Test Equipment Anonymous (TEA) group therapy thread

[Wolfgang]

FFS!

was away for a short time and return to find all the brits have turned into "statler and waldorf" from the muppet show!

and the dragon is now "the voice of reason".

screw this..........am going back down into my "fortress of solitude" and work some more on reducing noise in my 10 Ghz yig oscillator.  (actually managed to fry an lm317 yesterday.  did not expect it to fail short!)

ps  fortunately med is a never changing rock of stability.  (a blue assed NY bastard bitchin' about snow and the postal system).

LM317 is one of those devices with a lot of lore around it. Some of my favourites which I find terribly funny:

1. I can make a 1.2-30V regulator that will kick out an amp with one. Proceeds to then run it with a 40V input and 3.3V output with an inch across heatsink on it trying to absorb 25 watts or so. BOOM!
2. They automatically current limit if they get hot. BOOM!
3. They make a good charger IC for SLA batteries. Line power goes down, missing protection diodes. BOOM!
4. They are universally stable. Runs it with 30 foot of inductors between the regulator and load. BOOM!
5. They are short circuit proo.....BOOM!
6. You can strap a CC circuit behind a CV circuit and make a bench power supply. Oh it's oscillating like fuck. BOOM!
7. It fails open (a thing you found out isn't true). Nope the pass transistor goes C-E short almost always and dumps the entire input voltage across the load.

If you value anything that any linear regulator is connected to regardless of type or vendor or even if it's a commercial power supply from a big brand, you stick a crowbar in circuit.

Fully agree. All this LM317/337 stuff is completely overrated. It might be cheap, but all other properies are crap:
- regulation line/load is *evil*
- tolerance agains abuse (anything with "over" just kills it, may it be current, voltage, power, heat,strange looks, ...)
- can be used as a noise source (several MILLIvolts !!!). What if you dont need one ?

In my stuff the only three-legged thing is the TL431A. If I need serious power, its the LM723J. And lab grade its LM399/LT1028 or better.
I am so old that I got a license to use retro stuff as much as I want. 

[bd139]

Freaking engineers.  Don't you guys ever look at the data sheets? 

LOL who do you think we are? The SOA is better determined experimentally. Can't go trusting those datasheets! 

..may be subject to change without further notice..     

Datasheets are not binding in a legal sense.   
You get most fun of it when you're trying to build a logical network into a CPLD and the manufacturer changed timing properties and parts of the structure/paths. In some sense he improved it but according to the current datasheet I worked with, it was partially out of spec, resulting in a hard to trigger (and hard to observe) race condition that should not even have existed but did.   
Ok, I learned to improve my design to be not dependent on timing boundary values of the underlying hardware..
Was a student then and never used that know-how again.   

Yeah STM32 datasheets. Grrrrrr 

Fully agree. All this LM317/337 stuff is completely overrated. It might be cheap, but all other properies are crap:
- regulation line/load is *evil*
- tolerance agains abuse (anything with "over" just kills it, may it be current, voltage, power, heat,strange looks, ...)
- can be used as a noise source (several MILLIvolts !!!). What if you dont need one ?

In my stuff the only three-legged thing is the TL431A. If I need serious power, its the LM723J. And lab grade its LM399/LT1028 or better.
I am so old that I got  license to use retro stuff as much as I want. 

Agree with LM723. Still a good one that is. Plus easy to use current limiting unlike most other regulators. The LT1083 or LT3080 is my goto if I'm feeling rich 

[med6753]

Here it is. And it works, at least on DC volts. Haven't tried any other functions yet. Last calibration date is November 10, 2003. It's on been on for a little while and I'm going to let it cook for a few hours. But notice something? 9.99716V against the AD584 ref of 9.99691V. Not bad. In fact...fucking spectacular and the peanut gallery that keeps telling me that my assumptions about calibration/accuracy are wrong I not so kindly invite them to go piss up a rope. 



