Test Equipment Anonymous (TEA) group therapy thread

[med6753]

The final capacitor assemblies built up and ready to install in the Type 547. Rip out of the old cans and install tomorrow.

[Neomys Sapiens]

Well this is a pleasant surprise. The macOS version of LTspice is, well how shall we say, crap. Turns out the Windows one actually works better on arm64 macOS (with Crossover) than it does on x86-64 windows 10 with the high DPI shenanigans going on. Also crossover comes with a package manager so you can just run it in that 

Well I guess simulation is your best way to solve your real estate problem : no need for actual space hungry TE and component cabinets.... it's all in the laptop !!! 

Want to play with simulation too, later. Lab and house still aren't to that point just yet.

I think the first and last time I used Spice was at Uni in Hatfield between 1998/2001. Only very, very vague recollections...

I can only emphasise the usefulness of Spice not only in design, but also in repair situations.
I sat before a power supply for which I had even the somewhat convoluted circuit diagram, and which had a malfunction in the OVP circuit. I modeled a extract of the circuit in LTSpice and obtained by this way some pointers about what I SHOULD see. After that, it was a matter of minutes.

[bd139]

Well this is a pleasant surprise. The macOS version of LTspice is, well how shall we say, crap. Turns out the Windows one actually works better on arm64 macOS (with Crossover) than it does on x86-64 windows 10 with the high DPI shenanigans going on. Also crossover comes with a package manager so you can just run it in that 

Well I guess simulation is your best way to solve your real estate problem : no need for actual space hungry TE and component cabinets.... it's all in the laptop !!! 

Want to play with simulation too, later. Lab and house still aren't to that point just yet.

I think the first and last time I used Spice was at Uni in Hatfield between 1998/2001. Only very, very vague recollections...

I can only emphasise the usefulness of Spice not only in design, but also in repair situations.
I sat before a power supply for which I had even the somewhat convoluted circuit diagram, and which had a malfunction in the OVP circuit. I modeled a extract of the circuit in LTSpice and obtained by this way some pointers about what I SHOULD see. After that, it was a matter of minutes.

That is roughly how I operate too.

Plus it allows for stupid ideas to be tested before you invest the time in building them.

This was a stupid idea that worked!

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg1931791/#msg1931791

Resultant built here. Sorry YT link is dead;

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg1936213/#msg1936213

Oh and fuck buses. I’ll leave that one off the forum. Had to bail half way on my 40km walk due to thunderstorm and navigate zombie land buses. Has taken three hours to get home so far and I’m not even at Wimbledon yet.

[Specmaster]

Yep, that's why in zombie land car ownership is so high and also why they tend to be bigger cars than most decent city dwellers who have been spoilt rotten with their transport options. As I mentioned the other day, it took me about 70 minutes to go 2.6 miles by 2 buses and we are supposed to a city FFS  Oh and the cost if I had to pay for that would be about £5 

[bd139]

Yep, that's why in zombie land car ownership is so high and also why they tend to be bigger cars than most decent city dwellers who have been spoilt rotten with their transport options. As I mentioned the other day, it took me about 70 minutes to go 2.6 miles by 2 buses and we are supposed to a city FFS 

Indeed.

Back in reality now and this is more like it

[mnementh]

From the POINT OF KNOW RETURN Dept...



For good or evil, the prep work is done... All the patching chunked-out spots with epoxy, all the sanding... cleaning with alcohol and wet-wiping to raise the grain... now the stain is on there, and it's just a matter of waiting.

No idea how well it's going to take the stain... it's been hot and humid the last couple days, and a completely unforecast rain shower hit it this morning while I was eating breakfast between rounds with the jitterbug sander.



In related news... if you have any need to sand anything ever, and you can get to Horror Fraught before the 22nd... I can highly recommend this sander. Just top-notch all the way, and on sale for $20 it's a steal.

See me gush over it shamelessly here on the Wha'dja buy? thread:

https://www.eevblog.com/forum/chat/what-did-you-buy-today-post-your-latest-purchase!/msg4176511/#msg4176511

mnem
*shameless tool-ho dwagon*

[mansaxel]

No discord for me tonight. Out with the wife.

I'm posting this here mostly for the benefit of BD, who did express some envy on Discord perhaps 2 weeks back..



This was the second concert they played since 2017 (the first was on the same stage the night before), so it was really an event. Am really happy.

