Test Equipment Anonymous (TEA) group therapy thread

[mansaxel]

And, as is custom, I'm presenting the Page Instrument:



Some history: This is my first vintage TE.  It came to me, working, as a gift when I worked at the Royal Institute of Technology. I ran it with various inputs as a display, which through "being-left-on" cooked its PSU  ,  and made me repair it.

This then turned out to be my first TE repair, where one gets a badly scanned manual and finds new parts; IIRC a resistor that had gone O/C and a broken cap.  It's got the dreaded PH-163 input connector, and the 50Hz option. Which is good because: It uses, IIRC, the mains frequency as time base, and is as accurate as one can ask.

Bonus question: What are its neighbours? I've cropped the pic, but there are enough clues.

[Kosmic]

Well stranger things has finished now. There’s nothing else to do apart from stab each other.

Anyway on topic, I spent the last 2 hours filling up LTspice with bits of Farnell TM6 and comparing to measurements. It appears unfortunately that the chopper is likely to be faulty. It dumps a hell of a lot of current into the two LEDs and I think they have got dim and it’s not shuffling enough electrons through. Going to attempt to work out a way of proving this one way or another…. TBC…

Edit: also due to the stupid high impedance I need to make a bob pease FET probe…

I need to reacquaint myself with LT-Spice, I did start it a while ago and gave up because I found it had a steep learning curve and couldn't be arsed at the time, to much going on, but maybe the time is coming that I could get started again.

You should try micro-cap instead (http://www.spectrum-soft.com/download/download.shtm). Used to cost a lot of $$ but now free. Got a huge library of components (not just Analog Device IC). And I find it easier to use than LTSpice.

[factory]

Here are mine, both came from the US, the 'B' does at least have a 1MHz crystal fitted as standard, the 'A' needs some custom divider ICs, to get the reference working from 50Hz mains.



The choobs in these seem to get the rust/crack related failure of the pins.

David

P.S. hopefully won't be sniped on another counter, ending very soon in the US.

[bd139]

Beware the Rindsmilchübersättigungsübelkeit! I drank coffee with cow extract until I could not have gotten down another cup. Since then, 90% black and I'm fine.

This is now my favourite German compound word. Suffer badly from it to the point I drink soya milk.

Well, according to my Mum, Soya is no good for you and can actually be dangerous to humans, are you human or what 

How quickly some people forget:

Ook ook it is then 

Back to the library it is 

[Wolfgang]

Well stranger things has finished now. There’s nothing else to do apart from stab each other.

Anyway on topic, I spent the last 2 hours filling up LTspice with bits of Farnell TM6 and comparing to measurements. It appears unfortunately that the chopper is likely to be faulty. It dumps a hell of a lot of current into the two LEDs and I think they have got dim and it’s not shuffling enough electrons through. Going to attempt to work out a way of proving this one way or another…. TBC…

Edit: also due to the stupid high impedance I need to make a bob pease FET probe…

Probably capacitance is a problem too. Pease probes can be made really good, I used mine to check the voltages within high performance crystal oscillators without detuning or loading them. Properly built they can have < 1pF and several MOhms.

Indeed. I built one a few years back. Design reference here:



I have an absolute bucket load of 2N3823's VHF JFETs floating around waiting to be used so good timing

Whats still missing is the compensated input divider. Bobs mechanical setup is a bit fiddly and will pick up strays from everywhere ... And bandwidth can be souped up by using higher currents.

[bd139]

Well stranger things has finished now. There’s nothing else to do apart from stab each other.

Anyway on topic, I spent the last 2 hours filling up LTspice with bits of Farnell TM6 and comparing to measurements. It appears unfortunately that the chopper is likely to be faulty. It dumps a hell of a lot of current into the two LEDs and I think they have got dim and it’s not shuffling enough electrons through. Going to attempt to work out a way of proving this one way or another…. TBC…

Edit: also due to the stupid high impedance I need to make a bob pease FET probe…

I need to reacquaint myself with LT-Spice, I did start it a while ago and gave up because I found it had a steep learning curve and couldn't be arsed at the time, to much going on, but maybe the time is coming that I could get started again.

