Fluke 8505A vs HP 3456A

[kj7e]

Got the urge for a large LED bench DMM.  Between the Fluke 8505A and the HP 3456A, setting going price aside, which would you choose and why?

I do like the possibility for a shunt module on the Fluke, the A/C module is not a big concern but would be nice.  I know the HP is just dead solid stable, I've not found too much info on stability for the Fluke, but I suspect its close to the same.  The Fluke is not quite as deep, and would fit table where I would place it better.  Just want a new toy to play, no specific need. 

[pelule]

I am proud owner of both HP 3456A and Fluke 8506A (similar to 8505A except it offers thermal RMS conversion).
Simple said. It depends...(on the options installed at the Fluke).

HP3456A
HP3456A is a DVM
it has GPIB
it measures AC voltage (RMS by square root method), DC voltage 2w Ohms and 4w Ohms.
It does not offer current, neither DC nor AC (even I have seen offers claiming it does, but that may defekt  ).
HP 3456A offers a very high accuracy for a 6.5 digit meter and seems to have very low drift over time (see thread here).

8506A/8505A
I have all available modules (options) for my 8506A
so it has GPIB
it measures voltage DC & AC RMS (thermal conversion for 8506A only) 2w and 4w Ohms and DC Current.
It also offers a 7.5 digit resolution in the 10V DC range.
Since 2014 I have the equipment (HP 3458A) to measure the drift over time.
My 8506A has similar low drift as my 3456A.

In fact I never want to miss any of them - I am using both for GPIB bus controlled logging.
/PeLuLe

[pigrew]

My 8505A has a bit of 60-Hz mains hum from its transformer. It's annoying, but doesn't seem to effect the measurements. Mine has the DC options + GPIB (no AC measurements). The 8506A is probably better for AC, anyway.

When I bought it off of eBay, its calibration seemed a bit off when compared to a 34401A with a fresh calibration. I adjusted it to match the 34401A fairly easily. I don't trust any of my equipment to measure better than 10 ppm, so I can't say what it's been doing since then.

The user interface is a bit unintuitive. Without using it often, you'll forget how to set up filtering or other math operations (for example).

I read that some models of the 3456A have a fan while others do not, so they may or may not be noisier.

-Nathan

[kj7e]

Interesting about the fan on the 3456, Ill have to look in to that.

Took about 2 seconds and found some notes on the various series of 3456;
https://www.febo.com/pipermail/volt-nuts/2010-November/000626.html

[guenthert]

Only very early generations of the HP3456A had a fan.  Mine doesn't and it gets only modestly warm, not a big burner.  The HP3456A is very stable and has a very linear ADC, it can measure resistance up to 1GOhm and has a built-in calculator, which can be handy to remove offsets or gain errors, but otherwise offers few functions.  It is calibrated with externally accessible potentiometers, so there's no danger of calibration data loss due to a drained battery.  It's huuuge.  Compared to the case, the LED display is pitifully small, but bright and lasts for decades.  They truly don't make 'em, like they used to.  Watch out for US (60Hz, option -360) vs. European (50Hz, option -350) versions.

The Fluke 8506A with its thermal transfer ACV converter seems more desirable, but tends to cost more.
Both are generally overkill and of interest only to volt nuts. 

[kj7e]

I already have a 34465a and DMM7510 for the voltnutting.  Also have both an HP 3457a and HP 3478a.  Leaning toward the Fluke, not many to pick from on the used market right now, at least not clean with the options I would be looking for.  Where as the 3456a can be found with ease.

[AG7CK]

Get a new table with space for both. Forget all options except GPIB for the Fluke.

I have seen your videos and your posts. You do not need another DMM. Both the HP and the Fluke are voltmeters. Obsolete voltmeters for people with odd interests.

The soul of the 3456A is that nothing compares to the engineering packed around an LM399. And the Fluke uses a refamp (which when pampered enough is good enough to be the reference of 732B).

Get one of each, check that they are good, print out the manuals, and put the meters in range 10V DC never to be touched again.

[just my 10.00000 dollars]

[wn1fju]

One thing that hasn't been mentioned is calibration, if you are so inclined.  I have both the HP 3456A and the Fluke 8506A.  The 3456A is
quite simple to calibrate and involves turning a few pots (accessible from a removable section of the front panel).  The 8506A, however,
was a nightmare, at least for me.  You can do a "hardware" calibration, which involves turning pots, or a "software" calibration which stores
zeros and gains into nonvolatile memory.  The manual is clear as mud on the entire subject.  I ended up trying the software cal, but the
linearity seemed to suffer and was much worse than the 3456A.  So I ended up doing the hardware cal.  This restored the linearity of the
8506A to 3456A levels, and why not - after all, the hardware cal has you adjust a whole bunch of intermediate points in the A/D ladder. 
Anyway, I think I ended up doing both the hardware and software cals, maybe that was a mistake.  As I said, the manual was very confusing
in my opinion. 

Both DMMs are nice units.  But for serviceability, I like the HP better than the Fluke with its series of self-contained modules (for which you
need extenders).

