Test Equipment Anonymous (TEA) group therapy thread

[srb1954]

Managed to convince my TDS744A 500MHz scope to identify as a TDS784A 1GHz scope.
Just gotta calibrate it now, so I need to reassemble my SG503 levelled sine wave generator, which brings me to a question.

I need to bake a 1/8w carbon composite resistor, 80 degreesC for 8 hours. Any ideas on the easiest way to do so, that don't involve using the oven in the kitchen?
I wonder if my soldering iron can go that low......

Just hook it up to a suitable PSU and let it generate its own heat!

[Cerebus]

Managed to convince my TDS744A 500MHz scope to identify as a TDS784A 1GHz scope.
Just gotta calibrate it now, so I need to reassemble my SG503 levelled sine wave generator, which brings me to a question.

I need to bake a 1/8w carbon composite resistor, 80 degreesC for 8 hours. Any ideas on the easiest way to do so, that don't involve using the oven in the kitchen?
I wonder if my soldering iron can go that low......

I would put it in the sous vide. If you haven't got one yet, you now have the perfect excuse. Mmmm... pork chops with apples.

Even packed in a vacuum sealed nylon bag I think that might be very much not the way to bake components. Baking components is usually because you want to drive moisture off, that I fear would drive moisture in.

[mnementh]

Made it home alive; just finished making my dinner. Much breeze was shot and much good burgers consumed, and sweet potato fries. TE and gifts were exchanged. Pics and more details after father-son time.

mnem
 

[cyclin_al]

Made it home alive; just finished making my dinner. Much breeze was shot and much good burgers consumed, and sweet potato fries. TE and gifts were exchanged. Pics and more details after father-son time.

mnem
 

Same here, made it home safely.  That was a long day.  My vocal chords are tired.  Weather and driving was good.  Food was not good; rather it was excellent.  Now time to put the kid to bed and walk the dog.    I took a peek in the box and have some observations for later; good and bad...

[mnementh]

Yeah; my everything is tired. I think I'm gonna go to bed; pics can wait til tomorrow.

mnem
 

[Brumby]

But it already IS tomorrow here in Oz!!   

In fact, it's getting so late in the day here that it's almost the day after tomorrow.

[cyclin_al]

But it already IS tomorrow here in Oz!!   

In fact, it's getting so late in the day here that it's almost the day after tomorrow.

It is now tomorrow here too.

When kid went to bed, he informed the parental units that he wanted an alarm to wake up early to have play time (probably means computer games).
Then SWMBO says its all on me since she is going Montreal area early tomorrow morning (now today) 
It is way too late at night to take pictures of the low fuel warning on the car ... I pretty much used it up.  No photos of anything else either...
The trip to Montreal is going to be around 300km.  Local gas station is now closed (anyway, they prey on the desperate with crazy prices).
So now I am back from my garage with a couple of jerry cans of fuel 
That means supper is at 25h00, or is that 01h00?

[TERRA Operative]

Well, resistor is baked and coated in enamel and reinstalled.
I forgot I had a Chinese Uyue 946c hotplate hiding on my shelf, so it was put to good use.

Now that the SG503 is alive again, I now need to calibrate it. Buutttt I apparently need a differential comparator for my scope for some of the measurements.
Any ideas how to do it in-scope, or jury rig something up? I don't want to go drop $350-odd on an ADA400A just for one measurement (well yes I do, but my wallet doesn't... )

[Robert763]

Well, resistor is baked and coated in enamel and reinstalled.
I forgot I had a Chinese Uyue 946c hotplate hiding on my shelf, so it was put to good use.

Now that the SG503 is alive again, I now need to calibrate it. Buutttt I apparently need a differential comparator for my scope for some of the measurements.
Any ideas how to do it in-scope, or jury rig something up? I don't want to go drop $350-odd on an ADA400A just for one measurement (well yes I do, but my wallet doesn't... )

Invert one channel and use ChA + ChB math. Even old anlog 'scopes had this functonality. Or just ChA - ChB if your 'scope has that direct function.

[med6753]

Well, resistor is baked and coated in enamel and reinstalled.
I forgot I had a Chinese Uyue 946c hotplate hiding on my shelf, so it was put to good use.

