Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

Holy fucking asscrackers! 5 full sets of Fischer leads...? You could probably keep the 3 nicest and still hold a successful Vulcan kal-if-fee for the sloppy seconds right here on eevBlog...

mnem
 

[mnementh]

I would still assert that programming has nothing to do with mathematics. Maths was the one subject I didn't do well at in school. The teacher was brilliant (probably a genius - he could solve a randomly mixed Rubik's cube behind his back), but could not control the rather unruly class that I was part of and my maths education suffered for it.

Programming is about abstract patterns and the ability to translate sentences of human language specification into detailed descriptions so that a stupid computer can understand what you are telling it. That process is far more about description, analysis, patterning and art than it is about maths.

Again... it is not the math part I'm emphasizing; it's the intuition part. Maths are just part of it; the language in which the arts you're describing eventually are expressed. We're talking about the same thing, only getting hung up on semantics.

I agree with you; I'm just not a good enough programmer to say it right.

mnem
GOTO 13 

[Neomys Sapiens]

I still don't see what everyone finds in a book that was started, IMO, at best as a copy of established works with a bit of loose talk thrown in. Need more practical that Tietze/Schenk? Get yourself the big Nührmann (Das grosse Werkbuch while there are copies (3- or 4 volume edition).
And I could give a loong list of works that I consider indispensable. AoE is not among them.

BTW, EE is a science, not an art.

By page 4 of chapter one Tietze and Schenk is into a mathematical model of the diode. TAoE is for exactly the people who don't want that approach - let's blind you with mathematics and not actually tell you what a diode is (in any straightforward way), why you might use one, etc.

As for Nührmann? Erm perhaps it's because Das grosse Werkbuch.. is that, a German book. There appears to be no English version, or any other language for that matter. Hate to be the one to break it to you, but German's a bit of a minority language. (English 1132 million speakers, <10 other languages accounting for 3944 million speakers>, German 132 million speakers, Japanese at 128 million speakers and so on).

I am well aware about the distribution of the German language. So what my original post was missing is the observation that I specifically could not understand it for someone living in Germany and fluent in that language.

Also, if the list of books already missing was not so long, I might even add it to round out the library. (Correction: 3rd Ed. PDF already here) Even if it was for exclusive content that would fit in a pocketbook. In the attachment you can see what I keep around. So I know what I am talking about.

So why my original and rather sharp comment? It was directed more at the reputation that this work is held in than to the book itself.  Because I see too much Guru-ism there.

[Neomys Sapiens]

OTOH learning C is far from sufficient w.r.t. the art of programming. I'd argue an appropriate set of realtime/embedded design patterns would be a better starting point.

This is a very astute observation, as I have spend much time, in multiple projects, to teach far better programmers and SW engineers than I could ever claim to be about such things.
Like the processing of analogue inputs, ensuring the actuality and validity of data in heterogenous distributed control systems, how to propagate information between various levels of control, how to apply diagnostic functions to a physical thing under SW control, etc.

[tggzzz]

I have a Solartron 7061 on the way to me. With all the original test leads!! The Fischer connectors are probably worth what I paid for the meter!

If they are the connector shown in that picture (there is a variant), then ISTR someone (in Spain?) selling new 2+1 pin leads for ~£75 and 4+1 pin leads for £95.

Keep a couple for yourself; one for the 7061 and one for another Solartron

[tggzzz]

OTOH learning C is far from sufficient w.r.t. the art of programming. I'd argue an appropriate set of realtime/embedded design patterns would be a better starting point.

This is a very astute observation, as I have spend much time, in multiple projects, to teach far better programmers and SW engineers than I could ever claim to be about such things.
Like the processing of analogue inputs, ensuring the actuality and validity of data in heterogenous distributed control systems, how to propagate information between various levels of control, how to apply diagnostic functions to a physical thing under SW control, etc.