[kj7e]

One thing that hasn't been mentioned is calibration, if you are so inclined.  I have both the HP 3456A and the Fluke 8506A.  The 3456A is
quite simple to calibrate and involves turning a few pots (accessible from a removable section of the front panel).  The 8506A, however,
was a nightmare, at least for me.  You can do a "hardware" calibration, which involves turning pots, or a "software" calibration which stores
zeros and gains into nonvolatile memory.  The manual is clear as mud on the entire subject.  I ended up trying the software cal, but the
linearity seemed to suffer and was much worse than the 3456A.  So I ended up doing the hardware cal.  This restored the linearity of the
8506A to 3456A levels, and why not - after all, the hardware cal has you adjust a whole bunch of intermediate points in the A/D ladder. 
Anyway, I think I ended up doing both the hardware and software cals, maybe that was a mistake.  As I said, the manual was very confusing
in my opinion. 

Both DMMs are nice units.  But for serviceability, I like the HP better than the Fluke with its series of self-contained modules (for which you
need extenders).

That is a great point!

[edavid]

8506A/8505A
I have all available modules (options) for my 8506A
so it has GPIB
it measures voltage DC & AC RMS (thermal conversion for 8506A only) 2w and 4w Ohms and DC Current.

Clarification:  The 8506A has only one slot free, so you can install either the ohms converter or the current shunts, not both at the same time.

My 8505A has a bit of 60-Hz mains hum from its transformer. It's annoying, but doesn't seem to effect the measurements.

My 8506A has this problem also.  It's bad enough that I have to turn it off when I'm not using it.


Also, the 3456A has a small, hard to read display (better than the 3457A though).

[kj7e]

Pulled the trigger on a 3456a, looks clean, appears to work.  The serviceability issue was the kicker.

[AG7CK]

Very good. An amazing box. I bought one (a month ago) with HP calibration and anti-tampering stickers on the front panel and case screws.

I could have bought locally another 34401A just for the shipping cost of the 3456A. But I knew I wanted the 3456A for nostalgic reasons.

Since HP went Agilent around the year 2000, the meter hasn't been adjusted for some 17 years. When measuring references in range 10V DC, it tracks my best other instruments to within a few ppm (typically less then 20uV - no more than 50) from 20 - 30 degC.

An amazing machine. I'll take it to my grave 

[kj7e]

Had a few little issues off the bat with the new 3456a, consent resetting after power on.  Just needed the digital board to be reseated.  Literally blew out the cobwebs and gave it a good cleaning, came out not too bad looking.  Date code of 1984 on most of the parts.  Found the Ohms did not work, no current source, found F201 fuse broke (not blown), Ohms works fine now.  Had a Cal sticker from 2001 (edit: make that 1990) with the seals still in place.  Measures really darn close to my DMM7510 

[AG7CK]

  +  x = 

[AG7CK]

I looked at PhotoBucket, but couldn't find a picture comparing the meters.
If you have the time to spare, please post some data (10v / a few temp points).

I don't see a calibration sticker on any of the two front panel screws. Have you adjusted the meter behind the front panel, or do you suppose it is within spec after 17 years?

[kj7e]

I looked at PhotoBucket, but couldn't find a picture comparing the meters.
If you have the time to spare, please post some data (10v / a few temp points).

I don't see a calibration sticker on any of the two front panel screws. Have you adjusted the meter behind the front panel, or do you suppose it is within spec after 17 years?

I removed all the stickers during the cleanup, here is what It looked like as received (sorry, last cal 1990);




I was rather displeased with the packing, the sender just tossed the meter in an oversized box with white styrophome peanuts, the meter pulverized the peanuts and they where embedded in every single hole.  The first hour I spent cleaning styrophome out of the meter.

No adjustments made yet and it was about 10uV (1ppm) off my DMM7510 at 10v last night!  I left it running overnight, Ill take some comparisons vs my recently calibrated 34465A and DMM7510.

[kj7e]

All at 23 Deg C;
1mV


100mV


1V


10V


After running all night, about 8ppm high compared to the DMM7510. Not bad for not being touched in 27 years.

[guido]

Your 7510 seems a bit off 

As to the 3456A. The power supply caps probably need changing and also you could encounter a bad ROM (or more). One of my devices had that, replaced them with EPROMS.

[kj7e]

There, plenty of room 

[AG7CK]

The meter looks very nice. (So does the bench and the LTZ1000 box in your videos. Wish it was me - my stuff is nothing but unfinished mess ...).

I think you have been lucky with this meter. First - the stickers show that it has not been vandalized. Secondly - the "722" means port 7 HP-IB (GPIB) and address 22. Your meter has probably had its former life (before it went into storage) in a serious test/measurement system controlled by a HP computer with HP-IB controller. Something like this:

http://www.hp9845.net/9845/tutorials/hpib/

I am impressed - 8.4ppm or so on the native 10v ADC in 27 years - that is some 0.31 ppm per year in average. It is a pity they truncated the 7th digit (see the manual p. 8-12 Table 8.5) from the ADC (while they did include a 7th digit from GPIB on the less stable 3457A and 34401A.)

[kj7e]

Only been playing with the 56 for two days now, I'm starting to wonder why more people don't have these?  What a great meter for the price.

[bitseeker]

Shhh. Not too loudly, now. The prices will go up!

[lowimpedance]

There, plenty of room 

Nice bench there btw, reloading press on the right too !. Pistol cal ? judging be the size of the dies.

[kj7e]

Nice bench there btw, reloading press on the right too !. Pistol cal ? judging be the size of the dies.

On the Redding T7 press there, 338 Lapua and 458 SOCOM dies.  I also reload for 45/70 govt, 44mag, 475 Linebaugh, 45 ACP. I have two other presses not in view.

[lowimpedance]

Jeez must get my eye ball re called  . Some decent chamberings there. Fan of the 45/70 too.