Rigol DM3058 "Hold" trigger mode. What use is it?

[Macbeth]

The Rigol is my first "proper" bench PSU. 

I always wondered what the RUN/HOLD button was for, other than lighting up in RUN, extinguishing in SINGLE (which does actually hold its reading), and flashing when in HOLD (which doesn't seem to hold anything at all). Reading the Chinglish documentation made mention of some formula and the settings of 0.01% - 10% in the trigger setup.

Testing with a PSU it does indeed "hold" the reading on the display until the voltage falls outside the window. For example, set the trigger to 10% and measure 10V then lower the PSU voltage. The display won't change until the PSU drops below 9V (including removing the test leads) or rises above 11V. However, it still appears to trigger anyway and records the "false" held readings in the history log. I would have thought if it was any use at all it wouldn't record more than the initial decided on "held" reading and stop recording until a measurement outside of the held range is triggered.

I just put this down as a feature important to proper electronics engineers far more worthy and knowledgable than me. I mean Rigol dedicated a whole backlit button to it and made it flash!

I recently got a Keithley 2015 and decided to give it's Hold a whirl. What do you know? It is utterly brilliant. It has the same window settings of 0.01% - 10% along with a sample count (default 5). Not only does the trigger actually work as in it only records the held reading once, but when you take the leads off it only freaking actually HOLDS the reading on the display until you go and probe somewhere else. You just place your probes wait for the beep and go and look at the display with the probes removed. The Rigol just shows the 0 volt noise like it does in regular RUN mode here.

So... WTF is this function for on the Rigol, and why has nobody else brought it up? Now I know what it should do it's implementation is an utter joke. Unless someone can tell me otherwise? 

[Wytnucls]

I have the Rigol 3058E and the Keithley 2000.

On the Rigol, the 'Reading Hold' seems to work as advertised, displaying a new seed when a measurement falls outside the % set window.

On the Keithley, the 'Reading Hold' is slighly different. It also displays a new seed when a measurement falls outside the % set window, but, depending on the rate and number of samples settings, when the probes are removed from the DUT, the last measurement may remain on the screen for long enough to take note of it, the meter acting like a Fluke AutoHold.

[Macbeth]

But what is that used for? Also why does it hold it on the display, but carry on recording (like its not triggering at all - just running as usual) wasting data points in memory?

It clearly is something I am missing. I can only imagine something really really obscure for a "feature" like that. So why all the prominence of it being a front backlit button, flashing away to show its stuff. It must be a really important and popular function that everyone asks for to hold such prominence rather than hidden away in the trigger menu somewhere. Why is it in the trigger settings anyway? (I mean I could understand its useful for something if it stops filling memory until the trigger condition, like the Keithley, but it doesn't)

Also, does the DM3068 exhibit the same behaviour?

[Macbeth]

...when the probes are removed from the DUT, the last measurement may remain on the screen for long enough to take note of it, the meter acting like a Fluke AutoHold.

Indeed it does. In fact I left it there all day, plenty of time to take note. It also only used 1 datapoint in the memory.

Of course for a feature like autohold to work it is important to not operate the meter in super high impedance mode as when the leads are removed the stray capacitance causes the stray voltages to ramp up as I am sure you are aware with your Keithley. So when in trigger hold mode the Keithley switches to regular 10M ohm impedance.

The Rigol manual also states that when in reading hold mode the meter is put in low impedance. This can only be for a proper autohold function to work as it only applies when the leads are removed from DUT, so why do Rigol implement this step?

Thanks for your response wytnucls, because I feel like a loner losing his mind on this one 

[Wytnucls]

The Keithley doesn't always hold the last reading with the probes disconnected. Try 10V at 10% in Fast mode with 2 samples only.

I think the problem on the Rigol DM3058 is that the number of samples cannot be set for the Reading Hold. It is stuck at a very low setting.
On the Rigol DM3068, the number of samples can be set very high for the reading Hold and it probably behaves more like the Keithley.

I suspect the DM3058 was meant to work in the same manner, but Rigol screwed it up somehow.

Most bench meters have the Reading Hold feature with the percentage setting, sometimes calling it Auto Hold, like the Agilent U1231A.

