Keysight DMM - Yay or nay?

[Traceless]

Currently I looking for a better multimeter. A hot candidate was the Keysight 34465A, specs-wise my only complaint is the meager capacitance range. However after reading this I have second thoughts about going with Keysight. Unfortunately there don't seem to be too many good alternatives at that price point. The Keithley DMM 6500 looks nice but the touch screen is a deal-breaker. The Keythley 2000 is missing capacitance mode entirely, which is also a deal-breaker. The Rohde&Schwarz HMC8012 seems a bit outdated and overall too expensive for its specs. I also consider the Siglent SDM 3065X, it has a much better capacitance range. Otherwise it can't quite keep up with the Keysight 34465A but it is also considerably cheaper.

Is there any other contender I should consider? Are Keysight 34465As repair friendly or does one have to rely on the virtually non-existent support? Does the Keysight 34465A or the SiglentSDM 3065X have any quirks I should know about before making a choice?

[tautech]

SiglentSDM 3065X have any quirks I should know about before making a choice?

Mainly to get the latest firmware into it ASAP to fix the boot freeze the odd units have.

[Kleinstein]

The capacitance measurement on the DMMs is usually not really competative to a dedicated RCL instrument. This starts with an often not so well defined frequency and bias. With electrolytic caps 1% accuracy only makes sense if you also know the frequency.

The KS 34465 uses digital RMS and this way have a much faster response in AC mode than old style analog RMS chips. The stability can also be quite good.
The SDM3065 has a 20 V range with high impedance, while the KS meter only some 10(12?) V. This can make a difference.

Modern DMMs are rather reliable and a repair is not very likely. If a repair is needed it gets complicated, as for the newer meters there are little repair instructions or schematics provided. Custom parts, especially resistor arrays and special relays can limit what can actually be repaired.

[Cerebus]

Don't be distracted by capacitance measurement on DMMs, it's a gimmick that's there just to affect decisions like the one you're trying to make and tip your final decision between near rivals. Forget about choosing a DMM on its capacitance measuring and treat it as a bonus if and only if all the other measurement capabilities meet your requirements.

If you're lucky the capacitance range on a general purpose DMM will give you a vaguely accurate capacitance measurement by feeding the capacitor from a constant current source and timing the resultant slope between two voltages, it won't measure the voltage, current and phase produced by passing a sine wave at defined frequencies through the capacitor the way a proper LCR bridge will. It won't give you anything other than the raw capacity either, you won't get ESR, Q, \$tan \delta\$ or any other such parameters. A £100 LCR meter will give you better and probably more accurate measurements.

[bob91343]

I agree.  Don't try to measure capacitance with a DMM.  I can measure capacitance with any of a number of instruments.  I have two bridges, a VNA, a multi component Chinese tester, and a DMM with capacitance ranges.  Plus a couple of Q meters and an RX meter.

The bridges win, hands down.  The VNA does well at higher frequencies.  The Q and RX meters are great but they are boat anchors and a nuisance to drag out to use.  The DMM sits in the closet in a box.  The Chinese tester is okay but I always verify its reading with another instrument.

If I really care about a measurement I will always use a bridge.

[Traceless]

Thanks to everyone for the feedback so far.

@Kleinstein, @Cerebus, @bob91343: I do know that DMM capacitance mode is not the holy grail, for me it is merely a convenience function. For instance when I do replace caps I like to do a quick and dirty check on the replacement cap just to see if capacitance is in the right ball-park. I already own a proper LCR-Meter (Matrix MCR-5200), but for simple stuff like that I tend to just use the DMM and don't bother to grab the LCR-leads out of the drawer and hook them up.

[nctnico]

Don't be distracted by capacitance measurement on DMMs, it's a gimmick that's there just to affect decisions like the one you're trying to make and tip your final decision between near rivals. Forget about choosing a DMM on its capacitance measuring and treat it as a bonus if and only if all the other measurement capabilities meet your requirements.

