Test Equipment Anonymous (TEA) group therapy thread

[DH7DN]

1.5 years ago at the 2019 Maker Faire Hannover (Germany), a faire visitor (he was probably working in the audio/microphones sector) we never met before offered us a huge box of service manuals for free (they ought to be thrown away). I picked some Tek manuals, just for fun. The manuals are very comprehensive and may be used for circuit studying etc...

[...]

Greetings to our friends from Wellenkino!

Oh, my ass and belly is on your photo    The fellow offering the oversized box with manuals is the senior development  manager of Sennheiser, IIRC. Many thanks to him!

Regards, Bernd

Oh well, our world is really small ;-)))   Hi Bernd, I really had to laugh on this one 
Thanks for the hint, I think you're right 

I'm really afraid of deep rabbit holes - but that's the price of being a TEA   

73, Denis

[AVGresponding]

Is there a recommended manufacturer to look at to replace the originals?

Now that I have it all back together, it has changed behavior again. Now +130, -15, and -50 are up but the rest are still screwed up [edit], they appear to be shorted or at least with very low resistance inconsistent with the nominal values. The bizarre crt behavior has stopped too. Part of me wonders if there is something mechanical going on, on top of bad tantalums.

Any of the beaded tants from Mouser are fine. Usually AVX or Kemet.

Edit...what tautech said above. 

Yeah I was going to say AVX. Not about tants in particular, but in tests of nominally equivalent capacitors from several manufacturers AVX always seem to come out on top for performance in often unspecified qualities like dielectric absorption. It's no accident that you often see AVX parts specified for critical bits of circuitry where consistent, reliable parts are needed. Nothing wrong with Kemet, it's just that AVX always seem to come out on top.

AVX ones also seem to be fairly difficult to turn into a firey inferno. They just get really hot and fall off the board. I haven't been able to get one to make fire come out either in reverse current and overvoltage scenarios.

Reported to PETE (People for the Ethical Treatment of Electronics) for component abuse.

[cyclin_al]

Time to tighten the hatches down folks I reckon that as of midnight tonight or tomorrow, we will be in another lockdown which will be enforceable to remain at indoors unless you are a key worker, going out to get medication or food shopping, with police forces doing the enforcing.

SWMBO works in the local Crown Court, where life goes on as usual, juries are still sitting, court cases still going on, and they have already had something like 8 or 9 positive cases among staff there. How many people work in these regional courts?

Surely it's time to consider moving Crown Court sittings onto electronic platforms like Zoom, with 4 or 5 courts in operation per day, thats 48 or 60 members of the public being introduced into an already fairly cramped environment for several hours a day plus all the required officials. Similarly, with schools, universities etc is it any wonder that the graph for new cases is now flattening out, but only if you rotate the graph clockwise by 90 degrees.
Forget the so-called economy, put people first and foremost, the economy will take care of itself, as long as the people survive, the economy will pick up again.

Edit: The PM, Boris Johnson is to address the nation tonight at 8.00pm.

All sadly, quite predictable. There have been several things going on that anyone with two braincells to rub together ought to have predicted.

The insistence on getting kids physically back to school was always going to end in disaster. Anyone who has had children of school age, or knows anyone who has had children of school age, knows that if there's "anything going around" that kids spread it like wildfire. So insisting on cramming kids back into school (you know, the people who aren't yet mature enough for their to be any hope that they will follow infection control measures rigourously, either from lack of understanding or poor impulse control) was always going to create a wave of new infections. So, kids went back to school in September, the infections rates started climbing again almost exactly one incubation period later. It was completely predictable and completely avoidable.

Yesterday our glorious leader said that he "could not have reasonably predicted" the new, more transmissible variant of the virus that has emerged over the autumn. Weird then that I was able to predict that, on here, back in March. If I know that mutated strains are inevitable one would hope that the experts advising the government made that possibility very clear to them. Either SAGE are idiots, or the government is deliberately ignoring the advice they get from them, or are so incompetent that they can't understand the advice. Whichever it is, someone, somewhere has blood on their hands from not having the decency to say "I'm not up to this, put someone better in charge".

I completely agree as well.

Here, they claimed that the kids were much less likely to get sick and much less likely to have major illness when they re-opened the schools.  Later, the data has shown this to be incorrect.  Even if they were correct, kids still spread stuff like wildfire exactly as Cerebus described.  Okay, so maybe the kids themselves would not be significantly affected, but it was still using kids as a transmission vector to infect the families with kids, and so on. 

