Test Equipment Anonymous (TEA) group therapy thread

[Vince]

VINTAGE RECAPPING

OK found some courage this evening... I started the chore of ordering the XXXXXX caps I need to recap my 6 pieces of Vintage TE.

Using Mouser then as I said.

They are great for all the small caps, but the big ones are astonishingly expensive. 6 Euros / USD for a 2,200uF 100V cap, and I need x4 of them. Then many more bigger ones still for the Tek 7603 PSU.

I need a few axial caps and these are silly prohibitive as well. Need a couple 330uF 25V. 6 Euros a pop at Mouser ! OK it's a Vishay, but still.
Checked on Reichelt, local to me then, well inside the Euro zone anywho. They sell it for.... 0.28 Euro, 20 times cheaper ! OK, it's not a Vishay... you do'nt even know what it is... you probably wouldn't want to know anyhow.... but still....I can't believe the cap situation has gone  so bad that I am now having to weight quality vs price, for a simple recapping !
I guess this hobby is not in my budget, I should stick to playing Sudoku my computer....

I will have to rethink my recap plan....it's getting out of hand already.

I have only just started and it already looks like even when buying on a huge American site like Mouser... it's still far from affordable.

This is all very depressing.... I guess I should wait a few years to recap all my TE, the time for the war and COVID to be behind us and things to go back to normal.... but this will take years and I am not gonna tolerate all this TE sitting around waiting for caps for years, scattered i pieces all around the lab.


I don't see any satisfactory solution to this situation.... I am... severely depressed by this cap situation....   

[tautech]

Yep shotgunning caps is only really the option when you have the budget.
Better to invest in an ESR meter and make correct calls on each cap.

[factory]

Not got much further with the Tek 318 today, found the vertical deflection problem yesterday, none of the caps on the CRT board test bad with the ESR60, should probably have checked the vertical signal into the CRT board, but instead noticed the composite video out connector on the side.

A quick solution was the £1 CCTV monitor I got, after tweaking the monitor controls I now have a picture, sadly it tells me of more problems, including PRL fail (whatever that is) and T/H error, the second seems to relate to input boards A01/A02, great don't have the extenders needed. 





David

[Robert763]

What's the 10dB step at ~2GHz ?

It's the change from baseband to down converted coverage at about 3GHz and the Pre-amp is on below 3GHz.

Is it normal for such a step to be visible ?

Yes. Its perfectly normal. This is the noise floor of the instrument not a signal, the RBW is at maximum and amplitude on auto. Below is a picture off the web of a 8566A on a 2Ghz to 22GHz scan (i.e. all above the baseband) you can clearly see the step for each band of the converter.  The IFR analyser is flat to specification from 10 MHz to 18 GHz (Checked R&S microwave sig gen with output level conformed by power meter)

[Neomys Sapiens]

What's the 10dB step at ~2GHz ?

It's the change from baseband to down converted coverage at about 3GHz and the Pre-amp is on below 3GHz.

Is it normal for such a step to be visible ?

On the Agilent E44xx in my company lab it is slightly smaller, but visible too. Occurs somewhere in the same region.

[tautech]

What's the 10dB step at ~2GHz ?

It's the change from baseband to down converted coverage at about 3GHz and the Pre-amp is on below 3GHz.

Is it normal for such a step to be visible ?

Yes. Its perfectly normal. This is the noise floor of the instrument not a signal, the RBW is at maximum and amplitude on auto. Below is a picture off the web of a 8566A on a 2Ghz to 22GHz scan (i.e. all above the baseband) you can clearly see the step for each band of the converter.  The IFR analyser is flat to specification from 10 MHz to 18 GHz (Checked R&S microwave sig gen with output level conformed by power meter)

Thanks, not seen such before and I quite understand it would be negated with a Normalize.
Hmmm, compare that to this full 26.5 GHz sweep in the new SSA5000A where no step is visible with these settings.

[TERRA Operative]

@vince, the trick with axial capacitors is to not use axial capacitors.

Solder a bit of tinned copper wire to the +ve lead to extend it a couple inches, bend it back along the can and cover the lot with some clear heatshrink.
Form the leads appropriately and pow, you have converted a radial capacitor to axial, with the benefit of having your choice of the newer tech and better parameters in a cheaper part (no-one is investing in making nice low-ESR / low-ESL radial caps...)

