Siglent Bench DMMs SDM3065X, SDM3055 and SDM3045X

[et328]

Is there a way/hack/hw-modification to make the SDM3045X power-on when AC power is connected?

[Pinkus]

There is nothing in the OS/firmware but you may 'hack' the button in a way that it is either pulled up or down to GND (you shall measure to find it out) for a while after powering the device (simple electrolytic capacitor and a resistor should do it).
Hints:
· by looking on the teardown pics I guess, J15 will probably be the connector where the power-on-button is connected.
and/or:
· See the teardown here: https://lygte-info.dk/review/DMMSiglent%20SDM3045X%20UK.html
Then search for "Q3", you may proceed from there on.

If you are successful, please add the needed information here.

[et328]

There is nothing in the OS/firmware but you may 'hack' the button in a way that it is either pulled up or down to GND (you shall measure to find it out) for a while after powering the device (simple electrolytic capacitor and a resistor should do it).
Hints:
· by looking on the teardown pics I guess, J15 will probably be the connector where the power-on-button is connected.
and/or:
· See the teardown here: https://lygte-info.dk/review/DMMSiglent%20SDM3045X%20UK.html
Then search for "Q3", you may proceed from there on.

If you are successful, please add the needed information here.

J15 goes to the USB connector. Looks like the power button controls some wakeup line and is one of those flat cable wires.

[et328]

Did more checking... Front panel power switch line goes to this point and grounds it when the power button is pressed. So, looks like now it needs just something to pull it to ground for few seconds when AC power is connected. Capacitor-resistor, relay or 555 timer.
Edit: I put 470 ohm resistor and 1000uf capacitor in series to ground from that point and now the unit powers on when AC is connected. Power switch still works and allows to power off and on again.

[Gridstop]

Did anything ever happen with a user cal procedure that a third party cal lab (other than transcat) could use for any of these meters?

I still keep holding out hope that these SW issues and cal procedure get sorted out for these. From what I can see Instek and BK have the same issues around calibration at least so that would be a nice differentiator.

[tautech]

Did anything ever happen with a user cal procedure that a third party cal lab (other than transcat) could use for any of these meters?

Yes there is a user adjustment process available for SDM3055 and SDM3065X but it is only available on request and not for public release.
If owners can demonstrate they have adequate references to calibrate to/against the procedure is made available.

[kcbrown]

I'm considering one of these meters, but I have a concern about one of its features not working properly, specifically the hold feature.

TRX Lab did a review of the SDM3045X some time ago.  In it, he tested the hold feature in comparison with his Agilent U1272A handheld meter.  The review of that feature starts here: https://youtu.be/K09BLgyk9Hg?t=2005

Now, the problem is that, for him, when there's nothing being tested, the meter will randomly believe that there's a sample that it needs to capture, with the end result being that, firstly, your real sample will not be held for long as it detects another bogus one, and secondly, the history of such samples will fill with bogus samples fairly quickly.  This is with the meter in autoranging mode but that might not make any difference, and in any case that would be my primary use case.

My question is: has this issue been resolved in the current firmware?  Seeing how Siglent didn't (and perhaps still doesn't) seem to be treating these meters as first class citizens, I'm reluctant to purchase any of these meters if functionality that I'd really want would still be in a broken state, since under those conditions I may as well just pay a few hundred more and get an Agilent 34461A and be done with it.

If, on the other hand, the basic features of the meter have been dealt with, then the price of these would make them quite compelling.

One other question: the SDM3045X is a 60K count meter.  The SDM3055 is a 240K count meter.  The SDM3045X seems to retain its significant digits at medium speed, whilst the SDM3055 loses a digit and becomes the equivalent of the SDM3045X under those conditions.  My question is: what is the count value of the SDM3055 when in medium speed mode?  Is it still 240K counts, or is it 24K counts?  If the latter, then the SDM3045X is actually better suited for my needs.

[tautech]

Back in 2017 a Hold bug was addressed: Modify Hold function arithmetic.
You can find all the improvements listed in the Revision notes that are with each FW update.

There is also a chance not applying Rel in an EMI noisy environment could result in false measurements in Hold mode.
Readings/s settings allow the meter to settle for more precise measurements.

SDM3045X is fanless but has some different range specs to its bigger brothers so you need check the datasheet KC.

[kcbrown]

Back in 2017 a Hold bug was addressed: Modify Hold function arithmetic.
You can find all the improvements listed in the Revision notes that are with each FW update.

There is also a chance not applying Rel in an EMI noisy environment could result in false measurements in Hold mode.
Readings/s settings allow the meter to settle for more precise measurements.

