Test Equipment Anonymous (TEA) group therapy thread

[Specmaster]

SSD's do indeed rock, I've been using them for years on my desktop and laptop but they are still quite expensive in real term so sometimes a hybrid could be a better option cost option.

Not sure I like the idea of hybrids. I haven't used a rust disk since I got a Samsung 470 sometime in 2010. Not one. The amount of time I've saved measures in weeks or months. One of the few devices that the ROI is quite frankly insane on.

SSD's do indeed rock, I've been using them for years on my desktop and laptop but they are still quite expensive in real term so sometimes a hybrid could be a better option cost option.

They cost bugger all nowadays. A terabyte for not even 100 quid is just amusingly cheap.

Bloody hell. Just looked on amazon and they have bombed in price.

I haven't actually had to buy one yet. I keep getting given discarded 256 gig ones for nothing because people are less tidy with their data than I am and had to buy bigger ones ... Only 117Gb used here.

I've managed to get a 1TiB G-drive, 5x Samsung 256Gb 850 pros, 2x Samsung 128Gb 850 EVOs and a 128Gb 840 pro for nothing in the last year. The G-drive is £150 if anyone wants it

I just checked them on Amazon, and they are a hell of a lot cheaper now than 5 years ago that's for sure. I got a 500Gb Crucial drive and I think it was over £100 at the time, now its around £60 and so far I've only used 150Gb of that space but I do also have 2 normal drives, 1 x 2Tb and 1 x 500Gb that I store my programs and data on with 6Tb of network storage drives and another 2 to 3Tb of spare drives that I can just plug into my hot swap bay as required for back ups etc.

[mnementh]

That's not PCIe4. Well, it is, in a way; it shows just how much faster the baseline for PCIe4.0 is than PCIe 3.0. That little difference is because yours is a 1TB drive, while mine is 1/2TB.

These are the absolute cheapest $59 half-TB drives I could find using  the same Phison E12 controller as the Samsung EVO Plus. That I got 2 of for $49 each, because they were out of stock on the next model lower.    They're not just meeting spec of 3100/1900, they're blowing it out of the water and approaching 970EVO Plus speeds for half the price. And this was a preliminary benchmark literally JUST after installing Windows on my 950EVO SATA SSD; I forgot to install chipset drivers from AMD until AFTER I'd done these and gotten my memory so it behaved. 3800 at 16-17-17-38 is pretty damned good, IMHO.

More to come.

mnem

Larger SSDs are faster as a rule as they interleave on more "channels". Whatever the case, benchmarks have already confirmed current gen PCIe 4 SSDs aren't any faster in real life applications and in a remarkable number of cases actually slightly slower.

Not that it matters to the end user as he's not ever going to notice a difference one way or another.

I already knew this about larger SSDs being faster. So does bean.

This is absolutely not true. I see a difference in mine immediately: Booting is slow (25-28 seconds) now due to long RAID POST time, but apps load noticeably faster. As in Firefox loading in a couple seconds, even with 50 tabs open from the last session. That is a real-world everyday improvement.

I'm working on my post; it'll be up soon. As with everything, there are tradeoffs. But I'm glad I did the experiment.

mnem
*clicky-clicky-clickety-click*

[0culus]

That sure looks clean enough, hopefully you'll cure the tick and it will give many years of service.

Yeah, it's very clean. Obviously was kept nicely in a professional lab for all of it's life. I'm hoping someone didn't sabotage it...because it was working at item browsing as my prior pic shows.

[med6753]

Working on the triggering issue on the 485 and I suspect I have a bad IC but I'd like to get a confirmation before I make a potentially bad assumption. The condition is that the triggered light is always on unless you switch to single sweep then it goes off. But in auto trig and normal trig there is absolutely no sweep lock. The level control has no effect. So I started with the trigger light itself and worked my way back. Here is the partial schematic.



