Test Equipment Anonymous (TEA) group therapy thread

[factory]

The 8840A and 8840a/AF are definitely a mess. One looks like it was left in a damp place and symptoms indicate a bad keypad but it lights up and passes all the self tests. That's the one with the banged up front. The display on the other one didn't light up. Opening it up revealed why:

I paid fifteen bucks for the two Flukes, so if I can make one good one out them, I will have a pretty good deal.

Does the keypad strip just need cleaning, same as old TV remotes do when they get an insulating oily residue on the contacts?

David

[worsthorse]

I had no plans to buy any test equipment at the hamfest. My goal, in fact, was to sell eight or ten pieces of gear.

I didn't sell most of the gear I took, though I did get rid of a 2M radio, assorted books, odd repeater parts, and a Type 109 pulse generator that was beyond repair, at least at my hands.  And I just couldn't resist buying a couple of things:

and these:



The impedance bridge is in very good condition and a quick test, using a bunch of resistor at hand, suggests it works. I had it open this afternoon and the inside is clean. Whoever owned it did not leave 4 D cell batteries in it to wreak havoc and a test of the precision capacitor indicates it is still in good condition. More tests on this later this week, after I finish a couple of other projects.  There were actually two of these at the hamfest; the other one went for $110, which was crazy expensive.

The 8840A and 8840a/AF are definitely a mess. One looks like it was left in a damp place and symptoms indicate a bad keypad but it lights up and passes all the self tests. That's the one with the banged up front. The display on the other one didn't light up. Opening it up revealed why:



Someone removed most of the interesting ICs. Now I have to figure out if they did so because there was something wrong with the meter. First stop, the power supply... all the voltage outputs appear to be within specification and I can't find anything obviously wrong with it.  That said, I was getting cross-eyed so I cleaned up the bench and called it a night. 

I paid fifteen bucks for the two Flukes, so if I can make one good one out them, I will have a pretty good deal. 

If you get 1 of those 2 working for $15 , you will have the bargain of the century I'd say, wish I could come across some more like those, I love my 2 Fluke bench meters. 

Yup, it will certainly be a bargain if I can get one good one out of these. I have an 8840A on my bench and it is my go-to meter.  I did some more work on them this morning, which I will post separately about, as I need to hear the tribal wisdom before proceeding.

[bd139]

Couple of random discoveries over the last couple of days which are test gear related:

1. "The Expanse" season 3 contains a couple of older Fluke multimeters and what appears to be an HP distortion analyser. The Fluke multimeters are perched on the comms console on the bridge of the UNN flagship.
2. Unigine superposition allows you to throw nixie based test gear and a couple of scopes around in a consequence free virtual environment: https://benchmark.unigine.com/superposition?lang=en (requires a beefy graphics card)

Aaaaand that's me done. Boring I know 

Edit: highly recommend The Expanse. It fits the Firefly hole perfectly. If you have Amazon Prime it's free on Prime Video. Also available from the usual warez vendors for those who don't grace Bezos with more wonga to make giant space cocks.

[worsthorse]

so... tribal wisdom time

I've torn down the two Flukes. They appear to have come from two different locations. They are marked as the same rev (D) and there are no discernible layout or parts differences, but the main PCBs are rev A and rev N.

Fluke A (an 8840A/AF with an air force sn sticker) powers up, passes the self-test, has a transformer that looks like it cooked itself in the potting, and most of the switch functions are non-working or flaky. It does measure a 5.00 VDC supply correctly. There is rust on the external rivets, so it looked like it was stored in a damp place.  The display is bright. I haven't checked the PSU on this one yet.

Fluke B (an 8840A) has been stripped of parts... all of the socketed parts (microcontroller, UART, EPROM, etc), along with an IC used in a analog filter and a Fluke chip that provides DC scaling control logic.  It is much cleaner inside and out, the transformer is in good shape. All of the PSU voltages are in spec and there doesn't appear to be any problems (ripple) on any of the DC supplies.

I could go three ways on this...

Option A. Given that the power supply is working is working in Fluke A, I could move the socketed parts from Fluke A into Fluke B and see what happens. A search of the forums here suggests that may not work because the microcontroller programming is tied to the board rev but it doesn't sound like it will blow anything up.

Option B. I could check the Fluke B power supply to be sure it is in spec in spite of the apparent damage to the transformer. If it checks out, I could tear down Fluke B and see if the button problem is can be fixed by cleaning. Low probability of success, I think.

Option C. After checking the Fluke B PSU, I could try swapping the front panel from B into A. High PITA factor.

I am leaning toward Option A as that DMM is just in much better shape overall and this would be an easy fix. If I got the screen to light and it passed the self test, I could scavenge the control and filter ICs from the beatup Fluke.

Any opinions? Am I missing an obvious approach? Anyone torn one of these down and done this kind of repair?