When I removed it from the box I heard stuff rattling around inside. Found these pieces of plastic.



After I check it all out going to give it a good cleaning and I'll do a teardown.

[worsthorse]

That is a dangerous amount of bench space.

I like it!!   

yes.. and it is making my twelve square feet with a couple of shelves look more paltry than usual.  very nice setup!   

[Specmaster]

Forgot to take a picture but I had a Random smoke generation this afternoon but it headed straight for the bin. The Sonoff POW that has been looking after my 13A coffee machine for the last few years failed into a smelly blob for reasons unknown. Likely culprit was the relay but the ECLB and separate circuit breakers weren't tripped so 

I have a spare Sonoff POW but maybe some hot wiring to a better relay will be in order before dropping it into a new enclosure. Meanwhile I will have to stagger out to the switch 40-50 minutes prior to getting the morning heart starter instead of doing it from the comfort of my Bed. Clearly this is a first world priority problem to sort out but for tonight a plug change for manual.

What was plugged into the POW? Check for a DC to ground fault as this can cause some ECLBs not to trip as the detection transformer core can be saturated by a very small DC fault to the point where it cannot detect any load imbalance and thus fail to trip the relay off.

[Specmaster]

This supply nearly 30 years old. Designed and built by yours truly. It has 3 - LM317's and 1- LM337. And it's in daily operation powering scanner accessories. Never an issue.

 

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

[med6753]

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

Well of course. Isn't that what you're supposed to do when designing equipment? 

[Mortymore]

This supply nearly 30 years old. Designed and built by yours truly. It has 3 - LM317's and 1- LM337. And it's in daily operation powering scanner accessories. Never an issue.

 

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

Yap! I bet med6753 read the datasheet prior to build the PSU.

Freaking engineers.  Don't you guys ever look at the data sheets? 

[med6753]

This supply nearly 30 years old. Designed and built by yours truly. It has 3 - LM317's and 1- LM337. And it's in daily operation powering scanner accessories. Never an issue.

 

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

Yap! I bet med6753 read the datasheet prior to build the PSU.

Freaking engineers.  Don't you guys ever look at the data sheets? 

As a matter of fact I did. And that's how I knew that the pin out on the LM337 was different from the LM317. 

Freaking engineers. 

[factory]

I avoid games of chance where possible

Just seen my next car on ebay: https://www.ebay.co.uk/itm/313016881099

Great for burnouts or burning down a neighbours fence, or even in the Top Gear style destroying a caravan but utter shit and lugging home loads of TEA or Ham items 

You forgot the Datsun, I've had this on my PC for a long time, low quality screenshot taken when I was at college.

David

[Specmaster]

LM337... the pins are different to the 317. Guess how I discovered the latter 

I've discovered that the hard way too, fortunately I figured it out before the magic smoke escaped. Although there's only about £0.10 worth of magic smoke in an LM337 its still nice to not have to replace the only one you had to hand at the time. (Just checked, it's actually £0.60-0.70 worth of magic smoke, but the principle still applies.)

That is why the makers produce data sheets. 

[Specmaster]

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

Well of course. Isn't that what you're supposed to do when designing equipment? 

Yes, but as always some people will always say that its near enough and then run at its limit or slightly beyond and then BOOM.

[Specmaster]

This supply nearly 30 years old. Designed and built by yours truly. It has 3 - LM317's and 1- LM337. And it's in daily operation powering scanner accessories. Never an issue.

 

I suspect that is because you designed it to be used only within certain safe parameters and have never attempted to push it beyond its envelope of reliable operation.

Yap! I bet med6753 read the datasheet prior to build the PSU.

Freaking engineers.  Don't you guys ever look at the data sheets? 

As a matter of fact I did. And that's how I knew that the pin out on the LM337 was different from the LM317. 

Freaking engineers. 

Yep, sometimes its best the RTM.

[bd139]

Try that with software 

[factory]

Hmmph, just got around to powering up the HP 1120A. I now know 2 things.