If you're lucky the capacitance range on a general purpose DMM will give you a vaguely accurate capacitance measurement by feeding the capacitor from a constant current source and timing the resultant slope between two voltages, it won't measure the voltage, current and phase produced by passing a sine wave at defined frequencies through the capacitor the way a proper LCR bridge will. It won't give you anything other than the raw capacity either, you won't get ESR, Q, \$tan \delta\$ or any other such parameters.

But still it is very handy to have. I use the capacitance measurement function on my bench DMM (in most cases not a Keysight though) regulary to check the value of a capacitor quickly so see what value I mounted on a board while tweaking a circuit. This works from a couple of tens of pico-farads since capacitors commonly used are mostly in the E6 values range.

@the OP:
Choice of DMM between the Keysight 34465 and Keithley (Tektronix) DMM6500 is a though one. A couple of years ago I bought the 34461 but back then the DMM6500 didn't exist yet. Nowadays I might have choosen the DMM6500. Don't get too winded up about a touch-screen; when the interface is designed around a touchscreen from the ground up, then it works way better than buttons. What is important IMHO where it comes to precission instruments is pedigree; knowing how to build a proper 6.5 digit DMM. Both Keysight and Keithley have a long history with developing & producing higher end DMMs.

Forget about repairing though but then again test equipment has a lot less parts then in the old days so there is a lot less than can get broken. If you are worried about repairs costing a fortune, you can get an extended warranty contract.

[Traceless]

Choice of DMM between the Keysight 34465 and Keithley (Tektronix) DMM6500 is a though one.

I already pretty much ruled out the DMM 6500, the meter is probably good but I very often wear work gloves like this one which might be a problem with a touch screen plus I really prefer buttons over touch interfaces.

Forget about repairing though but then again test equipment has a lot less parts then in the old days so there is a lot less than can get broken. If you are worried about repairs costing a fortune, you can get an extended warranty contract.

That option would indeed be interesting but from what I understand Keysight does no longer support non-business customers. Do you know a way to get such a contract as hobbyist?

[tautech]

Thanks to everyone for the feedback so far.

@Kleinstein, @Cerebus, @bob91343: I do know that DMM capacitance mode is not the holy grail, for me it is merely a convenience function. For instance when I do replace caps I like to do a quick and dirty check on the replacement cap just to see if capacitance is in the right ball-park. I already own a proper LCR-Meter (Matrix MCR-5200), but for simple stuff like that I tend to just use the DMM and don't bother to grab the LCR-leads out of the drawer and hook them up.

Generally that approach works but not always as you can have a cap measure within 10% of its nominal value however its ESR can be miles off.
I used a borrowed bridge for a few years but the pissing around nulling it drove me nuts and a dedicated LCR meter I wasn't keen on and as I was getting into SMD a decent set of SMD tweezers were acquired instead which are much faster and more convenient to use plus it can quickly measure all those unmarked SMD caps. Still, I probably use it more on TH components than SMD.

[nctnico]

Choice of DMM between the Keysight 34465 and Keithley (Tektronix) DMM6500 is a though one.

I already pretty much ruled out the DMM 6500, the meter is probably good but I very often wear work gloves like this one which might be a problem with a touch screen plus I really prefer buttons over touch interfaces.

Forget about repairing though but then again test equipment has a lot less parts then in the old days so there is a lot less than can get broken. If you are worried about repairs costing a fortune, you can get an extended warranty contract.

That option would indeed be interesting but from what I understand Keysight does no longer support non-business customers. Do you know a way to get such a contract as hobbyist?

Buy it through a distributor. Just give a reputable one a call and ask.

[thm_w]

I already pretty much ruled out the DMM 6500, the meter is probably good but I very often wear work gloves like this one which might be a problem with a touch screen plus I really prefer buttons over touch interfaces.

Marco reps is that you? 
Touchscreen gloves are a thing as well, or stylus: https://www.uvex-safety.com/en/product-group/uvex-phynomic-airlite-touchscreen-safety-gloves/
But yeah if you want a button instead I get it.