Tggzzz, if I could trouble you to open up your 485 and verify some voltages for me around U780. I suspect mine may be defective. I would greatly appreciate it. Here is the scope set up:

1kHz calibrator to channel 1
Vert – DC - 1V/div
Int Trig to channel 1
Time/Div – 1ms
Horiz display  -A
Sweep Mode – Norm trig
Level – 0
Slope - +
A Triggering coupling to AC and source to INT

Here are the measurements I got on U780

Pin 1   +0.496V
Pin 2   +0.078V
Pin 3   +0.30V
Pin 4    +4.9V
Pin 5    -4. 9V
Pin 6   +0.028V
Pin 7   +0.39V
Pin 8   +0.64V
Pin 9    0V
Pin 10  +0.14V
Pin 11  +5.5V
Pin 12  +0.01V
Pin 13   0V
Pin 14  +4.9V
Pin 15  +0.86V
Pin 16   0V
Pin 17  -0.47V
Pin 18  0V
Pin 19  +2.6V
Pin 20  +4.9V

All voltages seem to be OK except pin 19. According to the schematic that pin should float to -2.52V when open. This one goes to +2.6V when open. That's what leads me to believe this IC is toast.

[med6753]

That sure looks clean enough, hopefully you'll cure the tick and it will give many years of service.

Yeah, it's very clean. Obviously was kept nicely in a professional lab for all of it's life. I'm hoping someone didn't sabotage it...because it was working at item browsing as my prior pic shows.

Hopefully the manual has the resistance readings of all the output voltages so you can quickly track down which one is giving you trouble. Tek was very inconsistent in providing that info. The 7904 manual has it. The 465/475/485 manuals...forget it.    If I recall the 2465 manual has it.

[0culus]

That sure looks clean enough, hopefully you'll cure the tick and it will give many years of service.

Yeah, it's very clean. Obviously was kept nicely in a professional lab for all of it's life. I'm hoping someone didn't sabotage it...because it was working at item browsing as my prior pic shows.

Hopefully the manual has the resistance readings of all the output voltages so you can quickly track down which one is giving you trouble. Tek was very inconsistent in providing that info. The 7904 manual has it. The 465/475/485 manuals...forget it.    If I recall the 2465 manual has it.

It does provide them, but I can't find the damn test points it calls out. 

[malagas_on_fire]

When I'm fiddling around with electronics (with the tongue at the right angle), my prefered music in the background is (in no particular order)
- Pink Floyd
- The Alan Parsons Project (from "Tales of mystery and imagination" up to "Vulture Culture")
- Kraftwerk
- Jean Michel Jarre (occasionly)
- Barclay James Harvest (preferred "Eyes of the universe")
- Jeff Wayne's "War of the worlds"
- Ganymed "Future wolrds"
- to be continued ...

For background music I prefer something soothing like this:

Nice choice for music. I'd go with this band and it's new singe... thanks TOol . It helps avoid kernel panics

[0culus]

That sure looks clean enough, hopefully you'll cure the tick and it will give many years of service.

Yeah, it's very clean. Obviously was kept nicely in a professional lab for all of it's life. I'm hoping someone didn't sabotage it...because it was working at item browsing as my prior pic shows.

Hopefully the manual has the resistance readings of all the output voltages so you can quickly track down which one is giving you trouble. Tek was very inconsistent in providing that info. The 7904 manual has it. The 465/475/485 manuals...forget it.    If I recall the 2465 manual has it.

It does provide them, but I can't find the damn test points it calls out. 

Found them, they just weren't labeled in the way I expected. 

That sure looks clean enough, hopefully you'll cure the tick and it will give many years of service.

Yeah, it's very clean. Obviously was kept nicely in a professional lab for all of it's life. I'm hoping someone didn't sabotage it...because it was working at item browsing as my prior pic shows.

Hopefully the manual has the resistance readings of all the output voltages so you can quickly track down which one is giving you trouble. Tek was very inconsistent in providing that info. The 7904 manual has it. The 465/475/485 manuals...forget it.    If I recall the 2465 manual has it.

Attached is the table from the manual. I do believe the + and - 15V rails are screwed.

