Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Seeking advice on a precision DVM
|
Hi All, ? I am wondering what sort of DVM the collective wisdom here might suggest for precision work. ? I am working on a project for work that involves designing a precision voltage controlled current source.? But due to the current pandemic, I am working from home and assembling and testing stuff in my own well equipped basement shop.? ?The current source is more or less a design out of Horowitz and Hill's "The Art of Electronics".? ?I am working to keep the full scale and linearity of this current source to within 0.1% or better.? ? ? My current best DVM is a Keithley model 173 DMM "4 and 1/2" digit, 4-wire meter (the leading digit reads to 3).? It's accurate to ~0.009%+1 digit over 24 hours on the V scales and ~0.035%+1 digit on the Ohms scales.? ?I think it's good enough for what I need, but it has a few limitations, and I am wondering if it is not time to upgrade to a real bench instrument. ? ?First of all, my DVM probably hasn't been calibrated in decades.? ?How do I know it's correct without some externally calibrated transfer standard?? ?It *may* be a possibility for me to use a meter at work (sent regularly to calibration services) and calibrate a portable voltage reference and bring it home.? ?But I don't have a resistance standard.? ? ? The other problem is that even with 4 wires on the lowest Ohms setting, I don't have enough resolution to measure a 0.100 Ohm sense resistor to within 1%.? ? ? ?So, I think I could use a better meter.? ?The HP34401A DVM seems like a nice meter to have.? It's even used as a circuit design example in H&H's text.? But they seem to go for several hundreds of $$ on e-bay.? ?Remember, this is coming out of my hobby budget. ? So, I looked at the? HP3455A / 3456A on e-bay.? These seem to be significantly cheaper, and I love vintage instruments.? BUT, do I want to take up a new project?? How reliable are these old beasts?? Are they easy to repair and calibrate?? ?What tends to go wrong with them? ? ?Even though it's more expensive, one major strong point of the HP34401A is in addition to GPIB, it's got a SERIAL interface on the back!? That is a whole different ballgame in terms of interfacing to a PC or laptop (and a whole lot easier for me, for sure).? But I do think I want to have a meter I can use, rather than one I will have to repair.? ? ? Are there any other models I am missing?? ?Are there any that come standard with an RS232 interface?? Please let me know and thank you in advance for your suggestions. ? Dan ?? |
|
The 3456a is probably the best deal you can find. I haven't ever had a
toggle quoted message
Show quoted text
problem with mine and it's 6.5 digits. It's not really hard to find them under $100 in working condition (I paid $75 for mine and have seen references for as low as $20). The only complaint is it does take up a little space. As for serial data, here's a thread on eevblog about building a serial adapter: Not the world's smarted gpib adapter, but trivial to implement and it gives you the serial data through a USB connection. A proper gpib connector will cost you more than the processor, but the picture in the post shows a cringe worthy budget approach. Calibration of any 6.5 digit meter on a hobbiest budget is a challenge, but if you have access to a recently calibrated 6.5 digit or better meter you can at least compare it and get an idea of how close it is. This is starting to get into volt-nuttery and I'll leave calibration advice to those who know more than me. Suffice it to say I have an ongoing project restoring a Keithley 2001 7.5 digit meter to tell me if my other meters are close to spec (note this one cost me $500 for a broken, maybe repairable example, jury's still out on that). Paul On Thu, Apr 09, 2020 at 05:28:53PM +0000, Daniel Koller via groups.io wrote:
Hi All, --
Paul Amaranth, GCIH | Manchester MI, USA Aurora Group of Michigan, LLC | Security, Systems & Software paul@... | Unix & Windows |
|
I recmmend the HP 3457. Good voltmeter and relatively inexpensive -
toggle quoted message
Show quoted text