[Zucca]

34465As repair friendly

Not really, no schematics. It depends on your brain and luck. That said if you post in the repair section well there is an EE army ready to help.
For free. KS pissed me off too with the hobbist thing, but sometimes a good community can compensate.

[mawyatt]

We have and use all three DMMs, 2 KS34465A, DMM6500 and SDM3065X.

The SDM3065X is the weakest of the 3, it doesn't have many of the features of the others and doesn't seem in the same league, something mentioned about the long history of KS (heritage with the HP34401A) and Keithley with DMMs seems to be present in their usage. However the SDM3065X also is ~1/2 the cost, so from a value standpoint it's quite good.

The KS34465A just does everything well, and is very easy to use and quickly get to what's important.....accurate measurements!! Being older still prefer the buttons to menus of the DMM6500.

The DMM6500 excels in the graphics area, almost like having a low sweep rate scope. Also has the lower ohms ranges, but doesn't have the lower current range of the KS. When long term measurements are required we turn to the DMM6500.

Both the KS and DMM6500 have switchable rear inputs, we've found these handy and hang thermistors from and record local ambient temperature without having to remove leads from the front inputs. The KS uses a pre-aged/selected LM399 reference, believe the DMM also uses a pre-aged reference but don't think the SDM3065X uses a pre-aged LM399.

We purchased the first KS34465A a couple years ago, then the SDM3065X, then the DMM6500 about 6 months ago. When we needed another quality DMM a month ago, it was between the KS and DMM, and finally decided on the KS.

Edit: The cap measurements are OK, more accurate on the KS and DMM. However, consider a dedicated LCR meter like the DE-5000, much more accurate, versatile and useful in every respect.

Best,

[Domitronic]

There is also a bug in the resistance measurement function of the DMM6500 that might be considered when choosing a multimeter. The DMM6500 is not able to measure resistance of coils and transformers in certain measurement ranges properly. In my opinion this is surprising and somehow embarassing for Keithley. But since it is acutally a hardware issue i guess it can't be fixed in software. As far as i remember the current source for measurement starts to oscillate when the multimeter is in 100 Ohm range and an inductive load is connected. This also screws up the autorange function. There is a thread about the issue in this forum.

[Traceless]

Generally that approach works but not always as you can have a cap measure within 10% of its nominal value however its ESR can be miles off.

You're right, I know. I like to give "new" replacement caps the benefit of a doubt, based on the (probably inaccurate) assumption that if they were stored too long/under wrong conditions or have a manufacturing defect I'll see a significate deviation from the nominal capacitance. For diagnosis on old/used stuff I always double-check with my LCR-meter.

Buy it through a distributor. Just give a reputable one a call and ask.

I'll give that a try.

Marco reps is that you? 

*grin* I wish I were, I don't think Marco would bother with buying "petty little 6.5 digit DMMs" . Among others he owns an HP3458A, a Solartron Schlumberger 7081, a Keithley 2002 and an Advantest R6581T.

Not really, no schematics. Sometimes a good community can compensate.

Good point the community here is awesome. Still I dislike it when vendors become this arrogant in which case I like to vote with the wallet. In this particular case there don't seem to be many alternatives though.

The SDM3065X is the weakest of the 3 ... However the SDM3065X also is ~1/2 the cost, so from a value standpoint it's quite good.

Yes that was my feeling as well. Currently the SDM seems to be good enough, and delivers the best bang for the buck but that's what I thought about my current DMM as well and now I'm occasionally frustrated with it (reason explained below). So I'm trying to avoid making the same mistake twice.

Edit: The cap measurements are OK, more accurate on the KS and DMM. However, consider a dedicated LCR meter like the DE-5000, much more accurate, versatile and useful in every respect.

Yeah as mentioned, I regard the DMM-capacitance mode as a convenience feature not as a precision-thing. I already own dedicated LCR-meter, the Matrix MCR 5200.