Yeah I don't blame you.

LTspice is quite frankly fucking horrible. It's written by engineers with absolutely no clue whatsoever about consistency or design or usability. The documentation is crap and it craps all over your filesystem like a rabid pigeon showing absolutely no respect or acknowledgement for anyone. Every part of it adheres to "not invented here".

And then they made a Mac version which is basically the same but completely entirely misses the entire Mac UI guidelines leaving you with twisted mutilated hands trying desperately to use the PC oriented keyboard shortcuts which barely even make sense on a PC. And it's buggy. So so so fucking buggy. Eventually you learn where all the bugs are but you have to hop around the software like someone trying to navigate through a cow dung filled field at night.

But there's nothing even remotely as good as it out there. Which is SO FUCKING DEPRESSING it makes me want to cry. So I will begrudgingly suffer.

Honestly if AD wanted to hire me I'd gladly rewrite both versions so they aren't absolutely cancerous and cage the engineers in the simulation engine dungeon where they belong, provided minimal sustenance and a black coffee drip directly into the stomach.

Hope that puts you off some more because it's actually painful using it. You just have to be that special kind of obtuse and stubborn to use it 

[25 CPS]

OT: Commercial FM radio over here is indeed crap. All the "Classic Rock" stations for example are owned by large media companies and they have a set playlist and never deviate from it which results in the same set of songs played over and over and over ad nauseam.  The DJ's are never given the freedom to play anything else.

TE content: BOM has been prepared for the 2235 re-cap. Also prep'ed the cabinet for re-spray which I'll probably complete later today.

It's the same here with slight differences due to the Canadian content requirements that are part of CRTC licensing.  It wasn't always this bad and there were several notable things I'd tune in for.  One of the radio stations in Toronto used to have a show in the evenings where they'd play a whole album side and they'd quite often pick some good records for that, and I went out and bought quite a few that they played that I enjoyed.  I have a couple of tube tuners that I got when I started learning electronics before the wheels fell completely off, which took place around the time the media industry was becoming consolidated and largely owned by telecommunications companies. 

Every format became worse than Top 40 repetition of whatever genre it was, the all Christmas music starting practically on November 1 started showing up and stuff like that.  The Toronto classic rock station, such as it is now, is now mired in a scandal involving the morning host's treatment of women and he and the company that own it have parted ways but the dust hasn't settled on that yet.  The classic rock station in Niagara was great until they changed hands.  It's still good but they've been adjusting the music selections not for the better which has me wondering if corporate's doing a frog-in-slowly-warming-water format change.  The classic rock station in Hamilton got kicked from one end of the dial to the other and the format got changed pretty severely by the company that bought them out which also owns the scandal plagued one in Toronto.

There's nothing in that 20 MHz of bandwidth I showed on that spectrum analyzer the other day, no compelling reason in there at all to go out and invest in a top end FM tuner or even a very good one.  The FM presets in the truck are for the aftermarket CarPlay to punch in on, the two classic rock stations, the jazz station, and the CBC.

[Wolfgang]

Well stranger things has finished now. There’s nothing else to do apart from stab each other.

Anyway on topic, I spent the last 2 hours filling up LTspice with bits of Farnell TM6 and comparing to measurements. It appears unfortunately that the chopper is likely to be faulty. It dumps a hell of a lot of current into the two LEDs and I think they have got dim and it’s not shuffling enough electrons through. Going to attempt to work out a way of proving this one way or another…. TBC…

Edit: also due to the stupid high impedance I need to make a bob pease FET probe…

I need to reacquaint myself with LT-Spice, I did start it a while ago and gave up because I found it had a steep learning curve and couldn't be arsed at the time, to much going on, but maybe the time is coming that I could get started again.

