UNI-T UT8805E or Siglent SDM3055 - which one should I choose?

[nctnico]

For this budget you can't buy something new without some compromise. I'd suggest to look at a used DMM from Keysight,  gw-instek or Keithley. Good deals can be found at used equipment dealers every now and then and they typically offer warranty as well. I bought some next to new units this route.

[TylerPeppy]

The OP is the one who mentioned "calibration" of a power supply in the beginning of this topic, so metrology aspects should be considered.

I have enough DMMs and calibrated equipment in general to definitively state that published accuracy specs are completely overrated.  Some manufacturers are 10x or even 100x overly cautious with their specs compared to real-world/observed performance, and others are so optimistic it becomes clear that they are just being outright deceptive in order to drive sales.

The 3458A is bit of a handful, as seen from multiple threads on its various failure modes as well as a recent Marco Reps video where he attempted a marathon repair session of six 3458A units.  I personally would never own one, but of course it's a staple for calibration labs.  You would need a few of them as well as a few references in order to be able to confidently state what is what, especially the further right you go on the display.

In another thread I mentioned that due to the equipment I have from various independent sources (K2010, PDVS2mini, etc), I was able to tell a reference provider that their 3458A was approximately 70uV off and sure enough, after they sent it in for calibration the report came in that it was 75uV out.  So a single calibrated piece of equipment is not adequate if you're going to play in this space.

We also do know that some of the brands like Siglent do not age their references (or not long enough), so some drift after purchase can be expected.  And older, well-aged units are prized for their aged stability.  Also, unless you continue with a service contract, the warranty is a bit overrated as well.  Siglent's warranty is three years, but if the unit drifts out of calibration, they will only cover that for one year after purchase.  So you should plan to have it calibrated yearly, which is going to be expensive over time.  So now the used equipment that is calibrated once at purchase but is stable over many years becomes attractive.

If acquiring such used equipment in a specific country is a problem, then that is fine to point out, but I don't see how it negates the premise of older equipment being preferred for a metrology-leaning hobbyist.

And all of that is completely not related to OP question.
Talking about missing the point.

OP mentioned adjusting his homemade PSU he is making. How did you come from that to metrology is anybody's guess.
To adjust PSU, 1mV accuracy is more than enough to call it precision power supply. So setting it to ppm level is a waste of time.
Even those like Rigol DP831 that have specified 0.01%/°C+5mV will drift more than that when under load, and have that random 5mV mixed in for fun. And then voltage drops on output cables (unless you have PSU with sense) will be a lot too.

How did we get from "I want to adjust my homemade PSU" to "person is metrology-learning enthusiast" ?
He is not. He just wants to have benchtop multimeter that is "slightly better" than his 2 rather good handhelds.
He is not into playing with and collecting meters, he wants to use it to do job with them.
Personally, I think his 2 handhelds are perfectly enough to adjust any PSU he might make.
But there are many pro & contra to both benchtop and handheld form factors, so having one benchtop is useful, and also to have 3rd opinion. So if he wants one, good.

From options OP asked for, Siglent is better by a large margin. We also pointed out that they can be had for better prices than he initially stated.
With Siglent, one must take into account that LeCroy rebadges many of Siglent devices. With literally no changes except sticker and BMP for logo. So you basically get LeCroy for less money.

DaneLaw mentioned Hantek meter. I have not tried it, so can not say, but UNI-T, Hantek, Owon and the likes are not in the same tier as a company. General fire and forget nature of the devices and basically no support is not the same as standing behind your product. Price is very low though, but quality might be proportionally lower too. I don't know. Another one "cat in a bag scenario".

Yep, I think J-R is completely missing the point here. OP is not looking for Metrology grade equipment. One can own a high-precision piece of equipment without wanting / needing that at all.

To OP: You can choose the SDM3055, but have you considered the SDM4055? It's significantly newer and might be better bang-for-your-buck? I am still not super clear as to your budget, but anyone here is just trying to point out, that you could get a more accurate, more precise DMM for the same price you are already paying. If you REALLY cannot go with used equipment, the Siglents will probably be pretty good for your use case.

I second the GW though, their equipment seems to be pretty impressive.

Tyler

[TUMEMBER]

You didn't make me feel any better with the latest Siglent software.
I found information online that the GDM-9052 can measure up to 320 measurements per second.

