technology and I may be over simplistic in my thinking, but it seems to me, to make sense. I'm assuming that the village had a powerful Wi-Fi service and was not a hardwired service? Also, the article makes no mention about the resident whose TV was causing the problem, but I'd like to think that somehow or other, they were given another TV which did not cause the problem.

[capt bullshot]

Learned a lesson about nA measurement:

This:



Gave the correct result, but I had to switch the K238 to the 15V measurement range. Using the 1.5V range, the result was not useful, as it went to a flickering "Compliance" LED. As some of you noticed, this setup is nothing but an ugly bodge neither using an appropriate Triax nor shielding. Since the Triax hasn't arrived yet, I slightly rearranged the setup to a short connection from K238 to that resistor.



Et voila, I'm getting reasonable results even at lower currents:

[bd139]

On ADSL/OR, I suspect this is a load of steaming bollocks. 612KHz is one things that seems to have some significance like water divining or organised religion. There is no way that it is a deterministic way of working anything out. The ADSL signal roughly consists of a bunch of tones in channels transmitted over a differential pair (note differential, which should eliminate common mode noise!!!!). Each channel is numbered. 621KHz is "tone 144" which in the UK is allocated to downstream data. Very little significance other than it's a downstream channel. If there is a problem with that channel through RFI, the line isn't going to fail out unless the equipment is faulty at either end and can't negotiate around the noise which results in less tones/channels and therefore less bandwidth. It'd still work. Unless something is massively fucked.

But there are two conditions where it can cause problems. Firstly that is the front end of the ADSL modem is overloaded to fuck which causes the differential signal to exceed the capability of the input of the router. The input side is usually a little transformer and a couple of fast opamps that drive an SDR type thing these days apparently. The second one is there isn't a differential signal present!

I reckon it's mostly the latter that's the problem i.e. a shitty pair because I had some fun with a scope on a duff line that kept dropping and found out that one of the pair conductors was eaten through by bird shit. That meant there was no common mode rejection at all. You could see it on the line clearly that one conductor in the pair was a nice antenna. I'm surprised it worked at all and it's a tribute to the guys who design the front end and filters and digital signal processing in these.

Why would it affect so many people? Go look inside the average MDF in the exchange or the box on the street. Total shit show.

Basically I think to reason with it, you need to stick a spectrum analyser on the end of the line and terminate it at the DSLAM end and see what is left. Anything else is probably water divining or moses grade shit. Now I know from dealing with OR for years that the first thing they do is just swap pairs around in the street and at the exchange and see if that makes it any better. There is no science involved until you get fairly high up. I reckon they must have about 5 capable people running the country these days.

But really at the end of the day this is really all down to the fact we're hacking massive amounts of bandwidth out of signalling systems that were designed 100 years ago. Bring on fibre. Mine is queued for installation - 900mbits here I come.

[capt bullshot]

You wouldn't even require an SA. This is what my ADSL modem shows (it's the popular Fritz Box):

[bd139]

You would need an SA because the Fritzbox (I have one as well) doesn't show any impulse noise. The picture is a lot more dynamic than that. If they showed error bars on there per tone it'd be good but I don't think the protocol works like that.

[med6753]

As promised and delivered right at noon as UPS consistently does. Parts for the Type 310A. 

[AVGresponding]

I have not encountered this term before, which I find suspicious. EMC/EMI is a topic I really strive to keep current in.
Citation:
'The TV was found to be emitting a single high-level impulse noise (SHINE), which causes electrical interference in other devices.'

The remainder is here:
https://www.bbc.com/news/uk-wales-54239180

As far as other news are concerned, I am feeling quite sick  Seems I have lost in the preservability fight yesterday. WTF, it did not taste as bad as it feels!

If it's related to openreach it says their PR team or road monkeys made up the term on the spot.

It took 18 months to fix. That's about right for them 

According to el Reg it was longer. They took an unspecified period of time from when the issue was first reported to rip out all their cabling and start again, then another 18 months from there to recognise and diagnose that it was an EMI problem. Knowing BT that unspecified initial period was probably another 18 months. So, it's not just BT that hire monkeys, it looks like the BBC news also does now, to have swallowed the spin and publish it as only 18 months in total.

The original Bodge-itTelecom Openretch press release is here.

Local Openreach engineer, Michael Jones, was determined to find the Aberhosan broadband smoking gun – so he decided to call on the assistance of colleagues from Openreach’s Chief Engineer team – the company’s telecoms equivalent of the ‘SAS’.

Rough translation: "We finally handed the problem over to 'engineers' who can actually read and write, and seldom use crayon to do so."

