Test Equipment Anonymous (TEA) group therapy thread

[tggzzz]

Point I found with SMD inductors is that the SRF can be much lower than the equivalent leaded unit.

On the Farnell Avnet website, there are >6000 SMD inductors with SRF>25MHz.

[bd139]

I need to open a trade account with them... 

[tggzzz]

I need to open a trade account with them... 

..."trade"... No you don't need to. You may choose to, of course.

[bd139]

It's trade account or get buggered £4 a go for small order charge.

Plus I can run it through my ltd

[mnementh]

Found it. Worth a look. And I paid for the extra outputs on mine after watching it

https://youtu.be/chrzrod3tQY

Whuf... and that's why I don't like making videos; I'm so scattered most of the time I sound like a total tool. I either have to shoot a dozen takes, or I keep going back to revisit something I forgot like this guy does; only like 10 times worse. 

Not best impressed with his construction techniques; he caps off perfectly serviceable vent holes and drills new ones that vent in the wrong place (out the sides, not the back), and all the assache he went to with that PIC.  If he just used a power switch with a built-in indicator, then he could repurpose that LED as a dedicated freq lock indicator.    A thermally-controlled fan, then after locating hot spots with IR thermo or FLIR, relocate the sensor, and all is golden.

Alternately, mount the standard to one of those areas of double-walled Al with lots of thermal paste, and again SEE how hot things get; it could operate quite nicely with 100% passive cooling. The Extron itself generates only a few watts; there's LOTS of thermal dissipation in that heat-sink case. 10.5 watts for the Rubidium should be easily dissipated by that huge aluminum mass.

Also, Extron tends to use off-the-shelf wherever possible; the PSU module already in there is probably a commodity-type unit you can look up the specs on AliEx or Taobao. It will probably have a higher-wattage cousin available cheaply that's a drop-in replacement, or it may in fact have plenty of overhead to operate that FE5680A which only needs 15-18VDC/700mA. Just eyeballing it, my guess is it's probably 800-1500mA on each output; so again, could very well be capable of powering everything. I'd certainly trust it to deliver 90% of rated power over using those cheap LM2587 buck converters he's got in there.

The guy really thought himself into a corner rather than properly leveraging what he already had, IMHO.

Okay... all done armchair quarterbacking. 

mnem

This message approved by the BoPD.

[Kosmic]

Anybody has an IEEE account ? I would like this document https://ieeexplore.ieee.org/document/6500517

I miss the time when I was in school and was able to access everything with the university accounts 

[bd139]

Some test gear for once!

Top one is Racal OCXO. Bottom one is Bliley OCXO. Bottom one is "correct" AFAIK from my limited ability to check with the relatively simple QRP Labs "almost a GPSDO" system. Really need a proper GPSDO now.

Racal OCXO takes at least 20 minutes to warm up and get stable. Bliley about 60 seconds. After that neither are budging a Hz past the odd digit flicker. Will poke them on 10s gate time later and check again once they're nice and toasty.



Edit: both units are 43 years old and were sealed when received. No repairs required other than the OCXO replacing in the bottom one because it had drifted past its setting range.

[tggzzz]

Whuf... and that's why I don't like making videos; I'm so scattered most of the time I sound like a total tool. I either have to shoot a dozen takes, or I keep going back to revisit something I forgot like this guy does; only like 10 times worse. 
...
The guy really thought himself into a corner rather than properly leveraging what he already had, IMHO.

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

[med6753]

Some test gear for once!

Top one is Racal OCXO. Bottom one is Bliley OCXO. Bottom one is "correct" AFAIK from my limited ability to check with the relatively simple QRP Labs "almost a GPSDO" system. Really need a proper GPSDO now.

Racal OCXO takes at least 20 minutes to warm up and get stable. Bliley about 60 seconds. After that neither are budging a Hz past the odd digit flicker. Will poke them on 10s gate time later and check again once they're nice and toasty.


Edit: both units are 43 years old and were sealed when received. No repairs required other than the OCXO replacing in the bottom one because it had drifted past its setting range.

I feel your pain. That 3Hz delta would make me nuts too.  In my volt-nut world that is a BIG difference. 

