Test Equipment Anonymous (TEA) group therapy thread

[xrunner]

Exactly.

And floppies are still way cheaper and faster and more convenient to move data around, than HAM radios data links or smoke signals. I will have a floppy any day thanks !

Vince here's how they will tell you to move data around your lab with a floppy disk -

[bd139]

USB sticks is where it's at 

As for a Nano, don't think I could sub one in there. PROM has 30ns turnaround. Even a simple C LUT can't touch that I don't think.

Yes why bother programming a micro when you can just program a ROM instead and be done with it, anyway ?!

I guess as an exercise for kids why not, but as a practical and efficient solution not really 

Well it's always a matter of what tools you have available. And 256x4 PROMs are probably more difficult to get hold of than microcontrollers and are more difficult to program.

Consider this which probably does a fairly good job of being a ROM:

Code: [Select]


uint_8 lut[] = {
    0x01,
    ....
    0x09
}

void main()
{
    // set up ports here
    ...
    // pretend to be a ROM
    while (1)
        PORTA = lut[PORTB];
}
Unfortunately each tick is 62.5ns at 16MHz and the LUT is a few instructions worth of indirection and a load-fetch cycle so waaaay too slow.

[capt bullshot]

Just use the correct disk - and you've got one thing less to care about

[Vince]

USB sticks is where it's at 

As for a Nano, don't think I could sub one in there. PROM has 30ns turnaround. Even a simple C LUT can't touch that I don't think.

Yes why bother programming a micro when you can just program a ROM instead and be done with it, anyway ?!

I guess as an exercise for kids why not, but as a practical and efficient solution not really 

Well it's always a matter of what tools you have available. And 256x4 PROMs are probably more difficult to get hold of than microcontrollers and are more difficult to program.

Consider this which probably does a fairly good job of being a ROM:

Code: [Select]


uint_8 lut[] = {
    0x01,
    ....
    0x09
}

void main()
{
    // set up ports here
    ...
    // pretend to be a ROM
    while (1)
        PORTA = lut[PORTB];
}
Unfortunately each tick is 62.5ns at 16MHz and the LUT is a few instructions worth of indirection and a load-fetch cycle so waaaay too slow.

Yes if you must use a micro because you have it at hand, sure, if it can do it why not.

But if you have to make a board for the micro, and have to buy the micro, then whyt not just put a ROM on that board instead of a micro ?
A ROM replacing a ROM is simpler and makes more sense. 4 bit ROM trouble again not a problem you can use an 8 bit one just as well, and an 8 bit ROM is still way closer to the original 4 bit ROM... then a micro-controller, so more likely to '"just work".

Why try to replace a bicycle with a semi- truck, when you can just a bicycle that might different tyre sizes, but still is a bicycle ! 

Regardless of the solution you end up adopting, as long as we get pictures and reports, that's fine with me ! 

[tggzzz]

Ok some quick lunch time debuggery today. I have located the 465B failure and for once it's not a tantalum capacitor. Well that's not true, there is one shorted tant but that's not what is killing the 5V rail. It's much much worse!

I eventually chased it around and found out that it only borked the power supply if channel 1 was selected. So I traced that down to the channel switch. And yep, it's a tek ASIC has gone short. But it's not an ASIC! Part is labelled Tektronix 160-0204-00 which is actually a 63S141J fuse programmable 256x4 PROM. Manual confirms this.



Fortunately it's documented!



Now the IC is available for £23 which is a bit much so I am considering what other options I can use here to replace it. The actual original blank PROM is fairly unobtainable and I'd probably fuck it up programming it.

Will mull on it for a few days...

What's the clock rate for U1705? Given that A3-7 are probably constant, the limiting speed appears to be the t_su of the Q-PROM-D loop on U1705. If slow enough, then bitbang an arduino class processor.

Failing that:

• PAL/MACH/CPLD
• duplicate the logic in SMD TTL/CMOS on a PCB carrier plugged into the PROM's socket
• EPROM/EEPROM

[Vince]

Exactly.

And floppies are still way cheaper and faster and more convenient to move data around, than HAM radios data links or smoke signals. I will have a floppy any day thanks !

Vince here's how they will tell you to move data around your lab with a floppy disk -

" No worries boss, will use your USB stick as soon as you replace that old TDS5XX with its 20K dollars USB modern equivalent ! Oh you don't want to cough the cash ? Well give me my floppies back then and STFU ! "   

[Vince]

Ok some quick lunch time debuggery today. I have located the 465B failure and for once it's not a tantalum capacitor. Well that's not true, there is one shorted tant but that's not what is killing the 5V rail. It's much much worse!

I eventually chased it around and found out that it only borked the power supply if channel 1 was selected. So I traced that down to the channel switch. And yep, it's a tek ASIC has gone short. But it's not an ASIC! Part is labelled Tektronix 160-0204-00 which is actually a 63S141J fuse programmable 256x4 PROM. Manual confirms this.



