Fluke 8000A: makes me doubt my high opinions of Fluke DMMs

[tggzzz]

As mentioned elsewhere, I picked up a fully working broken Fluke 8000A for £5 at Dunstable Downs hamfest. I've got it working, but in the process my long-standing high opinion of Fluke has been dented.

After a while it started showing signs of plausible life, but with completely unsteady readings and no decimal point (shades of Siglent FFTs )

So I took it apart and irrigate the many switches with IPA. Much better, decimal points had reappeared, readings were mostly good but there was noise and the 20V scale was offset by up to 0.5V.

Flipped the board and there was clear evidence of simian involvement by the IEC inlet. ...

IMG_9092.JPG

Couldn't they have soldered the PCB to the connector properly?! Tried my Metcal with a beefy "ground plane" tip - barely touched it! WTF?

Flipped the board again and found why: each connector pin is a solid slab that is self-supporting all the way to the PCB. No shroud, apart from the small external Fluke-blue bit of plastic that slips off?! WTF again.

IMG_9095.JPG

Look more closely, especially at the other side of the connector (with Fluke blue shroud reinstalled) and see there is a shroud made of - wait for it - perforated card?! Involuntary shudder, as I remember Bulgin connectors.

IMG_9096.JPG

Remaining problem is the 20V offset. What could be causing that? Remember the concept of select-on-test resistors? Yes, they were a horrible practical necessity at the time. But they were tolerable when soldered in, or if selected by snipped wire links as in the bomb proof Fluke I tore down ant The Other Plaice https://www.eevblog.com/forum/testgear/ ... msg2540730

But, oh no, that would be too difficult to implement. Guess how Fluke did implement it...









... go on, guess ...












... using my favourite technology, a mini solderless breadboard.

IMG_9098.JPG

Waggle those resistors up and down to dislodge the oxide, and the 20V offset "magically" disappears.

Job done.

Now I just need a beer to overcome the shakes.

[MED6753]

You're not done. The +5V supply is barely +4.7V and not regulated. Results in lots of short term drift. Solution? Another small transformer and a proper 7805 regulator. I have pictures somewhere and I'll see if I can find them.

[25 CPS]

Look more closely, especially at the other side of the connector (with Fluke blue shroud reinstalled) and see there is a shroud made of - wait for it - perforated card?! Involuntary shudder, as I remember Bulgin connectors.

Bulgin plugs. Memories of Quad ESL 57 speakers there.

Anyways, the Fluke 8000A multimeter is a disappointment. This is not the sort of product you'd expect to see the Fluke name on. Micronta or some other Radio Shack branding, sure, but not fluke. I guess every company releases a basketcase of junk to the market at one time or another. I have two Fluke 8000A. One works but the other just displays nonsensical values. I pulled the working one and put it on the shelf once I got my second Agilent 34401A. I have a ton of projects that I inch along on at a glacial pace. The Fluke 8000s aren't on the bottom of the list - they aren't on the list at all.

[MED6753]

Rather than searching through my records I opened up the 8000A and took a fresh photo. It's a tight squeeze but the separate +5V supply is in there. I also modified the +15V and -15V supplies with 7815 and 7915. The result was a much more stable instrument.

[AVGresponding]

Funnily enough I just mentioned my 8000A in Discord in relation to fitting some new NiCd D cells. Naturally Chris wanted me to set fire to it instead, but I prefer to try Med's mod

EDIT: Just occured to me to observe that the 8125A which I have more interest in resurrecting the battery option in, is far more sophisticated than the 8000A, at least in terms of the front end and adc, but the battery charger is exactly the same; a bloody incandescent blub!

[tggzzz]

You're not done. The +5V supply is barely +4.7V and not regulated. Results in lots of short term drift. Solution? Another small transformer and a proper 7805 regulator. I have pictures somewhere and I'll see if I can find them.

Job Done. Full stop

I've little appetite for polishing a boring 3.5 digit turd - and improving this is definitely whack-a-mole territory.

[tggzzz]

Funnily enough I just mentioned my 8000A in Discord in relation to fitting some new NiCd D cells. Naturally Chris wanted me to set fire to it instead, but I prefer to try Med's mod

EDIT: Just occured to me to observe that the 8125A which I have more interest in resurrecting the battery option in, is far more sophisticated than the 8000A, at least in terms of the front end and adc, but the battery charger is exactly the same; a bloody incandescent blub!

