Test Equipment Anonymous (TEA) group therapy thread

[Specmaster]

Personally I disagree with all of this nonsense that says only valve (tube) amplifiers are any good, same all the guitar add-on foot boxes. There is no reason why a decently designed transistor based circuit cannot replicate the so-called warm tone of a tube amp. It's a bit like (to me at least) some these UK voters who voted leave because they believe that good old days were the best ever in their lives and think that be leaving EU, we can recapture those times again. The reality is that those days were nowhere near as great as they remember them and the same goes for the audio quality, its all in the mind, and they can naff of and stop destroying good test gear.

[AVGresponding]

I have just spoken to someone I know who may be sending me some "desirable audio tubes". They are however knackered i.e. open filament or degraded but none are gassy so it looks like they're good. That may have the desired trolling effect 

edit: I'm going to just chuck them on the ground outside 

Perhaps you could fake remove them from a piece of audio gear, and toss them on the ground. Why? Because you need the casing to make a beer cooler ofc!


@menm, w10 has failed to solve the graphics problem     next step is to swap the 1060 out for an old 210 and see what happens
   

[BU508A]

This whole discussion about tubes, audiophoolery and what not the tubes can do to your ears are a bit silly, IMHO.
The main problem is that tubes are high-impedance devices, and speakers are just the opposite. 

But now this problem has been solved - by a long forgotten russian tube that can drive your speakers like nothing (45kV, 90A).
Its the GMI-2B, a must have. It adds glamour to your shack, leaves your fellow audiophools gasping with envy and it will also heat your flat.

You need it. Badly. 

Oooooh! What a beauty!     


I need one! Checking ebay ...

2 minutes later:   

They aren't that cheap, wow ...

USD 490.- for this one, plus shipping ...
https://www.ebay.com/itm/232982331328



This one from Bulgaria is cheaper, but still USD 260.-
https://www.ebay.com/itm/GMI-2B-2-Russian-Pulse-Tetrode/223478955081



Here are some more nice pictures and the datasheed can be found there as a pdf.
http://lampes-et-tubes.info/tt/tt025.php?l=e

Edit:
checked the data sheet.
weight: 5.5kg
Heater: 25V, something between 15 to 20A 
Ua = 32kV
Ia = 140mA

[worsthorse]

* worsthorse wandered off to make a bowl of popcorn about twenty-two posts ago...

i do love me those moments when you all get goin' on one topic or another. keeps me entertained for hours.   

[Wolfgang]

... but the pulse power for the base ! When you discharge this into a speaker the membranes can reach supersonic speed, I am so sure. 

[mnementh]

This whole discussion about tubes, audiophoolery and what not the tubes can do to your ears are a bit silly, IMHO.   The main problem is that tubes are high-impedance devices, and speakers are just the opposite.     But now this problem has been solved - by a long forgotten russian tube that can drive your speakers like nothing (45kV, 90A).  Its the GMI-2B, a must have. It adds glamour to your shack, leaves your fellow audiophools gasping with envy and it will also heat your flat.

You need it. Badly. 

Perfect for using to upconvert Dark Side of the Moon into analog uWAVE and beam it to Alpha Centari with your govt-surplus satellite dish...

mnem

[mnementh]

I have just spoken to someone I know who may be sending me some "desirable audio tubes". They are however knackered i.e. open filament or degraded but none are gassy so it looks like they're good. That may have the desired trolling effect     edit: I'm going to just chuck them on the ground outside 

Perhaps you could fake remove them from a piece of audio gear, and toss them on the ground. Why? Because you need the casing to make a beer cooler ofc!

@menm, w10 has failed to solve the graphics problem     next step is to swap the 1060 out for an old 210 and see what happens    

I'm currently arguing with Win10 over the drivers on my brand-new GTX2080 Super; ever since I installed it sleep/resume is totally borked. 

Of course, all I get back from Googs is page after page of fuckwits complaining that their PC crashes during gaming because they OCed their 2080 in stupid ways or hits for sleep/resume issues with 10-year-old video cards. 

mnem

[med6753]

I've calmed down enough to do some more capacitor replacement on the 535A. 

This go round 2 cans under the HV cover. One can is a dual 20uf/450V and the other a triple 10uf/450V. For some reason Tek wired the dual 20uf in parallel for 40uf equivalent. Don't know why they did that. Maybe they couldn't source a single 40uf/450V? Anyway, I decided to replace it with a single 47uf/450V.

Here's the before shot.



After.



And of course all is well.



There are 5 more cans to do. So that makes an even dozen in this scope. And it figures I only have 4 terminal strips left. I may be able to double up 2 cans on one terminal strip.   

[mnementh]

* worsthorse wandered off to make a bowl of popcorn about twenty-two posts ago...

i do love me those moments when you all get goin' on one topic or another. keeps me entertained for hours.   

