Test Equipment Anonymous (TEA) group therapy thread

[Ice-Tea]

Now, if you'll excuse me, I need a change of underwear.

Meh, very little power. Will just tickle a bit.

Edit: No, won't touch it (the HV output) anyway.

What makes you think the 0.6µA is a current limit, rather than the current?

The control at the top tight shows the overcurrent trip is 360µA, i.e. 10W at 30kV.

Correct. Even though I'm not sure it'll deliver that at 30kV. The 0.6uA was "lost" in electrostatic discharge, some cracklin' in the air, hence the change of underwear.

Sold damn quick. Was a bit afraid, but banked on the "Megger" name.

[Specmaster]

3 digit? Meh. That's not far off analogue meter accuracy

My 34410A is currently showing agreement with my Transcell voltage standard (as measured by a PTB hp3458a that had been calibrated the previous day) to 6 digits (9.999577 vs 9.999571).

I wonder how long that will last?

I learned, perhaps way too late, that when you're doing normal electronic design, that the odd 10% here and there doesn't matter that much.

Exactly that, you nailed it right there.

[bd139]

Booorrrrriiiiinnnngggg!

I learned very early on that the real trick is to get predictable operation with components that show hideous parametric variation, transistor h_FE V_be being the canonical examples. And then to extend that to system-level behaviour.

Of course I never selected a resistor's value based on how pretty the colour code looks. Oh no, never ever ever.

Nor have I ever checked a PSU rail's voltage by merely noting how fast an analogue meter starts to move.

The concept of measuring relative charge using an slow moving ammeter is completely foreign to youngsters.

Oh yes for certain. I honestly get quite grumpy and annoyed when I see a shitty circuit these days and quite ranty as this comment is probably going to turn into. It's actually really easy to get very predictable operation from crappy parts but you have to understand more than Forest Mimms barfed out into his books to do so. Not a lot more!  Ok that didn't get too ranty.

Going to be 100% honest here and one of my favourite books on discrete design is the Wes Hayward book Experimental Methods in RF Design. There is so much good stuff in that book, even if you don't give a crap about RF. Available on Library Genesis if you want a copy. That and TAoE are my desert island books

My favourite coloured resistor is 4.7K

.... added for edit ....

Again people don't think about charge. There is no conceptual modelling. MOSFET gates is the one that makes me laugh usually. MOSFET explodes? Oh that's because it's in the linear region for most of the conduction cycle because the gate is a capacitor and the thing filling it up is a current source. BLANK STARES, DROOL. FFS.

There's 100% more clue in EE than software though.

[mnementh]

OTOH: on the hoarding side, an opportunity to double up on crucial equipment just surfaced. I love listings that aren't compatible with the string search in sales platforms, because of speling erorrs. Makes the audience smaller.

Don't we all, don't we all. The only difficulty with this strategy is coming up some of the more creative misspellings that people manage to make. It's easy to do vowel transpositions, adjacent key errors, the commoner errors such as f/ph, but some people's spelling is in another lexicographic universe entirely.

I actually wrote some de-stemming and idiot simulation code which comes up with creative misspellings based on key proximity and known autocomplete fuck ups based on what I'm interested in. I've done well out of it.

Do you want some HealthKit knobs?

https://www.ebay.co.uk/itm/174444372380

Why reinvent the wheel when its already been done, unless you can make it better 

http://fatfingers.com/

Some things... a dirty mind is much better: fatfuckfingers.com fuckfingers.com fingfuckers.com   

mnem
"A dirty mind is a terrible thing to waste." ~mom

[bd139]

Why reinvent the wheel when its already been done, unless you can make it better 

http://fatfingers.com/

Oh it is better. It does the searches for me every 5 minutes and sends me an email if something unrealistically priced or entered by an idiot appears. Well it did until eBay started adding scraper protection and ramped up their API pricing. 

I'm going to replace it with selenium webdriver when I get around to it so it drives and scrapes off a real chrome instance.

[Cerebus]

I learned, perhaps way too late, that when you're doing normal electronic design, that the odd 10% here and there doesn't matter that much.

There are edge cases, but yeah, 10% isn't merely "good enough" most of the time, a lot of the time it counts as "unnecessary precision".

Most of my analogue design work is initially done on paper, with junior school long arithmetic to no more than 2 digits of precision. Like you, it probably took me longer than it should have to reach the point where I realised this was quite enough precision, and where it wasn't there is "feedback" to make it so.