[Macbeth]

Yes. I think they royally screwed up on this one. The keithley defaults to 5 samples, the Rigol fixed to 3. But there is more to it than that.

I reckon someone from the marketing dept. decided that the 3058 can't possibly have the same nice features as the 3068... "what can we cut? other than the 1 digit? Oh.... well, we can get rid of the trend feature.... just leave the histogram. MmmmKayy. Oh not the live histogram - make the plebs keep pressing the update button... Whats that? Allow them to trigger on a measurement window? Nononono - yes that is built into the Pass/Fail whotsmebob, but we have to remove at least 1 trigger so that will have to go... What's that - the engineers are saying its fundamental to the Reading Hold?? What do they know? Marketing knows best... it has been ordained"

...and so the 3058 "Reading Hold" is a total fail - biggest button on the meter and light flashing but utterly useless. I mean I can hold a reading just by pressing "Single" - I just need 3 hands but at least that doesn't fill up the memory buffer with useless readings. 

Ok. I think I have ascertained that this is a proper fail despite your best will in the world to prove me wrong. I'm going to moan at Rigol! 

[Macbeth]

I've just tried the Keithley in fast mode with no filtering (iinsanely fast) and 3 samples like the Rigol according to the user guide - 10V and 10% and it holds the reading perfectly for a good minute or so then drops and holds another "noise" reading... I think the 5 sample default is probably a good setting - the Keithley sits there all day with a reading if you let it.

Any DM3068 owners willing to explain how the hold functions on their meters? I don't recall it being shown on the Signal Path review - I'll have to watch it again. I noticed Shahriar recently reviewed the Siglent SDM3055 5.5 digit meter which is surely the competitor to the aging DM3058. He demonstrates its hold function and how it records readings into a kind of spreadsheet. Just a shame the meter is so slow! (which doesn't make sense as it is using the same 24 bit ADC as the DM3068!)

[Wytnucls]

If you want an Auto Hold on the Rigol, use the Single mode with slow speed with a trigger after 10 samples. That should give you about 5 seconds to take a measurement, leaving the probes on the DUT until the time is up. Press Single before each new measurement.
Cumbersome, but better than nothing.

[JimStar]

If you want an Auto Hold on the Rigol, use the Single mode with slow speed with a trigger after 10 samples. That should give you about 5 seconds to take a measurement, leaving the probes on the DUT until the time is up. Press Single before each new measurement.
Cumbersome, but better than nothing.

One more, some faster working method - is just using "Max" (or "Min") function. This way there is no need to wait for 5 seconds or be in rush to hit probes into the right places under the microscope in just 5 seconds.
Unfortunately there is no "one press" button to reset the result before the next measure, but pressing "REL" twice works for that purpose (you just need to open the menu first, 'cause in "Min/Max" mode it's hidden by default).

Or just attach some handheld to it by a tape, that has this function working properly...

I've had a conversation with Rigol about it (quoted below), but looks like they are completely unwilling to fix this ridiculous bug (show-stopper in my workflow). Despite the fact they confirm that this is a bug and they even have this bug-ticket in their bug-tracker system.
Absolutely ridiculous thing: for many years still continue advertising the "Hold" function sitting on the big lit dedicated button in the center of device, that does NOTHING when activated in comparison to when it's not activated...
Such a shame, 'cause I really like this device other than that...

Hi guys,

I've just bought your DM3058E multimeter, and I like it, it's really good device for the money.

One important thing only that I can't get working properly: the hold function. No matter what percentage for "Hold" I set in the settings, after switching on the "Hold" function (pressing "Run/Hold" button and it starts blinking) - it never holds the last stable measured value on the screen after disconnecting the probes: as soon as I disconnect the probes from the board the number on the screen jumps back to "close to zero volts" (same for resistance). So visually there is no difference in readings on the display whether the "Hold" is active or not - when it is active it reacts to disconnecting the probes absolutely the same as when it is not active: the value jumps back to "close to zero volts".