Yeah I don't blame you.

LTspice is quite frankly fucking horrible. It's written by engineers with absolutely no clue whatsoever about consistency or design or usability. The documentation is crap and it craps all over your filesystem like a rabid pigeon showing absolutely no respect or acknowledgement for anyone. Every part of it adheres to "not invented here".

And then they made a Mac version which is basically the same but completely entirely misses the entire Mac UI guidelines leaving you with twisted mutilated hands trying desperately to use the PC oriented keyboard shortcuts which barely even make sense on a PC. And it's buggy. So so so fucking buggy. Eventually you learn where all the bugs are but you have to hop around the software like someone trying to navigate through a cow dung filled field at night.

But there's nothing even remotely as good as it out there. Which is SO FUCKING DEPRESSING it makes me want to cry. So I will begrudgingly suffer.

Honestly if AD wanted to hire me I'd gladly rewrite both versions so they aren't absolutely cancerous and cage the engineers in the simulation engine dungeon where they belong, provided minimal sustenance and a black coffee drip directly into the stomach.

Hope that puts you off some more because it's actually painful using it. You just have to be that special kind of obtuse and stubborn to use it 

I beg to disagree about LtSpice. In my opinion,
- the user interface is nonstandard but efficient once you learned the ropes (painfully, I agree).
- the numerics are world class. Mike Engelhard is a multi-decade experienced simulation pro who absolutely knows what he is doing.  The many specialized numeric models for seemingly single components (voltage sources, caps, coils, ...) all with their parasitics are there because the more complex models facilitate the numeric treatment due to less artefacts and close-to singularities coming from idealized components.
- Simulation speed is awesome in most cases. The reason is that the whole solution procedure is compiled into assembly/machine code and then run, without any interpretation. The fastest Spice I know, by orders of magnitude.
- There are video tutorials with Mike where he shows why LtSpice is by far the most used simulator worldwide. The bag of tricks he uses is quite impressing.

Mike is a Klingon not respecting people with minor intellectual abilities a lot. His software was never meant to pamper the weak, but to solve, as much as possible, all simulation problems the user of LT chips might ever encounter, even if specialized code for complex functional blocks had to be specially developped. After the AD takeover of LT, Mike retired (rumors say due to style incompatibility with managers). Ergonomics never was high on his priority list, I am sure. Anyway, show me a *really* better Spice, and I will try it.

Recoding LTSpice with its present functional contents is a bet you will probably loose by any standards. Its weird, but the current code is several million LOCs and has absorbed the (Klingon style documented) black magic of decades of solved user problems and gotchas. Much fun to replicate that ...

https://www.analog.com/media/en/technical-documentation/lt-journal-article/ltjournal-v24n4-01-df-spicedifferentiation-mikeengelhardt.pdf

[dl6lr]

-hp- 4952A update:

The disc drive is broken. electrically and mechanically. Besides scrapeing the surface, it does not read from the lower head. The resistance of the head coils are OK, but one cannot swap the flex cables of both heads as they have mirrored  layout and are too short. So this is the cause why it will recognize no disc and fails to format a disc. The drive used is an old Sony drive, nearly unobtainium. So I prepared a Gotek floppy emulator with the latest FlashFloppy firmware. I prepared a special ribbon cable that cuts the wires that power the drive and rerouted to a power supply connector. Pin 1 is extracted to go to a pin header on the drive.
Fortunately I had an older Gotek with STM32F105, newer models use the Artery chip AR32F415and the cheaper ones use a QFN32 package that do not have all pins.

The config was tweaked to fit the disc layout 80 cylinders, 2 sides, 16 sectory a 256 bytes) of the 4952A. I had to use the logging firmware to find out some more parameters, see below.

Inserting a USB stick, the 4952A seeks to cylinder 78, then back to zero and indicates "Non LIF disc". Unfortunately formatting does not work, it formats cylinders 0 to 76 but gets stuck at cylinder 78 and reports "Bad disc". But prepared images from real discs are working, I can read and write to those.