[bd139]

Yes paper here too.

There's a good trick to that actually that I learned. Try and design your circuits with E3 values only rather than having a panic and having to order some random resistor value. Quickly things start to add up to "meh it's good enough"

Another rant: Lore. 240 ohms for an LM317 between out and adjust 

[Cerebus]

My favourite coloured resistor is 4.7K

Mine is 2K2 2%, easiest to spot in a bucket full of mixed resistors.

[Cerebus]

Yes paper here too.

There's a good trick to that actually that I learned. Try and design your circuits with E3 values only rather than having a panic and having to order some random resistor value. Quickly things start to add up to "meh it's good enough"

Another rant: Lore. 240 ohms for an LM317 between out and adjust 

I use the complete C4 series: 1k 2k2 4k7 10k. 

[Specmaster]

I learned, perhaps way too late, that when you're doing normal electronic design, that the odd 10% here and there doesn't matter that much.

There are edge cases, but yeah, 10% isn't merely "good enough" most of the time, a lot of the time it counts as "unnecessary precision".

Most of my analogue design work is initially done on paper, with junior school long arithmetic to no more than 2 digits of precision. Like you, it probably took me longer than it should have to reach the point where I realised this was quite enough precision, and where it wasn't there is "feedback" to make it so.

Agreed, most capacitors are rated at 20% even so it could be argued that 10% is even too high for most items anyway.

[mnementh]

It's "Columbus Day" here and schools are closed as well as some other places. This is the day we supposedly celebrate Christopher Columbus "discovering" America. Despite the fact that people were already here, what we call "Native Americans". Over the last few years there have been calls to rip down all the statues dedicated to Columbus which has got the people of Italian descent all up in arms because he was Italian working for Spain.   

That's pretty funny actually. Human race is a complete shit show.

Well I didn't get my day off. Turns out someone fudged something so been called out

Edit: fire out. Thinking about snagging one of these. This is mostly to hold boards while I'm working on them because the infernal things keep disappearing due to the weight of test leads etc.

   Anyone got any better solutions because it's the best part of 100 quid? 

https://www.amazon.ca/Aven-17010-Adjustable-Circuit-Holder/dp/B00Q2TTQEE/

This one works well for MOST PCB work, but it has a few limitations. It is mostly plastic, so it will weaken and embrittle with age. Mine hasn't YET, but I know it will. The base parts are steel tube, so it is more rigid than it looks and bottom-weighted JUST enough to be stable for most PCB/assembly work. The rotation clamps work well as long as you use BOTH; use one and the work will spin pretty easily. You just have to be careful every time when tightening them down, especially if you have big ol' hamhands like mine. This my expected point of failure: the plastic cracking out from overtightening the thumbscrews, or just plain repetitive stress fatigue.



That said... I also have that PanaVise... and for the quality, it's a fucking bargain. It's a tool you'll use the rest of your life and never find it wanting, plus the No 303 mini-vise that fits in the same base is just handy as a pocket on a shirt for only $40-ish.

Compared to the above plastic fantastic, it had better bottom-weighting, holds the board much more solidly, and the big thing... you can set the PCB ANY compound angle, so you can get EXACTLY the right angle of attack at a hard-to-reach component or solder pad on densely populated boards.



https://www.amazon.ca/Helping-Soldering-Workshop-Non-Slip-Weighted/dp/B07TTZCZPR

Add to that these little gems, and you have a lab-grade soldering workstation with wiring holders and iron management in a closet-stuffable portable package.



https://www.amazon.ca/PanaVise-324-Electronic-Work-Center/dp/B000SSPNBU

This is the Panavise station made for production assembly work... I've used one, but don't own it. It's all about quick turnover of repetitive assembly work, and worth every penny for that application.

Trust me... you will NEVER regret the money spent on PanaVise... what you'll regret once you have it is not having done it decades ago.

mnem
 

[bd139]

Dang it that's a good sell that. I shall mull on it overnight. Thanks for the encouragement 

[Specmaster]

Dang it that's a good sell that. I shall mull on it overnight. Thanks for the encouragement 

I've also got the Aven one and it is good but a bit fiddly mounting the board and tightening the clamps, other than that it works great.

[BU508A]

@bd139

How about something like the "helfende dritte Hand"?

Example:

https://www.ebay.co.uk/itm/362658381673

[Cerebus]

Don't forget the one Zucca pointed out a few pages back - the one from the kickstarter (or whatever crowd funding platform it was). Looked interesting just from the photos Zucca posted.