Like this:

[doktor pyta]

Something for Tek lovers

[Robert763]

What's the 10dB step at ~2GHz ?

It's the change from baseband to down converted coverage at about 3GHz and the Pre-amp is on below 3GHz.

Is it normal for such a step to be visible ?

Yes. Its perfectly normal. This is the noise floor of the instrument not a signal, the RBW is at maximum and amplitude on auto. Below is a picture off the web of a 8566A on a 2Ghz to 22GHz scan (i.e. all above the baseband) you can clearly see the step for each band of the converter.  The IFR analyser is flat to specification from 10 MHz to 18 GHz (Checked R&S microwave sig gen with output level conformed by power meter)

Thanks, not seen such before and I quite understand it would be negated with a Normalize.
Hmmm, compare that to this full 26.5 GHz sweep in the new SSA5000A where no step is visible with these settings.

That scan has a signal. The gain is reduced and the effect is less visible.  Note the floor is about 10dB lower on the IFR with no signal.
Additionally the Siglent, being newer, may cover the range without switching hardware. This results in a smoother rise in noise floor
You could normalise the bands so it looks right but you would be reducing the performance by doing so.

[tautech]

Actually I thought they were very similar sweeps to use as a comparison, 10dB/div, 3 MHz RBW/VBW and similar width noise floor ~60dB down and 26 GHz full sweep. It was just one I found on the Siglent website by chance.

Obviously the bands aren't being stepped like in your SA so maybe I'll go back and again watch the video Shahriar did on the SSA5000A recently where he takes us through all the RF stages and frequency conversions.

[Ice-Tea]

Re: stepped noisefloor. Seen it on a lot of SAs. Always makes me a bit uncomfortable but should be normal.

Had to take a Yokogawa VT 3100 monitoring station apart to replace the CMOS battery so you get a few pictures. There's a lot of memory in this machine. Well engineered, too. This was not a cheap machine back in the day.

[capt bullshot]

Had to take a Yokogawa VT 3100 monitoring station apart to replace the CMOS battery so you get a few pictures. There's a lot of memory in this machine. Well engineered, too. This was not a cheap machine back in the day.

Yeah, build quality (and most probably component choice) looks very good, as with the WT1600 power analyzer.

[capt bullshot]

Wonder if DHL understands this address label 

[BU508A]

Wonder if DHL understands this address label 

 

ECCP¹?



¹Electrolytic Capacitors Care Parcel

[Vince]

Wonder if DHL understands this address label 

I am sure they will find me ! 

Well I hope they do, because it looks like a pizza box so I would rather for ti to be still warm when it arrives, cold pizza doesn't taste as good  ! 

I like them with lots of goat cheese, bacon and mushrooms please ! 

[Vince]

@vince, the trick with axial capacitors is to not use axial capacitors.

Yeah sure, you did a really nice job there.. but to me an axially mounted radial just looks like a bodge 
However I guess I will have to get used to it, since old TE have lots of axial caps and their price is so ridiculous, and choice of value and voltage so limited.. yet as you say quality is not their goal so....
I guess I will start to get over it and take a deep breath and... start doing like you do. Being able to use good quality caps is I guess my best argument to justify it to myself... after all it's about functionality before looks, it's an electronic device, not a fancy luxury hand bag 
So I guess my wallet is forcing me to do what's technically best.... I guess.

I don't have clear tubing like you have, but I guess I should start buying some stock of the stuff now....

[TERRA Operative]

When the case is screwed back in place and you can't see the caps, and the gear is working perfectly, you won't care how they look inside anymore.

[Vince]

Yeah uh.... I am not like that sadly, I do care about what's not to be seen... just like girls care about putting sexy underwear even when they don't intend to let anyone see it on that day. 

But yes, I agree with you anyway.... I just hope I can get get used to such bodges, since it's based on good technical and financial grounds.... I will try to make it look as good as you, to help ease the pain...

[bsdphk]

Obviously the bands aren't being stepped like in your SA [...]

The june 78 and august 79  HP Journal articles explains very comprehensively how the 8568 and 8566 works and why the steps are there in the latter.