less expensive to ship (lighter) - and more reliable than the 3456. Cheers! Quoting Paul Amaranth <paul@...>: The 3456a is probably the best deal you can find. I haven't ever had a |
|
Majority of working 34401A units usually sell for $160-$290 which is really not greatly more expensive than 3457
That's why I decided to keep mine and not sell it! yes, some people also paid >$350 for it but those are rare and they will not be able to sell it at that price. just look at the completed items on ebay... dont pay too much attention on asking prices...but you may need to be a bit patient calibration of 34401A is usually not expensive. Once I got a quote from local tektronix lab here in Toronto and I think it was around $120-$150 Canadian $ and they keep their calibration for really long long time ... I never calibrated mine and it is still excellent except on low range AC current |
|
On 2020-04-09, at 20:24, Bruce <bruce@...> wrote:
Yes. Needs good front lighting though, as the display is an unilluminated LCD. But then, it doesn¡¯t degrade like the VFDs on newer instruments. (This is all mostly moot if you collect the measurements on a PC via HPIB.) As long as you have reasonable lighting and no vision problem, the 3457A is a slam dunk. (Like all meters, it does have some weak points, but it is exceptionally long-term stable; a 20-year-old calibration may still be < 10 ppm off.) One more problem with this vintage of HP equipment: The Schaffner IEC inlet with the famous RIFA madness (actually WIMA in this case). Some assembly required if you don¡¯t like things in your lab to explode at night. (Explosion more likely in 230-V-land.) This is true of a lot of instruments from the late 1980s. 3456A is a bit on the upward curve of the reliability bathtub by now; if you have 20 meters, it makes a good 21st, but I wouldn¡¯t use it as one out of two. I love the LED display on it; it does drift about way more than the 3457A though. (I don¡¯t have experience with the 3455A, but that seems to have left the bathtub already.) If you don¡¯t need 3000000 counts of the 3457A, the 3478A also is a nice meter. Outside the HP range, but cheap and 6 digits+, look for Solartron 7150 (incidentally, same Schaffner issue) or Philips PM2534/2535 (now, that display is actually a bad example for LCD). Gr¨¹?e, Carsten |
|
The?HP34401A DVM is a very good choice, and I have see them go for very reasonable prices. I bought three of them locally a couple of years ago for a low price. They are very stable and small. The don't contain much magic, so they still can be serviced. Documentation is excellent.
I have a very stable 10.000000 volt test source, that I bought new on eBay and had rechecked after a year. I also have a Fluke AC calibrator and a couple of DC calibrators, so I can test with good confidence. My reality check is a pair of Fluke 8506A DMMs. These can often be found at very low prices, along with their near twin, the Fluke 8505A. They however are much larger meters. Stuart K6YAZ Los Angeles |
¿ªÔÆÌåÓýBut you can buy 34401A displays fairly readily and for not too much cost, whereas the Keithley ones are like hens teeth and very many show which segments have been lit longer than others.I have a 3456A, 3465A, 34401A, Datron 1271 and a Keithley 2000 and have to say I prefer the Datron and Keithley displays the best - apart from the 3465A (lower class of DMM) there isn¡¯t much to choose from the others for what I need. From: [email protected] <[email protected]> on behalf of Mark Bielman <mbielman@...>
Sent: 09 April 2020 19:26:59 To: [email protected] <[email protected]> Subject: Re: [HP-Agilent-Keysight-equipment] Seeking advice on a precision DVM ?
I'm not a big fan of the LCD's HP used for a while. Hard to read at times.
The vacuum florescent displays are nice until they die. I prefer LEDs. The 34401A looks nice. Beware of failing displays! Mark |
|
I'd vote for the HP3455A and HP3456A. I have about three of each, and I did have to repair some of them initially (which is why they were quite cheap) - mostly the usual bad power supply cap replacement. I prefer the 3455A user interface, even though the 3456A is supposedly a little newer and better.