HP used two different copy protection schemes in their instrument:
1. An unusual disc layout where cylinder 78 head 1 uses sector ids starting at 17 and cylinder 79 using sector ids starting with 97. This was used to disable copying discs with other machines.
2. Application programs from their original discs can not be copied to another disc in the 4952A. Doing so will copy the application file, but the copied disc refuses to load the application with "Application denied". As the instrument does not approach the disc before emitting the error, this should be some information in the directory part of the LIF.

Unnecessary to say, the disc images that are available for the 4952A are made with Teledisk from already copied discs (You can see the unused tracks on cylinder 77 heads 0 and 1 and cylinder 78 head 0 are in DOS format with 9 sectory a 512 bytes with contents of DOS programs.

Currently disassembling the FDC ROM...

[tautech]

Rant mode = ON
Simulators 

Way back in 2016 when for me this was the future trying to pick up on it was exceedingly frustrating and I figured there must be others in the same boat only looking for some guidance but it seemed no one wanted to stoop low enough to do a Simulation for Dummies.......that I still am !

In 2016 I even tried to engage with those wiser but it seemed no one wanted to hold hands and walk us through the basics if only to get us started with some simple example we can compare with to know we're on the right track. 

Doesn't seem much has changed in the subsequent 6 years as the cost of admission is still significant. 
However if someone up to speed wants take on the challenge please feel free to pick up with this thread to help lower the cost of admission:
https://www.eevblog.com/forum/beginners/circuit-simulation-(sticky)/

Rant mode = OFF

[Wolfgang]

Rant mode = ON
Simulators 

Way back in 2016 when for me this was the future trying to pick up on it was exceedingly frustrating and I figured there must be others in the same boat only looking for some guidance but it seemed no one wanted to stoop low enough to do a Simulation for Dummies.......that I still am !

In 2016 I even tried to engage with those wiser but it seemed no one wanted to hold hands and walk us through the basics if only to get us started with some simple example we can compare with to know we're on the right track. 

Doesn't seem much has changed in the subsequent 6 years as the cost of admission is still significant. 
However if someone up to speed wants take on the challenge please feel free to pick up with this thread to help lower the cost of admission:
https://www.eevblog.com/forum/beginners/circuit-simulation-(sticky)/

Rant mode = OFF

I bear with you, but there is little hope for a general purpose, intuition-supporting, error-forgiving DWIM simulator.

A simulator can save you some rounds of development cycle, but can also cause some extra ones, if
- your models are idealized, incomplete, inaccurate or both, not representing *all* physically possible behaviour of your parts.
  Example: parasitic oscillations, extremely high speed circuits, varactive pumping (charge in MOSFETS), very high currents
- Your models are fine, but the numeric dynamic range is insufficient.
  Example: very high Q circuits, regulators with vatsly different time constants, oscillator startup, artificial oscillations caused by numeric instability, ...

The approach that you dont need to know how and why your simulator does something is asking for trouble for a lot of circuits.
But you will find out after you finally build your stuff

Paranoia rules surpreme ! Dont believe yourself too much 

[25 CPS]

The truck needed some work on Friday.  I did get a few pictures of some test equipment in use even though it's not comprehensive, but I figured I'd post what I have.

I was planning to use the Hantek CC 65 clamp on it for the first time so I got it set up but it wouldn't null nicely at first so that caused some pre-troubleshooting troubleshooting in the basement.  The manual, if you're foolish enough to believe the manual, says to dial up a null by pushing and turning the knob, so it's like the Fluke current clamps instead of like the OTC or Agilent I have where you push a button and it does it automatically.  I bought the Hantek specifically to do lower current work since all those large clamps run from 400 to 2000 amps, and low current measurements are almost in the noise at the bottom of the range on those.

