HP 3478A Multimeter Repair and Test

[@radio655]

Man, what a lucky find! I was looking for ages for this type but never found one.

[@benmodel5745]

Have the same one, love it. Got it for $35 USD broken, a few replacement caps later I had a nice accurate bench meter

[@DarianCabot]

That meter has some miles on it! Good to see it live another day 👍

[@buildstoys]

Been pumping out the videos lately, thanks!

[@NearFarMedia]

Finally working through the backlog of older videos. :) Glad you like them!

[@juanolmosmendoza5361]

thank you very much for sharing your project

[@markgreco1962]

Changed my battery THANKS

[@hightttech]

Excellent video 👍. I have some maintenance to do.

[@ydonl]

It may be that the voltage reference in your older 3478a has drifted, leading to most of the values tending toward "low". Calibration of that meter is quite easy, especially if a nice 6.5 digit meter to use as a standard. I'd say "go for it," or maybe even test the process with some range you don't think you're likely to use very often, and get familiar with the process. You can do ranges individually, and you could put it back where it was if you think that's the best. You would need references a little more stable than some of the values shown in the video, so you're going to have to do something about shielding, or at the last, twist the leads, and turn off all the intruders in your lab that's producing all the noise -- e.g., get that huge transformer away from it! In my lab, I turn off the cheap LED lighting with its noisy power supply when I need "quiet." In general, it's possible you would have been better off swapping parts into the newer 3478a you bought; if you look at the serial numbers, and check the updates list in the manual (which you can easily get online), you'll see that the newer meters have an updated voltage reference that is more stable over time. There are other updates, some of which were made for cost and such, but some that make it "a little nicer," and assuming it was more or less functional, or easily fixable, you might have come out with a nicer meter overall. (Except that chassis... where the nice older die-cast chassis with all those metal parts and screws and such was replaced by sheet metal and plastic!) The display swap is pretty easy; of course I don't know what else was broken in your newer meter, so this all might be way off base! And you have already put in all the effort for the capacitor swaps... eh. Don't know. In any case, enjoy -- that's (still) a very nice meter!

[@hightttech]

Indeed, these are closed case calibration and I absolutely LOVE THEM.

[@ydonl]

@@hightttech I have three. :)

[@ebb2421]

while you have it open, replace the 2 RIFA caps (C765, C766) under the power switch and the 2 (C720, C721) under the HPIB ribbon cable.

[@hombre44567]

You are soldering the backup AA battery pack in parallel with the existing 3V battery without a diode. If the battery voltages differ (they always do), that's the equivalent of a short circuit with currents that are only limited by stray resistance inside the batteries. For example, if the AA battery pack is at 3.2V (fresh alkaline AA are usually around 1.6V each) and the lithium battery is 3V, you have a 0.2V difference creating a current that needs to go somewhere... Other 3478A or 3468A battery replacement instructions put a diode in series with the backup battery and make the connection to the PCB at the cathode end of diodes CR500 or CR764. You could use a simple 1N4148 diode for this. The forward voltage drop across the diode from 3V to 2.4V wouldn't be an issue because the standby voltage for the SRAM can go as low as 2V.

[@jackhoffman6120]

100% - I could believe what NFM was advising his audience to do. Other 3478A or 3468A battery replacement instructions suggest a schottky diode (hell is paved with good intentions) with its forward voltage drop of 0.2 to 0.3 volts, but you are 100% on the mark to suggest an ordinary 1N4148 is adequate for the job. . NFM provided a Digikey shopping list of parts (capacitors) to buy, but given his demonstrated competency to instruct his audience to place dissimilar batteries in parallel negates my confidence in his capacitor choice. How can you trust someone to choose the correct replacement capacitors, when they demonstrate elementary faux pas with regard to placing different batteries in parallel. You should only place identical batteries from the same manufacturing batch in parallel. Just because something in electronics works does not mean it is right. I was going to buy his recommended capacitors from Digikey, but I feel safer if I choose them myself.

[@anandarochisha]

​@@jackhoffman6120 jeez dude..no animals were injured during the making of this video. Confucius say: "Perfection is a comb that you are left with, after you have pulled all of your hair out"

[@tuom86]

Hi. I just bought this same one. Is there possibility to get shopping list for the parts ?

[@hamishdavidson3368]

Im looking for the original leads being Banana no shroud?

[@tiromancino_tt]

this DMM besides the capacitors you changed, also has at least 5 tantalum capacitors, which are worse than the plague. Also you haven't changed the 3000uF capacitor the bigger one. However you did a great job

[@t1d100]

Because the magic IC was connected mechanically, I wondered if the connections just needed to be cleaned. Did you try cleaning them? If not, you may want to test the old IC again.

[@NearFarMedia]

I actually decapped the ceramic lid and found a couple of the gold bonding wires had melted open like fuses. If I was a little more careful in opening it up, I was thinking I could alllmmoossstt repair it. :D

[@RB-lt8kt]

Technically "calibration" is testing an unknown item / device against a known value and recording the value of the unknown. It is not verification, which is what you referenced as "the calibration", against a specification / allowable error.

