Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

   Had to Tweak the battery holder to MK2 for the 419A and make the arms a few mm longer so I could move the batteries nicely clear of the PCB and the rest of the mess behind it. The 1/2" frame hole spacing pushed the batteries to far back with MK1.

mmmm.... tasty! 

You should paint that bracket "phenolic brown"... sprinkle a little dust on it all, and nobody would know it didn't come that way as a super-secret undocumented special-order high-capacity option. 

mnem
*Instigating-ily*

[worsthorse]

Just bought another heathkit GDO. That’s 4 this week 

Each one contains one 500uA panel meter, One HT/filament transformer, One very nice variable capacitor, One socketed EC92 / 6AB4. Paid 1/4 the price for each one than the parts value. Plan is to get one museum grade restoration, one parts mule and build a regen RX with the left overs. Win.

That's pretty darn cool. Please post the restoration. I picked up one of the solid state heathkit GDOs and its in the rehab pile. And the regens, too. There's a stack of them here. It seems like whenever I get bored I build another one. 

[PrecisionAnalytic]

Personally, since they are still lumped components, I don't regard them as magic. I reserve "magic" for designs where you need finite element analysis of EM waves in order to design the PCB Examples:

...with some still alien technology to me.   

Neat link from on the Tekscopes posts with some excellent work with photos too: https://www.qsl.net/ct1dmk/wbond_ex.html

Wire bond?  I'll have to read into sometime.   

I'm just getting confident I can handle 0603 size components though need to practice desoldering some first.  Figured I'd have to use a hotplate and oven method if not a heat gun or IR heater.

Anyone ever made a custom plug-in module for a TDS-8000 series or repaired one?

[tautech]

I'm just getting confident I can handle 0603 size components though need to practice desoldering some first.  Figured I'd have to use a hotplate and oven method if not a heat gun or IR heater.

Get a Knife/Skew tip for your iron and with one you can bridge both SMD passive pads at once then either pick the passive off with tweezers or just wipe it off the pad with the iron.

This is what the Hakko K tip looks like:
https://www.hakko.com/english/tip_selection/series_900m.html#type_k

[beanflying]

Batteries and replacement wiring done as it was properly murdered by Blackwire death. IEC plug fitted triple checked the 240V switch was flicked and no smoke was removed and it is in the ballpark on batteries or mains. Shows signs of needing Capacitor replacements as it likes to drift. Still to fit the 1.35V reference. I really should have brought an Aneng POS

Bean Roasting and Beer time in the shack 

[tggzzz]

Neat link from on the Tekscopes posts with some excellent work with photos too: https://www.qsl.net/ct1dmk/wbond_ex.html

Wire bond?  I'll have to read into sometime.   

I'm just getting confident I can handle 0603 size components though need to practice desoldering some first.  Figured I'd have to use a hotplate and oven method if not a heat gun or IR heater.

Neat link indeed.

Don't be frightened of SMD; I now prefer it to PTH. 0.063mm pitch ic leads and 0603 are easy enough, provided you have magnification (I prefer a cheap head mounted visor over my glasses) and tweezers, and no sign of Parkinsons. I haven't needed to try smaller yet, but wouldn't hesitate. For really small I might want to try vacuum pickup on a divide-by-two pantograph.

My construction technique is a dedicated saucepan with a glass lid, a gas hob for instant heat input control, and a thin layer of sand to spread heat and decouple the PCB from the saucepan base. Do a couple of sacrificial PCBs first, use a non-contact thermometer to monitor the temperature, and watch the flux/solder change though the glass lid.

My rework technique, e.g. replacing one SMD resistor in the middle of a Tek 2445B A5 board, is to use a hot air gun. Loosely placed sticky polyimide/kapton tape over adjacent components deflects the hot air and protects them.

Touching a soldering iron tip to a ceramic SMD component runs the risk of thermal shock.

[bd139]

WRT to thermal shock on SMD ceramics, this is a big problem actually. However you can work around it well enough with home construction with a reasonable amount of flux being applied and keeping the dwell time under half a second or so. Experimentally I've been using AVX brand 500V 1206 ceramics as RF capacitors in power chains with 50W going through them and no explosions (yet) from hand soldering!

They are considerably cheaper than the other option of silver micas!

If you get the shakes, swap coffee for beer and they go away 

[tggzzz]

If you get the shakes, swap coffee for beer and they go away 

Ah, the professional dart player's "reason" for drinking.

Trouble is, if you have too much beer then all the components go away onto the floor - never to be found until the room is deep cleaned (i.e. never).

[bd139]

At least they're cheap now! Back in the old days, I'd hunt around on the floor for 10 minutes to find a single Tandy cruddy 5% carbon film resistor as it cost serious money for me back then between the zero income and Tandy's muggings. Now I just pop another one off the reel.