We have been hearing from health officials here that if there are early signs of rising numbers, the appropriate action is to shutdown for the 2 week duration of the incubation period.  Instead, we are getting the same thing here.  The economy got propped up for Christmas shopping period; now that the shopping period is over we are immediately thrown into a 6 week shutdown.  The people complaining about economic impact are still complaining anyways; nothing makes them stop.  Looking longer term, I suspect 2 weeks would have been less painful overall than 6 weeks, but who in politics thinks about anything except the immediate short term?

My work situation also appears to match with

deliberately ignoring the advice they get from them, or are so incompetent that they can't understand the advice

.  Thanks for all the advice on the topic.  It mainly confirmed a few things for me.  I am very well versed in CYA (Cover Your Ass) in terms of keeping notes of my own and keeping notes officially within the company system.  Select management people have also been advised with a record.  The situation is a parallel to my work, so not part of my job.  If a situation directly affected me and was intolerable, I would walk.  In fact, I have done so in the past and happens to correspond exactly with commencing adventures such as cycling across the continent or self-employment as a ski racer.  This case is not at all extreme as that.

Ah-ha, maybe something like this is needed: 

[Cerebus]

Have you overlooked that the LM1078 has a symmetrical supply, +15 V, -15 V? There should be no problem with the negative input. The zener is just there in case the current sense resistor opens up (again). I try to learn from my mistakes.

No, the schematic shows it taking its negative supply from common, output LO, but you seem to have realized this:

Ouch. The boo-boo is in the schematic, not the design. Pin 4 is indeed at -15 volts. The original schematics are hand-drawn. The kicad ones were done post-building. Obviously, I made some gross mistakes transferring them.

[snip stuff about slew rates]

But doesn't that apply to rising the voltage only? Lowering is accomplished by Q1/Q3, which can certainly pull more than a few milliamps. This seems to me to be the important direction: In case of a sudden short at the output, get those MOSFETs shut off quick!

Yup, it's rising slew rate only. It's not wrong, it just instinctively feels like "not a lot of gate drive current". As I said, as long as there's enough practical transconductance there (which I can't be bothered to work out all the necessary conditions for) you've got a decent slew rate on the output. Your falling slew rate is going to be determined entirely by your load characteristics, because even if you can turn off the current through your pass transistors instantly you've no active down-regulation.

[med6753]

Revisiting to re-cap or not on this Sansui Amplifier.

Upon closer examination it appears that the 6800uF caps are epoxied in place with that brown stuff that is known for going conductive. Plus I'm seeing some corrosion of the copper traces in that area. So that makes the decision easy. They have to go. Currently they are Nichicon 85 C. All getting replaced with Nichicon 105 C. BOM drawn up and ready to be ordered.

[factory]

Nice choice with the HP 130C, my Heathkit curve tracer gets used with the ex-British Rail HP 140A*, the X-Y 1405A module I use with it came from the US years ago, it needs a bit of warm up time to stop the trace drifting.



More replacement tunnel diode testing for the HP 5248M. 


*Maybe I'll put the original CRT back in it one day, currently has the 140S CRT fitted.

David

That's a nice looking stack of HP gear there!  What was the reason for switching out the CRT?  I'm kind of curious if you happen to know what kind of work the lab at British Railways was doing that it came from.  It's always interesting to hear what the history of vintage equipment like this is, where it came from and a rough idea of what it was being used for but it's pretty uncommon, at least in my experience, to know what the background is since older equipment's usually changed hands several times by the time it gets to me.  Unfortunately that's the case with the 130C and I don't know what the story behind it is.

I have a Tektronix 7704A that came from a medical equipment repair business that replaced it with a Siglent and a Tektronix 7633 that came with a scopemobile that was in the maintenance shop at the Kellogg's factory down in London, Ontario that was left behind when they closed the plant down and move it to Mexico.

Not been on here for a couple of days, got some catching up to do.

The CRT swap in the HP 140A was done when the 140/141 based spectrum analyzers were still quite expensive, oddly 140 oscilloscopes weren't (hence why I have a few). I had some crazy idea of using a 140A with the spectrum analyzer CRT and finding the modules to go with it, I think I got an RF module & broken IF module at the time, but the multi-way connector in the IF module turned out to be buggered so it never got used.
About two or three years ago I found a complete working 141S for not too much instead, since then I have acquired a 141T with the tracking generator. And another 141T with the job lot of TE last year.