Yeah, this is almost as pornographic as watching OPVN... 

mnem
 

[Wolfgang]

... And now the ultimate hifi porn tube :

see picture.  .It wont get a lot better than that.

[bd139]

Ok pulse generator take two. I was digging in the cupboard of doom and found my old JW pulse generator (2n2369 avalanche mode). Alas I couldn't find a power supply or enough things to plug together that had plugs to get the required 100V to give this thing a kick (70V my ass). Quick breadboard boost converter was developed with an IRF510 running dangerously over Vds(max) (fingers in ears) driven by the 3312A. It worked though. Voltage regulation was provided by dragging gate to ground using a 2N3904 and a resistive divider on the output.



This was plonked into the pulse generator that lives in the end of a Pomona box:



Getting busy on the bench now



And the results are pretty good!:



I make that ~296MHz equivalent bandwidth... suspiciously high? Maybe not. Seems to be within the realms of possibility for this little box: https://www.eevblog.com/forum/beginners/oscilloscopes-to-stay-away-from-(spec-tek-2337-question)/msg46880/#msg46880

Edit: also have to say this. I'm not sure why I didn't buy one of these scopes before. It's just perfect.

Also some circuit notes. MOSFET = IRF510. Standard boost topology. Inductor is from junk box around 6.8mH (way too high but it'll do). 10R in series with gate or it oscillated like mad. 470k/2.2k divider. 100uF decoupling cap. 1uF film across output.

[tggzzz]

I make that ~296MHz equivalent bandwidth... suspiciously high?

The pulse you see there is, mathematically speaking, the convolution of the input signal and the oscilloscope's time domain response. So, to work out the scope's response, you have to deconvolve the displayed trace with the input signal. Unfortunately you don't know the input signal, so you can't do that, and therefore cannot use the method to measure bandwidth. (Special case approximation: if you know the input signal's peak amplitude, and that its pulse width is much shorter than the displayed signal, then you can make useful estimates of the scope's response.)

Of course, the rule-of-thumb based on the step response risetime (/t_r) is f_3dB = 0.35/t_r. But you don't have a step response, so you can't use that rule-of-thumb

One unusual source of a step waveform is a Tek 1502 TDR - the tunnel diode generates a clean 50ps 0.4V step. You could have saved yourself some trouble by buying one when you had a chance

[Specmaster]

This whole discussion about tubes, audiophoolery and what not the tubes can do to your ears are a bit silly, IMHO.   The main problem is that tubes are high-impedance devices, and speakers are just the opposite.     But now this problem has been solved - by a long forgotten russian tube that can drive your speakers like nothing (45kV, 90A).  Its the GMI-2B, a must have. It adds glamour to your shack, leaves your fellow audiophools gasping with envy and it will also heat your flat.

You need it. Badly. 

Perfect for using to upconvert Dark Side of the Moon into analog uWAVE and beam it to Alpha Centari using your govt-surplus satellite dish...

mnem

Can you imagine the heartbeat with that powering the speakers, ear bleeds 

[grizewald]

bd, just get the Bodnar pulser and be done with it!

[grizewald]

The saga of the Zener TEAcosy continues...

Although I don't have my GPIB interface up and running yet, it occurred to me that the 7061 is a computing multimeter! (D'Oh). So I enabled the MaxMin and Statistics programs and ran the meter for 1.5 hours or so with the leads shorted both with and without my little cotton tea cosy for the reference.

Settings were: auto range (it'll stick at the lowest range with shorted inputs anyway), filter off (to expose as much internal noise as possible), DC Volts (of course) and 7 digits (2 second integration time).

 The results are quite interesting...




(text version for cut and pastery)

MAX +3.166496e-7 +2.495944e-7
MIN -3.241002e-7 -5.178153e-7
PP +6.407499e-7 +7.674098e-7
MEAN -1.678598e-8 -1.631235e-7
SD +6.158931e-8 +1.010704e-7
VAR +3.793245e-15 +1.021523e-14
RMS +6.383583e-8 +1.918972e-7

[bd139]

I make that ~296MHz equivalent bandwidth... suspiciously high?

The pulse you see there is, mathematically speaking, the convolution of the input signal and the oscilloscope's time domain response. So, to work out the scope's response, you have to deconvolve the displayed trace with the input signal. Unfortunately you don't know the input signal, so you can't do that, and therefore cannot use the method to measure bandwidth. (Special case approximation: if you know the input signal's peak amplitude, and that its pulse width is much shorter than the displayed signal, then you can make useful estimates of the scope's response.)

Of course, the rule-of-thumb based on the step response risetime (/t_r) is f_3dB = 0.35/t_r. But you don't have a step response, so you can't use that rule-of-thumb

One unusual source of a step waveform is a Tek 1502 TDR - the tunnel diode generates a clean 50ps 0.4V step. You could have saved yourself some trouble by buying one when you had a chance

Yes indeed. It’s a finger in the air. The next step is the step response. I’ll get to that in a bit.