Fortunately it's documented!



Now the IC is available for £23 which is a bit much so I am considering what other options I can use here to replace it. The actual original blank PROM is fairly unobtainable and I'd probably fuck it up programming it.

Will mull on it for a few days...

What's the clock rate for U1705? Given that A3-7 are probably constant, the limiting speed appears to be the t_su of the Q-PROM-D loop on U1705. If slow enough, then bitbang an arduino class processor.

Failing that:

• PAL/MACH/CPLD
• duplicate the logic in SMD TTL/CMOS on a PCB carrier plugged into the PROM's socket
• EPROM/EEPROM

... or just program a freaking memory to begin with ?! 
If I have to put a square peg in a square hole, I first try to see if I have square pegs laying around, before I resort to try to make a round peg fit 

[Specmaster]

More 465B shenanigans. So I figured I'd just do a short check on the IC and see if it was pin specific. Well it was.  Only the A3 line pin 4 was short to ground which is the CH1 select line.



Figured the best option here would be to just bend that pin out and see if the rest of the scope works. So here it is giving me the finger:



Of course I pulled the shorted tant out of the HV supply as well just to see if that would come up, but without a suitable fuse I'm not willing to power it the HT up in case I blow out the inverter transformer.



So next step is to find a suitably priced fuse and stick it in and see what else is broken and if it's worth spending £23 on an IC for 

The scope does power up now but without HT. Progress!

Well you beat me to it, I was going to say, just isolate Channel 1 for the time being and see what turns up 

[Kosmic]

An auction score from the FB group Old Tek Scopes. He didn't state how much in total he paid but he did state that the plug-in's were as little as $8/each. Given the total it still adds up to a significant chunk of change but the total amount of gear is unreal.   

And I thought two Type 547's, five plug-in's and two carts for $200 was a score.   

And now, the video 

[mansaxel]

Maybe you both misunderstood what I mean when I said I found FDD equiped TE "cool"... I thought it was obvious I meant it in an irrational / vintage interest kind of "cool", like we find our old glowing Tek scopes "cool"... 

See ? I can stir it too when I want ! Hey I might morph into a Dragon one day who knows !

I get it, and I'm probably too fed up with that GENERAL ERROR dude to not bite. 

Here goes:

To me, the floppy, as an ornament on TE, is kindasorta both a good and a bad thing. Because it mostly implies the onset of times when TE started being an embedded computer with an acquisition front end. That could have been a seriously cool era, or it could have been times littered with bad engineering decisions in the interest of vendor lock-in.  And it as things go is a bit of both.  The good grandchildren of this are the China scopes that have a web server, a real operating system, and a SSH daemon running.  The less stellar ones are all once-really-cool pieces of TE that still could be really-cool pieces of TE were it not for the criminally outdated Windows version and cobbled-together software on top making assumptions that mean upgrades are impossible. 

That the M68K (IIRC) scopes with floppies like the ones discussed earlier feel fresher than later Windows versions speaks volumes.

IMNSHO, floppies were just something that people without IEEE-488 had to use, and while they were le cool back then compared to a photo-chemical registration process, they still sucked. But in a forest of shrubbery, the cat is a lion..

Today, we've got Ethernet, and even if all we do is emulate serial or similar on top, it still is much better. There are edge cases mostly related to timing when other systems are better, but Ethernet is eating their lunch too with PTP for sync, and with some clever tricks you can tell Ethernet to do priority for some kinds of traffic.  None of this of course will be possible on the UNI-T's and Anengs of networking, like Netgear or D-Link, but there are things that obey and control much better than them.

[scl]

More 465B shenanigans. So I figured I'd just do a short check on the IC and see if it was pin specific. Well it was.  Only the A3 line pin 4 was short to ground which is the CH1 select line.

(snip)

74S287A looks to have the same pinout, and I see a datasheet claiming 30ns. Maybe also look at N82S129 and AM27S21.

I don't see any of those in my parts pile, and with postage from the US you are probably still better off finding one in the UK. But, if you can't, I will take another look. If did find one I could program it here too.

[AVGresponding]

The point is it's not the 1980s anymore; floppy is no longer "the convenient choice".

It is if you like to play with your Amiga 500 now and again.

[AVGresponding]

And speaking of overheating computers, my PC shut down on me an hour or so ago; almost certainly an overheat. It's been a long while since I busted the dust...

I noticed while I was in there that the header tank is only about half full now, so in another 5 years or so, it'll be empty...   