The whole purpose of the "Hold" function is to be able to probe some microscopic pins when you looking into the microscope and are not able to look at the multimeter display at the same time (to not risk to short something when losing visual control of the probes). So it is expected (and other even handheld multimeters do this autohold correctly) that after you had attached the probes under the microscope and heard the multimeter's "beep" notifying you that the stable reading within the given percentage had been made, then you removed the probes, and now you can look at the display of the multimeter and the last stable-measured value (in the range chosen in the settings) is frozen on the screen. But this never happens with my new DM3058E: as soon as I disconnect the probes and can move my eyes away from the microscope without the risk to short some lines - there is already no last stable "held" reading on the multimeter's display - it is back to "close to zero"... Even though the "Hold" function is active and the "Run/Hold" button is blinking...

So the question is: how can I make the "Hold" function work the same as with other multimeters that hold last stable reading on the display? So that I could measure small pins under the microscope without looking at the display of the multimeter at the same time and the last stable reading is HELD on the display after I had disconnected the probes and finally can move my eyes from the microscope to the display of your multimeter?


Thank you and kind regards,
Jim

Hi Jim,

We do have a bug report against this. The product lines says it will be fixed but I don’t have an estimated date when that will happen,

When was the unit purchased?

Best regards,


Steve Huss

Thank you for the answer, Steve.

I hope it gets fixed ASAP, because as I can see using this multimeter for a couple of days - this is THE ONLY big problem with it. I don't see any other drawbacks, without this bug the multimeter is ideal for tasks like mine...
But this bug makes is absolutely unusable in a very common workflow: searching a PCB fail reason under a microscope. You can't look at the multimeter display while looking into the microscope, so the properly working "Hold" function is absolutely essential.
Many proffessionals suffer from this and discussing this ridiculous Rigol fail, e.g. on well known EEVblog:
https://www.eevblog.com/forum/testgear/rigol-dm3058-_hold_-trigger-mode-what-use-is-it/

I'm a programmer myself (https://otoy.com), and I know that such a bugs do not take long to fix and rebuild a new release... Please fix it ASAP, please please please..

BTW, I've tried to download the firmware update from your site, and the version that is available currently is 01.01.00.02.02.00.02 from 03.12.2015. But my multimeter reports it has the 01.01.00.02.03.01 version. If it is a newer version - then it is strange that your site has such an old version as an "update" while the much newer is already available... And the fact that this bug is still not fixed since the last 2015 firmware version makes it even worse... Such a good-potential multimeter, and such badly broken by this trivial bug..


Kind regards,
Jim

Hi Jim,

Yes we do need to correct our site. I’ll ask that that be done.

Unfortunately the fix could be well out in the future. Considering this I should note that we do have a 30 day refund policy for any reason. How long have you had the DM?


Steve Huss

Hi Steve,

I have it for a couple of days, but I bought it 3-4 weeks before that - the parcel just took some time to arrive in New Zealand from China (from AliExpress). So I think I'm out of 30 days... And I've already attached it to my work bench, so too much hassle. Will just use my cheap handheld multimeter in addition that has this function working like a charm: shame on your expensive "professional" device, Rigol.

I understand how the support-lines work: the "well out in the future" in response to such a show-stopper bug report should be translated as "we don't have plans to return back to work on this device" - the latest firmware's date of 2015 speaks for itself... Taking into account such devices' life is kinda long, I'm sure in 2025 people will still get the "well out in the future" from Rigol support..
At least have a courage and state in the device description that your advertised and big-button-dedicated "Hold" function is nothing to do with a real "hold" and does not "hold" anything like even cheap handhelds do...

Thank you for the answer anyway, at least now I understand the situation. I understand it's not your guilt, Steve - you're not a decision maker... It's just really frustrating that such an otherwise good multimeter is so broken by this just one important thing...


Kind regards,
Jim

Hi Jim,

I can appreciate your comments. We will continue to encourage the product line to address this issue.

At the start of our conversation I didn’t realize you were located in New Zealand. I’m in the North American office located in Oregon. I will forward this request to our International support team so they are aware of your request.

Best regards,


Steve Huss

Have not heard from them anymore. Will keep this updated if some news...

[emartine]

Hello! Was this ever fixed? Thanks!