Siglent SDM3045X mA current measurement - weird behaviour

[ahope]

My Siglent 3045X is behaving odd when measuring  < 600mA.

I was trying to measure current consumption of an STM32F030K6 when I discovered some behaviour. Instead of the expected 3-6 mA current, it displayed about 10% of that - 5-800 uA.
I started to investigate. My two other multimeters showed the expected 3-6 mA, so I was starting to worry my siglent was totally out of spec.
Using my bench supply and a test setup with the 3045X, the current display on the bench supply and a handheld multimeter and a load I started testing from 3A down.
All meters were in agreement until I went below 600mA - then the 3045X would switch from the 6A scale to 600mA scale. This is where the really weird thing happens.
It will show the correct value, say 0.053 A, then switch to the 600mA scale and show 53.4 mA for about a second, then I hear a relay click and suddenly the display will be 6.58 mA.
Playing around and switching scale manually this is totally repeatable for small currents below a few hundred mA. The 600mA and lower scales will show values about 1/10 of actual current. The 6 and 10A scales are ok.

What is really freaking me out is that when manually changing scales down, it will show the correct value and scale for a second before a relay kicks in and ruins everything.

Anyone have any idea what is happening and how to fix?

The meter is a bit over a year old. A few months ago I blew the internal 10A fuse which I replaced myself. Otherwise the meter has been sitting on my bench since new.

[tautech]

My Siglent 3045X is behaving odd when measuring  < 600mA.

The meter is a bit over a year old.

Welcome to the forum.

Since then there has been a firmware update to V5.01.01.03.
Can you confirm the firmware version you're running ?

[ahope]

The firmware version was one of the first things I checked. I am running the latest  5.1.1.3. I was hoping for something more recent, its over a year old.

It could be firmware - in fact, it does look like it could be a firmware bug since it first displays the correct range and value before kicking in that relay and ruining everything.

But still looking for a resolution..

[tautech]

The firmware version was one of the first things I checked. I am running the latest  5.1.1.3. I was hoping for something more recent, its over a year old.

It could be firmware - in fact, it does look like it could be a firmware bug since it first displays the correct range and value before kicking in that relay and ruining everything.

But still looking for a resolution..

OK, I'll shoot your findings through to tech support. Might take a day or so for an answer.

[ahope]

Thanks. This one has me totally baffled.

[tautech]

After some discussion and trials in conjunction with tech support we have a suggestion for you to try.

PM sent.

[ahope]

I now tried full reset to factory, reinstall software (latest) to no avail.

With any current < 2-300 mA it will display the same behaviour.
-With manual scale it will  show correct current (checked with 2 other multimeters) on the 6A scale. Switching to 600mA scale or below it will show correct value for about 1 sec, then you can hear a "click" and it will display a value of about 12% of correct value.
-With automatic scaling this will happen when you drop below 600 mA, it will switch to 600mA scale, show the correct value for a second, then "click" and off she goes.

Without any good ideas next step is return to siglent or rip the darned thing open and troubleshoot myself.

[BillB]

I've also got a 3045X running v5.1.1.3.  I replicated the tests with a 3055 inline as well and experimented with both auto and manual ranging.  I couldn't get it to fail selecting the proper range, or the proper value when manually selecting another range. 

It does take that second to switch ranges.  I'm wondering it you are seeing the correct value in the old range (6A) before the click and the switch to the new range?

[tautech]

I've also got a 3045X running v5.1.1.3.  I replicated the tests with a 3055 inline as well and experimented with both auto and manual ranging.  I couldn't get it to fail selecting the proper range, or the proper value when manually selecting another range. 

Thanks heaps for checking this Bill. 

[tautech]

I have one more thing for ahope to try; this time a hidden reset that I will share by PM.

It should not produce a different result to the one sent previously but we're running out of suggestions. 

