Siglent SDM3065X

[eurofox]

Hi guy's,

I would like to buy a 6 1/2 digit multimeter with trending capability without breaking the bank 

Of course there is Keysight, Keithley, PeakTech and Siglent....maybe some other. 

When I look the the datasheet the Siglent SDM3065X seems to be a winner when I look to the price/performance.

Service from Siglent is maybe not yet on the level of Keysight.

Thank you for sharing your experience if you have such a multimeter. 

[tautech]

Thread about all the SDM3k models:
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/

I've got one with the 16ch logging card if there's anything you want me to check for you.
I sold another one to member Defpom and he uses it for his repairs on his YouTube channel....easy to find him if you look.

[eurofox]

Thread about all the SDM3k models:
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/

I've got one with the 16ch logging card if there's anything you want me to check for you.
I sold another one to member Defpom and he uses it for his repairs on his YouTube channel....easy to find him if you look.

I check already the Defpom 2hour video but it does not show any test about accuracy, is more relate to use the the multimeter.

Nobody connect it to a calibrator ar compare to a Keysight 34465A?

[tautech]

Thread about all the SDM3k models:
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/

I've got one with the 16ch logging card if there's anything you want me to check for you.
I sold another one to member Defpom and he uses it for his repairs on his YouTube channel....easy to find him if you look.

I check already the Defpom 2hour video but it does not show any test about accuracy, is more relate to use the the multimeter.

Nobody connect it to a calibrator ar compare to a Keysight 34465A?

You only need compare the datasheets, on my one the Cal cert specifies it was calibrated with a Fluke 5522A which was 3 months inside its Cal period so should ensure the specified datasheet accuracy. There's also a specified 1 year accuracy listed.

Member kj7e had one but I think he updated it for 7.5 digits or more.
He had a rough start with his and got another one from Siglent USA.
His first impressions are here:
https://www.eevblog.com/forum/testgear/siglent-new-bench-dmm-sdm3055/msg824341/#msg824341

[nctnico]

The cal sheet doesn't really guarantee anything. Peope doing tests with calibrators and way better instruments found some accuracy related bugs in the Keysight 3446xA series DMMs for example (which got fixed by a firmware update after Keysight reproduced it).

[kj7e]

Ive had both the SDM3055 and SDM3065, both where calibrated from the factory bang on and worked quite well.  They are good meters for the price.  My experience has been, Siglent seems to do a proper job with their factory calibrations.

As pointed out, my first SDM3055 had an issue, I never even removed the plastic wrapping as I could hear something rattling inside the meter when I unboxed it.  The distributor I bought it from was not as helpful as I would have liked, but Siglent USA stepped in and quickly resolved the issue.

[Kleinstein]

For the accuracy and longer time stability it is very difficult to tell how good the new meters are actually. There is a little more trust in the way Keysight or Keithley calculate the specs, but time will tell.  Just measuring a few new meters against a more or less accurate source does not give much information. The old established companies have there experience with older models.  At least the LM399 is about the same for all.

[kj7e]

^Agree, I think the Siglent specs are honest, but Keysight specs are conservative.

[JDubU]

^Agree, I think the Siglent specs are honest, but Keysight specs are conservative.

From the Keysight 34401A DMM service manual (see attached):

"The Agilent 34401A is designed and tested to meet performance better
than mean ±4 sigma of the published accuracy specifications."

[eurofox]

I just order one from Batronix.

I will post results ..... 

[analogRF]

Ive had both the SDM3055 and SDM3065, both where calibrated from the factory bang on and worked quite well.  They are good meters for the price.  My experience has been, Siglent seems to do a proper job with their factory calibrations.

As pointed out, my first SDM3055 had an issue, I never even removed the plastic wrapping as I could hear something rattling inside the meter when I unboxed it.  The distributor I bought it from was not as helpful as I would have liked, but Siglent USA stepped in and quickly resolved the issue.

I am having difficulty in choosing between SDM3065X and Rigol DM3068. Any advice? are you happy with it? any pros or cons?
I hear many people complain about the siglent being very slow and sluggish. I also have two specific questions: is the capacitance resolution 4 decimal digit or 3 at lowest range?  In reading hold mode can you adjust the sensitivity of capturing a new reading?

[tautech]

I am having difficulty in choosing between SDM3065X and Rigol DM3068. Any advice? are you happy with it? any pros or cons?
I hear many people complain about the siglent being very slow and sluggish.

The earliest releases were troublesome until firmware became mature and this only applies to early 3055 FW.
You cannot judge the 3065X on 3055 old commentary, one they are entirely different units and two, they run different FW.

I also have two specific questions: is the capacitance resolution 4 decimal digit or 3 at lowest range?

