Test Equipment Anonymous (TEA) group therapy thread

[GregDunn]

One NiCd is well past its sell-by date and the battery voltage with power applied only totals 4.3VDC - not even 4VDC when running on batteries alone, which trips the LO BATT indication.  I'm pretty sure this is not helping the meter's accuracy at all.  I know I shouldn't power it up with the batteries out, so I'm pondering my options for additional testing.

It may be simplest to just replace the NiCd cells, depending on size, capacity and whether or not they are tagged. Much to my amazement, that turned out to be the case for my TDR.

Another alternative might be NiMH, but be careful with both the charging current (NiMH needs <C/20) and claimed capacity.

Not sure that is the case now. Decent nicd cells have a price premium since 2016. Not sure when you did your TDR.

I would find a way of regulating the output of the transformer myself. LT do some very low drop out regulators for not much. Pd will be quite low.

That's part of the quandary - NiCd cells are getting more expensive, and for something I'll never use it seems wiser to just replace them with a stable power supply while I'm in there checking out the meter.  But I will have to do some probing of the existing supply, which means trying to test its output with some kind of load and the meter circuitry disconnected so as not to fry anything.  Good though on using a LDO reg, but I'll need to be sure what my input/output voltages have to be.  I'll probably open a separate thread for this.

[med6753]

Yes unfortunately. Due to the size of it and the lack of size of my house, it turned out to be wedged under the desk most of the time at risk of damage. Also it caused a little contention in the household as my bench is in the living space. I am currently trying to stockpile some cash to buy a new much more compact SA so other things will be going soon too. About half way there now.

Plan is to keep the 475A and the TDS210 as they are diametric opposites and both have extreme utility.

Most of the gear you see me with is passing through generating cash to pay for the nicer and newer bits. My net expenditure on electronics is currently zero you see which is something I want to maintain when I have no income in the future.

How ironic. I'm on the cusp of greatly expanding my TEA capability and you on the other hand have to look at down sizing. 

[GregDunn]

Before I forget, some porns:

Good score - looks pretty clean!

You would have been tempted sorely at our fest yesterday; it seems like people are selling decent Tek gear all over the place lately.  I have no need for another scope, and was still intrigued by some of the bargains.  Moreover, people seem to take pretty good care of their scopes in this region and proudly have them turned on for display at the fests...  My "other" scope is a T922 and I literally haven't opened it since I bought it used in 1988.  I can't bear to get rid of it because it's still ticking along and gets used a fair amount.

[bd139]

How ironic. I'm on the cusp of greatly expanding my TEA capability and you on the other hand have to look at down sizing. 

Indeed. I suspect my entire house’s floor space is probably the size of your new area and there are 5 of us in it. Live on a tiny island in one of the busiest cities in the world.

Plan is to move out of here ASAP (tried twice already!) and bail out to a larger place in the middle of nowhere.

Before I forget, some porns:

Good score - looks pretty clean!

You would have been tempted sorely at our fest yesterday; it seems like people are selling decent Tek gear all over the place lately.  I have no need for another scope, and was still intrigued by some of the bargains.  Moreover, people seem to take pretty good care of their scopes in this region and proudly have them turned on for display at the fests...  My "other" scope is a T922 and I literally haven't opened it since I bought it used in 1988.  I can't bear to get rid of it because it's still ticking along and gets used a fair amount.

Keep that T922. Has personal value for sure

[med6753]

One NiCd is well past its sell-by date and the battery voltage with power applied only totals 4.3VDC - not even 4VDC when running on batteries alone, which trips the LO BATT indication.  I'm pretty sure this is not helping the meter's accuracy at all.  I know I shouldn't power it up with the batteries out, so I'm pondering my options for additional testing.

It may be simplest to just replace the NiCd cells, depending on size, capacity and whether or not they are tagged. Much to my amazement, that turned out to be the case for my TDR.

Another alternative might be NiMH, but be careful with both the charging current (NiMH needs <C/20) and claimed capacity.

Not sure that is the case now. Decent nicd cells have a price premium since 2016. Not sure when you did your TDR.

I would find a way of regulating the output of the transformer myself. LT do some very low drop out regulators for not much. Pd will be quite low.

That's part of the quandary - NiCd cells are getting more expensive, and for something I'll never use it seems wiser to just replace them with a stable power supply while I'm in there checking out the meter.  But I will have to do some probing of the existing supply, which means trying to test its output with some kind of load and the meter circuitry disconnected so as not to fry anything.  Good though on using a LDO reg, but I'll need to be sure what my input/output voltages have to be.  I'll probably open a separate thread for this.