The basic idea is that instead of controlling the IF going into the mixer precisely, which was impossible at the time, there are two spectrum analyzer channels running in parallel, one with the input signal, the other with the output from an internal 20MHz comb generator, making it possible to derive the precise IF frequency from the peaks in the comb generator spectrum.

One advantage of this design was the high spectral purity of the IF signal.

The steps appear because the the 8566 works in four bands, each using a different harmonic from the LO.

Modern instruments control the IF frequency numerically and therefore avoid all the double-plumbing of the 8566, the downside is that most of them have much poorer spectral purity as a result.

[BU508A]

Here are some pictures of a little project of mine - The wireless LEDs Rubik's Cube (or, in short TWLEDRC)

A friend of mine collects these Rubiks Cube and their variations. His wife suggested that I build a cube with some electronics for his collection. One of his hobbies is electronics and I gave him some help to get started.

Yesterday evening the cube was finished. It is mainly based on these wireless LEDs that BigClive presented on his YouTube channel. However, I ordered a larger primary coil with a much bigger diameter on AliExpress so that the cube has enough space in the magnetic field.

Here are some pictures:


Yes, all LEDs are working but the ones placed in the middle field of the cube are having a smaller coil because of space restrictions. Therefore they are not as effective as the LEDs with the bigger coils, but as long as they are on top as shown in the pictures, they are working perfectly. The LED shown in one of the pictures has a big coil (diameter 6mm, height ca. 7mm plus PCB and LED). I've glued them with hot snot into the pieces of the cube. The idea was to use the hot snot as a diffusor but I managed somehow to use the transparent variation of it. Well, next time ...

[AVGresponding]

Here are some pictures of a little project of mine - The wireless LEDs Rubik's Cube (or, in short TWLEDRC)

A friend of mine collects these Rubiks Cube and their variations. His wife suggested that I build a cube with some electronics for his collection. One of his hobbies is electronics and I gave him some help to get started.

Yesterday evening the cube was finished. It is mainly based on these wireless LEDs that BigClive presented on his YouTube channel. However, I ordered a larger primary coil with a much bigger diameter on AliExpress so that the cube has enough space in the magnetic field.

Here are some pictures:


Yes, all LEDs are working but the ones placed in the middle field of the cube are having a smaller coil because of space restrictions. Therefore they are not as effective as the LEDs with the bigger coils, but as long as they are on top as shown in the pictures, they are working perfectly. The LED shown in one of the pictures has a big coil (diameter 6mm, height ca. 7mm plus PCB and LED). I've glued them with hot snot into the pieces of the cube. The idea was to use the hot snot as a diffusor but I managed somehow to use the transparent variation of it. Well, next time ...

Usage of RGB LEDs would enable some "I solved it!" "Oh no you didn't!" shenanigans...

I'm planning to use clear resin to encapsulate some LEDs into a shape suitable for a backlighting project, and have considered a couple of things I might add to the resin to make the light diffuse better. Glitter, and baking soda (sodium bicarbonate). Some experimentation is in order...

[TERRA Operative]

just like girls care about putting sexy underwear even when they don't intend to let anyone see it on that day. 

Don't we all? Right....? Guys?.....


Nah, me neither.

[McBryce]

Wonder if DHL understands this address label 

I once had a letter delivered that had nothing but my name and "Cologne, Germany" written on it. Admittedly I am probably the only person in Germany with my name (as I am Irish), but I was still amazed that they went to the bother of filling in the blanks.

McBryce.

[Robert763]

Yeah uh.... I am not like that sadly, I do care about what's not to be seen... just like girls care about putting sexy underwear even when they don't intend to let anyone see it on that day. 

But yes, I agree with you anyway.... I just hope I can get get used to such bodges, since it's based on good technical and financial grounds.... I will try to make it look as good as you, to help ease the pain...

You will just have to get a lathe, spinning tools and mandrels and spin your own axial cases to hide modern capacitors in 

[TERRA Operative]

Ooh, some thin aluminium pipe and epoxy for the ends.
Get some heatshrink for the label maker and you might have something there.
Even print a paper label all fancy like and then clear heatshrink it on.


In other news, I just won a HP 85032B Type N cal kit for the coming spectrum analyser.
(Pics from listing)
Also, got a HP brand Type N to BNC adapter and a Huber Suhner 50ohm Type N terminator coming too.