My Black Star Jupiter sig gen is useless for square waves above a few hundred kHz, and wow, this thing sucks power like a damn immersion heater!

Around 60mA on the -12.6V rail, and just short of 90mA on the +15V rail. The probulator actually gets warm to the touch!

Seems like a lot of power to me, though there were no bad smells or emissions of magic smoke.

Sooo... top of my shopping list is now a decent AWG   

I'm not surprised the probe gets warm, it has a hybrid circuit crammed into it, as shown in the Dec 1969 HP journal page #14*, it gives power drawn as 110mA for the 15V rail & 70mA for the -12.6V rail.
https://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1969-12.pdf
*oddly pages 12 & 13 are missing.

I need to add a power supply circuit & sockets to my HP 1725A to use the active probes I have here, attached is the information from the service manual.

David

[Specmaster]

Got my car back from the dealers today, all fixed and then popped into Tesco and Lidl on the way home. Both were almost empty of people and reasonable stocks of the more basic food items, TP still missing and both were limiting people to just 3 of anything so when those essential workers get off work, they will be able to get some provisions so well done to them for placing sensible limits  on things. That said, I did hear one heated discussion between a customer and staff member. The customer had 6 of everything and was arguing that 3 were for them and other 3 were for their 76-year-old mother was not able to get out of her house to get her shopping. I wonder how they got on, I left before the end of the discussion, but it was one that I can concur with because I had to do all of my mums shopping in the 2 years that she was bed bound towards the end. There is never an easy solution to these problems is there?

On topic, I have finally been able to transfer my photos of the tear down of the Advance J2B AF generator, from the phone to the PC, some buggy app was blocking it, so I'll be uploading these later.

[Specmaster]

Try that with software 

Nice with some salt, chips and peas, mmm 

[AVGresponding]

Hmmph, just got around to powering up the HP 1120A. I now know 2 things.

My Black Star Jupiter sig gen is useless for square waves above a few hundred kHz, and wow, this thing sucks power like a damn immersion heater!

Around 60mA on the -12.6V rail, and just short of 90mA on the +15V rail. The probulator actually gets warm to the touch!

Seems like a lot of power to me, though there were no bad smells or emissions of magic smoke.

Sooo... top of my shopping list is now a decent AWG   

I'm not surprised the probe gets warm, it has a hybrid circuit crammed into it, as shown in the Dec 1969 HP journal page #14*, it gives power drawn as 110mA for the 15V rail & 70mA for the -12.6V rail.
https://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1969-12.pdf
*oddly pages 12 & 13 are missing.

I need to add a power supply circuit & sockets to my HP 1725A to use the active probes I have here, attached is the information from the service manual.

David

Ah, thanks for that. It doesn't mention power consumption in the manual.

Yeah, that thick film hybrid     is where the only IC in the thing lives, and the manual does mention it's not repairable and is replaced as a (no doubt unobtainium) unit.
The IC is probably unobtainium as well, so genuinely not repairable...

[Cerebus]

Try that with software 

Nice with some salt, chips and peas, mmm 

Cabin fever getting to you? 

[mnementh]

Hello mnem,

I recently purchased an EDS-88A.  It seems pretty nice and since it is for my business, it is a tax deduction.  I didn't mind the cost and it seems real nice.  I see you also link to the ESR70 too.  I may order one of those as these type of devices fascinates me, for what ever reason - simple minds are easily amused LOL.

Do you have much experience with the ESR70? It looks good.  I find them for anywhere from $95 (shipping from the US) to $140 (also shipping from the US).  I don't know why there would be a $50 difference in where you purchase from.  Of course, my brain says buy the cheapest; my gut says to investigate further.

The EDS system machines rely on precise gold plated probe.  The other just have regular tin plated clips.  Perhaps this is why my experience with ESR machines has been so-so, and actual replacement is the best cap test.

