EEVblog #791 - Ebay Fluke 45 Multimeter Teardown

[EEVblog]

Dave scored a Fluke 45 on ebay, the world's first dual display multimeter
Does it work?
What's inside?

Service Manual: http://oldradio.tesla.hu/rajzok/muszerek/fluke45.pdf
HD6303 Processor: http://www.cryptomuseum.com/crypto/philips/px1000/files/hd6303rp.pdf

[Moshly]

+1  for the Austen Tayshus (random interlude)

I was just watching that the other week on YT ->

https://www.youtube.com/watch?v=StcXGhuliRk&feature=youtu.be

[Huluvu]

nice Score 

I received mine in a almost new and unused condition which is very impressive considering that this unit was probably build around 2003/4

[retrolefty]

I've had mine for decades now. I've used the RS-232 output, it's simple to interface/convert for just about anything.

Too bad the volt-nuts will pass on the voltage reference used. 

Looks like it might have had a long production run, anyone know the facts. I do seem to recall a $799 retail price.

[Fungus]

Still $100 for a slightly beat-up, "as-is" meter after all those years? Jeez.

(And that's at "Dave" price. Everybody else will have to pay $200.)

[dentaku]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

[Muttley Snickers]

I honestly took it for granted and automatically assumed that he would have acquired a few of these over the years and always had the impression that he didn't think much of them or that they were simply to old. The average price is generally around $200 to $300 and sometimes a lot more, this is why I haven't given them a second look but also most listings either show a dim VFD or the unit powered down and I did keep an eye out for a good one for many years but gave up trying, that may change after seeing this video.

Thanks Dave for a fantastic insight....

Muttley

[Macbeth]

Clearly in need of a new VFD. Unobtanium is my guess?

[Tek_TDS220]

At 31:15 the copyright is 1986, not 1996.    Any chance the display is dim because the high voltage supply has sagged a bit? 

I would have liked a bit of explanation about the AC input buffer.  It looked more complicated than I expected.

I always learn a few things from these videos - thanks!

[Fraser]

I own thee. 8842a bench multimeters. They still have superb accuracy and all are working well. I have not regretted buying them for a second. IIRC I paid around £60 each for my units. I owned the 8842a before buying an 87 III 

The VFD is a weakness in heavily used units. It was still available from FLUKE when I last checked.

FLUKE made some top notch bench meters, the 45 is no different but I have found it to be overpriced on eBay. Personally I would avoid a dim display unit though as it isn't normally the HT, it is the Tube that has degraded. Some people have had success in overdriving the display to 'refresh' it..... Not a reliable process though.

Aurora

[linux-works]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

[linux-works]

I've had mine for decades now. I've used the RS-232 output, it's simple to interface/convert for just about anything.

Too bad the volt-nuts will pass on the voltage reference used. 

I have 2 of them I bought a few yrs ago when they were below $200 (dirty and took a chance on them but both ended up being just fine).

the voltage ref is not 'volt nut certified' you say?  its not a 6.5digit meter, though; wonder why a fluke design from that (good, old) era is not acceptable to those folks?

the build quality is great; very robust and reliable and not much to go wrong.

if the display starts to fade on me, I'll probably upgrade it with my own display and controller (the rs232 control port on the back is the key and I'll be able to get the data I need and send it to some other device for display or maybe retrofit my own display inside the fluke, should the real vfd not be buyable anymore.

btw, one trick to restore the shine to the plastic display is toothpaste and cotton cloth.  an old tee shirt works great.  use your finger and dab some toothpaste (a fine polishing compound, really) on the display and rub with fingers.  use a cotton cloth (keeping it clean frequently) and use that to polish the plastic lens.  on each bit of gear I've had to do that to, it restored the shine and can remove some surface scratches.  careful over painted areas, but on the smoked bezels, the toothpaste trick is a pretty well known trick.

[steve30]

The Polyputthekettleon capacitors is definitely one of my new favourite quotes .

[Macbeth]

Too bad the volt-nuts will pass on the voltage reference used. 

Oh I don't know, the Solartron 7081 8.5 digit uses a zener diode as its reference. The volt nuts quite like that meter. I think it's a 1N829?

Daves Krohn Hite voltage standard

 

also uses (the same?) plain through hole zener diode reference and that thing is amazingly spot on compared to a cal lab 3458A (and without a Kelvin Varley divider too!)

[boffin]

Dave; is the dim VF display due to a bad cap on that -30v line?