[mansaxel]

Well this is a pleasant surprise. The macOS version of LTspice is, well how shall we say, crap. Turns out the Windows one actually works better on arm64 macOS (with Crossover) than it does on x86-64 windows 10 with the high DPI shenanigans going on. Also crossover comes with a package manager so you can just run it in that 

Cute! Are you running it in Parallels or how?

Edit: Oh, yeah, "Polished Wine" -- of course! I'll have a look-see at that very soon, I think..

[factory]

OK so I am done with the TO3 packages now.

23 of them in total. 3 that I could not identify, if people can help, you are most welcome. If not... straight to the junk bin, since an unidentified anything is as useful as a rock. Maybe I could saw them open out of curiosity...

Anyway as you can see they are :

- S-1828 made by " SI " in 1974. Or so I understand.

...snip...

Well the S-1828 gives nothing, but 1110715 cross references to NFR8117 diode on the parttarget NSN stock list website and searching for NFR8117 finds a Digital parts list here http://vtda.org/docs/computing/DEC/Handbooks/EB-21850-75_SparesKitHandbook_1982.pdf
The DEC book lists it as PIV 75, 20A in T03 package, under part# 11-10715-00 (also cross-references back to the NSN list). Does it test as a diode?


David

Thanks a lot ! 

Yes it does indeed register as a double diode when I shove it into the chinese tester : common cathode on the metal can. The two anodes on the two pins.
However what worries me is that the tester also displays a 5 ohm resistance between the two anodes ?! 
That's weird... maybe it's toast...

I would get a second opinion from some other TE, in case something funny is going on with the cheap tester.
Here is what my recently acquired Peak tester says when I test two random used diodes, as a pair with common connection, I tested them both ways round with slight variation in Vf values. No leakage value given, maybe I should have added a resistor to give a fault condition. I could test some T03 double diode, but would need to locate them in the mess.






David

[Cerebus]

Yep, that's why in zombie land car ownership is so high and also why they tend to be bigger cars than most decent city dwellers who have been spoilt rotten with their transport options. As I mentioned the other day, it took me about 70 minutes to go 2.6 miles by 2 buses and we are supposed to a city FFS  Oh and the cost if I had to pay for that would be about £5 

We're far from out in zombie land and our Friday evening journey (2.9 miles both ways) to go out for a birthday meal for SWMBO and a few friends, took fifteen minutes, was made in silent, air conditioned comfort for the grand expenditure in £0.88 in electricity.   Even had free off-street parking at the remote end. There's actually an almost direct bus service (1/4 mile walk at our end) that we have used before for the same restaurant when I've wanted to have a drink, but it goes around the houses so takes at least 40 minutes each way and would cost £1.65 each way. Having the luxury of a choice, I know which one I'd make.

[factory]

- S-1828 made by " SI " in 1974. Or so I understand.

Then a couple made by RCA :

- CVT.......... 418
- CXK ....... 3M

I think you got the RCA types wrong, because the '64595' appears in the same place and has the same number of places as, for example '40319' etc., which are IDENTIFIABLE RCA part numbers. That said, I have two '64115' devices from RCA (also TO-3), on which I came up empty not only from the internet, but every databook which I have in print or PDF. Nada, zero, nil, rien, gornischt. Oh yeah, one of those totally unspecific 'parts search' sites associates the number with RCA, but that could just mean than someone has tried before me.

No RCA or Harris discrete / power databook and no other semiconductor data or substitution tables list ANY 64xxx number which could beeven remotely connected to the components in question, be it mine or yours.

Hmm, in agreement with Factory then !  OK looks hopeless... I think I will put all the un-indentifiable parts I stumble upon, in a box.. then decide later what to do with them...  I guess I could use them when I am angry or bored... I could just pick a part in that box and saw it open to see the die, and play with it on the bench with high current and voltages until the magic smoke escapes. Seeing the die literally melt, before my eyes...

This thread on the radio forum, suggests that the five digit code printed on RCA parts is a drawing number and that you could have contacted RCA years ago, or look it up in a parts index.
https://www.vintage-radio.net/forum/showthread.php?t=189905

RCA tended to mark many things with a "drawing number" , which I believe is what is marked on the transistor. (I can't read the numbers in the pix).
Then you had to either call your local RCA parts dist., or have a "Drawing Number to Part Number Index".
No wonder I hated working on RCA products. I -think- I might still have one in storage, but am physically unable to move the boxes around to get to it.

David

[Vince]

A "drawing number" ?!  

Well that's quite original, interesting, thanks for that....