Talking of which, another big win recently. £3 inc delivery for a new unused roll of these, dated 2015: https://uk.rs-online.com/web/p/tantalum-capacitors/1360262/

[bd139]

On subject of SMD, I got these made up a couple of weeks back (10 of them) by JLCPCB and they arrived today. Not my design and I can't find the page they were on (will dig it out later):



SMA edge mounts fit on the edge conveniently and I have some cheap copper foil from China as well which you can cut and use as a ground plane if desired.

Edit: quick solder test. 100Hz LPF. Ugly, as expected:

[mnementh]

Neat link from on the Tekscopes posts with some excellent work with photos too: https://www.qsl.net/ct1dmk/wbond_ex.html

   

That is some hardcore "dark arts" shit right there.    Go up one level and you can see how he built the tools to do that work. Certainly NOT your typical "crank 'er up till she burns!!!" or 67 pages of variations on a single balun design HAM operator bloggery.

mnem
moo! 

[bd139]

First GDO arrived. Already have it in bits. Proper crudded up with shit and paint splashes as always.

Happy as a pig in shit today cleaning and refurbing

(this is all happening because it takes 40 minutes for the current thing I'm working on to compile  )

[LapTop006]

Had one of those annoying cases where you buy a nice bit of kit for repair, in this case an Agilent high performance DC power supply module and it just works :-(

So I knocked off a quick little project that had been sitting around. From another thread:

I personally miss the most physical LED for trigger event.
To the point that I connected a low power LED and resistor to trig out and sticking it out in the front.

Oh that's really tempting. The trig out on my keysight is on the back so I'd need to do something a little more than just putting a LED in a BNC.

In the end I picked up a cheap gooseneck LED light on eBay and replaced the PCB in the head with an 0603 LED (and no resistor).

Sadly it isn't anywhere near as bright as I was hoping, I guess due to the low duty cycle of the trigger out signal on my scope (which, annoyingly I don't have another scope to use to confirm that).

Possibly switching the white LED I used for a blue would give brighter results, and I almost can't believe I *want* that after years of taping over LEDs, blue ones in particular.

[tggzzz]

So, which TDR should I keep and which should I sell?

They are both showing the ~70mm of 50ohm stripline between the sampling diodes and the magic shorting BNC connector (the latter discharges any static in a cable before it it connected to the tunnel diode). The risetime is ~50ps, the falltime slightly longer.

The differences are:

• the top one has the best flat "internal cavity" response I've seen, but is calibrated in feet - which makes the *10 range setting difficult to calculate mentally.
• the bottom one has slight ripple, but is calibrated in metres - which is much more usable

I would like to keep the top one because of the flat response, but I don't do engineering in feet anymore (except 1ns = 1ft!).

Option 1: simply swap cavities or the front panel. But either of those would disturb the cavity and almost certainly bugger the response.
Option 2: change a few resistors/capacitors to rescale it, and stick a new range scale over the anodised scale.
Option 3: keep the bottom one.

Ah, decisions, decisions. That is a pretty good example of a "first world dilemma"!

[bd139]

Both 

[mnementh]

So, which TDR should I keep and which should I sell?

They are both showing the ~70mm of 50ohm stripline between the sampling diodes and the magic shorting BNC connector (the latter discharges any static in a cable before it it connected to the tunnel diode). The risetime is ~50ps, the falltime slightly longer.

The differences are:

• the top one has the best flat "internal cavity" response I've seen, but is calibrated in feet - which makes the *10 range setting difficult to calculate mentally.
• the bottom one has slight ripple, but is calibrated in metres - which is much more usable

I would like to keep the top one because of the flat response, but I don't do engineering in feet anymore (except 1ns = 1ft!).

Option 1: simply swap cavities or the front panel. But either of those would disturb the cavity and almost certainly bugger the response.
Option 2: change a few resistors/capacitors to rescale it, and stick a new range scale over the anodised scale.
Option 3: keep the bottom one.

   Ah, decisions, decisions. That is a pretty good example of a "first world dilemma"!

Option 4: Keep the better one just as it is, write a few macros for your sci-calc. 

Option 4A: Also add a secondary range scale with range substitution markers in metric. Should be very spreadsheet-doable.

Estimation isn't really that bad... 3.3 feet/meter or 0.3 meters/div (3.281/0.3048 for usable precision calculation) directly from that screen.

[EDIT] Are you certain the difference in response isn't entirely due to different scaling?

Also, is it possible that loading from the X-Y plugins are the cause of the variance? (Easy to test)[/EDIT]

mnem
"Shut up, fat dwagon!" That's actual WORK!!!

[tggzzz]

Both 

Get thee behind me, devil.

[worsthorse]

Nice dilemma! Did you restore both of those? Keep the one you want and use the money from the other one to buy a slide rule you can mount on the top to do conversions.   

And, yes, as soon as I saw your photos I headed over to ebay to see what was available because, of course, I need to add another item to the soon-to-be-rehabilitated list. 

