Agilent/HP 34401a Question

[sonicj]

i have the opportunity to pick up a used HP 34401a for about $350usd. i was considering the BK Precision toolbox, but i already have a Fluke 12 and I think a nice bench meter might be more useful. any thoughts? any specific questions i should ask the seller? apprec!
-sj

[mikeselectricstuff]

That's a very good price for a 34401a - it's a good high-precision meter, and also has RS232 which can be handy for datalogging - Agilent have an Excel plug-in that allows logging directly into a spreadsheet.
4-wire ohms can be useful for tracing PCB shorts as it has 0.1mR resolution.
I think I have the service manual for it if anyone's interested.

[saturation]

All the HP/Agilent's I know, particularly the HP branded ones, have a power on self-test function POST, or a full separate self test.  This is one good reason HPs are good on the second hand/eBay market, because there is a simple way to find out all is well, for the most part.  Depending on age of your HP device, a POST can test 75-100% of a device's functionality.

For the 34401a, on power on, it does an essential self test, not the complete one, which takes 15-30 seconds more.

To perform the complete front-panel self-test is as follows:
Hold down Shift as you press the Power switch to turn on the
multimeter; hold down the Shift for more than 5 seconds. The selftest
will begin when you release the key.

It will then report full status.

This meter is an industry "standard", and it has a lot of 3rd party software and accessories.  But it is obsoleted by its newer models, all are still sold by Agilent.  I think its the 34410a and 34411a, they are priced identical if bought new: you get more bandwidth, functions, and higher sampling speed in the new models, but the 01a is well entrenched since the new models haven't been battled tested in real world use as much as the 01a.

Is it worth $350, yes.  That's the going rate for a fully functional used one, I've been looking for one myself.

Its main competitor is the older bench model HP 3456a, which has a slight edge in absolute accuracy in Vdc, however it has no amp function which the 01a has.  It typically goes between $50-100, if working status is unknown, and $100-300, if working.  I bought 3 3456a, and planning to buy a few more.  The reason for having multiple is to calibrate them against each other, use one as transfer reference, and backup each other since they are 20-30 years old, and I can find them dirt cheap over the 01a.  Lastly, older bench DMM are easiler to self service in terms of parts, whereas the 01a has many custom ICs that can only be repaired by buying another 01a and strip it for parts.

[sonicj]

great info guys!

thats incredible that it can be used to trace pcb shorts!

hmm, the HP 3456a looks nice as well! another dude in my area has a HP 3466A and a HP 3467A along with some other meters for sale on the cheap. no 3456a's though...

the self diagnostic feature sounds handy. calibration was one of my main concerns as that seems expensive to have done. do they ship with leads included? any important questions i should ask the guy prior to purchase? cheers!
-sj

[alm]

hmm, the HP 3456a looks nice as well! another dude in my area has a HP 3466A and a HP 3467A along with some other meters for sale on the cheap. no 3456a's though...

3466/3467 are in a lower class than the 3456, probably 4.5 digits or so, no communication interface. The 3456A is a nice piece of equipment, but a lot larger and heavier than the 34401A, and clearly older. It will only do GPIB, no RS-232 (lat alone USB/LAN), and no features like current or frequency. Digital features like stats will probably also be less advanced, although I haven't compared them side-by-side. It was in a higher class in its time (like the 3458A today).

the self diagnostic feature sounds handy. calibration was one of my main concerns as that seems expensive to have done.

Self diagnostic checks if it works, not if it's within spec. Calibration is a different topic, and pretty hard to verify with these kind of tolerances. A real calibration at a reputable lab (which IMO doesn't include most Ebay sellers) is indeed expensive. But I've not yet encountered a high-quality meter that was fully working but grossly out of cal, they're either within spec or pretty close, or have a defect. Drift tends to stabilize with time, so ten year spec is usually much better than ten times one years spec. And for something like a 34401A, even ten times the one year spec would be pretty good for a hobbyist.

do they ship with leads included? any important questions i should ask the guy prior to purchase? cheers!

Used meters don't usually ship with any accessories, and sometimes even lack standard parts like fuse holders, handle or bumpers. I would ask the usual stuff: history, condition, calibration, good pictures (to judge physical abuse and missing parts), if you'll be able to return it. I believe $350 is close to the usual Ebay price, it's not exceptionally cheap, so I wouldn't buy it 'as-is'. There are various good older bench meters that are usually cheaper on the used market, I've mentioned them before on this forum.

[saturation]

Agree fully with alm's assessment. The 3456a was the top of the line DVM, digital voltmeter, not a multimeter.  It can only do Vac, Vdc, and resistance but can read to 10uV, and typically accurate at 100nV level.  Its mostly for lab style work where accuracy is more important than convenience and speed.

Its very difficult to calibrate meters at this level, to do so cost as much as one of these 3456a itself, but I'm currently on the volt-nuts forum, and its clear you can make a DIY transfer reference and maintain accuracy to 1-10uV by controlling the environmental conditions a cheapo voltage reference, like the Geller, is exposed too.

Two of the 3456a I bought were last calibrated 20 years ago, and one was calibrated 4 months ago.  As alm says, the 20 year old one's are still in spec at Vdc compared to the newly calibrated model, that's comparing the Vdc range from 10mVdc to 600Vdc.

[Ronnie]

I think a nice bench meter might be more useful. any thoughts?

You can also consider Agilent U3402A 5 1/2 digit dual display for US$600 (brand new). Highest sensitivity is 1µV at 120mV range and up to 0.012% basic DC accuracy but there is no communication ports like GPIB, RS-232, or USB for Agilent low cost multimeter U3400 series including model U3401A 4 1/2 digit dual display.   
http://www.alliedelec.com/search/productdetail.aspx?SKU=8350456