With the mainframe circuits connected:

+50 is at 1.82 kohm
+15 is at 73 ohm
+5 is at 3 ohm
-15 is at 84 ohm
-50 is at 4.494 kohm

With the mainframe circuits disconnected:

+50 is 9.28 kohm
+15 is 154 ohm
+5 is 7.53 kohm
-15 is 8.55 kohm
-50 is 30.2 kohm

[bd139]

Thinking about hitting this. https://probemaster.com/design-your-own-probe-kit/

Anyone got any probemaster probes to review?

[0culus]

I have two sets of their retractable banana plug shield multimeter probes, plus some accessories. Extremely nicely made.

[tggzzz]

Tggzzz, if I could trouble you to open up your 485 and verify some voltages for me around U780. I suspect mine may be defective. I would greatly appreciate it. Here is the scope set up:

A reasonable request, but unfortunately I'm away from home for large parts of next week (or more ). Can I ask you to nag me in a week?

I had a trigger problem with a 485 I've since sold. It was a bugger to track down, but I worked forward rather than backwards. Now I have a suspicion it was in the B trigger (unlike your A trigger), but, transposed to the A trigger, the problem was "going up" the U730, U738, U740 chain - but not in them. ISTR one of the schematic's "horizontal" components e.g. R729 was faulty.

A few semi-random thoughts...

If U789.19 is the collector of an internal differential pair, or similar, then the voltage you see might simply be a consequence of the circuit being permanently triggered.

Can single sweep operation give any clues, especially related to changing voltages around U780?

Can looking at the V-t voltages on U780 give a clue about their function?

Looking at your voltages, the +-5V is a little low; that might affect the tunnel diode biassing.

Presuming the front panel controls and signals are as defined in the manual, are the voltages/signals coming out of the "north" of U740 correct?

Have you gone through the procedures for setting the TD biassing; i.e. the trimpots between the north of U740 and the north of U780?

If they are socketed, consider reseating any of the above ICs and TDs, and JFETs, optionally with a drop of Caig DeOxit D100 on the end of a jeweller's screwdriver.

If you suspect U740 etc, I suspect you can swap the IC from the B trigger channel.

Is the "A sweep stop" circuits near Q822 working?

And don't forget to nag me.

[med6753]

Attached is the table from the manual. I do believe the + and - 15V rails are screwed.

With the mainframe circuits connected:

+50 is at 1.82 kohm
+15 is at 73 ohm
+5 is at 3 ohm
-15 is at 84 ohm
-50 is at 4.494 kohm

With the mainframe circuits disconnected:

+50 is 9.28 kohm
+15 is 154 ohm
+5 is 7.53 kohm
-15 is 8.55 kohm
-50 is 30.2 kohm

Look at your results again. The +/- 15V are OK. It's your +5V rail that's porked. 3 ohms? It should be at 50 ohms.

[med6753]

Tggzzz, if I could trouble you to open up your 485 and verify some voltages for me around U780. I suspect mine may be defective. I would greatly appreciate it. Here is the scope set up:

A reasonable request, but unfortunately I'm away from home for large parts of next week (or more ). Can I ask you to nag me in a week?

I had a trigger problem with a 485 I've since sold. It was a bugger to track down, but I worked forward rather than backwards. Now I have a suspicion it was in the B trigger (unlike your A trigger), but, transposed to the A trigger, the problem was "going up" the U730, U738, U740 chain - but not in them. ISTR one of the schematic's "horizontal" components e.g. R729 was faulty.

A few semi-random thoughts...

If U789.19 is the collector of an internal differential pair, or similar, then the voltage you see might simply be a consequence of the circuit being permanently triggered.

Can single sweep operation give any clues, especially related to changing voltages around U780?

Can looking at the V-t voltages on U780 give a clue about their function?

Looking at your voltages, the +-5V is a little low; that might affect the tunnel diode biassing.

Presuming the front panel controls and signals are as defined in the manual, are the voltages/signals coming out of the "north" of U740 correct?