[ArthurDent]

I suspect that when the meter switches current ranges there may be different value current shunts switched in. Say on the 10A range it is .1 ohm, on the 600Ma range it is 1 ohm, and on the 60Ma range it is 10 ohms. (I just guessed, I don't know what your instrument has for range and shunt values). If you are putting a resistance in series with your power supply and your load, the voltage drop across the meter may be too high on the lower ranges to allow your circuit under test to receive the proper voltage. Different multimeters will have different values so one might work while another might not.

Use a second voltmeter and try reading the actual voltage the device is receiving after the it goes through the ammeter. Switch the current ranges and see what effect that has on the voltage.

[tautech]

The meter is a bit over a year old. A few months ago I blew the internal 10A fuse which I replaced myself. Otherwise the meter has been sitting on my bench since new.

I've heard of this once before, ie. the internal fuse opening before the one on the back panel.

Do you remember the style of internal fuse, eg. fast blow, glass or ceramic ?

What is the fuse in the rear panel ?

From the datasheet:
http://www.siglentamerica.com/wp-content/uploads/dlm_uploads/2017/10/SDM3045X_Datasheet_DM0545_E01A.pdf

Internal :12 A, 250 V slow-melt fuse
Rear panel : accessible 10 A, 250 V fast-melt fuse

Which makes me question why the internal fuse should open first. 
Could they be installed incorrectly ?

[Performa01]

If there was a probem with the fuses, there would be no reading at all. More specifically, we would not see a correct reading in the Amp ranges.

The problem has to be with the shunt resistors:
0.01 Ohm for the Amp ranges (6A, 10A)
0.1 Ohm for mA ranges (60mA, 600mA)
1 Ohm for the lowest ranges (600µA, 6mA)

@ahope, have you tested the 6mA range and does it fail as well?

If just the 60mA and 600mA ranges are off, then something might be wrong with the 0.1 Ohm shunt resistor.
If the 6mA range fails as well, then I would suspect a problem with the relays for the range switching and/or some serious contamination (lump of solder?) on the PCB.

I do not think this is a calibration issue because the error is so massive.

[tautech]

OK so ahope tried what I sent him with zero result. 

So where we are behind the scenes in discussions with Siglent.
This has been escalated to the product manager and technicians and they are checking possible scenarios.

There is suspicion of an analog board problem, yes of course and I have some ideas of just what.
Now it's Sat at the factory so we won't get a reply until Mon (your Sun).

2 immediate possibilities in my mind; damage to relay contacts or one of the shunts.
Thoughtful testing should easily confirm.

One suspects the over-current/blown fuse event caused this...........but it shouldn't have ! 
The 10A FB rear panel fuse should open first......not the 12A SB on the PCB. 



ahope, make another post or 2 to get your count above 5 so full member privileges become available.

[ahope]

Hi guys,

Thanks for all the suggestions.

I ran through some more tests now, and the error is persistent from about 250mA down.

If I set it to 600mA and then gradually reduce the current it will show correct values until i go below approx 230mA. Then the relay clicks, display is reduced with about 88% and will stay that way all the way down to microAmps.

I am quite sure its a relay kicking in another shunt resistor that causes this. Question is what will cause the relay to trigger at the wrong time/current?

Still working on this, will have a look inside tomorrow.

[tautech]

Just try and find/identify the problem......relay contacts/operation or damaged shunt.
It will be good for the Siglent engineers and other members to know what it is so we know if to repair or it's a warranty claim.

[ahope]

Did some more investigation.

First, took the lid off and did a visual inspection.
Then I tried to locate which relay was kicking in. It seems like all of them are kicking in simoultaneously, but at least the big white in the read box is kicking in.

After looking around I found one resistor that looked bad - look inside the red box.

In the next picture I have a close-up. Its  R408 marked "SMT F0705 R010".
From the picture it looks like a burn-mark on one side and you can see it is bulging up in the middle.
Resistor R407 next to it looks like it may be a little non-flat, but not so easy to see.

Both these resistors are close to the big white relay, and next to that is the 12A fuse that blew a while ago.

Could this be the culprit?

[ahope]

Another thingvthat bothers me: why did the internal fuse blow and not the external? and why didi it not protect the internals?