I don't know exactly what the 140A was used for at British Rail, but the guy I bought it from said his son had rescued it & some other equipment that was being thrown out many years at BR in Derby, I do know that BR Derby was where a lot of research & development took place for new trains & well as being one of the sites where a lot of new trains were built, it could have also been used on the shop floor for testing parts of the trains being built, or once complete before they went into service.
It came with a 1400A & 1402A vertical modules and a pair of 1422A timebase modules. Apart from a missing vacuum bulb in one of the timebases all the modules were in working order, the timebase was also working once I found a replacement.
Apart from fitting that CRT the only other thing I've done is refitted a PH163 socket, as I nearly dropped the scope when I tripped on the fixed cable the previous owner had fitted.

A couple of pictures of the modules that came with the 140A.



David

[factory]

Mother of God...

Thats actually quite mild compared to some I've seen. Most likely clean up quite well, some soapy hot water and an old toothbrush, fibre glass pen, IPA and then a light smearing with petroleum jelly should keep it going for a while, unless you can swap out the battery holder?

Heh, of the batteries has just disintegrated. Like, only the plastic holds the remains together.

Can't replace the compartment, integral part of the housing.

Quite a mess, hopefully it'll clean up OK, got the larger version of those batteries in my shed radio & a torch.
I've cleaned my brother's front bike light twice now (picture taken after I started to wash them), last time was a nice way to remember Maplin's. 


David

Those Cateye Halogen bike lights were quite good back in their time.  That model must date back to 1993.  So 25 years at least and still going. 
Is he still using the original halogen bulb?
Newer designs switched to LEDs, but the plastic became much more fragile.

That one isn't quite that old, maybe early 2000's when LED lights were quite expensive or didn't exist  I can't remember.
Yes it still has the original halogen bulb, I don't think it gets used too much in the winter, hence why old batteries have made a mess twice over the years, the last time it was still working despite the leakage. 
These days my brother tends to use the train or bus more than the bike.

David

[Specmaster]

Today, I thought I'd try my hand at some bread making, using a bread making machine, I made a couple of 2lb loaves last week and as I'm the only one who really eats bread, a 2lb loaf is a little too big and some of it gets thrown away. So today I thought I'd get the other machine out, one that has 2 x 1lb tins in it, so I could freeze one and eat one. All went well, bread was rising well and proving nicely and had all indications of being 2 first class loaves, until the mince switched into cooking mode and the house was plunged into bleeding darkness. It had knocked out the main RCD and every time I reset it, the machine took it out again 

Long story short, the element has a 1M leakage to ground, no visible sign of damage anywhere, it just seems to have absorbed some moisture over the years of being in storage in a cupboard. The element is it seems totally unobtainable now, and the annoying thing is that if I didn't have that main RCD installed on the complete consumer unit, it would have worked perfectly

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

[Specmaster]

Revisiting to re-cap or not on this Sansui Amplifier.

Upon closer examination it appears that the 6800uF caps are epoxied in place with that brown stuff that is known for going conductive. Plus I'm seeing some corrosion of the copper traces in that area. So that makes the decision easy. They have to go. Currently they are Nichicon 85 C. All getting replaced with Nichicon 105 C. BOM drawn up and ready to be ordered.

Yeah, that brown epoxy stuff is a real bastard, and it can be a bastard to remove as well.

[cyclin_al]

You might well have been a model student when you went,

  You talk like a man who doesn't know me at all. 

Well you clearly speak with a forked tongue then for you said and I quote

Quote from: Cerebus on Today at 05:46:47 pm

    Indeed. The difference is that every community has several schools whereas the universities are relatively isolated, as are the students, who go home once a term, whereas schoolkids go home every day (and to their Gran's and their neighbours, and the local shops and so on). Even in the local community, there were times I went weeks at a time without leaving campus when I was at university. Universities compared to schools, are a trivial vector - that doesn't mean they should be ignored, but in terms of what's most likely to be damaging to the population as a whole it's the schools that need tackling first.

So which one is it then, a rebel or a model student were you 

There's sometimes no reason to leave the campus. There's computers, booze, usually a shitty SU nightclub, one or more shops, local pizza place and floozies. Plenty of time to to cause trouble there 

Likewise.  How else would one know that it takes 40,000 crumpled campus newspapers to fill an on-campus residence room?     Girlfriend of said resident shouted "Help, I can't swim!" on returning from the on-campus pub, and being pushed by very considerate friends into the sea of newspapers.... 