I have an Agilent 4362B and the Kelvin Clip assembly (!!!).  I found any cheap ESR meter so much easier to use and rarely fire up the Agilent 4362B.  Perhaps it is the Kevin 16089C only using tin plated clips?  We are talking very low resistance when measuring ESR.

I have also ordered everything needed (Case, PCB, parts kit, and gold plated probe) to build the EDS-89A Short finder.  Do you have any experience with it?  There is a youtube video showing the use of the older EDS-88A short finder.  It seems amazing.   You get a short anywhere on a power rail, it can be impossible to find the short without cutting traces and these days with 6 layer boards, that is mostly impossible.  We are talking milli-ohms, or less, in trying to track down shorts.  So, I am anxious to try that out too.

Take care and stay safe from the Covid-19 virus.

Dave

Yeah, the EDS-88A does SEEM amazing... actually, too good to be true, especially for a decades-old design. I highly doubt it will be nearly as effective in modern complex CMOS circuits where half a volt can turn on multiple junctions and completely bork any in-circuit results.  A cap analyzer that REALLY works in circuit even most of the time has been an engineering Holy Grail since LONG before I started; if these people actually had "the magic bullet", I highly doubt they'd be marketing the product the way they do. As you have one, I'd LOVE to be proven wrong, however.

I have used the ESR70+ since it was brammy-spankin' new. In fact, I was in on the first round of preorders, and waited a month and a half for mine. It was def worth the wait, and def worth the extra $30 at the time over the ESR70. I would characterize it as the best such DIAG TOOL I've owned; I get meaningful results in-circuit about half the time, and OOC every time.

Your characterization of the contacts is misinformed; the ESR70+ uses a proper Kelvin bridge, with two points of the bridge soldered to each chrome-plated alligator clip, so at most 10-12mm from the point of contact. There is nothing wrong with this arrangement, or with chrome-plated clips in this application, and actually, it yields a more reliable connection in general for quickly moving through a board. Remember, the ESR70/70+ is intended to be a DIAG TOOL; it just happens to be pretty damned accurate for OOC measurements too.

Don't get too hung-up on gold-plating for your contacts; outside of a laboratory environment where you can specify exactly how a connection is made, higher contact pressure is more important and a few electroplated layers of gold doesn't hold up well. Gold-plating is pretty, but it doesn't pay the bills.

For properly characterizing caps OOC, (and sometimes in-circuit as well) I find the DER-EE DE-5000 as close to laboratory-grade as you're going to get under $300; at around $110 delivered with a full kit of tweezer & alligator probes it is one of the best values in TE out there, bar none. It gives you BOTH ESR and Tan Delta; which means you can use the figures it gives both against the familiar tables technicians have developed for ESR, as well as the Tan Delta figures you will actually find in manufacturer's datasheets.  The inherent isolation of battery-operated design coupled with excellent accuracy across operating range (with the DE-5000, very comparable to much more expensive lab-grade gear from the big brands) is why I recommend both for anybody's bench; be it repair or design.

mnem

[Specmaster]

Try that with software 

Nice with some salt, chips and peas, mmm 

Cabin fever getting to you? 

Not yet, been out today collecting car from the dealer and shopping, so feeling good.

[Specmaster]

Meanwhile, in other news, here is what some of those troops have been doing.

[mansaxel]

For properly characterizing caps OOC, (and sometimes in-circuit as well) I find the DER-EE DE-5000 as close to laboratory-grade as you're going to get under $300; at around $110 delivered with a full kit of tweezer & alligator probes it is one of the best values in TE out there, bar none.

I've read in some threads that various suppliers of said instrument have been more or less dependable. Any recent purchasers who can recommend a good seller?

[xrunner]

We'll all be on a stay-at-home order as of midnight tonight. No going anywhere except for groceries and "essential" needs.

[Mr. Scram]

We'll all be on a stay-at-home order as of midnight tonight. No going anywhere except for groceries and "essential" needs.

Let nobody tell you a TE run isn't essential for the tealot!