Or even absolute basics such as how and when to use Finite State Machines, plus several different coding strategies. FSMs are useful for more than "just compilers" believe it or not

[mnementh]

3 Hours In... *wibble*

   

3 hours, and it's still only come up to 120°F. This thing is a monster, almost 10 lbs; it was all I could manage to cram it in there with 1 lb of carrots underneath and 4 cups of veggies on top. I felt like some horror show killer trying to stuff the vic into a 35 gallon drum. 

When core temp hits 130°F, I'll turn it down to low until it hits pull-apart tender at 195-200°. The hard part is the waiting; the house stinks so-o-o-o goood... *wibble* I may or may not carve off a few fat slices and put under the broiler covered with some Sweet Baby Ray's...

mnem

[worsthorse]

Face facts you old farts. You need glasses to see anything within 3 feet. Bite the bullet and get glasses with progressive lens. That way you wear them all the time and don't have to worry where you left them or have them dangling on a strap like an old spinster librarian. 

Yes, they take a while to get used to them but I got them years ago and never looked back. 

oh it is even worse than that... i have two sets of progressive lens glasses, one for every day and one for staring at the computer. i am always looking for the other pair as i am always wearing the wrong ones at the start of any particular activity. 

[worsthorse]

OTOH learning C is far from sufficient w.r.t. the art of programming. I'd argue an appropriate set of realtime/embedded design patterns would be a better starting point.

This is a very astute observation, as I have spend much time, in multiple projects, to teach far better programmers and SW engineers than I could ever claim to be about such things.
Like the processing of analogue inputs, ensuring the actuality and validity of data in heterogenous distributed control systems, how to propagate information between various levels of control, how to apply diagnostic functions to a physical thing under SW control, etc.

Or even absolute basics such as how and when to use Finite State Machines, plus several different coding strategies. FSMs are useful for more than "just compilers" believe it or not

Finite State Machine?  That made me smile and is a phrase I don't think I've heard uttered aloud since 1984, which is when I last used them to design a couple of memory and graphics boards.   

[kj7e]

Got a new toy to play with, had the covers off before I plugged it in.

Yaesu FTDX-101MP top side, Antenna tuner and dual direct sampling FPGA's

[kj7e]

Bottom side,

200watt PA, RX input RF filtering and shaping

[tggzzz]

OTOH learning C is far from sufficient w.r.t. the art of programming. I'd argue an appropriate set of realtime/embedded design patterns would be a better starting point.

This is a very astute observation, as I have spend much time, in multiple projects, to teach far better programmers and SW engineers than I could ever claim to be about such things.
Like the processing of analogue inputs, ensuring the actuality and validity of data in heterogenous distributed control systems, how to propagate information between various levels of control, how to apply diagnostic functions to a physical thing under SW control, etc.

Or even absolute basics such as how and when to use Finite State Machines, plus several different coding strategies. FSMs are useful for more than "just compilers" believe it or not

Finite State Machine?  That made me smile and is a phrase I don't think I've heard uttered aloud since 1984, which is when I last used them to design a couple of memory and graphics boards.   

Any event based programming can benefit from thinking in terms of FSMs, since they clarify control flow and make it explicitly what can and should not happen at any given time - and what to do about it.

Many programs are FSMs even though the developers might not realise it; think of any GUI program reacting to inputs, e.g. a smartphone app.

[grizewald]

Holy fucking asscrackers! 5 full sets of Fischer leads...? You could probably keep the 3 nicest and still hold a successful Vulcan kal-if-fee for the sloppy seconds right here on eevBlog...

mnem
 

I wish it was five, but there are three Fischer terminated leads for the front or rear measurement inputs and the other two are for the rear ratio input which is a different type of connector.

Still, when I saw that it came with the original leads, it became impossible not to press the "Buy it now" button!

[VK5RC]

Got a new toy to play with, had the covers off before I plugged it in.

Yaesu FTDX-101MP top side, Antenna tuner and dual direct sampling FPGA's

Ooooooooooh
Lovely, that has the stepper controlled input filter doesn't it?
Rob
PS I am still trying to repair one of its grandfathers, an ft 101b, receive is down about 3s units and no Tx. I have a mint working 101ex and a rough but going 101E.