Now that the SG503 is alive again, I now need to calibrate it. Buutttt I apparently need a differential comparator for my scope for some of the measurements.
Any ideas how to do it in-scope, or jury rig something up? I don't want to go drop $350-odd on an ADA400A just for one measurement (well yes I do, but my wallet doesn't... )

Invert one channel and use ChA + ChB math. Even old anlog 'scopes had this functonality. Or just ChA - ChB if your 'scope has that direct function.

Correct. Looking over some of my plug-in's here the Type 1A1/1A2 for the 547 and the 7A26 for the 7904 have that ability. So do the 465/475 series. Set volts/div same for each channel. Invert channel 2. Set mode switch to ADD. 

[TERRA Operative]

Yeah, I thought so, but I tried it and it looked nothing like the picture in the manual.
Maybe the difference in what was shown was down to a digital scope vs an analogue one, or just user error....

[med6753]

Yeah, I thought so, but I tried it and it looked nothing like the picture in the manual.
Maybe the difference in what was shown was down to a digital scope vs an analogue one, or just user error....

Or get one of these. Differential inputs on both vertical and horizontal. 

[Robert763]

Yeah, I thought so, but I tried it and it looked nothing like the picture in the manual.
Maybe the difference in what was shown was down to a digital scope vs an analogue one, or just user error....

Did you use a PG506 as the reference?

[Specmaster]

Ok, I know that a lot of you agree with Apples policy on repairs etc and berate Louis Rossman videos and his company, but here is a video where he is agreeing defending Apple on some points, so I tend to believe that he is 100% on the level with what he says about Apple when he slates them. I have always been of the opinion that he was telling the absolute truth and that he seems to be a fair person to be business with, but I also understand that you may not agree with me, what ever, YMMV.

[TERRA Operative]

Or get one of these. Differential inputs on both vertical and horizontal. 

https://imagizer.imageshack.com/v2/xq90/922/7xbiXL.jpg

That age of gear is somewhat rare here in Japan for some reason, plus my wife will kill me if I start a whole new line of test equipment collection

Did you use a PG506 as the reference?

I just pulled a 1kHz square wave from my Leo Bodnar GPS reference.

I haven't got to restoring either of my PG506 modules yet... They are scheduled for the chopping block soon.

[capt bullshot]

Some news from the fun with TEA department:

Maybe you remember I've got fake AD587 references from china and somewhat later real AD587UQ from Mounty.
I've put them on a simple board (don't have a photo readily available right now) and put them into that white cardboard box. Connected four of them (two fakes, two original) to the HP34401s and log the data:



Here's the first results (another colourful grafana picture):



Traces G and H are the original AD587UQ, E and F are the fakes. At some point (about 11:33) I've switched the HP34401 to slow 6 digit aquisition mode, so this is the reason for the reduced noise. At about 14:07 I've had the serial to USB converter taken out by a power switching transient (nothing impressive, just turned on the computer power circuit by a remote controlled contactor), so I had to restart the aquisition. Another proof USB is just crap.

All plots are normalized to the actual reference output voltage at about 11:40, so the graphs show the difference to each individual voltage at that time.

[factory]

Well, resistor is baked and coated in enamel and reinstalled.
I forgot I had a Chinese Uyue 946c hotplate hiding on my shelf, so it was put to good use.

Now that the SG503 is alive again, I now need to calibrate it. Buutttt I apparently need a differential comparator for my scope for some of the measurements.
Any ideas how to do it in-scope, or jury rig something up? I don't want to go drop $350-odd on an ADA400A just for one measurement (well yes I do, but my wallet doesn't... )

Invert one channel and use ChA + ChB math. Even old anlog 'scopes had this functonality. Or just ChA - ChB if your 'scope has that direct function.

Correct. Looking over some of my plug-in's here the Type 1A1/1A2 for the 547 and the 7A26 for the 7904 have that ability. So do the 465/475 series. Set volts/div same for each channel. Invert channel 2. Set mode switch to ADD.

I have the hp 1401A (has A-B) & hp 1400A, hp 1400B (differential) for the hp 140A scope, or the hp 1750A (has A-B) & hp 1752A (differential A-B) for the hp 175A scope, other plug-ins have A+B.

David

[mnementh]

But it already IS tomorrow here in Oz!!   

In fact, it's getting so late in the day here that it's almost the day after tomorrow.

It is now tomorrow here too.