If you're moving from 4.5 to 6.5 digits, it will be like a breath of fresh air. You? may be able to just not worry about calibration yet, since even if old, they will likely still be much better than your old gear. During a design, you can probably get by with less than ideal equipment, and no standards to compare. For the time being, it appears you mostly need more resolution, not necessarily absolute accuracy. As you get to the end, then you can worry about the exactness. BTW I have one old DMM that's very cool, I think, because of its exceptional resistance measurement capability. It's a DANA 5900, which at first glance is just another old 1970s lab grade DMM, with 5 digits plus 60% over-range. What's cool is that it utilizes those digits well - it includes 100 meg and 10 ohm FS ranges, which are outside those of typical DMMs. Indeed, they're out of its own auto-ranging capability, and have to be selected manually. At the high end, it tops out at 160 megs with 1 k ohm resolution. At the bottom, it can read to 16 ohms with 100 micro-ohm resolution. I picked this up years ago at a flea market for somewhere around free to five bucks, as I recall, and it worked when I checked it out. I was considering junking it out, since it's old and beat up, and I have plenty of DMMs, until I discovered the ohms ranges. I saved it set it aside just for high megohm use. I recently had need to use it, and found it had crapped out. I've been working on it on and off over the last couple of months, trying to save it. It's almost OK now. If not for those ohms, it would be long gone. This goes to show that even if you have good, working stuff, it can still let you down, resulting in repair or replacement projects. This will of course tend happen when you need it the most. Anyway, I'd recommend not looking for "a" new DMM, but "some" new DMMs, if it's for a mission-critical need. I keep the HPs on the shelf in reserve or for high grade measurements, my bench DMMs are old Fluke 8400As (5+ digits), but mostly I use my pile of Harbor Freight freebie DMMs. Ed |
¿ªÔÆÌåÓýFor precision nothing beats a 3458A, 8.5 digit, but it¡¯s expensive. The 3456 is the better of the 3455/3456 pair. ?Our standard bench meter is a 3478A. Should be relatively cheap and they seem to be accurate and hold cal well. Downside is a 300V max input.?My 2 cents worth. Regards, ? Stephen Hanselman Datagate Systems, LLC On Apr 9, 2020, at 10:29, Daniel Koller via groups.io <kaboomdk@...> wrote:
|
|
Hi all
this is one of my favorite topics (i must own over 100 different models of digital multimeter) so it is hard for me to resist adding my 2 cents! I think the immediate requirement of higher accuracy measurements of resistance in the range of 0.1 ohm is a good place to start. the meters that the other folks have recommended are all pretty good candidates, the practical differences for the intended use really boil down most to things like display type, physical size, and expected reliability and stability. i would like to add one candidate (which did get a passing mention) the Keithley 2000 and its variants, especially the 2700 ¡°data acquisition unit¡± (which is more or less a model 2000 with additional multichannel data and signal input circuits) the reason this meter is noteworthy is that they are selling, in working condition, in the same range as the gold standard hp 34401a, around $300. the 2700 has very similar accuracy, slightly better drift specs (and in my experience markedly better drift performance) and a 1 ohm resistance range! The 34401a lowest range is 100 ohms, and the 3457 30 ohms. Of course with accuracies on the order of 40-80 ppm (90 day specs) the useful resolution for a 0.1 ohm measurement is 4 digits. All of these meters can do that. And actually price/performance might lead you to the hp 3457, they are ridiculously cheap. But they are full rack size units, like the 3456 also. And bench space is important, to me! so I recommend a Keithley 2000 or 2700 or one of their other ¡°integra series¡± 6-1/2 digit models.? |
|
Opps, I have to correct my statement about the lowest resistance range for the Keithley 2000 and the integra 2700, 2701 models. The only model in that line providing 1 (and 10) ohm ranges is the 2750, which is not as common on the auction site and probably not priced competitively (it is also a full size form factor).
