Used $100-200 Bench DMM ?

[lordvader88]

I have some cash and I should get a bench DMM. I know that on SMPS controllers I'm dealing with, they enter overcurrent protection if Isense is 5mV higher than Vsense for each rail. If I really want to measure 5 millivolts with confidence that it's not just noise or drift...what should I be looking at ? So 3-1/2 digits won't do.


I'm looking at used ones like
Keysight 3478A
fluke 8840, 8842

VC8145 is mid-priced but I know what most folks would say about them

But another thing is often the used 1's don't have cables, so factor in the price of them, and rare stuff that might fail like LCD display, and the old ones from the 80s seem hard to read from any distance.

I really don't want to go to $250

Any recommendations ?

[nctnico]

The VC8145 (I have two of these as my daily DMMs) is a good all rounder with several handy features (like fast continuity, 20A current and capacitance) the other DMMs you list won't have. However I wouldn't use it on mains powered circuits.

[lordvader88]

I can really see myself wanting uV accuracy if I get into some physics experiments I wish I had money for.

But for day to day I'm sure a VC would be great. And it has that huge screen. So what's the problem it has no legit safety certifications ? (Aswell as not made bullet proof like a fluke would be)

[mtdoc]

I have both an HP 3478A and a Fluke 8842.  Both are excellent values.

You’ll get up to 6.5 digits with the 3478A. The no backlight LCD puts off some but I don’t mind and like that I don’t have to worry about burnout.

You need to be careful with the 8842 that the display is still bright and that the AC board is included.

[lordvader88]

I'm also tempted to get a broken 1 if it was just the PSU, but then again I really don't want it sitting broken for months, I want 1 now.

[nctnico]

I can really see myself wanting uV accuracy if I get into some physics experiments I wish I had money for.

But for day to day I'm sure a VC would be great. And it has that huge screen. So what's the problem it has no legit safety certifications ? (Aswell as not made bullet proof like a fluke would be)

The VC8145's construction is decent and it has a CAT II 1000V rating with the proper clearances and overvoltage protection components (including HRC fuses). OTOH it is still a relatively cheap device. For mains measurements I use a Keysight U1241B handheld DMM. It is more that I wouldn't like to use a bench DMM to measure a source (like mains) which can send high energy pulses into it which may degrade the input circuit somehow.

[David Hess]

If I really want to measure 5 millivolts with confidence that it's not just noise or drift...what should I be looking at ? So 3-1/2 digits won't do.

Most meters whether 3-1/2 digits or higher have at least 0.1 millivolt resolution and not all 4-1/2 or 4-3/4 digit meters have better than 0.1 millivolt resolution.

Noise or drift of the meter will be insignificant at this level but noise from the source might be a problem.

[lordvader88]

"HP 3478A Fluke 8840/8842,  fluke 45"
Any idea of the failure rates of these over time ? They are as old as me or younger.

Right now I'm actually learning towards the VC8145, even tho it $220 vs a $190 HP3478 or Fluke8840 or fluke 45


If the VC fails, it's probably a black box chip or cheap part/bad design.

If a 1980s model fails, I buy another 1 hoping to fix it basically.

But I'm guessing overall the 1980s ones are still more accurate ?? If your job depended on it I guess most people would take the used 1980s models still

Both of these are bad options, I could always wait till I have +500 and try again....IDK

I'm very tempted to get the VC since it's brand new, and I'll still use my Brymen DMMs most of the time probably anyway.




For the money I rather go with over 3.5 digit, just to get a bit higher up the price ladder, and not have to buy another in the near future.

[coromonadalix]

 for 250$  is it in Cad or Usd  currency

I can get you a good working fluke 189 with a new super capacitor in it  50,000 count display, data logging  etc ... for lower than 220$

I'm in Canada Quebec city

[mtdoc]

IMO a HP3478A (with a fresh battery) or a Fluke 8842A is going to be more reliable than a new cheap Chinese meter like the VC. And if they do break down, they can very likely be repaired using off the shelf parts with the full schematics fully available versus the VC which will likely go in the trash.

[Nx-1997]

Go for the HP 3478A if the self-test passes, they are great, just replace the internal battery. They are very stable, don't drift much but don't have null function lol. The fluke 8840A will drift alot, but don't have internal battery, also need to make sure the AC option is installed. Data logging from these DMMs requires GPIB adapter and cables, they are expensive. The fluke 8840A's VFD part can fail due to aging, replacements are rare and expensive sometimes. 150$ for these DMMs is a good price, if over that price, then ignore them.

[lordvader88]

for 250$  is it in Cad or Usd  currency

I can get you a good working fluke 189 with a new super capacitor in it  50,000 count display, data logging  etc ... for lower than 220$

I'm in Canada Quebec city

I'm in Canada, so that's another handheld tho. I have a BM869 that should on paper read 5mV well enough...but compared to a big bench DMM with big reference parts, I'm looking for bench models.

What price to big name brands of bench DMMs start at ?