But, before then, yes, I need to service, or even rebuild the cooling system, mostly because I don't have any blue coolant left and have no plans to buy any just for a PC. I have plenty of red coolant, so I'll use that next time. Can't be arsed/don't want to take the risk of a failed flush, so I'll use all new components. I already have a spare pump with tank, and some silicone tubing left over from last time, so I just need a new radiator, and a new CPU block.

Naturally I shall investigate the old block, to see if it has filled with sludge as per Mnem's insistance.   

[bd139]

Ah watercooling. Never had any problems with that here because I never bothered with it 

Decided after doing a KT decision analysis that the best approach to the 465B is to just buy the fucking IC for £23 and be done with it 

[bd139]

More 465B shenanigans. So I figured I'd just do a short check on the IC and see if it was pin specific. Well it was.  Only the A3 line pin 4 was short to ground which is the CH1 select line.

(snip)

74S287A looks to have the same pinout, and I see a datasheet claiming 30ns. Maybe also look at N82S129 and AM27S21.

I don't see any of those in my parts pile, and with postage from the US you are probably still better off finding one in the UK. But, if you can't, I will take another look. If did find one I could program it here too.

Thanks for suggestions. Look at those and they are a good match. But again as always I don't have the programming capability available and that will cost more money and time than just buying the part.

[Robert763]

I quite like floppies and still use 3.5" ones.
There are several bits of TE that actually need them to work e.g. various HP logic analysers.
Ther are even some aircraft systems that still use them for data updates. However like some of the SAs most have been changed to USB floppy emulators.

[Vince]

More 465B shenanigans. So I figured I'd just do a short check on the IC and see if it was pin specific. Well it was.  Only the A3 line pin 4 was short to ground which is the CH1 select line.

(snip)

74S287A looks to have the same pinout, and I see a datasheet claiming 30ns. Maybe also look at N82S129 and AM27S21.

I don't see any of those in my parts pile, and with postage from the US you are probably still better off finding one in the UK. But, if you can't, I will take another look. If did find one I could program it here too.

Thanks for suggestions. Look at those and they are a good match. But again as always I don't have the programming capability available and that will cost more money and time than just buying the part.

Oh looks like I misunderstood... I thought yuo meant that the blank/brand new 4 bit PROM was available, and that you could program it because you had this table with its contents listed... but in fact you mean that you can get the pre-programmed part, so it's ust plug and play, no need to program anything ?!
Why are even talking about alternatives then... just get the freaking chip and report on your success...
23 quid to solve the problem in the most suitable and elegant way, as well as the quickest and easiest... sounds like good value for money to me ?! 

Hit that Buy button and type in that credit card number...

[Cubdriver]

Ah watercooling. Never had any problems with that here because I never bothered with it 

In the grand scheme of things, aren’t they ALL air cooled in the end?

-Pat

[bd139]

Ah watercooling. Never had any problems with that here because I never bothered with it 

In the grand scheme of things, aren’t they ALL air cooled in the end?

-Pat

Exactly. Why stick a Rube Goldberg machine in the process 

[bd139]

More 465B shenanigans. So I figured I'd just do a short check on the IC and see if it was pin specific. Well it was.  Only the A3 line pin 4 was short to ground which is the CH1 select line.

(snip)

74S287A looks to have the same pinout, and I see a datasheet claiming 30ns. Maybe also look at N82S129 and AM27S21.

I don't see any of those in my parts pile, and with postage from the US you are probably still better off finding one in the UK. But, if you can't, I will take another look. If did find one I could program it here too.

Thanks for suggestions. Look at those and they are a good match. But again as always I don't have the programming capability available and that will cost more money and time than just buying the part.

Oh looks like I misunderstood... I thought yuo meant that the blank/brand new 4 bit PROM was available, and that you could program it because you had this table with its contents listed... but in fact you mean that you can get the pre-programmed part, so it's ust plug and play, no need to program anything ?!
Why are even talking about alternatives then... just get the freaking chip and report on your success...
23 quid to solve the problem in the most suitable and elegant way, as well as the quickest and easiest... sounds like good value for money to me ?! 

Hit that Buy button and type in that credit card number...

Yup. Going to buy it once I’ve replaced the capacitor in the HT, put the case back on and tested it.

One thing I’ve learned when repairing stuff is do the cheaper repairs first before you find out it’s totally fucked somewhere you didn’t expect 

[bd139]

Bought some junk. Thanks to AVGresponding for pointing this out and apologies if I outbid you 

£17.49 including delivery gets me ... one HP 3776A PCM test set.



Why? Well I went through the service manual and it is absolutely packed to the brim with rather expensive analogue and TTL bits and a metric ton of standard HP parts. Absolutely no use whatsoever in this day and age as it's all for testing Bell/CCITT 4KHz channel telecoms crap which is completely obsolete so it'll be used as an organ supply for other projects. Only problem is it's a full 4U rack unit and weighs 15Kg

On a positive note it was cheap enough that I don't care if the courier leaves it on the doorstep and it gets nicked. And if anyone nicks it, LOL enjoy the hernia 

[Specmaster]

Ah watercooling. Never had any problems with that here because I never bothered with it 

In the grand scheme of things, aren’t they ALL air cooled in the end?