I once had an Openretch 'engineer' come to fix a problem on my line who did not know what I was talking about when I described the noise I was hearing on the line as "white noise".

We're a long way from the days of the likes of Tommy Flowers, sadly.

[mnementh]

The real worry is how long I can keep bringing home new orders with my current SUV before I need get a bigger one !  Best effort so far is 26 DSO's and a bit of other stuff all in one load.    No room for passengers ! 

   https://www.amazon.com/gp/product/B001G4Z5J0/

I highly recommend one of these; tripled the cargo space (behind the back seats) in our Rav4. Get the biggest Pro (60 x 24" base) model; the 18gal Rubbermaid Roughneck totes fit 4 wide, and the 136L (35Gal) ones fit 2 the long way or even on top of the 4.

Also get the cargo net; 1 for each layer you plan to stack (so 1 or 2) https://www.amazon.com/gp/product/B07BTLL99G/

My cost ( I already had the hitch) was US$150 for the carrier, $US140 for all 6 totes ( I paid premium brick & mortar prices so I could measure first) and another $30 for the nets. They kept EVERYTHING (And a ginormous suitcase) very securely in place all the way from Tejas to Toronto.

My Draw-Tite 75235 Class III receiver hitch was US$130 from Uhaul; they were cheapest by far at the time.

mnem
 

[Specmaster]

On ADSL/OR, I suspect this is a load of steaming bollocks. 612KHz is one things that seems to have some significance like water divining or organised religion. There is no way that it is a deterministic way of working anything out. The ADSL signal roughly consists of a bunch of tones in channels transmitted over a differential pair (note differential, which should eliminate common mode noise!!!!). Each channel is numbered. 621KHz is "tone 144" which in the UK is allocated to downstream data. Very little significance other than it's a downstream channel. If there is a problem with that channel through RFI, the line isn't going to fail out unless the equipment is faulty at either end and can't negotiate around the noise which results in less tones/channels and therefore less bandwidth. It'd still work. Unless something is massively fucked.

But there are two conditions where it can cause problems. Firstly that is the front end of the ADSL modem is overloaded to fuck which causes the differential signal to exceed the capability of the input of the router. The input side is usually a little transformer and a couple of fast opamps that drive an SDR type thing these days apparently. The second one is there isn't a differential signal present!

I reckon it's mostly the latter that's the problem i.e. a shitty pair because I had some fun with a scope on a duff line that kept dropping and found out that one of the pair conductors was eaten through by bird shit. That meant there was no common mode rejection at all. You could see it on the line clearly that one conductor in the pair was a nice antenna. I'm surprised it worked at all and it's a tribute to the guys who design the front end and filters and digital signal processing in these.

Why would it affect so many people? Go look inside the average MDF in the exchange or the box on the street. Total shit show.

Basically I think to reason with it, you need to stick a spectrum analyser on the end of the line and terminate it at the DSLAM end and see what is left. Anything else is probably water divining or moses grade shit. Now I know from dealing with OR for years that the first thing they do is just swap pairs around in the street and at the exchange and see if that makes it any better. There is no science involved until you get fairly high up. I reckon they must have about 5 capable people running the country these days.

But really at the end of the day this is really all down to the fact we're hacking massive amounts of bandwidth out of signalling systems that were designed 100 years ago. Bring on fibre. Mine is queued for installation - 900mbits here I come.

Well I knew that I didn't that kind of shit before, now I positively don't, that is way out of my league. I trained as an electrical engineer, then went onto the higher levels of electrical engineering which also included some basic electronics hence my main interest is in audio work and is directed towards repairing rather than designing something from scratch. Give me a load of relays, push buttons and contactors, overload relays etc and I'm in my element then designing motor control consoles with interlocks etc, each to their own eh.

[Cerebus]

We're a long way from the days of the likes of Tommy Flowers, sadly.

I had a friend, Rick, who used to work at Martlesham Heath who retired in the mid to late 90s. He died back in 2007 and I'm just having one of those "That long ago? Where the heck does the time go?" moments.

[SilverSolder]

The real worry is how long I can keep bringing home new orders with my current SUV before I need get a bigger one !  Best effort so far is 26 DSO's and a bit of other stuff all in one load.    No room for passengers ! 

   https://www.amazon.com/gp/product/B001G4Z5J0/

I highly recommend one of these; tripled the cargo space (behind the back seats) in our Rav4. Get the biggest Pro (60 x 24" base) model; the 18gal Rubbermaid Roughneck totes fit 4 wide, and the 136L (35Gal) ones fit 2 the long way or even on top of the 4.