Chris is right, in this case at least

The 8125 is more interesting than the 8000, particularly in the 8120 variant that I had.

[MED6753]

Funnily enough I just mentioned my 8000A in Discord in relation to fitting some new NiCd D cells. Naturally Chris wanted me to set fire to it instead, but I prefer to try Med's mod

EDIT: Just occured to me to observe that the 8125A which I have more interest in resurrecting the battery option in, is far more sophisticated than the 8000A, at least in terms of the front end and adc, but the battery charger is exactly the same; a bloody incandescent blub!

Chris also told me to set fire to it back in 2018.

[Zenith]

At the same event I paid £10 in cash money for a sort of assured working Fluke 8600A. You were there when I bought if.

It sort of worked but the bottom segment of each LED display didn't light. That was fixed by wiggling the flex connector to the display. It may be a dry joint. Then it played up with a lot of relay chatter. When it worked it agreed nearly exactly with a new 4½ digit Brymen. After a time it dropped into a strange state with the relay chattering.

Two of the test points specified for checking the power lines weren't there, and the documentation was lacking on where test points were in general.

The grotty axial electrolytics were replaced with parts from the junk box (not a job I am proud of) and three of the tants on the outputs of the three pin regulators were replaced. It now powers up and the relay doesn't chatter, but after a time it dies. The 5V line drops to 0.9V.

There's another tant that could be replaced, or the 5V regulator might be acting up - rare but it can happen. I'll have to investigate further.

These radio rallies are full of rogues selling rubbish to the unsuspecting. Shocking.

I have yet to investigate the BC348-H receiver that a rogue palmed me off with for £25. I doubt you could bung it into a B24 with any realistic expectation of it working.

[tggzzz]

Rather than searching through my records I opened up the 8000A and took a fresh photo. It's a tight squeeze but the separate +5V supply is in there. I also modified the +15V and -15V supplies with 7815 and 7915. The result was a much more stable instrument.

When I replaced a fried transformer in a Tek 1502, I simply removed the transformer and UJT-based(!) regulator, and replaced it with a small 240V_ac->12V_dc brick. Saved a lot of faffing about.

+5V, +15V, -15V example: https://uk.farnell.com/xp-power/ecl15ut ... dp/1860904

[MED6753]

Look more closely, especially at the other side of the connector (with Fluke blue shroud reinstalled) and see there is a shroud made of - wait for it - perforated card?! Involuntary shudder, as I remember Bulgin connectors.

Bulgin plugs. Memories of Quad ESL 57 speakers there.

Anyways, the Fluke 8000A multimeter is a disappointment. This is not the sort of product you'd expect to see the Fluke name on. Micronta or some other Radio Shack branding, sure, but not fluke. I guess every company releases a basketcase of junk to the market at one time or another. I have two Fluke 8000A. One works but the other just displays nonsensical values. I pulled the working one and put it on the shelf once I got my second Agilent 34401A. I have a ton of projects that I inch along on at a glacial pace. The Fluke 8000s aren't on the bottom of the list - they aren't on the list at all.

In the 1970's I worked Final System Test of IBM System370/3158 and we were required by engineering to use the 8000A to set the power supply voltages. The main voltages were +1.250V and -3.00V. No matter how much you twiddled the power supplies they would either sit at +1.252V or +1.249V. Never could figure out if it was the DMM or the power supplies. I suspect it was the Fluke based upon what I've seen all these years later fixing it's design flaws.

[tautech]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

[Zenith]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

We are talking about bench DMMs about 50 years old, which were probably designed with a design life of maybe 10 years in mind. Then they sold for $$$s painful and now are bought for the price of a beer or two. Some designs have stood the test of time better than others.

I'll help you on with your coat and shove your hat on your head.

[tautech]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

We are talking about bench DMMs about 50 years old, which were probably designed with a design life of maybe 10 years in mind. Then they sold for $$$s painful and now are bought for the price of a beer or two. Some designs have stood the test of time better than others.

I'll help you on with your coat and shove your hat on your head.

So are the prized by some CRO's of yesteryear yet their addition cost of ownership now is keeping them going.....

Not all of us have the time to fix tools in order to finish a project......15 yrs ago that hit me like a brick and moved entirely away from the old unreliable stuff.
Now being a reseller I don't have much time for projects....sometimes you just can't win.

[Zenith]

I've little appetite for polishing a boring 3.5 digit turd - and improving this is definitely whack-a-mole territory.