[Cerebus]

"There's nobody here but us chickens..."

[mnementh]

Those are great for holding wires... worthless for actually holding a PCB solidly to work on it. I've tried to "get by" with one before; that's why I bought the blue plastic Aven thing at the time.

mnem
 

[mnementh]

I know that I have mentioned about my bench LCR meter, the XLW01 which is really accurate and is inexpensive and can user calibrated using the resistors supplied.

This video provides a excellent guide to using it because it may or may not come instructions depending on where you buy it from, but it is superb alternative to the much liked DE5000 which many people here favour but it does a lot more than this unit and not any more accurate.

Edit, it is geared towards power supplies and audio type of work and NOT RF work.

They also offer DIY kits now        https://www.aliexpress.com/item/4001156913695.html?spm=a2g0s.8937460.0.0.56e72e0eaUaFKS

I shopped that one pretty hard when I bought my DE-5000; I decided against it for several reasons.

1) It doesn't use industry-standard test frequencies. Any values you get from the non-standard frequencies will not directly translate to any comparative testing EXCEPT that done with the same type meter; so not really useful for sharing tech reference or comparing against datasheets.

2) It does ESR and Dissipation, NOT ESR and  tan δ. At lower frequencies the stats are very comparable... however at high frequencies and low ESRs, they diverge greatly. This makes the tester not ideal for the main thing I was trying to get a handle on at the time: what values are "normal" for MMLCCs. Non-standard freqs and no tan δ made it a non-starter for me. Also, Dissipation is also not always given in the datasheets, while  tan δ is, so you can check against a datasheet to confirm if your suspect cap is healthy or not.

3) The DE-5000 comes calibrated, ready to use with actual published specs. Obviously, if you're doing the cal yourself, no such specs apply.

4) No DUT slots. Uggh.

4) And finally... it costs more than a DE-5000. I paid US$110 for mine, with tweezers and alligator heads. At the time, this thing was going for US$130. Even at the kit price, still a "Ehhhhh...." deal. US$62 for a kit from AliEx that you have to build and cal yourself... I'd still take the DE-5000 every time.

And before anybody says it... the DE-5000 is Kelvin connection on every test jig. The slots are true Kelvin with signal at one plate and measurement the other, while the tweezers & alligators are Kelvin to within a few mm of the actual contact point. There are more videos out there than I can count now, all demonstrating that the assumed inherent inaccuracy is not measurable; that consistency in test setup is far more important, which the DE-5000 configuration makes easily repeatable.

I really WANTED to like the XLW01; but right now, I'd for sure spend that money AGAIN on a DE-5000, no question. And before I'd buy the XLW01 kit "for funsies", I'd get me a NanoVNA and a TinySA. 

mnem
 

[bd139]

Was going to mention the nanovna. That is an excellent little LCR meter 

[mnementh]

3 digit? Meh. That's not far off analogue meter accuracy

My 34410A is currently showing agreement with my Transcell voltage standard (as measured by a PTB hp3458a that had been calibrated the previous day) to 6 digits (9.999577 vs 9.999571).

I wonder how long that will last?

I learned, perhaps way too late, that when you're doing normal electronic design, that the odd 10% here and there doesn't matter that much.

Booorrrrriiiiinnnngggg!

I learned very early on that the real trick is to get predictable operation with components that show hideous parametric variation, transistor h_FE V_be being the canonical examples. And then to extend that to system-level behaviour.

Of course I never selected a resistor's value based on how pretty the colour code looks. Oh no, never ever ever.

Nor have I ever checked a PSU rail's voltage by merely noting how fast an analogue meter starts to move.

The concept of measuring relative charge using an slow moving ammeter is completely foreign to youngsters.

Yup. Just like quickly diag-ing a board with only the continuity beep on your meter.

Flukes with latching continuity are especially nice; you get to where you know viscerally just how long a beep is normal for a 47uF vs 470uF vs 1000uF electrolytic cap, and can move very fast doing the "quick & dirty diag" bit of looking for shorts where they shouldn't be and caps that aren't caps just by listening. When something doesn't sound right, that's when you reverse the leads and/or look up to see if paralleled by a resistor, coil, semi junction, etc.