When kid went to bed, he informed the parental units that he wanted an alarm to wake up early to have play time (probably means computer games).
Then SWMBO says its all on me since she is going Montreal area early tomorrow morning (now today) 
It is way too late at night to take pictures of the low fuel warning on the car ... I pretty much used it up.  No photos of anything else either...
The trip to Montreal is going to be around 300km.  Local gas station is now closed (anyway, they prey on the desperate with crazy prices).
So now I am back from my garage with a couple of jerry cans of fuel 
That means supper is at 25h00, or is that 01h00?

Heh... similar sitch here; got home and carputer said I have 1 mile range left on the gas in the tank. 

I knew I was cutting it close, but just sed "fuggitt, I need that free hour sleep tonight..." so I can drive another 2 hours this morning for my son and his lightsaber battles/playdate on the other side of town. I'll get gas around the corner and suck it up for 4c/liter.

Woops... there he is, already piling his toys up at the door...   

mnem
*jedi-daddeh-dwagon*

[med6753]

Some news from the fun with TEA department:

Maybe you remember I've got fake AD587 references from china and somewhat later real AD587UQ from Mounty.
I've put them on a simple board (don't have a photo readily available right now) and put them into that white cardboard box. Connected four of them (two fakes, two original) to the HP34401s and log the data:

....snip

So what can I conclude from this experiment? To an average Joe blow like me would see no real world difference between the fake and real OR the dedicated volt-nut would have a shit fit over the fakes and throw them against the wall?

Which is it or is it basically both?   

[mnementh]

IMHO, Without knowing/subtracting the response of the 4 different meters over that time as opposed to measuring with the same meter/multiplexer, it's pretty hard to accurately characterize anything at this resolution. It appears the "fakes" have aboot 2x-3x the actual drift of the "genuine" ones, at least as compared against the nexus of normalization at ~11:40.

mnem
 

[Kosmic]

Some news from the fun with TEA department:

Maybe you remember I've got fake AD587 references from china and somewhat later real AD587UQ from Mounty.
I've put them on a simple board (don't have a photo readily available right now) and put them into that white cardboard box. Connected four of them (two fakes, two original) to the HP34401s and log the data:

....snip

So what can I conclude from this experiment? To an average Joe blow like me would see no real world difference between the fake and real OR the dedicated volt-nut would have a shit fit over the fakes and throw them against the wall?

Which is it or is it basically both? 

The fake one might be reject from the factory. They were probably supposed to be scrapped but someone decided to make a little bit of money. Sometimes they are also reclaimed parts that got de-soldered with a blow torch and got way too much heat.

either way you can expect the fake one to be out of specs.

[Cerebus]

IMHO, Without knowing/subtracting the response of the 4 different meters over that time as opposed to measuring with the same meter/multiplexer, it's pretty hard to accurately characterize anything at this resolution. It appears the "fakes" have aboot 2x-3x the actual drift of the "genuine" ones, at least as compared against the nexus of normalization at ~11:40.

mnem
 

Yeah, but picking an arbitrary time point to determine 'drift' maketh no logical sense. For instance, if the left of the graph is the actual turn-on time then the 'fake' references would seem to have less drift overall as determined by the difference between maxima and minima for each trace.  Looking at both, sometimes the 'fake' are better than the 'real' depending on the timescale you evaluate drift over. There's a reason that this kind of drift is normally evaluated by looking at Allan variance.

What you can tell from those traces is that the 'fake' references have much higher peak to peak noise over relatively short periods (minutes) than the genuine article.

Also, capt bullshot would have received a glancing blow to the back of his head from my school physics teacher for having failed to label neither the axes nor title the graph. 

[med6753]

What you can tell from those traces is that the 'fake' references have much higher peak to peak noise over relatively short periods (minutes) than the genuine article.

Yes, that much is obvious. But if I were to take the real and fake and use them as a reference in a typical 4.5 or 5.5 digit DMM would I see much, if any, difference?

[Specmaster]

Yes, that much is obvious. But if I were to take the real and fake and use them as a reference in a typical 4.5 or 5.5 digit DMM would I see much, if any, difference?

Probably not.

[dl6lr]

It's Friday evening, time for some Fun with TE!
Todays star of the show: a low noise preamplifier

Also, sorry for the big pictures Vince.

Good Lord.... have I sinned to deserve such a punishment ?!  18MP 3.5MB per pic 

I am confused. Image #1 of ch_scr here show to be 2400x1600px with abt. 800kB. med's images are way larger...