[Performa01]

Nothing wrong with R408. It is the 0.01 Ohm shunt that works flawlessly in the amps range. That "burn mark" at the side comes from some milling for trimming its resistance value - I'm actually baffled that Siglent goes to the effort to do this on a low cost instrument like this.

But the photos also show that my assumption was wrong and there are only two shunt resistors, just as stated (not very clearly) in the datasheet. The only other shunt resistor is R407 at 1 Ohm.
 
According to the data sheet, this 1 Ohm resistor would be active for the 600mA range, but the very same data sheet also states a burden voltage of <0.5V in the 600mA range, which cannot be true because 600mA * 1 Ohm = 0.6V (plus voltage drop across cabling, fuse and connectors).

Anyway, R407 looks good as well and also from the latest symptom description it is unlikely that the shunt is at a fault. If the meter works correctly in the 600mA range down to some 230mA, then the shunt resistor should be okay.

The real problem is the relay(s) switching below 230mA - whichever they are, in any case throwing off the measurement.

[bson]

Verify that it can measure low voltages correctly, like maybe 5-10mV down to whatever you can produce reliably (50µV maybe).

[ahope]

First,

Thanks for the suggestion to vheck volt measurement. I set up a simple voltage divider and verified it with two hand-held DMMs.
The 3045X measured the same values (+/- a few % since it should be way more accurate than the cheap DMMs). The values I tested were 5mV and 0.5mV.

The meter would work normal on both auto-range and manual range.  The measurements were made on the 600mV range.

So the voltage side of the 3045X seems to be operating normal and within spec.

Back to the R408. From what I can see it is clearly bulging up. There is no smooth top surface on it, more like it has been "blown" up. Is this normal? You should be able to see it if you study the pictures.

[tautech]

Hmmm, yes it looks like that shunt has been hurt. 

Hunting around I found a couple of images of 3045X and 3055 and enlarged (that didn't work. Grrr) them......attached.

[HKJ]

The shunt has been damaged, but if the meter measures correctly it is not that serious yet, but maybe the reading will change with time when air has access to the shunt element.

[tautech]

The shunt has been damaged, but if the meter measures correctly it is not that serious yet, but maybe the reading will change with time when air has access to the shunt element.

Your one is not trimmed. 

Thanks for the pic. 

[HKJ]

Your one is not trimmed. 

I do not believe in laser trimming in this meter, electronic calibration is much easier and cheaper.

[ahope]

Thanks for the pics Tautech. It is clear that my R407 is damaged.
From your picture it looks like a 1R shunt next to the 0R010 R407 shunt.
I am guessing, but going from the 6A scale to the 0600mA scale it will switch from using R408 to R407 as shunt.
So if R407 is damaged, the shunt value will be off, giving the wrong reading?

I am tempted to try to replace R407. Anyone know exactly what spec/type part I should get from digikey or other distributor?

[tautech]

Thanks for the pics Tautech. It is clear that my R407 is damaged.
From your picture it looks like a 1R shunt next to the 0R010 R407 shunt.
I am guessing, but going from the 6A scale to the 0600mA scale it will switch from using R408 to R407 as shunt.
So if R407 is damaged, the shunt value will be off, giving the wrong reading?

I am tempted to try to replace R407. Anyone know exactly what spec/type part I should get from digikey or other distributor?

First you need to identify the SMD size and then of course to maintain accuracy get a low tolerance/low tempco device.
Here's a chart so you can compare physical measurement:



But hold off on ordering anything just yet until we hear what the factory says.....just in case they are willing to send you a replacement PCB.

[ahope]

thanks a lot. 

i will take good time before replacing this shunt. its a precision part, so I want to make sure I make a good replacement if I have to.

Rough measurements puts it at 6.5x3.5mm or thereabouts. The «F» indicates a 1% tolerance part.  But I have no idea if it is a 1w, 2w or something else. 

[tautech]

thanks a lot. 

i will take good time before replacing this shunt. its a precision part, so I want to make sure I make a good replacement if I have to.