-Pat

Exactly. Why stick a Rube Goldberg machine in the process 

I do so because it removes at least 1 noisy fan and is generally far more efficient in dissipating heat and it also means of course that I generally don't have to lug my huge great full size tower that weighs a bloody metric ton out from under my bench and carry it downstairs into the garden to blow out the dust. I simply have to remove the front filter and remove the debris from that, and job done in seconds. My youngest son has an air cooled CPU and the cooler on that is always clogging up with thick dust and fluff and screams like a bleeding banshee until it gets cleaned or else it shuts down because of over heating.

[mansaxel]

Ah watercooling. Never had any problems with that here because I never bothered with it 

In the grand scheme of things, aren’t they ALL air cooled in the end?

-Pat

Since I have a geothermal bore, I can get chilled (compared to computer, but warm enough to heatpump from in winter) brine without hitting air, and then I can heat-exchange to a water-cooled computer if I want.

Of course, I'm then going to heat the house with that waste heat which means that the athmosphere is getting the last word, again. 

Pat's right; it just takes a bit longer in some scenarios.

[bd139]

Ah watercooling. Never had any problems with that here because I never bothered with it 

In the grand scheme of things, aren’t they ALL air cooled in the end?

-Pat

Exactly. Why stick a Rube Goldberg machine in the process 

I do so because it removes at least 1 noisy fan and is generally far more efficient in dissipating heat and it also means of course that I generally don't have to lug my huge great full size tower that weighs a bloody metric ton out from under my bench and carry it downstairs into the garden to blow out the dust. I simply have to remove the front filter and remove the debris from that, and job done in seconds. My youngest son has an air cooled CPU and the cooler on that is always clogging up with thick dust and fluff and screams like a bleeding banshee until it gets cleaned or else it shuts down because of over heating.

My mother always put it this way:

"I was fed up of scrubbing your father's shirt collars so I made him scrub his neck instead"

aka suck up the dust before it gets in the PC 

[mnementh]

WALLY WURLD TOOL ALERT: Hart SS DirtSucker for 30 Bux

   
https://www.walmart.com/ip/HART-8-Gallon-6-Peak-HP-Stainless-Steel-Wet-Dry-Vacuum-VOC812SW-3701/106550327

I've been missing my Shop-Vacs (I used to keep one for general use and one just for wet use) ever since Tejas; they were old and reeked and I just wasn't going to deal with them, so they went down the road for a tenner each just to liquidate the space they took up.

Well, they recently issued a price rollback on their most expensive Hart Shop-Vac; so when this came through my inbox yesterday, I jumped on it. They didn't have any in the store to buy in person (I drive right over and checked), but the website said my local store had 4 that could be purchased online for pickup next day, and the checkout promised they would substitute another unit of equal value if they didn't have it. So I bought and sure enough, got the email while I was out doing other errands that it was ready for pickup.

In case you don't know, Hart is WalMart's current "Value Line" of Chinesium housewares and tools; their answer to B&D repeatedly getting by Horror Fraught.

I'm not sure if they're disco-ing the whole HART product line, or if just their dirtsuckers. I've actually shopped similar-looking units (as in, I'm sure it's exactly the same thing with their colors and name on it) sporting the Stanley and Craftsman brands for $120-140.

This one has of course the primary attraction of a stainless-steel tank, but other things made me look twice too:

Tank Drain (so many of the budget Shop-Vacs omit this now) to dump water with some measure of control

Great big 140mm wheels on back so it actually rolls in the shop & on the driveway

Nice big tilt handle like real industrial equipment

2-inch hose: I consider this a reasonable compromise between 3-inch "real Shop-Vac" hose and the 1 1/8-inch hose usually provided with the budget models

Reasonable power for size: 6A Peak/LRA is pretty average; but rated 11.5A at 120V is pretty much exactly same as my "real" Shop-Vacs. Almost 2HP 

Tall format means okay for shop use, but should still be very handle-able for cleaning stairs.

Things I found once I got it home and fondled it:

Ooodles of suction, but moderate noise thanks to huge 7-inch dia x 6-inch pleated filter

Oversized rocker power switch with very light, smooth action

Good assortment of useful attachments, including a floor squeegee for the gulper nozzle that looks like it'll actually work

Assembly was very simple and straightforward even without RTFM; and all the bits for the wheels actually have screws to hold them in.


Bottom line of course is still the bottom line: How long will it last?  For $30, I think it'll be hard not to get my money's worth.

mnem
*suckage FTW*