[BU508A]

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

Perhaps cleaning it first with some IPA could help as well to get out the moisture first and then using the lamp to dry it.

[Specmaster]

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

Perhaps cleaning it first with some IPA could help as well to get out the moisture first and then using the lamp to dry it.

I will clean the exposed part of the element (the part where the connections are made) although the element is not truly exposed, it is still enclosed in some form of silicon or something, the rest is enclosed in a metal jacket. I think the last time it was used was about 15 years ago and has been stored on the top shelf in a cupboard in the centre of the house which is centrally heated, so it is not condensation type of dampness that is tracking externally.

[beanflying]

A couple for the AVO nuts eBay auction: #274631005969



eBay auction: #224300221485

[mnementh]

AVX ones also seem to be fairly difficult to turn into a firey inferno. They just get really hot and fall off the board. I haven't been able to get one to make fire come out either in reverse current and overvoltage scenarios.

Reported to PETE (People for the Ethical Treatment of Electronics) for component abuse.

I'm sure they'll all have a good laugh at your report over beers at bd's next BBQ.

mnem

[mnementh]

just spent couple hundred on a VU33P. *sniff* Gotta write if off as expenses.

Now let's see if I can do OpenNMT -> FINN as in https://xilinx.github.io/finn/

I think my mind just boggled.

This is to bring neural-net deep-learning to FPGAs...? Which seems like an oxymoron as FPGA logic is fixed-state...?

I get that FPGA is exponentially faster... but most of the really great breakthrough tech using FPGA right now is in autonomous robotics, wherein they serve as the central nervous system to a fast, usually parallel-processing CPU which is doing the deep-learning...

mnem
*gives look like a tree full of owls*

No, you train your model elsewhere and then deploy it on FPGAs.

Yes, I know that's how you're supposed to do it. But did you read the git Saskia posted...? It looks like the entire point of the project is to process the deep learning... ummm... processes with FPGAs in real-time. Or are they talking aboot doing the learning in conventional processor-space, and reprogramming the FPGAs on-the fly...?

Please, don't think I'm arguing here; I'm just trying to wrap my head around this project. What Saskia does with machine translation is a field that has long intrigued me, but I have enough trouble just grokking arrays and tables. This stuff is like doing actual processing work within the voodoo confines of a hash function; it ties my brain up in knots. If you can ELI5, I would be grateful.

mnem
 

[Cerebus]

just spent couple hundred on a VU33P. *sniff* Gotta write if off as expenses.

Now let's see if I can do OpenNMT -> FINN as in https://xilinx.github.io/finn/

I think my mind just boggled.

This is to bring neural-net deep-learning to FPGAs...? Which seems like an oxymoron as FPGA logic is fixed-state...?

I get that FPGA is exponentially faster... but most of the really great breakthrough tech using FPGA right now is in autonomous robotics, wherein they serve as the central nervous system to a fast, usually parallel-processing CPU which is doing the deep-learning...

mnem
*gives look like a tree full of owls*

No, you train your model elsewhere and then deploy it on FPGAs.

Yes, I know that's how you're supposed to do it. But did you read the git Saskia posted...?

Yes. On the very first page:



Oh, and since when did "read a git" become a phrase? That's like saying "read a gcc".

It looks like the entire point of the project is to process the deep learning... ummm... processes with FPGAs in real-time. Or are they talking aboot doing the learning in conventional processor-space, and reprogramming the FPGAs on-the fly...?

Please, don't think I'm arguing here; I'm just trying to wrap my head around this project. What Saskia does with machine translation is a field that has long intrigued me, but I have enough trouble just grokking arrays and tables. This stuff is like doing actual processing work within the voodoo confines of a hash function; it ties my brain up in knots. If you can ELI5, I would be grateful.

mnem
 

The ELI5 is:

• Build and train the model using PyTorch and your own training data. (that's the learning bit, whether it's deep or not, or whether that's a noise word added to make it sound fancier is highly debatable.)
  
• Use the tools from the FINN git repository to convert the model into a bitstream for a Xylix FPGA.
• Deploy the model on a Xilinx FPGA.

[mnementh]

Okay, I misunderstood the source of the data.  I wouldn't have asked if I did understand.    I thought the point of the project was a new learning model/methodology in and of itself.

Thanks for taking the time. Now I'm a little less ignorant.