Luckily my frequency/time measurement capability is somewhat limited. A Fluke 1900A, Sinometer VC2000, and Tek 2465 DMS. And they all agree. 

[Kosmic]

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

So true. For me youtube is more a source of entertainment than a source of information.

[bd139]

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

It depends on the channel really. There are a lot of them out there which are extremely well presented, have dense information and reference material. For example Alan Wolke W2AEW.  They are all mostly "poo length" too

But NOTHING, I repeat NOTHING has a lower information density than a BBC science documentary with Brian Cox. It's like watching a snail try and fuck a cheetah.

I feel your pain. That 3Hz delta would make me nuts too.  In my volt-nut world that is a BIG difference. 

Luckily my frequency/time measurement capability is somewhat limited. A Fluke 1900A, Sinometer VC2000, and Tek 2465 DMS. And they all agree. 

They've both been on for an hour now and are still the same distance apart which is good. The signal source I used to run them up into, my DG1022Z, hasn't budged either (or has tracked in the same direction as both of them). I think that has a TCXO in it. So relative accuracy is about the same. I will adjust the new one to match the older one now.

But that doesn't mean absolute accuracy even if my QCX which is GPS calbrated (not disciplined) agrees. Ergo, I need a proper GPSDO with 10MHz and 1MHz outputs now 

[tggzzz]

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

So true. For me youtube is more a source of entertainment than a source of information.

Unfortunately there are some good videos, so yootoob can't be completely dismissed. One of my favourites is a series on how to solder https://www.youtube.com/playlist?list=PL926EC0F1F93C1837 All but one is <7minutes, and each ends with "turn off the projector and read the handbook".

That's a sign that they have carefully chosen how to use the right medium for each task and have planned each lesson - probably because films were expensive to create and duplicate

[med6753]

Whuf... and that's why I don't like making videos; I'm so scattered most of the time I sound like a total tool. I either have to shoot a dozen takes, or I keep going back to revisit something I forgot like this guy does; only like 10 times worse. 
...
The guy really thought himself into a corner rather than properly leveraging what he already had, IMHO.

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

That's what a fast forward is for. There are guys out there who go into extreme detail explaining what they are doing. Paul Carlson is one and IMHO he does an excellent job. Now to you and I it seems long winded because we already have the knowledge/experience. But what Paul is doing is adhering to a premise I was taught in a technical writing class in college. Always assume that your audience has no knowledge of your subject matter.

And even once in a while this old coot DOES learn something.   

[med6753]

They've both been on for an hour now and are still the same distance apart which is good. The signal source I used to run them up into, my DG1022Z, hasn't budged either (or has tracked in the same direction as both of them). I think that has a TCXO in it. So relative accuracy is about the same. I will adjust the new one to match the older one now.

But that doesn't mean absolute accuracy even if my QCX which is GPS calbrated (not disciplined) agrees. Ergo, I need a proper GPSDO with 10MHz and 1MHz outputs now 

Uh-oh......



 

[tggzzz]

Whuf... and that's why I don't like making videos; I'm so scattered most of the time I sound like a total tool. I either have to shoot a dozen takes, or I keep going back to revisit something I forgot like this guy does; only like 10 times worse. 
...
The guy really thought himself into a corner rather than properly leveraging what he already had, IMHO.

And that's why I rarely watch yootoob videos!

I suspect, without having spent an hour and a quarter of my remaing life watching it, that the information in it could have been conveyed in 10 minutes - or 2 minutes if I had been able to read a blog.

That's what a fast forward is for. There are guys out there who go into extreme detail explaining what they are doing. Paul Carlson is one and IMHO he does an excellent job. Now to you and I it seems long winded because we already have the knowledge/experience. But what Paul is doing is adhering to a premise i was taught in a technical writing class in college. Always assume that your audience has no knowledge of your subject matter.

And even once in a while this old coot DOES learn something. 

There is no "fast forward". There is a "wait 10s while skipping to a random place". That gives me 3 samples before the "30s elevator pitch" is over, and in an hour long video that's not a very dense sampling!

Technical writing should define up front who the intended audience is, so that people quickly know whether they are looking at a useful - to them - document. For videos that ought to be in the text below the video (Dave Jones does that; kudos), but very very few yootoobers do that. When in a nasty mood I presume that's because their prime motivation is "look at me".