The Hantek's knob wouldn't spin so I couldn't get a nice null like I can on the Fluke clamps.  Take apart the Hantek and the knob doesn't spin like the manual says, the knob is actually a pushbutton and it is supposed to auto-null like the Agilent U1213A and the giant OTC clamp:





Put it back together and do a fast sanity test with a 100 ohm wirewound resistor and a bench power supply before heading outside.  It still doesn't null nicely so you have to do a bit of mental math to remove whatever lingering offset it has after it auto-nulls.  The slide switch also works between off and the 1 mV/100mA range.  The output doesn't change between the 1mV/10mA range and the 1mV/100mA range when you'd expect it to shift by a factor of 10 so I'm going to have to take it apart again later and see if I can fix that.  Between the badly written manual, the mediocre auto-null, and the balky range switch, I'm not happy with the Hantek so I'm open to suggestions for a replacement AC/DC clamp with a similar current range for smaller applications.

The U1252B is one of the multimeters that lives in the truck full time.  The plan now is to add the U1253B I got recently in the cold months when LCD displays can struggle but I'm keeping it inside until then to avoid ageing the OLED fast due to heat.  Anyways, the U1252B got put into action.  The remote logging and display functions aren't needed when the test leads are just long enough to let the meter sit propped between the hood (bonnet) and windshield while I sit at the controls in the driver's seat.



Mixed results.  A bit over 11 volts, which I can work with since it proves the output's there, but well down from 35 volts at the other end where there's a connector that's known to be a weak spot but happens to be ok in this case, but I don't have any pictures of that.  Part of that is test leads sitting on nasty looking contact surfaces by gravity only so the connection wasn't the greatest.



4.72 ohms where I'd expect 4.  Again, nasty looking contact surfaces in the connector plus fairly weak one handed holding the test leads onto it.  This would account for the slightly elevated resistance reading and the significantly reduced voltage on either side of that connector.  Put the two halves of that connector together and I can see why it went intermittent and then open.  A lot of MG Electrosolve and a lot of mechanical agitation got everything working properly again but it's most likely going to be a temporary fix.  Chances are that connector's going to have to be replaced if it goes bad again.

I ended up not using the Hantek clamp after all.  The original plan was to use it to see if current was circulating when voltage was present on this circuit without having to break the circuit.  However, the results of the first voltage check pretty much had me convinced there was an open downstream which made me decided to break apart that connector so I could inspect it and do a load side continuity test at the same time.  At that point, I didn't need the Hantek anymore between the voltage and resistance checks and visual inspection of that connector which turned out to be surprisingly bad inside.

That Hantek was the other thing besides that connector that was surprisingly bad.  I'm really disappointed with it.  I'm going to have to disassemble it and see if I can get the switch to work correctly at a minimum and give it a more thorough set of bench tests both AC and DC to see how useable or not it is, but as I said, I'm open to suggestions for a replacement.  It definitely leaves a bit of a bad impression and tends towards confirming my preference for Fluke and Keysight equipment even if they can be expensive both new and on the used market.

[vk6zgo]

Aaaaaaargh! Fuggitt! I just lost another Mini-Maglite to the leakage monster!
 

This situation is not news. Why are you still using primary cells in something you care about?   

Hope you can get them to replace the thing, no leak guarantee etc blah yadda

Because at least one of the bloody things has to be kept in a READYNESS state?

Yeah, it's a bummer if you have an extended power outage, & are relying on rechargeables!
I guess you could have Solar with big batteries & recharge from that, or from your car, but primary cells aren't going anywhere, soon!

It is easier to buy a bunch of "El Cheapo" primary cells, & a "Battery Daddy"* to keep them in.
The pack even included a battery tester, which worked once, then stopped until I fixed up the dry solder joints inside, but since then, it has "worked a treat".

* I think it was originally meant it to be "Battery Caddy", but the plastic moulding job got stuffed up, so they decided to "run with it"!