[Wytnucls]

Advert June 1989:

[Luddi]

direct RS232 output is nice.
I build a simple GPIB -> serial -> USB adapter for my HP 3455 with an ATMega, it has a "talk only" mode so i only have to acknowledge each byte, and they are plain text like the Fluke does.
Datalogging with a 30+ year old meter with basicly unlimited length

[D3f1ant]

I still have a Fluke 45, and until a year or two ago it was in daily use. The VFD screen is still bright, and it was still within spec accuracy. A nice feature if your into audio is DBm and audio power measurements.
My users manual was printed Jan 1989, and there is an errata insert dated June 1989. I think I got in about 1992 or there abouts.

[Tothwolf]

The 8V battery pack makes a lot of sense. From the photo, it looks like they are using 4x 2V 2.5Ah sealed lead acid cells, most likely Cyclon D size (p/n 0810-0004) from Enersys/Hawker/Gates. I've replaced tons of these batteries in other 1980s vintage portable instrumentation, and given their popularity, I strongly suspect that's what they used here. [Fun fact: It isn't uncommon to get a 10-15 year service life from a Cyclon SLA cell in a float charge configuration.]

I wouldn't consider the power supply design all that convoluted either. That particular switching converter arrangement makes a lot of sense given the rechargeable battery option.

[Someone]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

There are many variations on that idea, I found a useful recipe with 6% H202 in hot water with plenty of washing powder containing Tetraacetylethylenediamine. You leave the mixture out in the sun for a few days and plastics come out happy and white. The advantage is the ingredients are cheap so you can create large baths of it.

[XFDDesign]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

There are many variations on that idea, I found a useful recipe with 6% H202 in hot water with plenty of washing powder containing Tetraacetylethylenediamine. You leave the mixture out in the sun for a few days and plastics come out happy and white. The advantage is the ingredients are cheap so you can create large baths of it.

Is this 6% solution from the store, or 6% of the brew is H2O2? What powder contains... that mentioned compound?

[calin]

i got a Fluke 45 just few months ago with 40$ from craigslist almost new condition, looks and feels like new to me, with manual and all. This video is right in time . Mine has a super nice bright VFD .. heck I can't balem anything on it.


It quickly became my preferred bench meter especially because it does it just right on many things and yep the RS232 is ridiculously simple to read in a file over time. Simple to use, no fuss .. heck I got a pair of nice ProbeMaster leads just for it that were half the rice of the mete.  I alsso have his bigger brother 5 1/2 digit 8442A and both the 45 and 442A are still bang on on par with new 71/2 Keythley's i have at work. Pretty darn impressive from a Zenner i wold say

[teerosheyal]

Dave, Can you do a video on how to find and buy that kind of equipment on Ebay? I never find these deals you are talking about...

[Someone]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

There are many variations on that idea, I found a useful recipe with 6% H202 in hot water with plenty of washing powder containing Tetraacetylethylenediamine. You leave the mixture out in the sun for a few days and plastics come out happy and white. The advantage is the ingredients are cheap so you can create large baths of it.

Is this 6% solution from the store, or 6% of the brew is H2O2? What powder contains... that mentioned compound?

Aim for 3-6% in solution (the final mix). As for the washing powder you'll need to do some searching for yourself as the brands and products available worldwide vary considerably.

[Tothwolf]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

Unfortunately, this is only a temporary "fix". The yellowing eventually returns, even if the plastic isn't exposed to light.

Retr0Bright (or RetroBright) treated plastics re-yellowing even with minimal light exposure?

[krivx]

Since it works so well and there's no repair video needed, maybe you could do a video about cleaning yellowed plastic enclosures.

http://www.retr0bright.com/

Unfortunately, this is only a temporary "fix". The yellowing eventually returns, even if the plastic isn't exposed to light.

Retr0Bright (or RetroBright) treated plastics re-yellowing even with minimal light exposure?

Apparently there have been good results with using a UV-protective clear coat immediately after the Retr0Bright.

[Dave]

You completely derped your error calculations.
9.9974mA with a 10mA reference means it's 0.026% off, not 0.26%.  It's well within the 0.05% + 5 counts spec.

[German_EE]

There is a type of elevator that I used to service that has a VFD display for the floor indicator, a common complaint from customers was that the display was too dim. Our solution was easy, strip the display down and clean the inside of the smoked plastic cover, a design error meant that it got coated with a film of oil and then a layer of dust.