Also get the cargo net; 1 for each layer you plan to stack (so 1 or 2) https://www.amazon.com/gp/product/B07BTLL99G/

My cost ( I already had the hitch) was US$150 for the carrier, $US140 for all 6 totes ( I pad premium brick & mortar prices so I could measure first) and another $30 for the nets. they kept EVERYTHING (And a ginormous suitcase) very securely in place all the way from Tejas to Toronto.

My Draw-Tite 75235 Class III receiver hitch was US$130 from Uhaul; they were cheapest by far at the time.

mnem
 

Putting a hitch on my old Escape was one of the smartest things I ever did - it comes in handy so often.  Surprisingly the most common use case has been towing other people out of snow drifts etc.! 

[bd139]

On ADSL/OR, I suspect this is a load of steaming bollocks. 612KHz is one things that seems to have some significance like water divining or organised religion. There is no way that it is a deterministic way of working anything out. The ADSL signal roughly consists of a bunch of tones in channels transmitted over a differential pair (note differential, which should eliminate common mode noise!!!!). Each channel is numbered. 621KHz is "tone 144" which in the UK is allocated to downstream data. Very little significance other than it's a downstream channel. If there is a problem with that channel through RFI, the line isn't going to fail out unless the equipment is faulty at either end and can't negotiate around the noise which results in less tones/channels and therefore less bandwidth. It'd still work. Unless something is massively fucked.

But there are two conditions where it can cause problems. Firstly that is the front end of the ADSL modem is overloaded to fuck which causes the differential signal to exceed the capability of the input of the router. The input side is usually a little transformer and a couple of fast opamps that drive an SDR type thing these days apparently. The second one is there isn't a differential signal present!

I reckon it's mostly the latter that's the problem i.e. a shitty pair because I had some fun with a scope on a duff line that kept dropping and found out that one of the pair conductors was eaten through by bird shit. That meant there was no common mode rejection at all. You could see it on the line clearly that one conductor in the pair was a nice antenna. I'm surprised it worked at all and it's a tribute to the guys who design the front end and filters and digital signal processing in these.

Why would it affect so many people? Go look inside the average MDF in the exchange or the box on the street. Total shit show.

Basically I think to reason with it, you need to stick a spectrum analyser on the end of the line and terminate it at the DSLAM end and see what is left. Anything else is probably water divining or moses grade shit. Now I know from dealing with OR for years that the first thing they do is just swap pairs around in the street and at the exchange and see if that makes it any better. There is no science involved until you get fairly high up. I reckon they must have about 5 capable people running the country these days.

But really at the end of the day this is really all down to the fact we're hacking massive amounts of bandwidth out of signalling systems that were designed 100 years ago. Bring on fibre. Mine is queued for installation - 900mbits here I come.

Well I knew that I didn't that kind of shit before, now I positively don't, that is way out of my league. I trained as an electrical engineer, then went onto the higher levels of electrical engineering which also included some basic electronics hence my main interest is in audio work and is directed towards repairing rather than designing something from scratch. Give me a load of relays, push buttons and contactors, overload relays etc and I'm in my element then designing motor control consoles with interlocks etc, each to their own eh.

I wish I didn’t know sometimes. The inner workings and forward engineering of opaque bits of technology has been an unfortunate responsibility for many years, initially EE and now IT faff. The more you know the more you die inside.

We have built so much complexity into this world and it’s going to fall apart one day from the aggregate failure risk.

[Saskia]

well, what could possibly go wrong ...

[bd139]

Zombies

[mnementh]

Erm, not 'backbone' as these are last mile services, literally as far away from the backbone as it's possible to get. Pedantry about terminology I'll grant you, but as you're being pedantic about their terminology I declare this "fair game". 

K.

mnem

[bd139]

We're a long way from the days of the likes of Tommy Flowers, sadly.

I had a friend, Rick, who used to work at Martlesham Heath who retired in the mid to late 90s. He died back in 2007 and I'm just having one of those "That long ago? Where the heck does the time go?" moments.

Had one of them the other day as well. A accidentally stumbled on an obituary of a former colleague and friend of mine back in the mid 1990s at my first "proper job". Turned out he died in a car accident in 2017. I shall quote him here:

"One day we'll all have ARM supercomputers in our pockets that can do anything"

Someone bet him £20 that wouldn't happen. I hope the fuck he cashed that in before he popped it.

Edit: random one. I was so damn busy today untangling wads of shit that I actually ate dinner twice  . Looks like I'm going to be a fat bastard again 

[mnementh]

If you were a Hobbit, you'd be starving yourself; only eating 4 squares a day. 

mnem
 

[bd139]

Well that explains that! 