[ArthurDent]

A few years ago I bought this Fluke 8840A at a fleamarket. It was missing the display cover and was marked as defective. The problem was when you tried to switch to A.C. you got an error message. I made a display cover with some plastic I saved from a defunct HP counter which works just fine. The error was real because it didn't have the A.C. option installed, which didn't bother me one bit. As you can see it is pretty damn close on my voltage standard.

I talked them down to $16USD. 

[Neomys Sapiens]

I had no plans to buy any test equipment at the hamfest. My goal, in fact, was to sell eight or ten pieces of gear.

I didn't sell most of the gear I took, though I did get rid of a 2M radio, assorted books, odd repeater parts, and a Type 109 pulse generator that was beyond repair, at least at my hands.  And I just couldn't resist buying a couple of things:




The impedance bridge is in very good condition and a quick test, using a bunch of resistor at hand, suggests it works. I had it open this afternoon and the inside is clean. Whoever owned it did not leave 4 D cell batteries in it to wreak havoc and a test of the precision capacitor indicates it is still in good condition. More tests on this later this week, after I finish a couple of other projects.  There were actually two of these at the hamfest; the other one went for $110, which was crazy expensive.

I don't know what motivates you, but on my part there is a very firm decision: manual bridge measurements of ANY type and make will not re-enter my house! I think they are the metrological equivalent of a Ouija-board. All the guessing and trying whether the last action moved the pointer or not.  And in the rare cases when you get a plausible reading after the minimal number of switch operations, you can almost not believe it and do it again. Never again!

[FransW]

Absolute measurements for Volt-nuts and time-nuts.
Our efforts to reach the impossible are a joy for ever.

Frans

[0culus]

I have a 7.5 digit meter on my desk at work (Keithley 2010), and to be honest, for most applications I use thermal noise makes the last 2-3 digits pretty much useless. My best meter at home is my HP 3455A, and it's plenty good enough for my purposes.

[worsthorse]

I had no plans to buy any test equipment at the hamfest. My goal, in fact, was to sell eight or ten pieces of gear.

I didn't sell most of the gear I took, though I did get rid of a 2M radio, assorted books, odd repeater parts, and a Type 109 pulse generator that was beyond repair, at least at my hands.  And I just couldn't resist buying a couple of things:




The impedance bridge is in very good condition and a quick test, using a bunch of resistor at hand, suggests it works. I had it open this afternoon and the inside is clean. Whoever owned it did not leave 4 D cell batteries in it to wreak havoc and a test of the precision capacitor indicates it is still in good condition. More tests on this later this week, after I finish a couple of other projects.  There were actually two of these at the hamfest; the other one went for $110, which was crazy expensive.

I don't know what motivates you, but on my part there is a very firm decision: manual bridge measurements of ANY type and make will not re-enter my house! I think they are the metrological equivalent of a Ouija-board. All the guessing and trying whether the last action moved the pointer or not.  And in the rare cases when you get a plausible reading after the minimal number of switch operations, you can almost not believe it and do it again. Never again!

Oh... given the tools available today, there is nothing practical about the 1650B. I just think it is freakin' cool in the same way that a slide rule is cool.  It will never replace my DE-5000 but it gets the job done to a rough-and-ready level of accuracy and figuring out how to use it properly is a great way to learn.  But then I haven't really gotten sucked into the ten-digits-of-stable-accuracy-over-a-year-or-else alternative universe yet.

[xrunner]

I designed this earlier in the year to see what a 3D printed insulator could withstand. It's been out since spring and no signs of any damage or stretching. In fact it looks like it was just made. I didn't expect it to be in this good a shape being in the hot Texas sun all summer. 

[beanflying]

I have had Various Bridges on my evilbay list and then off again for a few years now. They are just a cool toy and nostalgic as we were made to use them in Secondary school in various forms. Do I need on absolutely not will I get on maybe 

Speaking of un needed 'toys' my Heathkit 3121 Curve Tracer turned up this morning and given it may have been made from a kit by a butcher I popped the hood for a look. Seems to have been a little rework on the board along the way as there is two fairly distinct solder looks. Overall nothing scary. Socketed IC's and some 1% resistors in the mix and when I plugged it in no smokes were released.

Very tidy overall just the knobs have a little oxidation on the inserts so I might print or cut some new ones. To repaint it in @med TEK Blue or not when it gets mounted to the stand 

Quaintly the seller included an entire CD with something on it but I will need to dig out the one PC I still own with a drive in it.

[mnementh]

Whoa bean... nice score there on that tracer; looks like you def won the Heathkit lottery. Not sure if it’s Jammy-worthy given the fair price you paid... but it does seem finding one at a fair price is pretty rare anymore.

I have a 7.5 digit meter on my desk at work (Keithley 2010), and to be honest, for most applications I use thermal noise makes the last 2-3 digits pretty much useless. My best meter at home is my HP 3455A, and it's plenty good enough for my purposes.