Try it with a Fluke 45.

[Tothwolf]

Unfortunately, this is only a temporary "fix". The yellowing eventually returns, even if the plastic isn't exposed to light.

Retr0Bright (or RetroBright) treated plastics re-yellowing even with minimal light exposure?

Apparently there have been good results with using a UV-protective clear coat immediately after the Retr0Bright.

Nope. Treated plastic re-yellows even when stored in a dark closet. Note the photos in the link I posted are followup photos to those in the original "Retr0Bright" link.

UV / light are not the reason the plastic returns to a yellow color, so a clear coat will not stop the chemical reaction. There are more free bromides below the surface layer, and a peroxide treatment can only work on the outermost surface of the plastic. Multiple re-treatments might eventually slow down the re-yellowing process (although that also risks damaging the plastic), but even that won't stop it completely. A clear coat over the plastic would actually be worse than no coating because once the plastic had been coated, you could no longer re-treat the plastic.

http://www.vintage-computer.com/vcforum/showthread.php?35227-Retr0Bright-Does-re-yellowing-occur-even-in-the-absence-of-UV-light&p=359591#post359591

"Retr0Bright" in its current form is only temporary, and in many cases, probably isn't worth the trouble since the results are not long lasting or permanent.

[krivx]

Unfortunately, this is only a temporary "fix". The yellowing eventually returns, even if the plastic isn't exposed to light.

Retr0Bright (or RetroBright) treated plastics re-yellowing even with minimal light exposure?

Apparently there have been good results with using a UV-protective clear coat immediately after the Retr0Bright.

Nope. Treated plastic re-yellows even when stored in a dark closet. Note the photos in the link I posted are followup photos to those in the original "Retr0Bright" link.

UV / light are not the reason the plastic returns to a yellow color, so a clear coat will not stop the chemical reaction. There are more free bromides below the surface layer, and a peroxide treatment can only work on the outermost surface of the plastic. Multiple re-treatments might eventually slow down the re-yellowing process (although that also risks damaging the plastic), but even that won't stop it completely. A clear coat over the plastic would actually be worse than no coating because once the plastic had been coated, you could no longer re-treat the plastic.

http://www.vintage-computer.com/vcforum/showthread.php?35227-Retr0Bright-Does-re-yellowing-occur-even-in-the-absence-of-UV-light&p=359591#post359591

"Retr0Bright" in its current form is only temporary, and in many cases, probably isn't worth the trouble since the results are not long lasting or permanent.

IIRC the theory of clear coat was to not to just to prevent UV damage but also oxidation. I can't remember where I saw this though, sorry.

[Macbeth]

"Retr0Bright" in its current form is only temporary, and in many cases, probably isn't worth the trouble since the results are not long lasting or permanent.

I disagree. It seems the discoloration comes back after YEARS. So really, pretty much just like the original equipment!

I can't think of a better way of restoring something than to then have its real genuine crap-factor to be restored too! as a purist surely this is better than an artificial cover-up?

I have to admit some of my PC PS/2 keyboards are in a sorry state and I tend to strip them down and give them a good scrub every few years, adding the Retr0bright treatment doesn't look like much bother.

[linux-works]

question to the plastics engineers out there: is it just a matter of cost, at manufacturing time, that could avoid the ugly yellowing, or was it that they simply did not have the ability to make plastic stable that long and not change color?

is it just about economics?  something tells me it was always the vendor's choice in how long they wanted the color to remain 'clean' and new looking.

admittedly, how many things are designed to 'look good' over decades?  if its commercial, who cares what it looks like; and if its commercial it will be expensed over time and rarely do you carry old equipment on the books for decades (does anyone, ever?)

so its only us collectors who would care; and I bet its not worth any money to the vendors to make the plastics last any longer than they currently do.  shame, of course.  but I assume its all about money and not due to it being particularly 'hard'.

[Muttley Snickers]

If it's any help I have a couple of function generators that are severely affected that came from a university and these were the two best out of all of the ones that they were getting rid of in regards to the discolouration. the underside of the units are pretty much unaffected and they show clear "bra lines" on the top where the handles had been left in situ for who knows how long.

The university electronics department had large surrounding windows so I would assume that depending on where particular units had been located dictated how severe the sun tan effect became. My intention was to dismantle them and use some cutting compound and then a light buff with a polishing wheel just to skim off a bit of the textured surface, I am more than happy to use these as guinea pigs if anyone has some suggestions that we could try otherwise they may just get a new coat of plastic paint unless I can find new cases, these will eventually be given away to a more needy person but I feel it is my duty to make them the best that they can be.