C measurement.
22pF, 50V 5% NPO disc ceramic.
ST3 Smart Tweezers measurement 26-26.1pF

SDM3065X, 5 digits available on 2nF range
4 wire measurement, leads nulled.

In reading hold mode can you adjust the sensitivity of capturing a new reading?

No, not while in a Manual range setting with the Range buttons below the display.
In the Hold GUI you must select; Probe Hold=OFF then the range selection buttons become available.
Then re-enter the Hold UI with a press of Shift and Single/Hold.

[skander36]

Hi Tautech ,
How can I modify files definition for K Type thermocouple in SDM3065X . It show a lower temp - 18* instead of 23* C .
Thank you !

[tautech]

Hi Tautech ,
How can I modify files definition for K Type thermocouple in SDM3065X . It show a lower temp - 18* instead of 23* C .
Thank you !

Didn't know so had a look.....
In the Temp UI select the Load softkey and the right arrow key to select the correct file for your probe.
In the Load menu after selecting the correct File, press Read to check your thermocouple measurement File is a correct match for your probe.
Or you can use Define for which there are 128 offset corrections available.

[skander36]

Hi,
Thank you , but how can I use values from Define ?
Everything I choose don't change anything .

Thank you !

[tautech]

Hi,
Thank you , but how can I use values from Define ?
Everything I choose don't change anything .

Thank you !

Not sure.....having trouble to do that too. 

From what you say you need ~+5*C correction, that fits well with a #40 or 41 adjustment but it does not seem to be saved. 

I'll ask tech support.

[eplpwr]

For the accuracy and longer time stability it is very difficult to tell how good the new meters are actually. There is a little more trust in the way Keysight or Keithley calculate the specs, but time will tell.  Just measuring a few new meters against a more or less accurate source does not give much information. The old established companies have there experience with older models.  At least the LM399 is about the same for all.

Yes and no. LM399 is the same for sure. What you get from the big names are burned-in and binned '399s. Also note that many brands have 10V range as opposed to Siglents 20V. This is great except that part of the accuracy spec is % of range.

My SDM3065 was spot on day 1, then drifted exactly as LM399 datasheet predicts. Buy the meter, you do the burn-in.

[DaJMasta]

I've gotta say, while the Siglent offers very good price/performance.... I've got a 34461A and I really like the meter.  The two are very similar in look and form factor, and the specifications are around about the same, but the price premium for the Keysight is not so crazy either, and they can be found (at least in the US) for only $100 more or so used in good condition.  It may be worth seeing some videos of each in action to verify UI responsiveness and such to get a feel for the differences between them.  The minor advantage of the Keysight is that the lowest range has half the scale (100mV DC vs. 200mV DC and comparable across the ranges), whereas the Siglent has a faster rated frequency measurement, though I just clocked my 34461A reading a 820kHz 1Vpp sine wave, so maybe not by as much as it appears.

Either meter is probably a good choice, but see if you can see the Siglent in action or use it yourself, could help make your choice or differentiate otherwise similarly specified instruments.

[Performa01]

The minor advantage of the Keysight is that the lowest range has half the scale (100mV DC vs. 200mV DC and comparable across the ranges)

How could this be an advantage?

What counts is resolution and dynamic range.

The resolution is the same 100nV per count on both meters. But the Siglent can provide this resolution up to 219.9999mV because it has nearly twice the dynamic range, whereas the Keysight falls back to 1µV above 119.9999mV.

I own neither of them, but prefer my Fluke 8842As (5-1/2 digits) and Keithly 2001s (7-1/2 digits) over my other DMMs for this exact reason. The higher dynamic range also means high impedance up to at least 20V, whereas the 6-1/2 digit DMMs with only 1.2 million counts (like the Keithley 2015THD) tend to already show a noticable non-linear increase in input current once entering their overrange, i.e. above 10V.

For old style analog precision work in +/-15 to +/-18V systems, found in many calibrators, the high impedance 20V range with good resolution (1µV in case of the Keithley 2001) is just indispensable.

[skander36]

I've gotta say, while the Siglent offers very good price/performance.... I've got a 34461A and I really like the meter.  The two are very similar in look and form factor, and the specifications are around about the same, but the price premium for the Keysight is not so crazy either, and they can be found (at least in the US) for only $100 more or so used in good condition.

Used Keysight , Fluke , or Keysight may be a good choice in US , but in Europe they are very rare and at a high price . Import from USA is not an option and you must add 100-200E for calibration , because there is no means to use this precision device without being calibrated. So for me a good source for meters and scopes are clearance from Rigol or Siglent .
(I've got the SDM3065X from a Batterfly promotion at 585 E -no VAT ).