A working 3.5 digit bench DMM is OK. It doesn't run out of battery. If it doesn't work, it's in dustbin territory. However the processor is the same as used in better models, so if you are inclined to chuck it, remember me. I have an 8800A which I think has a shot controller IC.

[Zenith]

So are the prized by some CRO's of yesteryear yet their addition cost of ownership now is keeping them going.....

They have a savage grandeur which many find fascinating. Valve scopes are probably now a bit too savage for most.

Not all of us have the time to fix tools in order to finish a project......15 yrs ago that hit me like a brick and moved entirely away from the old unreliable stuff.
Now being a reseller I don't have much time for projects....sometimes you just can't win.

TEA is about a fascination with the tools and how they were made, and how something that cost the price of a house, is now available for a few quid, but may need some TLC.

[MED6753]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

I have several Flukes here. 8600A, 8010A, 8050A, 8800A and they all perform well. In addition to hand helded 8021B, 8060A, and 87. The 8000A is the only dud.

[tautech]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

I have several Flukes here. 8600A, 8010A, 8050A, 8800A and they all perform well. In addition to hand helded 8021B, 8060A, and 87. The 8000A is the only dud.

Yup, all brands have their dog models .....

All well in your part of the world Mike ?

[MED6753]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

I have several Flukes here. 8600A, 8010A, 8050A, 8800A and they all perform well. In addition to hand helded 8021B, 8060A, and 87. The 8000A is the only dud.

Yup, all brands have their dog models .....

All well in your part of the world Mike ?

All's well here Rob as we move into warm weather. How's things in Kiwi land as you move in Autumn?

[MED6753]

They have a savage grandeur which many find fascinating. Valve scopes are probably now a bit too savage for most.

Give me savage or give me death.

[tautech]

I have several Flukes here. 8600A, 8010A, 8050A, 8800A and they all perform well. In addition to hand helded 8021B, 8060A, and 87. The 8000A is the only dud.

Yup, all brands have their dog models .....

All well in your part of the world Mike ?

All's well here Rob as we move into warm weather. How's things in Kiwi land as you move in Autumn?

Pretty okay Mike, been doing a little travelling to deliver $ $ products and to do a warranty repair.

Was pretty lucky to also get approved for the flight deck on the 1.5 hr flight and get to see daughter at her work at the controls....very proud Dad moment !
Got the grand tour through a couple of customers establishments and one of them had a professional X-Ray machine just for inspecting potted PCB's and the like. Pretty cool for a electronic repair shop !

Getting cooler but no white stuff here like you guys get.

8hr road trip to see daughter again and grandy planned for Kings Bday long weekend next month and I'll drop in on a few customer buddys while down them ways.

[Cubdriver]

They have a savage grandeur which many find fascinating. Valve scopes are probably now a bit too savage for most.

Give me savage or give me death.

Agreed - savage FTW!


-Pat

[tggzzz]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

Read the first sentence of my OP.

The key phrase is "my long-standing high opinion of Fluke".

It is also clear that consequently this construction was an unpleasant surprise.

[tggzzz]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

We are talking about bench DMMs about 50 years old, which were probably designed with a design life of maybe 10 years in mind. Then they sold for $$$s painful and now are bought for the price of a beer or two. Some designs have stood the test of time better than others.

I'll help you on with your coat and shove your hat on your head.

My other Flukes are stunningly good, much more so than I would have believed.

When I get one I expect to have replace an electrolytic and/or clean a switch. If that is necessary, thereafter they are still spot on.

[tggzzz]

Based on this thread, shouldn't we declare Fluke products as a crap shoot.....
Gimme a mo to grab my hat and coat....

We are talking about bench DMMs about 50 years old, which were probably designed with a design life of maybe 10 years in mind. Then they sold for $$$s painful and now are bought for the price of a beer or two. Some designs have stood the test of time better than others.

I'll help you on with your coat and shove your hat on your head.

So are the prized by some CRO's of yesteryear yet their addition cost of ownership now is keeping them going.....

Not all of us have the time to fix tools in order to finish a project......15 yrs ago that hit me like a brick and moved entirely away from the old unreliable stuff.
Now being a reseller I don't have much time for projects....sometimes you just can't win.

The use case for professionals means traceability and time to market are key factors.

Don't deliberately confuse that case with the inexperienced and impecunious amateur use case, where the tradeoffs can be very different.