This is precisely how I feel about my 3478A... and my 189 most of the time.

mnem
Waiting for med to chime in any moment now... “And my little 87 too!!!”

[beanflying]

Yeah not really a score or jammy git item. Fair 'current market' price for what turns out to be a good example.

I have a 7.5 digit meter on my desk at work (Keithley 2010), and to be honest, for most applications I use thermal noise makes the last 2-3 digits pretty much useless. My best meter at home is my HP 3455A, and it's plenty good enough for my purposes.

Surprised the Keithley would be so bouncy. My 34401A has moved 6PPM (last digit) for a 15C temperature rise against a heated reference I am currently logging. Last digit on a 2010 for sure will be floaty as it share the same VRef (LM399) as my Agilent 6 1/2 digit meters and really isn't up to 7 1/2.

[capt bullshot]

From look and feel, I'd always prefer the HPAK 34401A over the Keithley 2015 (a K2000 with THD), e.g. for a simple DC voltage measurement of a halfway stable voltage, the 34401A simply looks better and more stable at a glance. To get a similar looking result from the K2015, one has to fiddle with the filter settings etc. the 34401A does it simply in its default variety. With a stable reference, both perform similar - but to my eyes the 34401A still looks better.

[tggzzz]

I have had Various Bridges on my evilbay list and then off again for a few years now. They are just a cool toy and nostalgic as we were made to use them in Secondary school in various forms. Do I need on absolutely not will I get on maybe 

Yup, except we made our bridges in physics lessons.

Ingredients: NiFe cell, 1m steel rule, 1m resistance wire, galvanometer, Weston Standard Cell, slide rule, pencil and paper.

Benefits: elemental understanding of what's inside "real" equipment., a visceral appreciation of the limits of precision in measurements.

[mnementh]

Also a “visceral understanding” of what happens when you “accidentally” place the clips too close to each other on the resistance wire... or someone “helps you out” in such fashion.

mnem
“Hot potato” got nuthin’ on hot nichrome...

[Specmaster]

so... tribal wisdom time

I've torn down the two Flukes. They appear to have come from two different locations. They are marked as the same rev (D) and there are no discernible layout or parts differences, but the main PCBs are rev A and rev N.

Fluke A (an 8840A/AF with an air force sn sticker) powers up, passes the self-test, has a transformer that looks like it cooked itself in the potting, and most of the switch functions are non-working or flaky. It does measure a 5.00 VDC supply correctly. There is rust on the external rivets, so it looked like it was stored in a damp place.  The display is bright. I haven't checked the PSU on this one yet.

Fluke B (an 8840A) has been stripped of parts... all of the socketed parts (microcontroller, UART, EPROM, etc), along with an IC used in a analog filter and a Fluke chip that provides DC scaling control logic.  It is much cleaner inside and out, the transformer is in good shape. All of the PSU voltages are in spec and there doesn't appear to be any problems (ripple) on any of the DC supplies.

I could go three ways on this...

Option A. Given that the power supply is working is working in Fluke A, I could move the socketed parts from Fluke A into Fluke B and see what happens. A search of the forums here suggests that may not work because the microcontroller programming is tied to the board rev but it doesn't sound like it will blow anything up.

Option B. I could check the Fluke B power supply to be sure it is in spec in spite of the apparent damage to the transformer. If it checks out, I could tear down Fluke B and see if the button problem is can be fixed by cleaning. Low probability of success, I think.

Option C. After checking the Fluke B PSU, I could try swapping the front panel from B into A. High PITA factor.

I am leaning toward Option A as that DMM is just in much better shape overall and this would be an easy fix. If I got the screen to light and it passed the self test, I could scavenge the control and filter ICs from the beatup Fluke.

Any opinions? Am I missing an obvious approach? Anyone torn one of these down and done this kind of repair?

You say the transformer looks cooked but its working OK, I think you will find that that some transformers had the yellow tape under the resin and others had brown tape, so I would suggest that its not cooked at all and is perfectly fine.

I personally would just swap over the front panels and then you will have a perfectly good meter albeit restricted to DC only because I didn't notice any indication on the back panels to show that they had the AC module installed.

[med6753]

More calibrations. The next victim is the Fluke 8000A. When I joined our little group of TE misfits about 1.5 years go this was one of my first projects. This 8000A came to me with what I considered excessive drift when warming up. In looking over it's design it immediately became apparent that Fluke went cheap on the power supply. The +5V was unregulated. It consisted of a bridge rectifier and a capacitor. I installed another transformer with a 7805 regulator plus modified the +/- 15V supplies by pulling the zener/pass transistor combo and installing 7815/7915 regulators. The result was absolutely no drift and this 8000A has been a stellar performer.