@Domitronic: Thanks, that is a very interesting point. Speaking of "screwing up auto range". My number one complaint with my current DMM is that auto range is excrutiatingly slow, also AC current measurements tend to have a long delay to the point where I'm now convinced that my DMM is actively trying to kill me (sometimes it does not pick up mains voltage in a reasonable time frame). That's why I nicknamed it Christine.  Can I assume that, aforementioned bug aside, KS, Keithley and Siglent don't suffer from that problem?

[Kleinstein]

Analog true RMS is naturally slow. To get good accuracy at 20 Hz it needs a rather slow settling filter and this gives a naturally somewhat sluggish reaction. Some chips are worse than others in this respect (worst case things get even slower at low votlage), but even the good ones are slow. One should still see a first reaction reasonably fast, but the final accuracy may well need 5 seconds of waiting.
If implemented poorly AC autoranging could suffer quite a bit, though there are ways around this.

The DMM6500 and KS346x meters use digital RMS and can react quite fast in AC mode (stable final reading after some 200 ms - don't know the details on these meters). One can already see the difference with handheld meters - some of the newer, more low cost true RMS meters use digital RMS and offer surprisingly fast AC settling.
The SDM3065 uses a classical analog RMS chip.

At a  slightly higher price point there is also the Fluke 8845 DMM as a possible candidate - not my favorite, but just to add to the list.

Occasional failing AC is a potential risk. It could be from a relay problem - this is indeed bad and a good reason not to trust any such meter to check if there is no mains. At least German regulations wants you to use a dedicated tool for this and not a normal DMM, let alone a bench DMM.

[ch_scr]

My father has the HMC8012 as his bench DMM basically since it came out. It is not a bad little unit, but it came out around the time of the Hameg merger and it feels like they forgot to continue FW development. The last firmware is from 2017. And not because the FW on the unit is perfect, it's not Rigol-level-buggy, don't get me wrong, but the control concept feels off and is not totally consistent, some buttons seem almost unused on the UI. And a modern DMM with color LCD but without plotting? Seriously? There are basic statistics but no histogram or tracking view - soo much lost potential there alone! The interleaved voltage / current measurement without relay switching is a very nice feature for a general bench DMM, much appreciated, as is the fast boot time. It is not a bad instrument, works well in general, but it is really let down by what feels like abandoned development.

[Traceless]

But the final accuracy may well need 5 seconds of waiting.
...
Occasional failing AC is a potential risk. It could be from a relay problem - this is indeed bad and a good reason not to trust any such meter to check if there is no mains.

I don't think there is a defect with my DMM. I think what is going on is this:

When I take an AC measurement the DMM takes time - 5s sounds about right. The catch is: My DMM seems to display nothing until it has a stable reading (-> 5s delay plus the autoranging delay that varies depending on how many ranges it has to cycle through). Now when there is bad contact, due to oxidized or dirty joints or just because the probes are sliding a bit when applying pressure and contact is lost for a fraction of a second the measurement resets and the DMM starts over trying to get a stable value. So basically AC measurements are not constant time but vary depending on the current circumstance. So you can never be sure if you waited long enough. I ordered some new super sharp probes to see if that helps with the issue.

At least German regulations wants you to use a dedicated tool for this and not a normal DMM, let alone a bench DMM.

I had replaced the fuse in a 24V SMPs adapter and just wanted to check if I had mains at the input filter after replacing the fuse. I didn't get a reading in time on my bench DMM. All plugs were connected and my isolation variac was turned on, also I checked the fuse. Knowing about the delays I got suspicious grabed my Aneng 8009 and voila mains present. The Aneng is just a lot faster than the bench DMM. Workarounds are the following:

- Patience (and probably better probes)
- Select range manually
- Increase the trigger speed

The problem here is that the meter does not save the trigger setting when turned off, also the least significant digits become very jittery when increasing the trigger speed. Ultimately that means that I need to keep manually switching between low and high trigger speed, depending on if I want to measure presence of mains or do high precision measurements on the secondary side. All of this is doable and otherwise the meter is really good but those constant user interactions, switching between low and fast trigger and manually changing the range get really annoying. Luckily that only is an issue when I work on PSUs in other cases the meter does a very good job.