Once you get the current issues sorted out with that 8600A you should consider changing the reed relays as I described in the old Fluke multimeter thread. But once I fixed that issue I found another issue that was being masked by the relays. Essentially the high resistance contact problem with the relay caused a short term (with in a half hour) drift of almost +0.003 volts. Once that was sorted out I discovered a long term (1 to 1.5 hour) drift of almost -0.003 volts. An applied voltage of 9.99691V would start at power on at 9.999V and slowly drift down to 9.996V. It took quite a bit of sleuthing to find the culprit. Y1, the 1Mhz crystal, that feeds the custom DVM chip was drifting. I replaced it with a generic $10 USD 1Mhz crystal from Mouser and it fixed it. Now at power on it reads 9.998V for about 5 to 10 minutes and settles down to a correct 9.997V and NEVER moves, even after many hours.



If you ever come across a Fluke 8000A I discovered a long term drift issue with them too and it appears to be a design flaw. The power supply has a shabby regulated +/- 15V rail and the +5V rail is totally unregulated. I completely redesigned a new supply with all voltages tightly regulated and that made the difference. I'll freely share that design if you need it.

And finally, I have experience bringing a total basket case 8010A back to life. So if you find one of those I can help there too.

[bd139]

Also to note on the 8000a the power supply pass transistors wear out. Not sure how they managed that. That 5v rail is unforgivable.

[med6753]

Also to note on the 8000a the power supply pass transistors wear out. Not sure how they managed that. That 5v rail is unforgivable.

Exactly, which is why 7815/7915 regulators were installed in their place. And the 5V rail got a 7805.

[bd139]

Good job. That’ll not blow the ASICs out now.

Just found some major OCD TEA: http://www.vacuumtubeaudio.info/project10.htm

[med6753]

Good job. That’ll not blow the ASICs out now.

Just found some major OCD TEA: http://www.vacuumtubeaudio.info/project10.htm

Wow! That is hardcore and some nice work, if you chose to go to that expense.

[Specmaster]

@bd139, I'd keep the 2235 and tds210 if I had to chose.

[bd139]

I did consider that but the 475A has 250Mhz bandwidth which is a metric shit ton. Perhaps all three 

Talking of which I’m going to pull all the tunnel diodes out of that 475a and test them this evening. Currently enjoying some peace in KFC for a few minutes. Chicken power!

[med6753]

I did consider that but the 475A has 250Mhz bandwidth which is a metric shit ton. Perhaps all three 

Talking of which I’m going to pull all the tunnel diodes out of that 475a and test them this evening. Currently enjoying some peace in KFC for a few minutes. Chicken power!

KFC is utter dog shit  compared to the fried chicken my woman makes. 

[WastelandTek]

jeez

are you guys still in here?

[med6753]

jeez

are you guys still in here?

Yes, and why do you ask? You think our addiction is just a passing phase? 

[tggzzz]

Talking of which I’m going to pull all the tunnel diodes out of that 475a and test them this evening.

They may be functional, but over time they have become incorrectly biassed.

ISTR there are procedures in the SM for setting the bias correctly (or am I thinking of a 465 or 485?). I'd certainly tweak things in the specified controlled fashion before letting my fingers loose on them.

[bd139]

KFC is utter dog shit  compared to the fried chicken my woman makes. 

For sure but the plane tickets are expensive 

Talking of which I’m going to pull all the tunnel diodes out of that 475a and test them this evening.

They may be functional, but over time they have become incorrectly biassed.

ISTR there are procedures in the SM for setting the bias correctly (or am I thinking of a 465 or 485?). I'd certainly tweak things in the specified controlled fashion before letting my fingers loose on them.

I’ve eliminated that one unfortunately. Went through the whole procedure. Issue is it stops triggering after about a minute of power up. All voltages are fine. No sign of any rail or capacitor problems. 

[GregDunn]

Once you get the current issues sorted out with that 8600A you should consider changing the reed relays as I described in the old Fluke multimeter thread. But once I fixed that issue I found another issue that was being masked by the relays. Essentially the high resistance contact problem with the relay caused a short term (with in a half hour) drift of almost +0.003 volts. Once that was sorted out I discovered a long term (1 to 1.5 hour) drift of almost -0.003 volts. An applied voltage of 9.99691V would start at power on at 9.999V and slowly drift down to 9.996V. It took quite a bit of sleuthing to find the culprit. Y1, the 1Mhz crystal, that feeds the custom DVM chip was drifting. I replaced it with a generic $10 USD 1Mhz crystal from Mouser and it fixed it. Now at power on it reads 9.998V for about 5 to 10 minutes and settles down to a correct 9.997V and NEVER moves, even after many hours.