The calibration was a breeze. Just the 190mV needed a minor tweak. (189.9mV). Everything else was dead nuts. Next on the list to get checked is the Tek 2465 DMM option.   

[mnementh]

FLUKE sez BAM!!! 

mnem
*currently stalking the wily Smash Bros game cartridge through the wilds of New York*

[Mr. Scram]

I designed this earlier in the year to see what a 3D printed insulator could withstand. It's been out since spring and no signs of any damage or stretching. In fact it looks like it was just made. I didn't expect it to be in this good a shape being in the hot Texas sun all summer. 

What material is that? FDM prints are known to become very brittle when left to the elements.

[xrunner]

I designed this earlier in the year to see what a 3D printed insulator could withstand. It's been out since spring and no signs of any damage or stretching. In fact it looks like it was just made. I didn't expect it to be in this good a shape being in the hot Texas sun all summer. 

What material is that? FDM prints are known to become very brittle when left to the elements.

PLA. I didn't think it would hold up this good. Going to leave it for another year and see what happens. 

[Mr. Scram]

PLA. I didn't think it would hold up this good. Going to leave it for another year and see what happens. 

That's remarkable. Maybe the drought is actually helping as it's biodegradable. PLA usually isn't so resistant.

[factory]

I have had Various Bridges on my evilbay list and then off again for a few years now. They are just a cool toy and nostalgic as we were made to use them in Secondary school in various forms. Do I need on absolutely not will I get on maybe 

Speaking of un needed 'toys' my Heathkit 3121 Curve Tracer turned up this morning and given it may have been made from a kit by a butcher I popped the hood for a look. Seems to have been a little rework on the board along the way as there is two fairly distinct solder looks. Overall nothing scary. Socketed IC's and some 1% resistors in the mix and when I plugged it in no smokes were released.

Very tidy overall just the knobs have a little oxidation on the inserts so I might print or cut some new ones. To repaint it in @med TEK Blue or not when it gets mounted to the stand 

Quaintly the seller included an entire CD with something on it but I will need to dig out the one PC I still own with a drive in it.

Looks like that version is just a cosmetic update of the IT-1121 which looks identical on the inside.
Mine didn't need anything doing other than both fuses replacing with the correct values.

David

[Martin.M]

Bandwith 4.5kHz is now working, I have found the foul 

https://www.wellenkino.de/e108/test-2.mp4

[worsthorse]

so... tribal wisdom time

I've torn down the two Flukes. They appear to have come from two different locations. They are marked as the same rev (D) and there are no discernible layout or parts differences, but the main PCBs are rev A and rev N.

Fluke A (an 8840A/AF with an air force sn sticker) powers up, passes the self-test, has a transformer that looks like it cooked itself in the potting, and most of the switch functions are non-working or flaky. It does measure a 5.00 VDC supply correctly. There is rust on the external rivets, so it looked like it was stored in a damp place.  The display is bright. I haven't checked the PSU on this one yet.

Fluke B (an 8840A) has been stripped of parts... all of the socketed parts (microcontroller, UART, EPROM, etc), along with an IC used in a analog filter and a Fluke chip that provides DC scaling control logic.  It is much cleaner inside and out, the transformer is in good shape. All of the PSU voltages are in spec and there doesn't appear to be any problems (ripple) on any of the DC supplies.

I could go three ways on this...

Option A. Given that the power supply is working is working in Fluke A, I could move the socketed parts from Fluke A into Fluke B and see what happens. A search of the forums here suggests that may not work because the microcontroller programming is tied to the board rev but it doesn't sound like it will blow anything up.

Option B. I could check the Fluke B power supply to be sure it is in spec in spite of the apparent damage to the transformer. If it checks out, I could tear down Fluke B and see if the button problem is can be fixed by cleaning. Low probability of success, I think.

Option C. After checking the Fluke B PSU, I could try swapping the front panel from B into A. High PITA factor.

I am leaning toward Option A as that DMM is just in much better shape overall and this would be an easy fix. If I got the screen to light and it passed the self test, I could scavenge the control and filter ICs from the beatup Fluke.

Any opinions? Am I missing an obvious approach? Anyone torn one of these down and done this kind of repair?

You say the transformer looks cooked but its working OK, I think you will find that that some transformers had the yellow tape under the resin and others had brown tape, so I would suggest that its not cooked at all and is perfectly fine.

I personally would just swap over the front panels and then you will have a perfectly good meter albeit restricted to DC only because I didn't notice any indication on the back panels to show that they had the AC module installed.

The plot thickens, so I think I am going to move this project into a separate thread. When I sat down to the bench last night, the "working" Fluke started failing the self-test in a way that suggests a power supply problem but the power supply rails are all in spec.  I may try moving the controller chip set over to the other Fluke, just to see if the screen lights up and then figure out what to do next. 

sidenote: the AC measurement option was removed from both Flukes before I got my hands on them