Have you gone through the procedures for setting the TD biassing; i.e. the trimpots between the north of U740 and the north of U780?

If they are socketed, consider reseating any of the above ICs and TDs, and JFETs, optionally with a drop of Caig DeOxit D100 on the end of a jeweller's screwdriver.

If you suspect U740 etc, I suspect you can swap the IC from the B trigger channel.

Is the "A sweep stop" circuits near Q822 working?

And don't forget to nag me.

OK, thanks. In the meantime I do some additional checking per your suggestions. 

[0culus]

Attached is the table from the manual. I do believe the + and - 15V rails are screwed.

With the mainframe circuits connected:

+50 is at 1.82 kohm
+15 is at 73 ohm
+5 is at 3 ohm
-15 is at 84 ohm
-50 is at 4.494 kohm

With the mainframe circuits disconnected:

+50 is 9.28 kohm
+15 is 154 ohm
+5 is 7.53 kohm
-15 is 8.55 kohm
-50 is 30.2 kohm

Look at your results again. The +/- 15V are OK. It's your +5V rail that's porked. 3 ohms? It should be at 50 ohms.

Are you sure? .005 kohm = 5 ohm?

[med6753]

I suppose it could be correct. That would be about 1.7 amps. On the 7904 the 5V is 50 ohm minimum, not 5 ohms.

[0culus]

I suppose it could be correct. That would be about 1.7 amps. On the 7904 the 5V is 50 ohm minimum, not 5 ohms.

The manual for the 7104 could have a misprint. It's happened. 

[med6753]

I suppose it could be correct. That would be about 1.7 amps. On the 7904 the 5V is 50 ohm minimum, not 5 ohms.

The manual for the 7104 could have a misprint. It's happened. 

The other readings look OK. Yes, the +/- 15V are a little low but I doubt it would cause a shutdown.

[med6753]

Just for comparison I looked up 7904.

+15V  70 ohm
+5V    50 ohm
-15V   180 ohm

[bd139]

I would work from the shutdown /fault circuitry back here. If you look at the bottom right hand bit of the primary SMPS portion in the service manual, there's a voltage sampling network that should reveal some clues. I haven't LTspiced it but it looks like it's a divider that sums to zero if it's working properly. Anything that doesn't add up trips the fault sense on U1275. You should be able to work back from that to a rail.

If that's fine then the problem may be on the primary side of the SMPS.



Edit: sums to something other than zero by the looks or it's a current summing junction.

Edit 2: I was right. If fault input to U1275 is >200mV for more than 10ms it will shut the supply down. Start there and work back. Read fault protection section 3-55. Interesting design.

[med6753]

I just took a quick look at the 7104 schematics. It uses a whole shitload of +5V logic that the 7904 lacks. So it would seem that +5V reading of about 3 ohms is correct.

[bd139]

It's probably about right. 1.7A isn't a lot for 5V logic. I had a power supply in something that was a rack of 7400 TTL (not even LS) that took 45 amps

[bitseeker]

Using a thermal camera seems easier and safer than dicking around with flammable liquids and plenty of room for user error.

Plus,

Thermal camera = TEA

IPA ≠ TEA

 

[bitseeker]

Using a thermal camera seems easier and safer than dicking around with flammable liquids and plenty of room for user error.

Easier and safer be blowed! A thermal camera offers far more opportunity for the indulgence of TEA 

D'oh! I should've read ahead before posting. Well said, nfmax.

[bitseeker]

www.dresdencodak.com

The advice in the penultimate frame has been working here, too: "TEA-less conversation should pique interest but never dwell on the controversial."

[bitseeker]

Thinking about hitting this. https://probemaster.com/design-your-own-probe-kit/

Anyone got any probemaster probes to review?

I have two sets of the 8000 series probes with 8018S retractable-shroud banana plugs for my bench DMMs. I also got a pair (red & black) of 8050 sprung hooks for each set of probes.

They work great with both standard and safety banana jacks and the tips are super sharp (quite adept at drawing blood).