[mansaxel]

I'm not happy with the Hantek so I'm open to suggestions for a replacement AC/DC clamp with a similar current range for smaller applications.

Obviously there is a [hp] 428b and a matching probe somewhere, just waiting for you. It is my most used valve instrument. I whole-heartedly recommend it, in every lab. 1mA FSD is some resolution. For a current clamp.

Until such time as one will grace your bench, may I suggest the Uni-T UT-210e? (it being the "e" model is important, that's the one with the interesting ranges)  Resolution to 1mA on the 2A range, Ranges to 100A IIRC, does AC and DC, will null, and doubles as that 3rd simple DMM one sometimes needs. Price can't be beaten; I paid something like 50 US$ plus shipping.  It's a no-brainer. And none of that "interfacing to DMM" messiness. Display gives value. End.

Edit: Changed model number to be correct, and took actual max current range from datasheet instead of memory.

[Cubdriver]

Here’s my first one that looked like ten miles of bad road upon arrival:
 

Cute little buggers.

-Pat

[vk6zgo]

This is now my favourite German compound word. Suffer badly from it to the point I drink soya milk beans ground up.

FTFY. Nutritionally tolerable so long as it is fortified with several industrial chemicals microbial products.

Which reminds me; time to have my prescribed 1200mg Ca + vit D. I get my microbial output from normal food.

Really all this belongs in the https://www.eevblog.com/forum/cooking/ section. No mods action so far, but I've still got a couple of pages to go...

Coffee is a long-standing discussion point in TEA; it is a foundational topic in this memespace, as it is another Addiction common to most of the long-term members. We had another "Coffee Chronicles" thread for a while; it sortof withered when beanflying (miss you, buddy ) wandered off due to IRL getting in the way.

mnem

When I worked at the TV studio, they had an "endless refill" policy with the coffee machine.
If you were passing, you just grabbed a coffee.
OK, it was crap coffee, but it was caffeine!

I would often grab one, take it back to my bench, & continue with my work, ----I know, the O.H & S. people would have a fit!
The "wake up call" that it was probably not a good idea came when I dunked my soldering iron in the coffee cup!    OOOOPS!

The coffee did help when mulling*over some obscure fault on, mainly Picture Monitors, but also including some fairly weird stuff peculiar to TV studios.

*I did originally type "poring over", but thought better of it in this crowd!

[Brumby]

may I suggest the Uni-T UT-210e? (it being the "e" model is important, that's the one with the interesting ranges)  Resolution to 1mA on the 2A range, Ranges to 100A IIRC, does AC and DC, will null, and doubles as that 3rd simple DMM one sometimes needs. Price can't be beaten; I paid something like 50 US$ plus shipping.  It's a no-brainer. And none of that "interfacing to DMM" messiness. Display gives value. End.

Edit: Changed model number to be correct, and took actual max current range from datasheet instead of memory.

Indeed.  I have a UT210e and can attest to it being quite useful, mainly because it can do DC current. 

I have found the best way to get consistent DC current measurements (for me, at least) ... Put the clamp alongside the wire you want to measure in the orientation you will use for the measurement and press zero.  Then just open the clamp, put the wire inside and close - keeping the meter as still as you can.  Don't do any rotations or translations as this seems to change the pickup of ambient fields and will affect your readings.

From my experience, AC clamp meters are not so particular.

[Cubdriver]

Here are mine, both came from the US, the 'B' does at least have a 1MHz crystal fitted as standard, the 'A' needs some custom divider ICs, to get the reference working from 50Hz mains.



The choobs in these seem to get the rust/crack related failure of the pins.

David

P.S. hopefully won't be sniped on another counter, ending very soon in the US.

Yeah, those little inverted number tubes seem to be less robust than most - I’ve had a few dead ones, too.

Any joy on the counter you were going for?  I was watching a 5323A, but it ended while I was out and about, and went for more than I’d have been inclined to pay.