"The git" is just a term some of the programmers I've worked with on FC and ESC projects used to refer to any specific repository. I figured it was common slang.   

mnem
 

[AVGresponding]

Today, I thought I'd try my hand at some bread making, using a bread making machine, I made a couple of 2lb loaves last week and as I'm the only one who really eats bread, a 2lb loaf is a little too big and some of it gets thrown away. So today I thought I'd get the other machine out, one that has 2 x 1lb tins in it, so I could freeze one and eat one. All went well, bread was rising well and proving nicely and had all indications of being 2 first class loaves, until the mince switched into cooking mode and the house was plunged into bleeding darkness. It had knocked out the main RCD and every time I reset it, the machine took it out again 

Long story short, the element has a 1M leakage to ground, no visible sign of damage anywhere, it just seems to have absorbed some moisture over the years of being in storage in a cupboard. The element is it seems totally unobtainable now, and the annoying thing is that if I didn't have that main RCD installed on the complete consumer unit, it would have worked perfectly

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

Do you have a cooker socket you could use? Those shouldn't normally be rcd protected, or at least should be 100mA types because it is a known thing for heating elements to absorb moisture over time, and have quite low (I'd have to drag my IET PAT manual out for the exact figures) resistance to earth, certainly well below 1M.
In fact PAT testing (yes I know) is a bit of a ball-ache on any appliance of this nature as you are supposed (most don't bother) to test the earth leakage from cold and also once it's warmed up, which obviously takes time.

I once got a bollocking (which ofc I ignored) for condemning some ovens in a school food tech lab. One had a resistance to earth of 0 ohms. Yeah, no m8, I'm not turning that one on to warm up.

[mnementh]

Not sure 100W in series is gonna develop enough heat to do jack... might need to double 'em up with a 2-fer socket to get enough current flow. Still, that is probably how I'd do it.

Naaah, who am I kidding... I'd put it across the ol' Scariac at aboot 60% with a right scary-arse suicide cord connection.     2-wires, alligators, ungrounded of course.

mnem
tzzt.

[Neomys Sapiens]

Today, I thought I'd try my hand at some bread making, using a bread making machine, I made a couple of 2lb loaves last week and as I'm the only one who really eats bread, a 2lb loaf is a little too big and some of it gets thrown away. So today I thought I'd get the other machine out, one that has 2 x 1lb tins in it, so I could freeze one and eat one. All went well, bread was rising well and proving nicely and had all indications of being 2 first class loaves, until the mince switched into cooking mode and the house was plunged into bleeding darkness. It had knocked out the main RCD and every time I reset it, the machine took it out again 

Long story short, the element has a 1M leakage to ground, no visible sign of damage anywhere, it just seems to have absorbed some moisture over the years of being in storage in a cupboard. The element is it seems totally unobtainable now, and the annoying thing is that if I didn't have that main RCD installed on the complete consumer unit, it would have worked perfectly

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

Do you have a cooker socket you could use? Those shouldn't normally be rcd protected, or at least should be 100mA types because it is a known thing for heating elements to absorb moisture over time, and have quite low (I'd have to drag my IET PAT manual out for the exact figures) resistance to earth, certainly well below 1M.
In fact PAT testing (yes I know) is a bit of a ball-ache on any appliance of this nature as you are supposed (most don't bother) to test the earth leakage from cold and also once it's warmed up, which obviously takes time.

I once got a bollocking (which ofc I ignored) for condemning some ovens in a school food tech lab. One had a resistance to earth of 0 ohms. Yeah, no m8, I'm not turning that one on to warm up.

Yeah, my thought has been similar. I would have wired a temporary non-rcd outlet into my switchboard.

[Ero-Shan]

But doesn't that apply to rising the voltage only? Lowering is accomplished by Q1/Q3, which can certainly pull more than a few milliamps. This seems to me to be the important direction: In case of a sudden short at the output, get those MOSFETs shut off quick!

Yup, it's rising slew rate only. It's not wrong, it just instinctively feels like "not a lot of gate drive current". As I said, as long as there's enough practical transconductance there (which I can't be bothered to work out all the necessary conditions for) you've got a decent slew rate on the output. Your falling slew rate is going to be determined entirely by your load characteristics, because even if you can turn off the current through your pass transistors instantly you've no active down-regulation.

Actually, the drivers do pull down the output, albeit through the zener and 1.2 k resistor. If the output gets shorted, I don't really care about regulating the voltage down - my major concern then is to protect the pass transistors.
And this can also be said about most other lab supplies.