[@mariomionskowski6223]

@nfm Hello Jared, realy, realy good explained how to repair a HP-3478A, thank you for that, but one thing is wrong, the C761 have not a capacity of 30000µF it is only 3000µF.

[@juanolmosmendoza5361]

excellent video

[@dvhammer]

Would you please post the part numbers of the parts you used to replace the components? I have the same unit and want to make sure I get it right. Thank you!

[@NearFarMedia]

KZbin is being buggy again.... so rather than put a link to the Digikey shopping cart in the video description, I'll add it here: www.digikey.com/short/zndr39

[@tommyh4049]

@@NearFarMedia thanks a million for parts list

[@juanolmosmendoza5361]

The second question is regarding these multimeter models that do not have the continuity or diode measurement function. It is possible to modify the multimeter through a circuit to add this function.

[@NearFarMedia]

I don't know if this is possible, maybe easier to just have a separate meter for that. There is a project that added a few extra functions, but a diode check isn't one unfortunately. www.eevblog.com/forum/testgear/hp-3478a-rom-modification-relative-mode/

[@Homestead_Adjacent]

For diode test use a small constant current power supply . With high enough voltage you can also test zeners

[@markgreco1962]

Have one on the way won eBay 100$ working tested NEW SUBSCRIBER

[@stevenspmd]

Just picked one of these up myself. Now to do the same. What is your rule of thumb for knowing which caps to discharge to prevent getting zapped? I'n new to the hobby :-)

[@NearFarMedia]

If the capacitor's rated voltage is higher than around 50-ish volts, I'd start being careful.

[@stevenspmd]

@@NearFarMedia Great timing. I actually just scored a 3468a too ; apparently, just like the 3478a you can dump the calc ram for safety before swapping the battery.

[@abeleski]

Danced my ass off during the cap replacement

[@juanolmosmendoza5361]

How about I have 2 important questions that I would like you to answer me recently I bought a multimeter at a very good price but with a more or less serious defect the previous owner tried to change the battery without following the correct procedure and disconnected the backup battery that kept the readings calibration and this function was lost, my question is if you made a backup of the software contained in the ram u509 UPD5101L_NECElectronics when changing the battery, in case a low voltage occurred and the information was lost, did you have a backup of the software? if you had it, could you do me the favor of passing it to me or selling it to me to reprogram the ram and recover the multimeter

[@NearFarMedia]

Unfortunately I don't have a copy of the calibration data, I'll try to read it out over GPIB, if you have access to the GPIB interface on your unit? If you head to the eevblog forums, there might be someone there that can help you too. Although, a different unit's cal data won't make for accurate measurements, your unit really needs to be readjusted/recalibrated. Here's a link with some info: k6jca.blogspot.com/2017/12/save-and-restore-hp-3478a-calibration.html

[@soothcoder]

Maybe I missed it - did you do the GPIB thing to backup the calibration data first?

[@NearFarMedia]

Nah, I took my chances with the AA batteries. :)

[@nrcha]

Noice! If you are quick enough, you can just plop the new battery right in without any backups. Requires a bit of quick soldering, some faith and you’re all set for another eternity. btw - what do you do with methanol and vinegar?

[@NearFarMedia]

I thought about doing the quick swap, but knowing me, I would have taken half a second too long..... :D The methanol I use for removing stubborn stickers and ink from marker pens etc, the vinegar is for cleaning the alkaline goop left behind when batteries leak.

[@nrcha]

@@NearFarMedia Nooo, that was my attempt at making a joke. Although.. maybe the quick swap would work if you buy the battery at a church store. Is methanol better at removing stickers than ethanol? I should probably invest in a label printer before I drink the wrong stuff from my bench.

[@MitchellBumgarner]

😂 "The healing bench" 😂 😂 😂

[@Dont_Gnaw_on_the_Kitty_1]

Poking your dirty fingers into the precision measuring sections of this meter had me cringing. You left finger oil on pcb traces which will have affected its calibration. Please wear surgical gloves before you start poking around.

[@NearFarMedia]

It's all ok, it's all been cleaned. I probably should have shown that.. I always clean all boards I work on to remove grease, dirt and flux.

[@Dont_Gnaw_on_the_Kitty_1]

@@NearFarMedia The leakage paths on the pcb have now been altered and the calibration should now be re done to get this meter accurate. This is a precision instrument which is now inaccurate.

[@DavidSmith-zx7wz]

Too much talking, sorry

[@BruceNitroxpro]

Geeze... that zesty music and fancy editing with ten times the speed certainly showed nothing at all and taught me less. Pure crap.

[@NearFarMedia]

I apologize that I failed you, I'll give you a full refund on what you paid to view this video. :) Maybe you can show me how it should have been done with your own video? Oh, you haven't made a single video.... ;)

[@BruceNitroxpro]

@@NearFarMedia , No, I only owned a graphics/media company. Sorry that I put a damper on your humor.

[@DarianCabot]

Grow up Bruce.