-Pat

[AVGresponding]

Aaaaaaargh! Fuggitt! I just lost another Mini-Maglite to the leakage monster!
 

This situation is not news. Why are you still using primary cells in something you care about?   

Hope you can get them to replace the thing, no leak guarantee etc blah yadda

Because at least one of the bloody things has to be kept in a READYNESS state?

Ni-MH LSD FTW!

I use them in my Tesco fake Maglite, and change them maybe once a year? It always works, and being an LED torch/flashlight with a built-in boost regulator it works all the way past the point where you'd normally want to recharge them, ie below 1V/cell. Just loses a bit of brightness when they are really dead.

[AVGresponding]

GP ultra AA’s. Never had one leak yet. One did go open circuit once but that was it.

Interesting we had an exploding llama the other day.  This was a bunch of GP LR44’s in a light up llama toy my youngest had. Went with quite a bang. So don’t buy their LR44’s 

FTFY

[KG7AMV]

A Winter Project Precision Apparatus Co. 612 Tube Tester. Not Sure What Winter for Sure Not this Winter!

[bd139]

Well stranger things has finished now. There’s nothing else to do apart from stab each other.

Anyway on topic, I spent the last 2 hours filling up LTspice with bits of Farnell TM6 and comparing to measurements. It appears unfortunately that the chopper is likely to be faulty. It dumps a hell of a lot of current into the two LEDs and I think they have got dim and it’s not shuffling enough electrons through. Going to attempt to work out a way of proving this one way or another…. TBC…

Edit: also due to the stupid high impedance I need to make a bob pease FET probe…

I need to reacquaint myself with LT-Spice, I did start it a while ago and gave up because I found it had a steep learning curve and couldn't be arsed at the time, to much going on, but maybe the time is coming that I could get started again.

Yeah I don't blame you.

LTspice is quite frankly fucking horrible. It's written by engineers with absolutely no clue whatsoever about consistency or design or usability. The documentation is crap and it craps all over your filesystem like a rabid pigeon showing absolutely no respect or acknowledgement for anyone. Every part of it adheres to "not invented here".

And then they made a Mac version which is basically the same but completely entirely misses the entire Mac UI guidelines leaving you with twisted mutilated hands trying desperately to use the PC oriented keyboard shortcuts which barely even make sense on a PC. And it's buggy. So so so fucking buggy. Eventually you learn where all the bugs are but you have to hop around the software like someone trying to navigate through a cow dung filled field at night.

But there's nothing even remotely as good as it out there. Which is SO FUCKING DEPRESSING it makes me want to cry. So I will begrudgingly suffer.

Honestly if AD wanted to hire me I'd gladly rewrite both versions so they aren't absolutely cancerous and cage the engineers in the simulation engine dungeon where they belong, provided minimal sustenance and a black coffee drip directly into the stomach.

Hope that puts you off some more because it's actually painful using it. You just have to be that special kind of obtuse and stubborn to use it 

I beg to disagree about LtSpice. In my opinion,
- the user interface is nonstandard but efficient once you learned the ropes (painfully, I agree).
- the numerics are world class. Mike Engelhard is a multi-decade experienced simulation pro who absolutely knows what he is doing.  The many specialized numeric models for seemingly single components (voltage sources, caps, coils, ...) all with their parasitics are there because the more complex models facilitate the numeric treatment due to less artefacts and close-to singularities coming from idealized components.
- Simulation speed is awesome in most cases. The reason is that the whole solution procedure is compiled into assembly/machine code and then run, without any interpretation. The fastest Spice I know, by orders of magnitude.
- There are video tutorials with Mike where he shows why LtSpice is by far the most used simulator worldwide. The bag of tricks he uses is quite impressing.

Mike is a Klingon not respecting people with minor intellectual abilities a lot. His software was never meant to pamper the weak, but to solve, as much as possible, all simulation problems the user of LT chips might ever encounter, even if specialized code for complex functional blocks had to be specially developped. After the AD takeover of LT, Mike retired (rumors say due to style incompatibility with managers). Ergonomics never was high on his priority list, I am sure. Anyway, show me a *really* better Spice, and I will try it.