BTW, another possible cause for my "mishaps" occurred to me: the dangers of open source. What if the H6N100 wasn't overpowered but the source resistor failed open due to overcurrent? V_GDR may be 1000 volts, but with R=∞ it might approach V_GS, a mere 20 volts. That would explain gate-drain breakdown and subsequent bursting of the gate resistor!

[cyclin_al]

I have a failed SSD from the mother-in-law.  It is an ADATA SU650 from a UEFI system running Windows 10.
The drive no longer shows any partitions when connected to a computer 

In searching for recovery tools, I can run GNU DDRescue with DDRescue-GUI on the drive using my Ubuntu family computer.
Good idea?  Other options to consider?

--Please sir, can I spend just a little bit of time playing with TEA?

[Saskia]

Okay, I misunderstood the source of the data.  I wouldn't have asked if I did understand.    I thought the point of the project was a new learning model/methodology in and of itself.

Thanks for taking the time. Now I'm a little less ignorant.

"The git" is just a term some of the programmers I've worked with on FC and ESC projects used to refer to any specific repository. I figured it was common slang.   

mnem
 

No worries.

It will be a bit more difficult anyway.
This FPGA gizmo comes on a PCI-E FPGA mining card which means that it does not have the fancy connectors  that you would want for all those I/O functions you'd normally want to use.

This means that I have to find a way to repurpose the card to
1) use as a more or less generic accellerator card
2) find a way to access it from Vivado
3) pytorch is dead as far as I am concerned (and as NMT goes). The framework to be used is tensorflow.
4) hack OpenNMT to use tensorflow, integrate it into my toolchain and do a PoC tensorflow training for a popular language, let's say Klingon
5) set up some basic test suite to deploy the tensorflow model (and test it), ideally with a REST API
6) remodel that stuff and check if I can use the xDNN or the FINN toolkit to generate deployable FPGA models with it
7) test the POC on the FPGA
if successful run a real life test with a larger model (let's say English -> French with 15 million data sets out of a 275 million pool)
9) check if training can be ported to such FPGA (if not, well, too bad ...)
10) benchmark it against what we have now (4x Tesla P100 for training, 400 CPU cores for the deployed models)
11) find someone who pays for all of that ...
 

[Specmaster]

Today, I thought I'd try my hand at some bread making, using a bread making machine, I made a couple of 2lb loaves last week and as I'm the only one who really eats bread, a 2lb loaf is a little too big and some of it gets thrown away. So today I thought I'd get the other machine out, one that has 2 x 1lb tins in it, so I could freeze one and eat one. All went well, bread was rising well and proving nicely and had all indications of being 2 first class loaves, until the mince switched into cooking mode and the house was plunged into bleeding darkness. It had knocked out the main RCD and every time I reset it, the machine took it out again 

Long story short, the element has a 1M leakage to ground, no visible sign of damage anywhere, it just seems to have absorbed some moisture over the years of being in storage in a cupboard. The element is it seems totally unobtainable now, and the annoying thing is that if I didn't have that main RCD installed on the complete consumer unit, it would have worked perfectly

Thinking about cobbling up a 500VA isolating transformer wired directly to the element, via 100W bulb in my dim bulb tester with a 100W bulb fitted and let it cook for a while and hopefully restore the element (875W) to a workable state again once the moisture has been driven out, sound feasible?

Do you have a cooker socket you could use? Those shouldn't normally be rcd protected, or at least should be 100mA types because it is a known thing for heating elements to absorb moisture over time, and have quite low (I'd have to drag my IET PAT manual out for the exact figures) resistance to earth, certainly well below 1M.
In fact PAT testing (yes I know) is a bit of a ball-ache on any appliance of this nature as you are supposed (most don't bother) to test the earth leakage from cold and also once it's warmed up, which obviously takes time.

I once got a bollocking (which ofc I ignored) for condemning some ovens in a school food tech lab. One had a resistance to earth of 0 ohms. Yeah, no m8, I'm not turning that one on to warm up.

No I don't have any unprotected sockets since the LA replaced the consumer unit a few years from a Crabtree Starbreaker split load board for a Sq D twin RCD split load board in order to provide fire protection for the lighting circuits. I do have a 500VA (2A) variac that I could bring into play via a lead with croc clips and feed the element with about 100-120V and let it dry out without earthing the bread maker (as per Dwagons suggestion, which I thought about last night as well).

[BU508A]

*gives look like a tree full of owls*

Is it like this?


Or like this?


Or more like this?