But even with 100 Ohms as the lowest range, I highly recommend the 2000 and it's sisters. Just be careful that you don't buy one with a dim display! |
|
Ha!? I love my freebie Harbor Freight DMM!? Nice feature is the On/Off switch; why do I have to turn the darn dial back to Off every time, when 99 times out of 100 I'm just going back to Ohms anyway?!? Some of my other DMMs don't have that nice feature. I'm considering getting a couple of 34401A's since I've used them before at work, they are nice instruments, and they are all over the place.? I remember using the high impedance ohms mode when working down in the nanoamp range and getting weird results with a "normal" handheld DMM.? It didn't take me long to figure out the bias and offset currents of the normal DMM were higher than what I was trying to measure. I figure one 34401A for a dedicated connection to my 432A microwave power meter bridge circuit, GPIB to the computer, and let it calculate the RF power.? Another 34401A for general bench use.? Last time I checked the 'bay, there were seventy 34401A's for sale! Jim Ford
------ Original Message ------
From: "Ed Breya via groups.io" <edbreya@...>
Sent: 4/9/2020 3:24:06 PM
Subject: Re: [HP-Agilent-Keysight-equipment] Seeking advice on a precision DVM I'd vote for the HP3455A and HP3456A. I have about three of each, and I did have to repair some of them initially (which is why they were quite cheap) - mostly the usual bad power supply cap replacement. I prefer the 3455A user interface, even though the 3456A is supposedly a little newer and better. |
|
For a great many measurements that most of us make, I suspect the the "free" Harbor Freight Tools DMM serves the purpose just fine. It's great for quick checks and especially continuity testing.?
You probably don't need to fire up your 6 1/2 digit "wonder meter" to measure a nine volt battery or to confirm that a five volt bus in an instrument is still working. Each instrument has it's place. Stuart K6YAZ Los Angeles, USA? |
|
I have 2 HP3456's, picked up in the San Diego area for $20 and for $50. I like the fact that it is calibrated with multi-turn pots accessible from the front panel, rather than calibration constants stored in a battery backed up memory. Long out of cal and my DIY LM399 10 Volt source also long out, but checked recently against an average of 3 out of cal meters, so I'm eagerly waiting my turn for access the the EEV floating cal standard.
toggle quoted message
Show quoted text
<> Hope COVID19 doesn't break the chain. I've bought a pair of Aneng AN870 4 1/2 handheld multimeters and found them to be within spec, much better than a Harbor Freight meter. David, contact me privately and perhaps I can get some 0.01% resistors to you to check your ohmmeter. KK6IL@... John On 4/9/2020 10:28 AM, Daniel Koller via groups.io wrote:
Hi All, |
|
On 2020-04-10, at 22:05, victor smith <vcs3@...> wrote:
(Full disclosure: I just got three more :-) But they are full rack size units, like the 3456 also. And bench space is important, to me!Well, the 3457A are rack width (425mm (16.75 ¡°)), but only 292mm (11.5 ¡°) deep (compare to 3456A at 527.1 mm (20? triple-barleycorns)). Compared to a 3456A (10.49 kg (23.13 lbs)), they are also very light (5.05 kg (11.1 lbs)), so they can be lugged easily to the place they are needed. The robust plastic case and the low weight also mean that it¡¯s a bit harder to damage them while lugging them around. Disadvantage: the 3457A input resistance is 10G¦¸ only up to 3 volts, while the 3456a can do 12 volts at that input resistance. But then, in Ohms, 10 ?¦¸ resolution, where the 3456A is 100 ?¦¸. The 3457A can output up to 12 V in Ohms ranges, enough to fry many MOSFETs (so hello Harbor Freight(*) for that!), while the 3456A is slightly friendlier to your circuits at 9.5 V max. And so on. And don¡¯t forget the 3457A are 3030000 counts [6.48 digits], so they have more than one bit more of resolution than the 1200000 counts [6.08 digits] of the 3456a. I still love the 3456A display, though¡ Only the Keithley 192/196 are even more beautiful. [Well, OK, Datron 10xx can¡¯t be beat, but those are not LED.] Gr¨¹?e, Carsten (*) OK, ANENG. And those are incredibly accurate for what they are¡ |
to navigate to use esc to dismiss