Competition: guess which one a was geek and enjoyed looking out the window.

Note: please do not quote this post as it may be considered as off topic and I might relocate it to another thread if I can find a suitable one.

[Tothwolf]

question to the plastics engineers out there: is it just a matter of cost, at manufacturing time, that could avoid the ugly yellowing, or was it that they simply did not have the ability to make plastic stable that long and not change color?

is it just about economics?  something tells me it was always the vendor's choice in how long they wanted the color to remain 'clean' and new looking.

admittedly, how many things are designed to 'look good' over decades?  if its commercial, who cares what it looks like; and if its commercial it will be expensed over time and rarely do you carry old equipment on the books for decades (does anyone, ever?)

so its only us collectors who would care; and I bet its not worth any money to the vendors to make the plastics last any longer than they currently do.  shame, of course.  but I assume its all about money and not due to it being particularly 'hard'.

It wasn't about making the plastic stable. These bromine compounds were added to plastics as flame retardants.

I suspect earlier on no one realized what these bromine compounds would do long term. Later, bean counters probably didn't care if they turned plastics yellow/brown. Due to their toxicity and that they are now classified as cancerigans, many of these compounds have been phased out, so newer plastics might not be affected by this today.

Edit -- another link which explains the degradation process: http://deskthority.net/keyboards-f2/retrobrite-the-yellow-strikes-back-t6084-60.html#p228424

[Treehouseman]

Got a chance to quickly tear down one of my 45's to show some differences, namely the addition of a 440mA HRC fuse for the 100mA range. Tried to take a look around for a solid date code, the relays are saying 0884, but the asic is saying 4201. The latter seems more reasonable to me because of the fact mine still has a bright screen and no yellowing, but that seems awfully recent. These were "defective" from the local community college, I nabbed 3 of them, all had blown 440mA fuses because no one could be bothered to look inside the thing past the 500mA fuse in the jack itself.

I can't upload all the photos in this post due to size, but they're here in my dropbox https://www.dropbox.com/sh/d2uza7q90yfm3m7/AAC7xDoRl8RfCbU43NXhKk84a?dl=0

[nicknails]

Dave mentioned the electrostatic shield that's between the primary and the secondary. I'm wondering what this is and where I can find some? I'm assuming it's some sort of foil and there's different versions for frequencies, etc, etc. I did a quick search on Digikey and Mouser and couldn't find anything. Is it called something else?

P.S. The company I work for has 42 Fluke 45s!

[bigsky]

I really enjoyed this video - it shows the superb quality of a Fluke meter - 22 years old, still working and in spec. I wonder if today's products will last as long.

[linux-works]

is there a way to run the 45 with the display off and only doing i/o over the serial or gpib ports?

I thought there was either a front panel seq or a scpi cmd to turn the display off, but I'm not seeing it.

[Stephan_T]

Hi Dave,

regarding your isolation transformer and its electrostatic shielding you claim that this "basically cuts its capacitance completely down to zero". Well I challenge you on that one.

My experience is: "Nothing isolated has a capacitance of zero". Parasitic capacitance is every where :-)
I have a very similar isolation transformer (HP M1389A), which also has this kind of shielding (protective earth wire going into the toroidal transformer).
When i run it at 230V, the output voltages of the secondary contact measured with the 10M? DMM show 160V (at one side, 67V on the other). As far as i know, the VDE 0701/0702 (similar to PAT, portable appliance testing) specifies a 2k? resistor to ground for testing touch currents on the isolated parts of a device. These 2k? represent the internal resistance of the human body. The current limit is specified at 0.5mA.

The 160V @ 10M? drop down to about 200mV when grounded with 2,2k?.
0.2V/2200? = 91µA

230V / 91µA gives about 2.5M? impedance  (@50Hz).

With X_c = 1/(2*pi*50*C) I get an effective capacitance of about 1.27nF or 1270pF.
Close, but not quite zero ;-)
Measuring capacitance between secondary and ground also gives me a value of 1,15nF (all measurements done with a Brymen BM257)

BTW, have you seen this thread (replies #4ff) about isolation transformers? I was shocked to see that there seem to be pseudo isolation transformers, which have a direct ground connection on the secondary side of the transformer. Such devices should not be called isolation transformers.