[tggzzz]

So, which TDR should I keep and which should I sell?

They are both showing the ~70mm of 50ohm stripline between the sampling diodes and the magic shorting BNC connector (the latter discharges any static in a cable before it it connected to the tunnel diode). The risetime is ~50ps, the falltime slightly longer.

The differences are:

• the top one has the best flat "internal cavity" response I've seen, but is calibrated in feet - which makes the *10 range setting difficult to calculate mentally.
• the bottom one has slight ripple, but is calibrated in metres - which is much more usable

I would like to keep the top one because of the flat response, but I don't do engineering in feet anymore (except 1ns = 1ft!).

Option 1: simply swap cavities or the front panel. But either of those would disturb the cavity and almost certainly bugger the response.
Option 2: change a few resistors/capacitors to rescale it, and stick a new range scale over the anodised scale.
Option 3: keep the bottom one.

   Ah, decisions, decisions. That is a pretty good example of a "first world dilemma"!

Option 4: Keep the better one just as it is, write a few macros for your sci-calc. 

What is this "sci-calc" of which you speak? There's never one available when you need it in the middle of a field (battery powered:) and in the driving rain (nominally waterproof, but the waterproofing procedure is painfully complex, and there's 2kV on the the focus/intensity pots ).

Option 4A: Also add a secondary range scale with range substitution markers in metric. Should be very spreadsheet-doable.

Very easy calculation, e.g. 0.5ft/div is 152mm/div. The 5.6divs at 152mm/div can't be put on the front panel, and then it is back to the calculator

[EDIT] Are you certain the difference in response isn't entirely due to different scaling?

Absolutely sure: the ripple is the "suboptimality", not the pulse width.

Also, is it possible that loading from the X-Y plugins are the cause of the variance? (Easy to test)[/EDIT]

There's no significant loading there; all the problems are in the cavity. In the picture below, the cavity walls have been removed to allow you to see the internal stripline at the bottom of the picture. The tunnel diode is on the right, the shorting BNC connector on the left, and the sampling diodes between the two

[med6753]

Give me the one calibrated in inches/feet and keep the other one. Problem solved.   

[tggzzz]

Nice dilemma! Did you restore both of those?

I restored all four of them. It is difficult to distinguish from some kind of fetish.

Keep the one you want and use the money from the other one to buy a slide rule you can mount on the top to do conversions.   

Too late. I still have my (boxed) 12" slide rule and Sinclair Scientific that I used at school, and I've since added an HP35

And, yes, as soon as I saw your photos I headed over to ebay to see what was available because, of course, I need to add another item to the soon-to-be-rehabilitated list. 

There are plenty for sale, none with a visible trace, and some with an entertaining price tag. The tunnel diodes are fragile (hence the strange shorting BNC) and notoriously difficult to obtain.

Be aware that the PSU is strange in the extreme: there's an off-24V switcher based around a PUJT and an SCR, and it attempts to keep a constant mean current going into the battery. They don't even switch on unless there is a working NiCd battery in them (or NiMH but watch the currents, especially for ~400ms after you plug in the mains).

[tggzzz]

Give me the one calibrated in inches/feet and keep the other one. Problem solved.   

Will do. All you have to do is cross my palm with sufficient gold.

[med6753]

Boy oh boy these early DSO's can be a pain in the butt. The Tek 2430 is failing intermittently again on boot up. Same fail as the other day. 7000 CCD. If I drill down it stops at test 7221 CCD/Gain/FISO/Ch1-1. If I run self test it passes with the statement "Not Warmed Up". Same thing happens if I run self diags. Passes but that same message. If I hit "Menu Off" then "Acquire" the scope completes boot and seems to operate normally. If I got back say...10 minutes later..... and do self cal and self diags again they both pass with no warm up message. So something is flaky in channel 1. But since I can bypass the initial fail and get it boot and operate this one is low priority for now. Got other projects more pressing ahead of it.   

[med6753]

Give me the one calibrated in inches/feet and keep the other one. Problem solved.   

Will do. All you have to do is cross my palm with sufficient gold.

I notice you said "gold". That figures. I found out the hard way that you Brits won't accept USD.  Given that and the fact that I can't rob Fort Knox I guess it ain't gonna happen. 

[tggzzz]

Give me the one calibrated in inches/feet and keep the other one. Problem solved.   

Will do. All you have to do is cross my palm with sufficient gold.

I notice you said "gold". That figures. I found out the hard way that you Brits won't accept USD.  Given that and the fact that I can't rob Fort Knox I guess it ain't gonna happen. 

I'll accept platinum, but I'm not familiar with those payment devices. Gold is more readily available and tradeable; I suggest one Britannia or one Krugerrand would be sufficient, but others may be acceptable.

On second thoughts, make that Britannias or Sovereigns since I won't have to pay CGT on them when the pound falls. Buy them here:
https://www.royalmint.com/invest/bullion/bullion-coins/gold-coins/