Recoding LTSpice with its present functional contents is a bet you will probably loose by any standards. Its weird, but the current code is several million LOCs and has absorbed the (Klingon style documented) black magic of decades of solved user problems and gotchas. Much fun to replicate that ...

https://www.analog.com/media/en/technical-documentation/lt-journal-article/ltjournal-v24n4-01-df-spicedifferentiation-mikeengelhardt.pdf

I see a lot of typical engineering excuses for producing horrible software in your post. To clarify my position is producing software that isn’t horrible so I’ve heard all these before 

Certainly not talking about reimplementation which would be stupid but removing some of the sharp edges and clunk from the user interface.. Regardless of the simulation engine capabilities which are excellent, the product isn’t done until someone has made it usable. The last 20% needs sorting.

This is the same reason Linux on the desktop is a crap solution for most people. No one did the last 20% of the work.

[bd139]

Rant mode = ON
Simulators 

Way back in 2016 when for me this was the future trying to pick up on it was exceedingly frustrating and I figured there must be others in the same boat only looking for some guidance but it seemed no one wanted to stoop low enough to do a Simulation for Dummies.......that I still am !

In 2016 I even tried to engage with those wiser but it seemed no one wanted to hold hands and walk us through the basics if only to get us started with some simple example we can compare with to know we're on the right track. 

Doesn't seem much has changed in the subsequent 6 years as the cost of admission is still significant. 
However if someone up to speed wants take on the challenge please feel free to pick up with this thread to help lower the cost of admission:
https://www.eevblog.com/forum/beginners/circuit-simulation-(sticky)/

Rant mode = OFF

I bear with you, but there is little hope for a general purpose, intuition-supporting, error-forgiving DWIM simulator.

A simulator can save you some rounds of development cycle, but can also cause some extra ones, if
- your models are idealized, incomplete, inaccurate or both, not representing *all* physically possible behaviour of your parts.
  Example: parasitic oscillations, extremely high speed circuits, varactive pumping (charge in MOSFETS), very high currents
- Your models are fine, but the numeric dynamic range is insufficient.
  Example: very high Q circuits, regulators with vatsly different time constants, oscillator startup, artificial oscillations caused by numeric instability, ...

The approach that you dont need to know how and why your simulator does something is asking for trouble for a lot of circuits.
But you will find out after you finally build your stuff

Paranoia rules surpreme ! Dont believe yourself too much 

Yes good points here.

To clarify my use of simulation it’s only ever used in two cases:

1. Trying to understand something complex. I will build a model in it because it’s easier to deal with that than chowing algebra or laplace transforms for hours on end. Even if you do have to use LTspice.

2. When forward engineering using it as a “pre breadboard” phase to test ideas out before either committing to a dead bug prototype or ordering a PCB. That will be circuit fragments and interfaces between them. It will almost always take more than one design revision to knock the bugs out.

[Brumby]

I see a lot of typical engineering excuses for producing horrible software in your post. To clarify my position is producing software that isn’t horrible so I’ve heard all these before 

Certainly not talking about reimplementation which would be stupid but removing some of the sharp edges and clunk from the user interface.. Regardless of the simulation engine capabilities which are excellent, the product isn’t done until someone has made it usable. The last 20% needs sorting.

This is the same reason Linux on the desktop is a crap solution for most people. No one did the last 20% of the work.

Indeed.

It's like creating a high performing F1 racing car only to have skid-steer and a fruit box for a seat.

[Zoli]

That doesn't mean that the legislation they infect the rest of the world with won't affect you... it'll just take a wee bit longer.   

mnem
no further comment.

That's a reason why they are called MAFIAA.    And I'm still looking for my 76.8kHz Xtal generator(built by myself) .