Maybe if I e-mail them they will send me the drawing... a picture of a transistor die in x100 scale that I could frame and hang on the wall !

[mansaxel]

EDIT: Now that I finished watching it, 57 minutes and something worth every single second. Goose bumps.

Basically, he's built a system that is a polished, non-touring version of a modern PA system. Active crossover, line array, balanced interconnection. All of this is what one does when one means business, and is serious with the sciencey part.

Then it's overmuch in every direction, which is nice. In principle.

To  in the Bose discussion, this is very much what Bose is not about;

• There is a defined relationship between multiple drivers, keeping them coherent. Note how he's concentrating drivers closer in the horizontal plane -- this is proven to keep the wavefront coherent over a large frequency span.
• There is a low crossover point, 60Hz, to the subwoofer system, none of that HONK from the boom boxes.
• He's not afraid of having a crossover in the vocal range, which some people see as a problem (we all choose our personal favourite problem to focus on) because he's got good steep crossovers that can be made to sum well at the crossover point, and as a benefit of that freedom he can choose nice drivers that will work better than a bunch of overextended midrange 4" drivers.

[Vince]

Thanks a lot ! 

Yes it does indeed register as a double diode when I shove it into the chinese tester : common cathode on the metal can. The two anodes on the two pins.
However what worries me is that the tester also displays a 5 ohm resistance between the two anodes ?! 
That's weird... maybe it's toast...

I would get a second opinion from some other TE, in case something funny is going on with the cheap tester.
Here is what my recently acquired Peak tester says when I test two random used diodes, as a pair with common connection, I tested them both ways round with slight variation in Vf values. No leakage value given, maybe I should have added a resistor to give a fault condition. I could test some T03 double diode, but would need to locate them in the mess.

David

Second opinion I now have : I tested it with a DMM, my Fluke 11 but whatever.

It confirms what the chinese meter says, but is less ambiguous.

The resistance between the two anodes is not 4+ ohms, it's a dead short !
Which means the two diodes are in // ... which explains why the DMM measures the drop of each diode to be strictly identical, 0.453mV.

How did the chinese meter manage to measure slightly different voltage drops I don't know.... but just goes to show that basic testing with a DMM still rules !

So the diodes are good but their anodes are connected internally... so it's toast. So.... now I know it's hopeless, I can crack it open and have a look inside the can ! 

[factory]

Can we actually be sure it is a double diode and not just one diode in the package, with the two leads internally connected together?

Edit: just tried a single diode, with two test leads of the Peak DCA75 connected to one diode leg, it says those two leads shorted with the diode both ways round, so it would have been completely confused by the T03 package you have.


David.

[Vince]

Yes you are right, stupid me ! 

The info you found in that old document only said it was 75 Volts and 20 amps, it never said it was a double diode, it's only the chinese meter that fooled me into thinking that !

OK.. so this diode is just fine... no opening of the can anymore, its life has just been saved thanks to you ! 

20 amps is a good bit of current so I guess they decided to make use of both pins to spread the current, makes sense....

[mansaxel]

Agreed, but I tend to  think that the speakers are also way over the top, surely if you're trying to get as close to the original as possible, then shouldn't your drivers be at approx the height as the instruments would have been? His speakers were 8ft tall FFS. Also, of course, he did rely on a lot of processor power, which is how Bose and the like get such "fantastic" one note bass response from such a small driver and enclosure.

Some points;

• Those are line arrays. They put out a pie slice of sound, not higher than the column of speakers. It won't sound much below or above them. Therefore, about 2,4m is reasonable for a living room.
• There are 3 channels of amplification per speaker, plus subs. Just crossing over and distributing all those signals is going to require a major amount of signal processing. 
• Add in the requirement of surround processing where a 5.1 or 2.0 signal needs upscaling into something that can feed all speakers; even if you've got a full all-channels signal, it's going to need decoding because it comes as a Dolby-whatever bitstream from the TV tuner or DVD player.
• I do not think he's playing tricks with a bass boost circuit here; if you've got an active crossover, just working with the frequency band gains is going to be a very powerful tool in making the system sound right. Once one's tried that, going back to caps-n-coils bodges hot-snotted to a piece of chipboard inside the speaker is very hard.
  

[Robert763]

OK so I am done with the TO3 packages now.

23 of them in total. 3 that I could not identify, if people can help, you are most welcome. If not... straight to the junk bin, since an unidentified anything is as useful as a rock. Maybe I could saw them open out of curiosity...