Then if/when you don't find something, you can move on to the more time-intensive diag that requires looking up at your meter to see actual values. 

mnem
 

[mnementh]

Was going to mention the nanovna. That is an excellent little LCR meter 

Does it actually have a routine in the unit for that with ESR, etc or are you just talking about using the capacitance/inductance figures given as part of the load-characterization?

mnem
 

[Specmaster]

I know that I have mentioned about my bench LCR meter, the XLW01 which is really accurate and is inexpensive and can user calibrated using the resistors supplied.

This video provides a excellent guide to using it because it may or may not come instructions depending on where you buy it from, but it is superb alternative to the much liked DE5000 which many people here favour but it does a lot more than this unit and not any more accurate.

Edit, it is geared towards power supplies and audio type of work and NOT RF work.

They also offer DIY kits now        https://www.aliexpress.com/item/4001156913695.html?spm=a2g0s.8937460.0.0.56e72e0eaUaFKS

I shopped that one pretty hard when I bought my DE-5000; I decided against it for several reasons.

1) It doesn't use industry-standard test frequencies. Any values you get from the non-standard frequencies will not directly translate to any comparative testing EXCEPT that done with the same type meter; so not really useful for sharing tech reference or comparing against datasheets.

2) It does ESR and Dissipation, NOT ESR and  tan δ. At lower frequencies the stats are very comparable... however at high frequencies and low ESRs, they diverge greatly. This makes the tester not ideal for the main thing I was trying to get a handle on at the time: what values are "normal" for MMLCCs. Non-standard freqs and no tan δ made it a non-starter for me. Also, Dissipation is also not always given in the datasheets, while  tan δ is, so you can check against a datasheet to confirm if your suspect cap is healthy or not.

3) The DE-5000 comes calibrated, ready to use with actual published specs. Obviously, if you're doing the cal yourself, no such specs apply.

4) No DUT slots. Uggh.

4) And finally... it costs more than a DE-5000. I paid US$110 for mine, with tweezers and alligator heads. At the time, this thing was going for US$130. Even at the kit price, still a "Ehhhhh...." deal. US$62 for a kit from AliEx that you have to build and cal yourself... I'd still take the DE-5000 every time.

And before anybody says it... the DE-5000 is Kelvin connection on every test jig. The slots are true Kelvin with signal at one plate and measurement the other, while the tweezers & alligators are Kelvin to within a few mm of the actual contact point. There are more videos out there than I can count now, all demonstrating that the assumed inherent inaccuracy is not measurable; that consistency in test setup is far more important, which the DE-5000 configuration makes easily repeatable.

I really WANTED to like the XLW01; but right now, I'd for sure spend that money AGAIN on a DE-5000, no question. And before I'd buy the XLW01 kit "for funsies", I'd get me a NanoVNA and a TinySA. 

mnem
 

No No, its perfect for anything except RF because of the frequencies stop at 7.8KHz but it is ideal for audio and power supply and general power type of work and lets face it, most problems with caps are generally either Tants or Electrolytics, shorting to ground or high ESR etc, and both I and the reviewer both stated that it is really suitable for RF work where you are dealing higher frequencies and maybe filters etc. Not everyone is into that side of electronics and so for those people it is an ideal tool and it does indeed come pre-calibrated from the factory but you can also do it yourself at anytime without any need to send it away for an expensive operation.

As to the price, it is far less expensive than a DE5000. https://www.ebay.com/itm/XJW01-Auto-LCR-Digital-Bridge-Resistance-Capacitor-Inductance-ESR-Meter-0-3/124187892736

[mnementh]

$US60 vs $US110... looks like everybody else figured out what I did when I was shopping it.

Nope. Still not willing to give up the tweezers, DUT slots and factory cal for $50. Sorry. 

mnem
 

[bd139]

Was going to mention the nanovna. That is an excellent little LCR meter 

Does it actually have a routine in the unit for that with ESR, etc or are you just talking about using the capacitance/inductance figures given as part of the load-characterization?

mnem
 

You can measure it fairly easy as the L and C portion of the DUT cancels out at resonance leaving the ESR only so it’s probably more accurate.

I just missed two working HP 3456’s for £50 and £65 each elsewhere 

[mnementh]

   

https://www.amazon.ca/4-Digital-Precision-Adjustable-Regulated-Switching/dp/B07MFLYSWL

Currently looking at this 150W RF White Noise Generator Bench Power Supply as it's only CAD$85/US$65 with Prime Next-Day. How much extra fire insurance should I buy along with...?

mnem
What could possibly go wrong...?