Hmm... well, maybe that's what was going on in that video (did you watch it?).  But clearly, his Agilent wasn't doing the same thing.  And in any case, in a noisy environment, going into rel mode won't help you because it's apparently the amount of variation in the value that determines whether the meter re-arms the hold trigger, and after that its just a question of whether or not a given value remains within some defined amount of variation that determines whether or not the meter decides that the new value qualifies as a new measurement.

I think it would be really useful to be able to specify a value window within which the meter should consider to be "probes off".  That would take care of the problem nicely.  That might require an alternative hold mode implementation, however, depending on how hold mode is implemented already.

SDM3045X is fanless but has some different range specs to its bigger brothers so you need check the datasheet KC.

I already checked the datasheet.  It is utterly silent as to the counts of the meter in medium or fast speed modes.  It should be easy enough to test.  Do you happen to have an SDM3055?  If so, could you switch the meter into medium speed and see at what value it switches ranges?  My suspicion is that it'll be 2.4 something, which would make it a 24K count meter at medium speed.

[flash2b]

SDM3055

Fast/Medium


Slow

[kcbrown]

SDM3055

Fast/Medium
(Attachment Link)

Slow
(Attachment Link)

Does that show the point at which the range automatically changes?  That would be the point that's related to the "count" of the meter.  The specifications say that the SDM3055 is a 240,000 count meter, which means that the range would change on or near that value.  For instance, it would change from the 2V range to the 20V range when the voltage goes above roughly 2.4V.

In your case, you were at 7.5V, which would put you in the 20V range, and that's exactly what the display says.  Thing is, if the meter were, say, a 60K count meter when in medium speed, then the transition point would shift from 2.4V to roughly 6V.  But the 7.5 value you show would still put you in the 20V range even with a 60K count.   That's what I'd like to see tested: when at medium speed, where's the transition point?  I imagine it doesn't change based on the speed setting, but it's of interest to me anyway.

I may end up going for the SDM3055 regardless, just for the added flexibility.  But I'm undecided at this point.

By the way, does fast mode really have the same number of digits as medium mode?

Also, if you could test hold mode to see if the issue I mentioned persists, that too would be very helpful.

[flash2b]

Switching happens around the 2.3999 --> 2.4000V, sorry that I did the test with my Rigol DP832A didn't have time to warm-up my Burster Gernsbach reference.

Slow/Medium has the same number of digits.

Hope this answers your 1st question.

[kcbrown]

Switching happens around the 2.3999 --> 2.4000V, sorry that I did the test with my Rigol DP832A didn't have time to warm-up my Burster Gernsbach reference.

Slow/Medium has the same number of digits.

Hope this answers your 1st question.

Thanks, it does!

Be interesting to see what you see with hold mode.   You're on the latest firmware, I presume?

[kcbrown]

Any answers to the question about the hold functionality?

Right now I have an Instek GDM-8251A.  Its hold functionality is pretty much worthless for what I'm after, since it takes a new reading whenever the value read differs from the current value by the specified percentage, and that will obviously happen whenever you remove the probes from whatever you're measuring.  And its resistance mode is s....l....o...o...o....w for homing in on a resistance value.  It's a 5.5 digit, 120K count meter, and works pretty well, especially for the price I paid (just over $100).  But I'm looking for something significantly faster and it appears the Siglent qualifies.  It would be immensely useful if the hold functionality worked the way it does for the Fluke handhelds and for that Agilent handheld that TRX Lab was using in his video.

Note that he's not the only person who had problems with the hold functionality.  That was also seen here: https://youtu.be/kLfcEqmhwCE?t=615

The video was released in October of 2017.  It's possible that the version of the firmware he's using precedes the version with the changes to the hold functionality.

[tautech]

KC, I've looked through this thread trying to find Hold images thinking I'd put some up but haven't sorry however last time I played with Hold it worked fine and displays up to 8 measurements in a list.
If I get a chance today I'll get a unit out and pre sales check it and grab some shots of the Hold function.

[tautech]

Teaser more to follow:
Hold mode, DCV, Filter ON, 10 minutes sitting idle with no probes connected.
~20 mins later zero further measurements recorded however when attaching probes of course triggers a measurement.

Second attachment is random measurements taken from a STB3 Siglent PCB.

[kcbrown]

Teaser more to follow:
Hold mode, DCV, Filter ON, 10 minutes sitting idle with no probes connected.

Well, for probe mode to be useful, it has to be "stable" in the face of connected probes, of course, since the use case here is that you're probing various parts of the circuit and you want the meter to recognize the difference between when the probes are performing a measurement and when the probes are just being held in-hand (with perhaps just one end connected to the circuit).