[coromonadalix]

HP 3478A on ebay
They start around 80$ usd minus the shipping ... service manuals are available, i would go for this one
Seems to have an DC 30mv range, AC is 300 mv

Fluke 8842A, there's some service manual available .. some models not too expensive on Ebay too
You have one : https://www.ebay.com/itm/Fluke-8842A-09-Digital-Multimeter-SN-4282349/292539272207?epid=1300253118&hash=item441cb30c0f%3Ag%3AuSYAAOSwLfRbPjp3&_sop=7&_sacat=0&_nkw=Fluke+8842A+ac-09&_from=R40&rt=nc&_trksid=m570.l1313&LH_TitleDesc=0

You had Gw Instek GDM-8251A 120,000 count dmm, sold for almost nothing, but their price spiked, it was easy to enter them in cal mode ... no service manual leaked

thread : https://www.eevblog.com/forum/testgear/got-my-gdm-8251-from-ebay-today-one-question/


For starting prices, anything goes, you have  80$ up to 1.5K $$$  depending of what you need, and the calibration price may leave an sour taste

Personally i hate all the higher ends models with vfd's, since they tend to dim with usage, unless the display hacks for the 3458 and the recent : HP 34401A DMM with leaking segments   thread

For me the perfect one would be an Fluke 8846A

good luck

As stated previously, the parts availability may be an concern ...  unobtanium in some cases ...

[lordvader88]

Yeah on ebay $CAD150 total for broken ones, about the same for Fluke 45, 8840/42, and most don't have probes
 

and the VC at 220-230

[coromonadalix]

yep

I would skip the VC  tho, i did not see very good comments for it.

Fluke leads start at 8 to 20$ and up

[lordvader88]

I've seen firmware issues mentioned a lot with the VC and win7/8/10 problems...but then again 1980s stuff uses GPIB ? like on my Tek2430.




Whats more likely to be worn out between Fluke 8840/8842 and HP3478 ?

Looking up 'bench multimeter' 'bench DMM' on ebay leaves out a lot, of models I never heard of ( which is 99% of them)

They are all the same price basically, and VC is new....getting gifts is much easier than choosing stuff

[lordvader88]

there's lots like a Keithley 175 that looks fine and they are cheaper with 4 digits,


I literally know nothing about these, I just want to spend the money before I waste it on junkfood, alcohol bars, etc.

[nctnico]

IMO a HP3478A (with a fresh battery) or a Fluke 8842A is going to be more reliable than a new cheap Chinese meter like the VC. And if they do break down, they can very likely be repaired using off the shelf parts with the full schematics fully available versus the VC which will likely go in the trash.

But you'll need to have the equipment and skills to repair it. The fact is the VC8145 has way less parts and no parts that get hot so it is very unlikely to break. The build quality is also good. Mine have been working for many years already.

The reasons for me to buy the VC8145 and not an older used DMM is because the older DMMs have several drawbacks which make their value for money lower:
- VFD and LCD displays are small
- VFDs can get dim and the LCDs typically don't have backlight
- no features like continuity beeper, fast continuity, temperature and capacitance measurement
- usually limited to 3A (or so) current measurement capability
- usually single measurement display

[Mechatrommer]

Bench DMM? i've got an idea! already bought some stuffs for it. the cost is like $30 or something, excluding enclosure, PSU etc. assume the $30 as expendable cost, if it breaks, buy another, so its $60, if it breaks again, buy another again now its $90 and then $120, and then $150, still alot cheaper. i dont think its going to break 2 times in decades let alone 5, from previous experience. anyway, my handheld DMMs are standing by in any case.

[coromonadalix]

But lordvader wants to be able to measure 5mv changes,  the 3478A will be perfect with an 30mv DC range the Keithley 175 has 200 mv range

The VC8145 on Aliexpress are around 120$ usd and start at 80mv dc range, and has rs232 ports, more recent than gpib thingy
It would be cost effective and cheap, not complicated to run, and have recent technology ?

Agree on what nctnico says too.

An arduino gpib project exist
Emanuele Girlando: ARDUINO UNO as a USB to GPIB controller - version 6.1  : http://egirland.blogspot.com/2014/03/arduino-uno-as-usb-to-gpib-controller.html


There's a VC8145B who is an 33,000 count model,  but sometimes not advertised has the ''B'' model ...

[pigrew]

To measure 5mV on top of a 10V rail, you only need a 5.5 digit meter (assuming the last digit is noise), so a 3478a is perfect. Keithley 195A might be cheaply available, too.

But, isn't more digits better? If bench space is not a concern, I'd start looking for a 3457A or Keithley 193/196? Check the reading speed on these, they may be a bit slow. If you want to get into uV-level measurements, then you will start to worry about thermal voltages, want a unit with binding posts, and more digits.

The Fluke 8505A has a really weird interface and can't be configured with both AC and ohms (or something like that), but they claim 7.5 digit DC accuracy in averaging mode.

If you can get a good deal, the 34401A or 2000 may be in your budget. They are more modern.

I guess the real question is if you want a 1980s or 2000s-vintage meter. Older ones generally have slower read rates, but are repairable and you'll probably get more accuracy per dollar.

[Kosmic]

you can find a review of the VC8145 here: https://lygte-info.dk/review/DMMVici%20VC8145%20UK.html

It's pretty much a empty box 

You should definitly consider quality old DMM from ebay (or other source) any of them will probably be more stable and precise than the VC8145.

Got a Philips 2534 (6.5 digits) for something like 60$ recently. Was sold as used and working and it was calibrated in 2016.

Just be patient and you will find the deal you are looking for.

[coromonadalix]

oh oh  i did not tought of theses models, lm399 reference

service manual available:
http://www.ko4bb.com/manuals/download.php?file=Philips/Philips_PM2534_PM2535_System_Multimeter_Service_Manual.pdf

A teardown:

[lordvader88]

I went with a Keithley 177 for $80 total. It powers on but not tested.

It's 4.5 digit, and has 20u division, compared to 200u on pretty much all the other ones that price.


And at least most of the internals are old school repairable/replaceable.

I should get some high precession parts to test.

[rsjsouza]

Keithley 191 and Racal Dana 5001 are good options on a dime. My impressions:
https://www.eevblog.com/forum/testgear/bench-multimeter/msg1471414/#msg1471414