If you're not just looking for a "fresh-smelling" gadget with fancy displays, I'd also consider older models. Most "live statistical measurements from the display" can be easily obtained from data transferred by a multimeter via RS232 or GPIB. Processing data "for graphing" using Excel isn't exactly a genius skill. If you're looking for "on-screen" gadgets that cost a fortune, then that's your choice.


Read the linked topic.

Resolution isn't "accuracy." Stability isn't a "datasheet" on sale—stability is the change in accuracy over time, and should be monitored periodically to understand its evolution. You should have your own stable "reference standards" or use external calibration, but these are exponentially increasing costs.

https://www.eevblog.com/forum/testgear/how-accurate-is-a-good-40-year-old-dmm-today/msg6019491/#msg6019491


Contrary to appearances, "old is beautiful because it is old" - what was supposed to flow has already flowed, changes in accuracy will not occur as quickly as "in new products straight from the shelf".

[Kleinstein]

The SDM4055 should be essentially the same as the SDM3055, just with a different display / UI part.

I would avoid the slightly cheaper SDM3045 - the newer ones can no longer be easy changed to the chinese model 3055E. It would also be less accurate.
In the EU the supply of used instruments is rather limited. In the US one could consider a used HP3478 - these are often relatively cheap (e.g. $100-150) and usually pretty accurate for a 5.5 digit meter.

[DaneLaw]

DaneLaw mentioned Hantek meter. I have not tried it, so can not say, but UNI-T, Hantek, Owon and the likes are not in the same tier as a company. General fire and forget nature of the devices and basically no support is not the same as standing behind your product. Price is very low though, but quality might be proportionally lower too. I don't know. Another one "cat in a bag scenario".

Yep, none of the other that are mentioned UniT, Owon, Siglent, Fnirsi, Rigol are producing the same class of bench DMM instruments.
Hantek is really not a toy or low-tier brand as perse, they do also make some serious gear..

Producing 7.5-digit meters does take some dedication, dont recall I have seen 7.5-digit meters from any of the other Chinese brands?
Haven't tried either of the products from Hantek, though did look into their 6.5 count some years ago, as it was somewhat attractive on paper, not least for around 600 bucks, but I do lack the need

[tiger80]

To TylerPeppy - I've seen the SDM4055, but it's way over my budget. My budget is a maximum of 2500 PLN, or about $650 USD. I don't want used equipment; I don't want to buy on eBay. It might turn out that the used one will soon die, leaving me with neither a meter nor money. No, that's it.
I'm grateful for every suggestion, which is why I asked for your help, as I don't have experience with this type of equipment myself.
I'm getting closer to choosing the Siglent SDM3055, but I'm also considering the GDM-9052. Both meters are within my budget.

To TUMEMBER - I don't care about fancy displays. That's why I'm looking at the GDM-9052.

To DaneLaw - The Hantek HDM3065 was way over my budget.

[TylerPeppy]

To TylerPeppy - I've seen the SDM4055, but it's way over my budget. My budget is a maximum of 2500 PLN, or about $650 USD. I don't want used equipment; I don't want to buy on eBay. It might turn out that the used one will soon die, leaving me with neither a meter nor money. No, that's it.
I'm grateful for every suggestion, which is why I asked for your help, as I don't have experience with this type of equipment myself.
I'm getting closer to choosing the Siglent SDM3055, but I'm also considering the GDM-9052. Both meters are within my budget.

To TUMEMBER - I don't care about fancy displays. That's why I'm looking at the GDM-9052.

To DaneLaw - The Hantek HDM3065 was way over my budget.

Just doing some further research, looking at the Datasheets, and using AI tools to build me some comparisons; they both really have their advantages, and both are pretty similar. Here are my main bits:

1.) The GW seems to have a faster maximum update rate of 320rdgs/sec, whereas the Siglent has a maximum of 150rdgs/sec. Note that I am assuming these will only show up on the lowest measurement resolution.

2.) They both have 4.3" TFT displays, so they should suffice in the visual sense. I think the Siglent might have a slightly better UI with graphing and trend view, but nothing to write home about.

3.) The GDM9052 does not have a USB Host interface whereas the Siglent does. That means no datalogging directly to a USB stick on the GW, but you can on the Siglent. Personally, I would use the fantastic TestController app for datalogging all of this.

4.) It does seem like the 9052 has a *slightly* better basic DCV stated accuracy of 0.012% + 0.003% compared to the Siglent's 0.015 + 0.003%. GW wins here, but I am not sure of the "true" accuracy of these models anyways.