If you ever come across a Fluke 8000A I discovered a long term drift issue with them too and it appears to be a design flaw. The power supply has a shabby regulated +/- 15V rail and the +5V rail is totally unregulated. I completely redesigned a new supply with all voltages tightly regulated and that made the difference. I'll freely share that design if you need it.

And finally, I have experience bringing a total basket case 8010A back to life. So if you find one of those I can help there too.

Yes, I am following the old Fluke thread as well - I'll probably add some replacement relays and a crystal on my next parts order just so I'll have them if mine prove to be troublesome.  I've used both of my previous Flukes plus the Keithley to track my voltage references over a 24 hr period (doing an error analysis) and they settle quickly.  I've been fortunate so far with not having many weird quirks in my meters.

I'm probably sticking with 8600A and 8800A (if I can find another one with the updated battery-free Ω board) for the future, but you never know what a hamfest will drop in your lap.     

I've opened a separate thread for my new 8600A so as to not fill this one with repair details:

https://www.eevblog.com/forum/repair/fluke-8600a-battery-powered-multimeter-convert-to-line-only/

[grizewald]

After a couple of pleasant evenings crossing off nodes on a circuit diagram and doing 3D puzzling, a new scopething is about to be fed power for the first time.

What do you reckon smoke or bouncy ball?

[tggzzz]

After a couple of pleasant evenings crossing off nodes on a circuit diagram and doing 3D puzzling, a new scopething is about to be fed power for the first time.

What do you reckon smoke or bouncy ball?

Neither. You'll have hours of fun debugging stray capacitance, inductance, and crappy contacts.

Solderless breadboards are the work of the devil. Use manhattan or rats-nest techniques, like the masters.

[grizewald]

 

I knew that would upset someone!

There was no smoke, but it did absolutely nothing apart from push one side high and the other low. More than likely a misplaced connection, but there's work tomorrow morning, so I'll have to debug tomorrow.

[bd139]

I knew that would upset someone!

There was no smoke, but it did absolutely nothing apart from push one side high and the other low. More than likely a misplaced connection, but there's work tomorrow morning, so I'll have to debug tomorrow.

Good luck. I would test each section at a time, starting with the oscillators, then the integrators, then the quadrature oscillator. Then the summing amps.

It pulled around 3-5mA per rail according to my power supplies when it was working - worth comparing against that.

tggzzz: The original was on solderless board as well. And it had a note saying "only to annoy tggzzz" 


More solderless board fun to annoy tggzzz. Testing tunnel diodes out of that 475A. All test good.

Poor man's curve tracer. 33120A as curve source, scope ch1 as voltage (measuring voltage across D), ch2 as current (measuring voltage across R). Don't do this with a ground referenced function generator - won't work! Has to be floating, one reason I bought the thing!



Now for a tunnel diode output. This is why these things are rather cool. Notice the "blank" spot which is a zone of negative resistance.



And a 1N4148 just for fun.

[med6753]

Once you get the current issues sorted out with that 8600A you should consider changing the reed relays as I described in the old Fluke multimeter thread. But once I fixed that issue I found another issue that was being masked by the relays. Essentially the high resistance contact problem with the relay caused a short term (with in a half hour) drift of almost +0.003 volts. Once that was sorted out I discovered a long term (1 to 1.5 hour) drift of almost -0.003 volts. An applied voltage of 9.99691V would start at power on at 9.999V and slowly drift down to 9.996V. It took quite a bit of sleuthing to find the culprit. Y1, the 1Mhz crystal, that feeds the custom DVM chip was drifting. I replaced it with a generic $10 USD 1Mhz crystal from Mouser and it fixed it. Now at power on it reads 9.998V for about 5 to 10 minutes and settles down to a correct 9.997V and NEVER moves, even after many hours.



If you ever come across a Fluke 8000A I discovered a long term drift issue with them too and it appears to be a design flaw. The power supply has a shabby regulated +/- 15V rail and the +5V rail is totally unregulated. I completely redesigned a new supply with all voltages tightly regulated and that made the difference. I'll freely share that design if you need it.

And finally, I have experience bringing a total basket case 8010A back to life. So if you find one of those I can help there too.