Anyway as you can see they are :

- S-1828 made by " SI " in 1974. Or so I understand.

...snip...

Well the S-1828 gives nothing, but 1110715 cross references to NFR8117 diode on the parttarget NSN stock list website and searching for NFR8117 finds a Digital parts list here http://vtda.org/docs/computing/DEC/Handbooks/EB-21850-75_SparesKitHandbook_1982.pdf
The DEC book lists it as PIV 75, 20A in T03 package, under part# 11-10715-00 (also cross-references back to the NSN list). Does it test as a diode?


David

Thanks a lot ! 

Yes it does indeed register as a double diode when I shove it into the chinese tester : common cathode on the metal can. The two anodes on the two pins.
However what worries me is that the tester also displays a 5 ohm resistance between the two anodes ?! 
That's weird... maybe it's toast...

I would get a second opinion from some other TE, in case something funny is going on with the cheap tester.
Here is what my recently acquired Peak tester says when I test two random used diodes, as a pair with common connection, I tested them both ways round with slight variation in Vf values. No leakage value given, maybe I should have added a resistor to give a fault condition. I could test some T03 double diode, but would need to locate them in the mess.






David

This level of difference is probably due to small temperature differences. Even handling will cause a change. Typical Si Vf temperature sensitivity is about 2 mV per deg C.

[Neomys Sapiens]

I won a bid for an even smaller multimeter yesterday.
You might recall that my SOAR pen-type multimeter is inoperable due to a broken leaf spring of the battery holder.
In the kit where the SOAR was located, nothing larger can be accomodated. So I got this Benning for a temporary replacement, until I manage a fix or get another SOAR. It will make a nice gift or backup stowaway after that.

[Neomys Sapiens]

On Ebay.de:

https://www.ebay.de/itm/294976828863

A rather low-priced Bernstein Spannfix ending tomorrow evening!
Correction: ending Tuesday

[factory]

A "drawing number" ?!  

Well that's quite original, interesting, thanks for that....

Maybe if I e-mail them they will send me the drawing... a picture of a transistor die in x100 scale that I could frame and hang on the wall !

You are assuming RCA, or whoever took their component division over (Harris?) kept the drawings and didn't cremate or bury them, like a lot of evil companies do with their archives. 

David

[mansaxel]

TE report: Missed out on a Metrix MX56 in good condition yesterday. Would have been a nice add-on to the 4 ¹/₂ digit meter stable.

Non-TE win: Bought a Zoom multi-effect for the Middle Boy.

[factory]

I would get a second opinion from some other TE, in case something funny is going on with the cheap tester.
Here is what my recently acquired Peak tester says when I test two random used diodes, as a pair with common connection, I tested them both ways round with slight variation in Vf values. No leakage value given, maybe I should have added a resistor to give a fault condition. I could test some T03 double diode, but would need to locate them in the mess.




David

This level of difference is probably due to small temperature differences. Even handling will cause a change. Typical Si Vf temperature sensitivity is about 2 mV per deg C.

That's what I was thinking too, I'm sure I read Ge parts are even worse for this when testing, just don't have many NOS Ge trannys to test & compare with salvaged parts.
Which reminds me, I've got a bag of unlabelled NOS trannys somewhere, need to see what the DCA thinks of them.

David

[bd139]

Ge parts are terrible. Heat and current cause changes. Have to let them settle.

I built a transistor beta characteriser once to bounce 500 Ge transistors during peak guitar pedal garbage. Was an Arduino, UT61E and some resistors and a relay all hooked up to a laptop that ran a python script.

Made a bloody fortune with that machine 

[bd139]

Yep, that's why in zombie land car ownership is so high and also why they tend to be bigger cars than most decent city dwellers who have been spoilt rotten with their transport options. As I mentioned the other day, it took me about 70 minutes to go 2.6 miles by 2 buses and we are supposed to a city FFS  Oh and the cost if I had to pay for that would be about £5 

We're far from out in zombie land and our Friday evening journey (2.9 miles both ways) to go out for a birthday meal for SWMBO and a few friends, took fifteen minutes, was made in silent, air conditioned comfort for the grand expenditure in £0.88 in electricity.   Even had free off-street parking at the remote end. There's actually an almost direct bus service (1/4 mile walk at our end) that we have used before for the same restaurant when I've wanted to have a drink, but it goes around the houses so takes at least 40 minutes each way and would cost £1.65 each way. Having the luxury of a choice, I know which one I'd make.

I just got my dinner delivered. Even easier!