[grizewald]

Many programs are FSMs even though the developers might not realise it; think of any GUI program reacting to inputs, e.g. a smartphone app.

One of the most boring contracts I ever worked on was at Plessey in Liverpool. They made the old armoured coin payphones that were fitted in every phone box in the UK during the '80s. The phone's firmware was a state machine and was documented on some 15 A0 sized state transition diagrams.
My job was to write a test procedure which would prove to BT that the payphone behaved in accordance with the documentation.

I think the test procedure ended up being around 500 A4 pages, most pages starting with the sentence "Lift the handset".   

[GreyWoolfe]

I do that every time I walk into a room. "Why did I come in here?" 

I used to have senior moments.  Now, it is all the time.

Face facts you old farts. You need glasses to see anything within 3 feet. Bite the bullet and get glasses with progressive lens. That way you wear them all the time and don't have to worry where you left them or have them dangling on a strap like an old spinster librarian. 

Yes, they take a while to get used to them but I got them years ago and never looked back. 

I have been wearing them for 20 years now.  The first Rx, the Dr told me to give it 2 weeks befor I came back.  I'm glad I listened.  I had on and off headaches for the 2 weeks plus struggling with the different strengths.  For the past few years, I have Seiko digital progressives.  Much wider field of view and sharper vision.  They do cost more but from my point of view, absolutely worth the cost.  I will not go back to regular progressives ever.

Face facts you old farts. You need glasses to see anything within 3 feet. Bite the bullet and get glasses with progressive lens. That way you wear them all the time and don't have to worry where you left them or have them dangling on a strap like an old spinster librarian. 

Yes, they take a while to get used to them but I got them years ago and never looked back. 

Not in my case; I need glasses to see anything further away than my elbow. I've lived with this defective vision and Coke-bottle-bottom glasses (or contacts so high a scrip I have to special order them) all my life.

Wanna trade eyeballs?

mnem
There's more than one reason I've learned to "see" with my fingertips; it's not JUST that being a mechanic you're ALWAYS working BEHIND something that's in front of you...

I learned that skill back in the early 90's doing car stereo, alarm and mobile cell phone installs as a side gig.  I am not blind but I do have amblyopia.  Outside of the progressives, left lens is basically glass, right lens and sunlight and I can set the lawn on fire. .  Vision aside, the dexterity I learned now allows me to easily repair the granddaughter's cheap costume jewelry necklaces and their almost microscopic links with my penis fingers.  Mrs GreyWoolfe gets a kick out of watching me because it still amazes her that I can fix something that small with hands as big as I have.  We won't mention the precision tweezers, pliers and headband magnifier that help.

[kj7e]

Got a new toy to play with, had the covers off before I plugged it in.

Yaesu FTDX-101MP top side, Antenna tuner and dual direct sampling FPGA's

Ooooooooooh
Lovely, that has the stepper controlled input filter doesn't it?
Rob
PS I am still trying to repair one of its grandfathers, an ft 101b, receive is down about 3s units and no Tx. I have a mint working 101ex and a rough but going 101E.

Yes, it has a belt driven variable capacitance input pre-selectors on both RX units.  They are under the IF boards on the bottom side.    Such a great receiver, the audio quality is top notch and the AGC works great.

[bd139]

Ooooh nice toy. I had a play with one of them a few weeks back. Enjoy 

[tggzzz]

Many programs are FSMs even though the developers might not realise it; think of any GUI program reacting to inputs, e.g. a smartphone app.

One of the most boring contracts I ever worked on was at Plessey in Liverpool. They made the old armoured coin payphones that were fitted in every phone box in the UK during the '80s. The phone's firmware was a state machine and was documented on some 15 A0 sized state transition diagrams.
My job was to write a test procedure which would prove to BT that the payphone behaved in accordance with the documentation.