5.) Both meters do have a >10G\$\Omega\$ switchable impedance, but on 200mV and 2V ranges.

6.) The SDM3055 has a feature I really like; which is the statistics mode. It will do MIN, MAX, AVG, STD. DEV, SPAN, and # Of Samples. The GDM doesn't seem to have the avg, standard deviation, span, or samples.

7.) The Siglent seems to support more temperature measuring sensor types. For TCs: B, E, J, K, N, R, S, T. It also supports PT100 RTDs. Supposedly, newer firmwares also allow you to add in your own sensors... My 4065A has this and it is nice for using Arduino sensors. On the other hand, the GDM only supports J, K, and T thermocouples. No RTDs or custom sensors.

8.) For resistance measurement, a max of 100M\$\Omega\$ on both. For capacitance, the Siglent claims to do up to 10,000uF, whereas the 9052 does up to 119.9uF?

9.) The test for diodes on the GDM-9052 seems to use 1mA, 6V fixed. The Siglent has adjustability between 2V 1mA, and 4V 100uA. Not too clear on this guy though.

I hope this helps a bit more, anyone please feel free to jump in and correct my somewhat hastily collected info!

Tyler

[tiger80]

Everything points to the fact that I won't be able to find anything better than the SDM3055 at this budget.
There's also the RIGOL DM3058E  , a 5.5-digit model with 0.015% accuracy. It's cheaper than the SDM3055.

[TylerPeppy]

Everything points to the fact that I won't be able to find anything better than the SDM3055 at this budget.
There's also the RIGOL DM3058E  , a 5.5-digit model with 0.015% accuracy. It's cheaper than the SDM3055.

While I won't ever tell someone what to buy (unless there's a really good case for it), I think most things point to the Siglent being pretty versatile. I wouldn't really consider the Rigol, I just don't like the UI on it.

Tyler

[2N3055]

I have DM 3068. While it measures fine, display is horrible and really small.

[tiger80]

I haven't seen the DM3058E in person, but the display does seem small and unreadable. I prefer large numbers, preferably against a dark background. The DM858, on the other hand, has a large display, but it has so many hardware flaws that it's not even worth a quarter of the price.

[Martin72]

At Rigol, the line between useful tools and toys has always been a fine one.

[TylerPeppy]

At Rigol, the line between useful tools and toys has always been a fine one.

Although, some of their new high-bandwith oscilloscope and microwave instruments look very nice. I would be super curious to see how they stack up against the main rivals.

[J-R]

I brought up metrology because the OP specifically stated "laboratory power supply", "accurate measurements" and "calibration".  These are quite different from "I want to adjust my homemade PSU".  But maybe they didn't mean what they said...

[tiger80]

To J-R
You understood correctly. But that's assuming I'm just doing this as a hobby. I won't make any money from it. As I mentioned in the first post, I'm an amateur. Professionally, I do something completely different. I like playing with electronics, creating something new, but for my own needs.

[adam4521]

A few points that no-one picked up yet:

1. The Siglent has ethernet, GW Instek does not.
2. For GW Instek feedback, I've got an older model IDM-8342 (rebrand of GDM-8342). This is only a 50000 count, so lower class meter. However, I was pleased to find that it can do voltage/current dual measurement without click-clack relay on some ranges (it seems with voltage range 50V and up). Silence is golden, but it also means it maintains decent measurement speeds when the relay is not needed. Unfortunately this sort of detail isn't covered in datasheets so I have no idea how different meters compare on this. This meter also offers 0.05% accuracy on the mA current ranges. My impression overall is that it is competent, they didn't skimp on the hardware design.
3. However, on TestController, instrument profiles are well supported for Siglent, and sporadic for GW Instek. I'm having to build up my own definition file (never done one before so I have had to learn it). In addition, GW Instek use unique vendor ID and product ID codes for the USB serial interface. It will be ok on Windows with the manufacturer driver, but on Linux, I had to make a custom udev rule to recognise it (I've posted about how to do that elsewhere on the forum).
4. For reasons like the above, there could be some subtle differences in details between the meters, but overall I think you are going to find smoother compatibility and better forum support with Siglent.

[Martin72]

but overall I think you are going to find smoother compatibility and better forum support with Siglent.

I think that's an important point—the reach.
As I said before, GWInstek is an interesting brand—I don't know much about them, but what I do know is good.
Still, I'd lean toward Siglent because a) they're also good, and b) they're much more widely used in the community, which can and will pay off one day.