[tautech]

Another thing noticed with 3045X was Hold measurements were slow to be recorded in Slow speed measure mode but just fine in Middle speed measure mode.
Any further requests before I put my stickers on it, check FW is latest and rebox it ?

[kcbrown]

Another thing noticed with 3045X was Hold measurements were slow to be recorded in Slow speed measure mode but just fine in Middle speed measure mode.
Any further requests before I put my stickers on it, check FW is latest and rebox it ?

Thanks Rob!  Much appreciated.

I have one question about the hold measurements: did it end up recording measurements when you had the probes off the DUT?  It would be unexpected if it did.   I notice a couple of very small voltages that landed in the list, and I can't tell if those are true measurements or if they are false ones.  My concern here is false measurements polluting the list and, especially, pushing the current valid measurement to the list before I might get a chance to see it on the screen.

Also, how much time did you give it between measurements?

[tautech]

KC, low readings happened to be Gnd planes and measurements were taken just seconds apart like you would when in normal use. No plan and only poked at a few pads and when it beeped moved on to another one.

[kcbrown]

KC, low readings happened to be Gnd planes and measurements were taken just seconds apart like you would when in normal use. No plan and only poked at a few pads and when it beeped moved on to another one.

Awesome.  Thanks!

I went ahead and ordered an SDM3055.  Should be interesting to see how I like it, but from everything I've seen online, it should be very nice to use.

[MediaTech]

Did anything ever happen with a user cal procedure that a third party cal lab (other than transcat) could use for any of these meters?

Yes there is a user adjustment process available for SDM3055 and SDM3065X but it is only available on request and not for public release.
If owners can demonstrate they have adequate references to calibrate to/against the procedure is made available.

Hi Tautech,
you wrote:

Yes there is a user adjustment process available for SDM3055 and SDM3065X but it is only available on request and not for public release.
If owners can demonstrate they have adequate references to calibrate to/against the procedure is made available.

Can you please tell us to whom we have to send our request for the adjustment process?
Would be great to have just the right contact..
Thanks and best regards
Peter

[tautech]

Did anything ever happen with a user cal procedure that a third party cal lab (other than transcat) could use for any of these meters?

Yes there is a user adjustment process available for SDM3055 and SDM3065X but it is only available on request and not for public release.
If owners can demonstrate they have adequate references to calibrate to/against the procedure is made available.

Can you please tell us to whom we have to send our request for the adjustment process?
Would be great to have just the right contact..
Thanks and best regards
Peter

Welcome to the forum.

Latest firmware for 3055 and 3065X provided access to user adjustment but it is imperative that any reference standard used has been calibrated to high standards of accuracy otherwise your SDM will not meet its datasheet spec.
Any attempted adjustment must be carried out with great care after the SDM has been running for some time in a temp stable environment and using shielded cabling between any reference standard and the SDM.

Access to adjustment is in Utility>System Setup>Calibration although without a previously saved/uploaded ***cal.csv for the appropriate measurement type (VDC, VAC, ADC etc) will not allow selection of a new calibration adjustment to be loaded.
Therefore calibration adjustment tweak files must be first saved and then applied and this is initially done in Utility>Store Settings with a previously 'tweaked' Cal file in the correct format and file type.

The process while apparently complex is straightforward yet despite our pleas to Siglent they won't release it to the public domain.......yet.

Thanks MT as your post has served as a reminder to further discuss with Siglent to get easier access to User Cal procedures for owners of these bench meters that may have drifted out of spec.

[TheDefpom]

I actually think the calibration could be done in a MUCH better and simpler way than the existing method to make general user calibration easier (for those wondering, I have the required information to do the user cal as I was asked to test out the calibration procedure).

My suggestion is to employ the arrow buttons in a special calibration mode, you would inject for example, a DC voltage in the DC calibration mode, and press the arrow up and down buttons to trim the reading on screen until it matches the calibration source, this could be done for all of the required calibration points, that way a manual user calibration could be done very easily without the current additional steps or needing to use automated calibration setups.

[tautech]

I actually think the calibration could be done in a MUCH better and simpler way than the existing method to make general user calibration easier (for those wondering, I have the required information to do the user cal as I was asked to test out the calibration procedure).

My suggestion is to employ the arrow buttons in a special calibration mode, you would inject for example, a DC voltage in the DC calibration mode, and press the arrow up and down buttons to trim the reading on screen until it matches the calibration source, this could be done for all of the required calibration points, that way a manual user calibration could be done very easily without the current additional steps or needing to use automated calibration setups.

From high up a short while ago:
I think we should generate an app note to explain the (User Cal) feature.