And yes indeed the TDR would have done the job .

Well it turns out some nice little LVDS drivers and comparators seems to have <100ps rise and fall times so I’m cost balancing that now. I don’t need bodnar level performance so there are a few cheaper devices that are easy to manage hand soldering that will be tried now. Also being a comparator you’re not dependent on the avalanche repetition rate and can control the pulse width and duty easily. Thus I’ll knock up a little board via JLCPCB and drop a comparator on it. Drive it with a function generator and bench supply and see what I can screw out of that.

[bd139]

bd, just get the Bodnar pulser and be done with it!

I probably should but why be easy when you can be impossible and try and build something yourself

Edit: also sometimes it’s about the journey and not the destination.

[BU508A]

bd, just get the Bodnar pulser and be done with it!

Nah, bd can't do that, way too easy.   

Edit:
I did a comparison of the Jim Williams and the Leo Bodnar fast pulse generator here:
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg2659587/#msg2659587

In comparison to the pulse from bd139 my pulse is much wider and therefore more applicable for the rule of thumb.

Edit 2:
This is my 100V+ generator:


Based on this schematic:

[grizewald]

Edit: also sometimes it’s about the journey and not the destination.

I thought the 3D printed case would have tipped you over the edge.

Must. Try. Harder.

 

[worsthorse]

Ok pulse generator take two. I was digging in the cupboard of doom and found my old JW pulse generator (2n2369 avalanche mode). Alas I couldn't find a power supply or enough things to plug together that had plugs to get the required 100V to give this thing a kick (70V my ass). Quick breadboard boost converter was developed with an IRF510 running dangerously over Vds(max) (fingers in ears) driven by the 3312A. It worked though. Voltage regulation was provided by dragging gate to ground using a 2N3904 and a resistive divider on the output.



This was plonked into the pulse generator that lives in the end of a Pomona box:



Getting busy on the bench now



And the results are pretty good!:



I make that ~296MHz equivalent bandwidth... suspiciously high? Maybe not. Seems to be within the realms of possibility for this little box: https://www.eevblog.com/forum/beginners/oscilloscopes-to-stay-away-from-(spec-tek-2337-question)/msg46880/#msg46880

Edit: also have to say this. I'm not sure why I didn't buy one of these scopes before. It's just perfect.

Also some circuit notes. MOSFET = IRF510. Standard boost topology. Inductor is from junk box around 6.8mH (way too high but it'll do). 10R in series with gate or it oscillated like mad. 470k/2.2k divider. 100uF decoupling cap. 1uF film across output.

...the 184 that I accidentally turned into a parts mule has 200MHz and 500MHz outputs. All of the problems are in the countdown board; the front end frequency generation path is fine. I wonder if I could turn it into a pulse generator?   

edit: overkill? of course. otherwise, why bother?   

[grizewald]

I guess that the only problem with making a pulser yourself is if you don't have a sampling 'scope to actually measure the pulse?

[bd139]

I guess that the only problem with making a pulser yourself is if you don't have a sampling 'scope to actually measure the pulse?

Turns out it’s not rocket science to build your own 1GHz sampling head 

[tautech]

I guess that the only problem with making a pulser yourself is if you don't have a sampling 'scope to actually measure the pulse?

Turns out it’s not rocket science to build your own 1GHz sampling head 

(Attachment Link)

Or get a 500 MHz scope and hack it to 1 GHz ! 

[bd139]

Not as much fun. Plus I reckon with some parts subs you could squeeze some more juice out of that design.

[mnementh]

This whole discussion about tubes, audiophoolery and what not the tubes can do to your ears are a bit silly, IMHO.   The main problem is that tubes are high-impedance devices, and speakers are just the opposite.     But now this problem has been solved - by a long forgotten russian tube that can drive your speakers like nothing (45kV, 90A).  Its the GMI-2B, a must have. It adds glamour to your shack, leaves your fellow audiophools gasping with envy and it will also heat your flat.

You need it. Badly. 

Perfect for using to upconvert Dark Side of the Moon into analog uWAVE and beam it to Alpha Centauri using your govt-surplus satellite dish...

mnem

Can you imagine the heartbeat with that powering the speakers, ear bleeds 

Yeah... they could probably hear it on Alpha Centauri, vacuum of space be damned... 

Edit: also sometimes it’s about the journey and not the destination.

I thought the 3D printed case would have tipped you over the edge.

Must. Try. Harder.

 

mmmokay... bd - what's your uncertainty factor multiplied by the cost of BOM for these two projects and a couple afternoons just to save what... 50 clams for a known quantity in the LBP...? 

mnem
"You're a monster, Zorg."
"I know."