I think the test procedure ended up being around 500 A4 pages, most pages starting with the sentence "Lift the handset".   

But did any of the procedures include "turn it off and on to see if it restarts to respond"?

[grizewald]

But did any of the procedures include "turn it off and on to see if it restarts to respond"?

 

Absolutely not! That payphone's firmware and electronics was just as rock solid as the outer casing and the cash box.
The CPU was the venerable RCA 1802 (originally designed for the AIM-9 Sidewinder air-to-air missile).

[bd139]

yeaaaaah no. If that was the Plessey CT24 it had a rather well known vulnerability. Put 20p in, call someone, when you hit 9p credit hold down "follow on call". it'd have to give you change so out pops 10p BUT stuff a quid in and it'll go straight through the coin chute and register it while you're holding follow on call. Just keep doing this until you're done! Worked on about 50% of the units i tried it on so someone at least tried to fix it (you?  )

Edit: not going to get into my escapades nicking stuff out of BT vans  . Still have a liberated krone tool and linesman set somewhere 

[Kosmic]

OTOH learning C is far from sufficient w.r.t. the art of programming. I'd argue an appropriate set of realtime/embedded design patterns would be a better starting point.

This is a very astute observation, as I have spend much time, in multiple projects, to teach far better programmers and SW engineers than I could ever claim to be about such things.
Like the processing of analogue inputs, ensuring the actuality and validity of data in heterogenous distributed control systems, how to propagate information between various levels of control, how to apply diagnostic functions to a physical thing under SW control, etc.

Or even absolute basics such as how and when to use Finite State Machines, plus several different coding strategies. FSMs are useful for more than "just compilers" believe it or not

Finite State Machine?  That made me smile and is a phrase I don't think I've heard uttered aloud since 1984, which is when I last used them to design a couple of memory and graphics boards.   

In video games we use it a lot. Everything is built out of finite state machines. It's not working really well when you try to expose them (via visual scripting) to non programmers though. I had to create some sort of heavily parallel state machine where the states (actions in this case) can be expressed in their own vocabulary and concepts.

[WastelandTek]

gawd, we're into WWII already

this thread really moves

[Brumby]

gawd, we're into WWII already

this thread really moves

... and at the rate we're going, we'll be getting ahead of ourselves by the end of February.

[grizewald]

yeaaaaah no. If that was the Plessey CT24 it had a rather well known vulnerability. Put 20p in, call someone, when you hit 9p credit hold down "follow on call". it'd have to give you change so out pops 10p BUT stuff a quid in and it'll go straight through the coin chute and register it while you're holding follow on call. Just keep doing this until you're done! Worked on about 50% of the units i tried it on so someone at least tried to fix it (you?  )

Edit: not going to get into my escapades nicking stuff out of BT vans  . Still have a liberated krone tool and linesman set somewhere 

That's the one, the CT24. It wasn't me that found or fixed that bug. I guess it was already fixed as the phone had already been out for some time before I worked on validating the state transition diagrams. (You know how it is with those parts of projects sometimes; they only get done when the customer starts shouting!)

After doing the verification job, the contract got a lot more interesting and I ended up implementing the charging logic for the Hong Kong version of the CT24. That was even more complicated than the BT version due to Hong Kong's plethora of different rates depending on time of day, call destination, whether it was a public holiday (they had quite a few mobile ones) and several other factors. The phone's firmware had to completely understand the charging regime because it had to be able to switch to internal charging if the charge pulses from the exchange went missing due to hardware failure or people manipulating the incoming line to filter them out.

I really got to respect the 1802 when I wrote that stuff. The indexed load and compare instructions were really useful for the application The fact that it has 16 completely general purpose 16 bit registers on an 8 bit CPU was something I also admired. Program Counter? Nah, just a four bit register that lets you designate any of your 16 general purpose registers to be the program counter. It doesn't even have a stack! Clean and simple.

The boss was a good bloke too. He always made sure we got out to some pub for a couple of quick pints at lunchtime!

Those were the days!