Yes, I am following the old Fluke thread as well - I'll probably add some replacement relays and a crystal on my next parts order just so I'll have them if mine prove to be troublesome.  I've used both of my previous Flukes plus the Keithley to track my voltage references over a 24 hr period (doing an error analysis) and they settle quickly.  I've been fortunate so far with not having many weird quirks in my meters.

I'm probably sticking with 8600A and 8800A (if I can find another one with the updated battery-free Ω board) for the future, but you never know what a hamfest will drop in your lap.     

I've opened a separate thread for my new 8600A so as to not fill this one with repair details:

https://www.eevblog.com/forum/repair/fluke-8600a-battery-powered-multimeter-convert-to-line-only/

Okey dokey....and if you need help in how to fatality damage a priceless 8800A in one easy step just ask and I'll provide those details too.  Yep....after replacing the reed relays I got careless in replacing the board and in my haste miswired it and probably destroyed several components including unobtaniums. Not one of my finest moments. It currently sits forlornly in the closet for possible repair.   

[0culus]

Utilized both the EEVblog promo code and my .edu email at Tequipment and ordered me a Hakko 301 desoldering gun, with extra tips and stand. In order to support repair of TEA of course. 

[GregDunn]

Once you get the current issues sorted out with that 8600A you should consider changing the reed relays as I described in the old Fluke multimeter thread. But once I fixed that issue I found another issue that was being masked by the relays. Essentially the high resistance contact problem with the relay caused a short term (with in a half hour) drift of almost +0.003 volts. Once that was sorted out I discovered a long term (1 to 1.5 hour) drift of almost -0.003 volts. An applied voltage of 9.99691V would start at power on at 9.999V and slowly drift down to 9.996V. It took quite a bit of sleuthing to find the culprit. Y1, the 1Mhz crystal, that feeds the custom DVM chip was drifting. I replaced it with a generic $10 USD 1Mhz crystal from Mouser and it fixed it. Now at power on it reads 9.998V for about 5 to 10 minutes and settles down to a correct 9.997V and NEVER moves, even after many hours.



If you ever come across a Fluke 8000A I discovered a long term drift issue with them too and it appears to be a design flaw. The power supply has a shabby regulated +/- 15V rail and the +5V rail is totally unregulated. I completely redesigned a new supply with all voltages tightly regulated and that made the difference. I'll freely share that design if you need it.

And finally, I have experience bringing a total basket case 8010A back to life. So if you find one of those I can help there too.

Yes, I am following the old Fluke thread as well - I'll probably add some replacement relays and a crystal on my next parts order just so I'll have them if mine prove to be troublesome.  I've used both of my previous Flukes plus the Keithley to track my voltage references over a 24 hr period (doing an error analysis) and they settle quickly.  I've been fortunate so far with not having many weird quirks in my meters.

I'm probably sticking with 8600A and 8800A (if I can find another one with the updated battery-free Ω board) for the future, but you never know what a hamfest will drop in your lap.     

I've opened a separate thread for my new 8600A so as to not fill this one with repair details:

https://www.eevblog.com/forum/repair/fluke-8600a-battery-powered-multimeter-convert-to-line-only/

Okey dokey....and if you need help in how to fatality damage a priceless 8800A in one easy step just ask and I'll provide those details too.  Yep....after replacing the reed relays I got careless in replacing the board and in my haste miswired it and probably destroyed several components including unobtaniums. Not one of my finest moments. It currently sits forlornly in the closet for possible repair. 

   

My very first build-it project (I was about 16) was the simplest thing ever - hook a power transformer to an amplifier board which already had the filter caps and rectifiers on it.  Couldn't be simpler, right?  For some reason, I swapped the CT with one of the other windings and put full voltage across the upper half of the complementary outputs.  I'd already survived working with tube amps having hot chassis and bare electrical wiring on my stereo rig, but I never had a problem till I let the magic smoke out of that dinky transistor amp...

[med6753]

More solderless board fun to annoy tggzzz. Testing tunnel diodes out of that 475A. All test good.

Poor man's curve tracer. 33120A as curve source, scope ch1 as voltage (measuring voltage across D), ch2 as current (measuring voltage across R). Don't do this with a ground referenced function generator - won't work! Has to be floating, one reason I bought the thing!



Now for a tunnel diode output. This is why these things are rather cool. Notice the "blank" spot which is a zone of negative resistance.

I've heard that someone came up with an equivalent circuit that mimics the action of those unobtanium tunnel diodes but never saw said circuit. I think you'd probably have better luck getting stabbed in arse by a unicorn.