At a  slightly higher price point there is also the Fluke 8845 DMM as a possible candidate - not my favorite, but just to add to the list.

Good tip I added the Fluke to the list, though feature-wise the even more expensive 8846A is probably closer to the 34465A.

@ch_scr: Thanks for reporting these experiences. That confirms my suspicion that the HMC8012 does not offer a good value (any more) for its price.

[Markus2801A]

Hello

I just want to bring in the GW-Instek GDM-906x Series with two types namely GDM- GDM-9060 and the  GDM-9061

https://www.gwinstek.com/en-global/products/detail/GDM-906x

What do you guys think of them?

Kind regards!
Markus from Austria

[mawyatt]

We also discovered some charge injection on the DMM inputs, and the DMM6500 seemed to have slightly more.


https://www.eevblog.com/forum/testgear/dmm6500-glitch-on-dcv/msg3753929/#msg3753929

Best,

[Cerebus]

@Domitronic: Thanks, that is a very interesting point. Speaking of "screwing up auto range". My number one complaint with my current DMM is that auto range is excrutiatingly slow, also AC current measurements tend to have a long delay to the point where I'm now convinced that my DMM is actively trying to kill me (sometimes it does not pick up mains voltage in a reasonable time frame). That's why I nicknamed it Christine.  Can I assume that, aforementioned bug aside, KS, Keithley and Siglent don't suffer from that problem?

Cue George Thorogood and the Destroyers:

Da da da da da dum de-dum de-dum
Da da da da da dum de-dum de-dum

Bad to the bone
Da da da da dum
Bad to the bone
Da da da da dum
B-b-b-b-b-bad
Da da da da dum
B-b-b-b-b-bad

[Traceless]

@cerebus: Yep, pretty sure they played that song during the assembly of my DMM as well

[Traceless]

I just want to bring in the GW-Instek GDM-906x Series with two types namely GDM- GDM-9060 and the  GDM-9061
What do you guys think of them?

Hey Markus, thanks for bringing these meters to my attention. I took a quick look at the GDM-9061's specs and it seems that the Siglent SDM 3065X is more accurate, also the Siglent has 2200000 counts vs the 1200000 counts of the GWInstek plus the GDM-906x models seem to be more expensive. Unless the GWInstek has some trick up its sleeve that I missed while skimming over the specs the SDM 3065X would be the better option.

[blackdog]

Hi,

I have 2x a 34461A 1x 34465a and a Keithley 6500.
The reason I bought a Keithley 6500 in December 2021 is because I wanted to try how another brand does in terms of 10V stability.

My 34465a was a big disappointment, the Autocal is Markerting Bullshit. my instrument has been away from home for 6 months to Malaysia for verification of the problem that when you do an autocal it drifts out again within 20 minuter if the temperature in the measuring instrument remains stable.

The service here in Holland from KEySight was excellent, I got a loaner instrument and this loaner instrument had exactly the same problem.

Keysight Maleysia told me that the drift that occurred remained within the daily average and therefore they can't do anything about it...
Don't buy the 34465a for the autocal! but the instrument has more useful features that might make it worth buying this type anyway, like the ezxxtra filtering that is possible, just to name one aextra feature.

My Keitley DMM6500 sat now for almost two months and the deviation on the 10V range stays within +-1PPM at varying lab temperatures, this is very good, and better than my three Keysight 3446x series meters.

As far as the interface is concerned, at the moment my preference is slightly with the KeySight.
But I haven't had time enough to do e.g. scripting on the Keithley meter.

So you will have to make your own trade-offs as to what is best for you.
Also see if temperature measurements are important to you so you can connect accurate sensors to the DMM you choose.

Kind regards,
Bram

[Kleinstein]

The GDM906x meters seem to have rear terminals, that I don't see with the sigilent.
At least the GDM9060 seems to use a comparable SD ADC chip to the Sigilent.

There is also a Teledyn (T3DMM6-5), that seems to be essetially the same as the Sigilent - possibly an OEM build.