When I was young, the key skill was extracting all the information from the few available sources. Now the key skill quickly deciding what to ignore amongst the morass of stuff available. After that you can concentrate on what's useful.

As for learning something, I'm always open to that! Hence my deciding in 30s whether this post or that video will teach me anything.

[bd139]

They've both been on for an hour now and are still the same distance apart which is good. The signal source I used to run them up into, my DG1022Z, hasn't budged either (or has tracked in the same direction as both of them). I think that has a TCXO in it. So relative accuracy is about the same. I will adjust the new one to match the older one now.

But that doesn't mean absolute accuracy even if my QCX which is GPS calbrated (not disciplined) agrees. Ergo, I need a proper GPSDO with 10MHz and 1MHz outputs now 

Uh-oh......



 

Not too far down yet. I have sync'ed them properly now. Quick steps for anyone who hasn't done this with old stuff:

1. Connect reference output of known good counter to ch1 of a scope and trigger on it.
2. Connect reference output of unknown counter to ch2 of the scope.
3. Adjust the unknown counter until ch2 stops rolling or as close as you can get it.
4. Connect good counter ref output to input of unknown counter.
5. Connect unknown counter ref output to input of good counter.
6. Stick on longest gate time and leave until two cycles have triggered.
7. If you did a good job, then they will show the same count.



Can get it to within 100mHz on a 10MHz OCXO out of that with these counters with a shit scope and a twiddle stick.

[med6753]

Get an Amazon Fire Stick. It gives you much finer control over the "sampling" fast forward and back.

A lot of Carlson's videos deal with vacuum tube (valve) equipment. Now to old bastards like you and I we already know that theory and have experience with it. But think of all the young EXPERIENCED engineers/techs who won't know a 6AL5 from a 12AX7 because they never worked with it. So the detail we find snoozing. So should Paul premise his videos by saying "you old bastards are gonna be bored" vs "you young guys be careful because this stuff will bite". He correctly assumes the later and you and I can decide if we want to watch it.

[med6753]

Not too far down yet. I have sync'ed them properly now. Quick steps for anyone who hasn't done this with old stuff:

1. Connect reference output of known good counter to ch1 of a scope and trigger on it.
2. Connect reference output of unknown counter to ch2 of the scope.
3. Adjust the unknown counter until ch2 stops rolling or as close as you can get it.
4. Connect good counter ref output to input of unknown counter.
5. Connect unknown counter ref output to input of good counter.
6. Stick on longest gate time and leave until two cycles have triggered.
7. If you did a good job, then they will show the same count.



Can get it to within 100mHz on a 10MHz OCXO out of that with these counters with a shit scope and a twiddle stick.

OK, how do you know which one was the "good" counter??? Hmmmm??? 

[bd139]

The marginally cleaner one! 

That one was tested against QCX GPS cal.

Next stop GPSDO so rabbit hole indeed 

[tautech]

Think the backlight on my TDS210 is starting to give out. Took a hell of a long time to warm up last night. May get an LED retrofit at some point next month.

Check the backlight HV inverter frequency. IIRC either the service manual states it in Circuit operation and typically is a couple of 10's of KHz.
In a TDS1002 I had the RC network for the push pull bipolars on the primary, well the C had drifted into the pF's. 
But that's all it was and an easy fix.

Thanks for the tip. Will check that 

Oh and you know you can check the oscillator frequency with just a probe held close to it. 

Thanks guys for the tips here, my display is a little yellow I think judging by some other photos I have seen so I might just check these caps and see if that has anything to do with it. Mine is still eminently usable though, its just I feel the display could be brighter and clearer than it is                                                                                                                                                                         

For the TDS1002 it was just a single dipped MLCC that had gone AWOL so the switchers (push pull) had run higher and higher till they passed the bipolar transition frequency and ceased to operate.
For your older TDS210 the CCFL backlight tubes are probably past their use by and probably an LED backlight conversion could be worth considering.
IIRC Shahriar did a LED backlight conversion vid that might be useful to hunt out. 