[KungFuJosh]

I would avoid the slightly cheaper SDM3045 - the newer ones can no longer be easy changed to the chinese model 3055E. It would also be less accurate.

There's no longer any issue updating the newer models. The improvement thread has it covered with the latest firmware update.

[BillyO]

Just don't buy UNI-T.  I've had a UNI-T bench DMM and a scope.  Both had issues and UNI-T offered zero support.  I don't mean poor support, I mean zero support.  Nothing.  No firmware updates, no responses to queries.  Nothing.

They have supposedly opened an office in the US, except the alleged address has changed 3 times in the last year.  Total shit show.

Edit: Oh, and I don't mean they just moved down the street.  No, they moved to different states. 

I know you're not in the US, but it serves as an example of the kind of company they are.

Edit 2: Below is a screenshot of their search page asking for their address.   

[tiger80]

I won't buy a UNI-T. I'm seriously considering your suggestions. Since you write not to buy UNI-T, I'll take your word for it 

A year ago, I bought a UT181A handheld multimeter and regretted it because it wasn't cheap and it struggles to measure resistance below 1 ohm. The UNI-T probes are garbage. Even using brand-name Brymen probes with gold-plated tips, the readings on the UNI-T fluctuate. And on my old Sanwa PC5000a, the readings are stable.

[J-R]

Broad arguments about a specific brand or support can fall apart quickly, in my opinion.

There is zero support from Brymen for US customers, yet their DMMs are very popular here.  (I consider them to be disposable purchases.)

Some Uni-T products are terrible and don't make sense to buy, others are considered to be excellent value for money by a significant number of forum members.  (I have a few of the Multicomp Pro rebranded Uni-T UT256 fork meters that I got for 10c on the dollar; a very shrewd purchase.)

Forum users have complained about some of the top brands pulling back from hobbyist support.  (Personally,  I have had no issues here in the US with Keysight.)

So ultimately what to purchase is complex, as seen by how long this topic has been going on...

[nctnico]

Broad arguments about a specific brand or support can fall apart quickly, in my opinion.

There is zero support from Brymen for US customers, yet their DMMs are very popular here.  (I consider them to be disposable purchases.)

Some Uni-T products are terrible and don't make sense to buy, others are considered to be excellent value for money by a significant number of forum members.  (I have a few of the Multicomp Pro rebranded Uni-T UT256 fork meters that I got for 10c on the dollar; a very shrewd purchase.)

That is a good approach. With any piece of equipment you really don't want to have to wait until features materialise or bugs get fixed. It works out of the box as expected or it doesn't. The trick is to find good information (thourough reviews) though. Otherwise you'd have to do the testing yourself.

[Doctorandus_P]

Maybe have a look at the: GW Instek GDM-9052

[J-R]

A couple more thoughts with regard to some other things that were mentioned:

- For low resistance measurements, don't use a 2-wire setup, even if one DMM appears to act better than another.  It's just not a reliable method.  You can get great results from a Kelvin connection using a basic power supply and handheld DMM, even with what might be considered cheap junk, because the lead and contact resistance is essentially inconsequential.  You can get down to milliohm accuracy quite easily.  Consider that 1mV/100mA=0.01 Ohms, and 1.1mV/90mA=0.012 Ohms, for example.  Even if your handheld DMM or power supply was off by a mile, you'd still be solidly in the ballpark of 0.01 Ohms.  Also, in many cases you can beat a typical bench DMM with 4-wire capabilities because those can only output low currents.  Consider that 0.1mV/1A=0.1mOhms.  Best resolution on the SDM3055 is 1mOhms.

- Even if a DMM "passes" calibration or is otherwise stated to be within spec, that fact can be of limited use if accuracy is important, especially without seeing the recent and historical calibration data.  Consider the allowed 1 year "Error from Nominal" for the SDM3055 at 10V, for example, which is anything between 10.0023V and 9.9977V.  So plan to have the stability checked regularly, either by shipping it out for calibration or via your own reference standards.

[tiger80]

To Doctorandus_P - I was considering the GW Instek GDM-9052, but TylerPeppy in Reply #31 showed more advantages for the SDM3055.

To J-R - I don't have a multimeter for measuring resistance with a 4-wire (Kelvin) resistor. I only compared the readings of both my multimeters, and the UT181A's readings differed from the PC5000A's. Having a benchtop multimeter with a 4-wire resistance measurement will provide significantly better measurement accuracy than handheld multimeters.