Ok so here we go. med has given me a hankering for a Tek scope posting all those glorious innards. Let's see how it goes 

(I'm waiting for the ass kicking)

[mnementh]

CAD$179:   https://www.kijiji.ca/v-view-details.html?adId=1522100956

mnem
 

[factory]

I have not encountered this term before, which I find suspicious. EMC/EMI is a topic I really strive to keep current in.
Citation:
'The TV was found to be emitting a single high-level impulse noise (SHINE), which causes electrical interference in other devices.'

The remainder is here:
https://www.bbc.com/news/uk-wales-54239180

As far as other news are concerned, I am feeling quite sick  Seems I have lost in the preservability fight yesterday. WTF, it did not taste as bad as it feels!

If it's related to openreach it says their PR team or road monkeys made up the term on the spot.

It took 18 months to fix. That's about right for them 

According to el Reg it was longer. They took an unspecified period of time from when the issue was first reported to rip out all their cabling and start again, then another 18 months from there to recognise and diagnose that it was an EMI problem. Knowing BT that unspecified initial period was probably another 18 months. So, it's not just BT that hire monkeys, it looks like the BBC news also does now, to have swallowed the spin and publish it as only 18 months in total.

The original Bodge-itTelecom Openretch press release is here.

Local Openreach engineer, Michael Jones, was determined to find the Aberhosan broadband smoking gun – so he decided to call on the assistance of colleagues from Openreach’s Chief Engineer team – the company’s telecoms equivalent of the ‘SAS’.

Rough translation: "We finally handed the problem over to 'engineers' who can actually read and write, and seldom use crayon to do so."

I once had an Openretch 'engineer' come to fix a problem on my line who did not know what I was talking about when I described the noise I was hearing on the line as "white noise".

What I haven't seen mentioned in any of the articles about this but I'm really curious about:  The TV set itself.  What sort of TV was it and why was it radiating such powerful interference?

OK the BBC fake news channel article said they didn't want to be identified, but they must have been happy to tell the story to a grotty tabloid paper (Sun), apparently the second-hand TV was a 16" Bush flat-screen model, so not that old. I guess anything made over 6 months ago is old to those overpaid BBC morning news presenters.
Why they told the paper they disposed of it in a builder's skip I don't know (if that is the truth), clearly neither the owner or journalist are bothered that it hasn't been sent for proper disposal under the WEEE regs.  Maybe the WEEE police will send them a nice fine. 
Now we will never know if it was just faulty or missing interference filtering components.

David

[bd139]

CAD$179:   https://www.kijiji.ca/v-view-details.html?adId=1522100956

mnem
 

Ahh that can fuck off with its horrid power supply. I have spent tooooo much of my life having the shit scared out of me by exploding MOSFETs in the 2235 

OK the BBC fake news channel article said they didn't want to be identified, but they must have been happy to tell the story to a grotty tabloid paper (Sun), apparently the second-hand TV was a 16" Bush flat-screen model, so not that old. I guess anything made over 6 months ago is old to those overpaid BBC morning news presenters.
Why they told the paper they disposed of it in a builder's skip I don't know (if that is the truth), clearly neither the owner or journalist are bothered that it hasn't been sent for proper disposal under the WEEE regs.  Maybe the WEEE police will send them a nice fine. 
Also we will never know if it was faulty or missing interference filtering components.

David

Nailed it.

If it's the turd I think it is, it probably wasn't even the TV that was wonky. There is a standard 19V power brick for them which was probably made by the lowest bidder.

[med6753]

Well that explains that! 

Ok so here we go. med has given me a hankering for a Tek scope posting all those glorious innards. Let's see how it goes 

(I'm waiting for the ass kicking)

 

Bd and Tek scopes:

Buy it.

Look at it.

Like it.

Use it.

Wake up one day in a bad mood.

Hate it.

Talk trash about it.

Sell it.


......Rinse and repeat.   

[factory]

OK the BBC fake news channel article said they didn't want to be identified, but they must have been happy to tell the story to a grotty tabloid paper (Sun), apparently the second-hand TV was a 16" Bush flat-screen model, so not that old. I guess anything made over 6 months ago is old to those overpaid BBC morning news presenters.
Why they told the paper they disposed of it in a builder's skip I don't know (if that is the truth), clearly neither the owner or journalist are bothered that it hasn't been sent for proper disposal under the WEEE regs.  Maybe the WEEE police will send them a nice fine. 
Also we will never know if it was faulty or missing interference filtering components.

David

Nailed it.

If it's the turd I think it is, it probably wasn't even the TV that was wonky. There is a standard 19V power brick for them which was probably made by the lowest bidder.

I hope they didn't keep the PSU then, those are the kind of thing that gets saved and chucked in a box for future use.

David