Found
https://youtu.be/smrMoIpvYTk

[tggzzz]

Get an Amazon Fire Stick. It gives you much finer control over the "sampling" fast forward and back.

Is that free, and how is it used on a computer with, say, the videos referenced here?

My only experience with Amazon Prime is on a "smart" tv. There the fast forward is braindead to the point of dissuading me from using Amazon Prime; Netflix, OTOH, is perfectly usable.

With Amazon, to skip forward and stop you have to press buttons XYXZX. To zoom forward you press XYXXZX. If you are concentrating on what's on the screen then the tendency is to leave your finger on that second X, rather than preparing to stop at some time in the future by pressing Z. The, if I come to a bit that looks interesting, I instinctively press the button (X) and zoom forward. Getting back to the possibly interesting point is an exercise in frustration.

GUI usability? No problem; the designers can use it just fine.

A lot of Carlson's videos deal with vacuum tube (valve) equipment. Now to old bastards like you and I we already know that theory and have experience with it. But think of all the young EXPERIENCED engineers/techs who won't know a 6AL5 from a 12AX7 because they never worked with it. So the detail we find snoozing. So should Paul premise his videos by saying "you old bastards are gonna be bored" vs "you young guys be careful because this stuff will bite". He correctly assumes the later and you and I can decide if we want to watch it.

Curiously, I know little about valves. Since I've now got a partially working Tek 184 which multiplies a 10MHz crystal output to 20MHz, 40MHz, 100MHz, 200MHz, 500MHz, I'm going through the pain of how 7487 nuvistors are enabled/disabled in the context of "random" tweakable inductors and "parasitic" couplings. Using 3 transistors (plus caps and resistors) to divide a period by 5 is comparatively obvious

Ideally Carlson (to unfairly continue your example) would state something like "this video takes you through the basics of X, and will be of interest to people that don't know the difference between a 6AL5 and a 12AX7". Or something.

[bd139]

Minor scope update ... 2235 dug out and new multiplier swapped in.



Much much brighter but still some intermittent flickering and focus problems  ... will check ALL the resistors in the HT side of things i think.

[med6753]

Get an Amazon Fire Stick. It gives you much finer control over the "sampling" fast forward and back.

Is that free, and how is it used on a computer with, say, the videos referenced here?

My only experience with Amazon Prime is on a "smart" tv. There the fast forward is braindead to the point of dissuading me from using Amazon Prime; Netflix, OTOH, is perfectly usable.

With Amazon, to skip forward and stop you have to press buttons XYXZX. To zoom forward you press XYXXZX. If you are concentrating on what's on the screen then the tendency is to leave your finger on that second X, rather than preparing to stop at some time in the future by pressing Z. The, if I come to a bit that looks interesting, I instinctively press the button (X) and zoom forward. Getting back to the possibly interesting point is an exercise in frustration.

GUI usability? No problem; the designers can use it just fine.

Amazon Fire Stick. Can be used on a smart and non-smart TV. Requires a vacant HMDI slot and Amazon Prime.

[med6753]

Curiously, I know little about valves. Since I've now got a partially working Tek 184 which multiplies a 10MHz crystal output to 20MHz, 40MHz, 100MHz, 200MHz, 500MHz, I'm going through the pain of how 7487 nuvistors are enabled/disabled in the context of "random" tweakable inductors and "parasitic" couplings. Using 3 transistors (plus caps and resistors) to divide a period by 5 is comparatively obvious

Ideally Carlson (to unfairly continue your example) would state something like "this video takes you through the basics of X, and will be of interest to people that don't know the difference between a 6AL5 and a 12AX7". Or something.

I attended 2 yr community college from 1971-1973 with an Associates degree in Electronics. At that time the courses were roughly split 50/50 vacuum tube theory and solid state theory. And I was very familiar with tubes prior to attending school.

Within a few years after that vacuum tube theory was almost completely eliminated from the courses.   

[med6753]

Minor scope update ... 2235 dug out and new multiplier swapped in.



Much much brighter but still some intermittent flickering and focus problems  ... will check ALL the resistors in the HT side of things i think.

Keep plugging away at it.  Tomorrow going to start cleaning/checking the 475A and generate a replacement tant list.