[DaJMasta]

you must add 100-200E for calibration , because there is no means to use this precision device without being calibrated.

It's certainly true with less of a used market the value proposition goes down, but I would pretty strongly disagree with this statement.  It your application requires significant measurement accuracy, sure, but you still get all the linearity and resolution of an out-of-cal instrument, and an LM399 is no slouch.  While calibration is required for traceability and specific measurement tasks, the majority of uses for a meter don't depend on it, and the likelihood that the meter is going to be out of cal even years after its certificate expires is low, given the architecture.


Regarding my 'range advantage' comment, yes, resolution would be the pertinent thing, and while I have not seen a counts equivalent metric for the 34461a and don't know how the ADC ENOB between that and the SDM3065X compares, but if the ADC portion is equivalent, a range setting that is half a competing meter would have double the resolution.  Yes dynamic range is important, but with equivalent converters, dynamic range is the same, so the same dynamic range on a full scale measurement that's twice as large means your minimum resolution step is also double the size.  Now I don't know if the ADC portions are comparable and don't know if there's enough of a noise floor in the bottom of the range to negate any potential resolution benefits, but while the larger range may be preferable in some applications, the smaller lowest range means the higher resolution on very low measurements.

[skander36]

but you still get all the linearity and resolution of an out-of-cal instrument, and an LM399 is no slouch.

Can you argue this ?

[DaJMasta]

What is your measurement requirement?  Can you argue that the inherent drift of the reference, the resistors, or the integration cap or something used is enough to realistically cause a problem with your measurements?

It takes a demanding measurement requirement to really need it and the majority of instruments that go in for a cal need no adjustment.  Your application may require extreme precision, in which case adhering to the calibration cycle and expecting only the specifications on the data sheet are acceptable, but the vast majority of measurement situations don't require that.  Keysight has a long history (at least, as Agilent and HP) of instruments that live up to their specification even after their calibration cycles are up, and the 34461a has been around for plenty of time to have its aging characterized.

None of this is a slight at the Siglent, which from what I can tell uses very similar parts to maintain its stability and specification, but unless your application requires traceable calibration or absolute accuracy, I don't think there's a compelling argument for keeping an instrument regularly calibrated at a lab.  Have some other references as sanity checks and understand how the instrument works, then if anything seems awry you can address it at that time (cal or fix as needed).  Sending it in to cal every 2 years when it's not even going to be adjusted unless its out of spec.

[skander36]

I understand you point of view about the necessity or not about accuracy, but if you can focus on your statement maybe will be more useful for me , i mean technical.
I think that if you don't need precision you don't buy 5,5 / 6,5 digit mutimeter . Instead Fluke 87 or even a more cheaper Brymen 869 will be enough .
And if you decide to buy SH 6,5 multimeter  , you will verify against one with the same specification to find if it still maintain his accuracy .
You don't tell me that you will buy a 5,5 digit Agilent and you don't care if it is specifications . Why dont you buy a Brymen then ?

[DaJMasta]

I have a 6.5 digit meter because I want the precision, even though the accuracy of a 4.5 digit meter would be fine for me.  You can use a 6.5 digit meter to measure small scale drift trends, and even when well outside of calibration, it can happily demonstrate that down to at least 5.5 digits or so accurately.  That said, since my measurement requirements don't require 6.5 digits of accuracy, letting it drift and be out of cal for a decade or two will still likely be within 5.5 digit meter accuracy spec, which still exceeds my requirements, but will show me all of the fine detail of the nicer meter.

I can use my 6.5 digit meter to measure the tempco of a thick film resistor accurately at room temperature - just a few degrees of fluctuation, and that is a useful/fascinating measurement to me that would be tough with a 5.5 digit or less meter.  I don't particularly mind if the resistor I'm measuring starts at 101.256k Ohms or 101.254k Ohms (for example), so the calibrated accuracy spec far exceeds my requirements and I trust enough in the architecture, component choice, and testing done by the manufacturer to not be dramatically out of calibration for no reason even well beyond the calibration period.

Maybe others don't have the same trust or do demand accuracy or traceability, and it's their choice to make, but for my use-case, a meter that's over-specified for my requirements but is out of cal shows me fine detail lesser meters can't, and is very accurate, even if it's out of its specification.  Then if I get my hands on a reference that's been checked against a real standard, or a different 6.5 digit meter in passing or something similar, I can use that as a transfer standard to check my instrument out and verify that at the level of detail I can't otherwise verify on my own, it at least agrees with other instruments.  As I've said, It's not the approach for every application or every person, but I think it suits a lot of use cases well.  A calibrated instrument does have some extra value, but I don't think it's an inherent requirement for a purchase of a high resolution meter for most people/situations.