Rough measurements puts it at 6.5x3.5mm or thereabouts.

= SMD 2312 (imp) or 6032 (metric)

[nctnico]

Perhaps it is also a good idea to put two big diodes anti-parallel to the shunt. That is the recommended way to prevent an overcurrent (surge) from damaging a shunt resistor. Also looking at the size of the resistor it doesn't look like it has been properly sized for handling 10A continuously (= 1 Watt of heat dissipation). The resistor may be rated for 1W (or even more) but those ratings are always in ideal circumstances and running the resistor at the absolute maximum temperature. Derating 3 to 4 times is not a bad idea. A through hole current sense resistor made of thick wire is much more forgiving when it comes to abuse compared to an SMT resistor (AFAIK there are SMT current sensing resistors made entirely from metal but these aren't cheap).

[rf-loop]

Perhaps it is also a good idea to put two big diodes anti-parallel to the shunt. That is the recommended way to prevent an overcurrent (surge) from damaging a shunt resistor. Also looking at the size of the resistor it doesn't look like it has been properly sized for handling 10A continuously (= 1 Watt of heat dissipation). The resistor may be rated for 1W (or even more) but those ratings are always in ideal circumstances and running the resistor at the absolute maximum temperature. Derating 3 to 4 times is not a bad idea. A through hole current sense resistor made of thick wire is much more forgiving when it comes to abuse compared to an SMT resistor (AFAIK there are SMT current sensing resistors made entirely from metal but these aren't cheap).

Of course it can not handle 10A continuously. No one have made it for continuous 10A and no one have promised it can handle 10 continuously. So why this straw man argument?

Specifications for SDM3045X:
Current:
30 seconds OFF after 30 seconds ON is recommend for the continuous current that higher than DC 7 A or AC RMS 7 A.

Also in based to image image and other info it looks like 2W but also this same size some are rated for 1W. measuring from image without good reference for dimensions is just guessing.  But then @ahope: "Rough measurements puts it at 6.5x3.5mm or thereabouts." So perhaps size is 6432 metric size (2512).

Now if we look Siglent recommendation in data sheet and use 10A. 30sec on 30sec off. (1W 30sec and then cooling 30sec) average 500mW.   Then if look continuous 7A. It means 490mW. 

[nctnico]

Perhaps it is also a good idea to put two big diodes anti-parallel to the shunt. That is the recommended way to prevent an overcurrent (surge) from damaging a shunt resistor. Also looking at the size of the resistor it doesn't look like it has been properly sized for handling 10A continuously (= 1 Watt of heat dissipation). The resistor may be rated for 1W (or even more) but those ratings are always in ideal circumstances and running the resistor at the absolute maximum temperature. Derating 3 to 4 times is not a bad idea. A through hole current sense resistor made of thick wire is much more forgiving when it comes to abuse compared to an SMT resistor (AFAIK there are SMT current sensing resistors made entirely from metal but these aren't cheap).

Of course it can not handle 10A continuously. No one have made it for continuous 10A and no one have promised it can handle 10 continuously. So why this straw man argument?

Specifications for SDM3045X:
Current:
30 seconds OFF after 30 seconds ON is recommend for the continuous current that higher than DC 7 A or AC RMS 7 A.

Also in based to image image and other info it looks like 2W but also this same size some are rated for 1W. measuring from image without good reference for dimensions is just guessing.  But then @ahope: "Rough measurements puts it at 6.5x3.5mm or thereabouts." So perhaps size is 6432 metric size (2512).

Now if we look Siglent recommendation in data sheet and use 10A. 30sec on 30sec off. (1W 30sec and then cooling 30sec) average 500mW.   Then if look continuous 7A. It means 490mW.

Well it says 10A on the spec sheet I pulled from a website quickly. And it also says 10A on the front of the SDM3055 DMM without a time limit (at least on the pictures I've been able to find). Usually when there is a time limit this is also printed on the front. Either way adding the two diodes is still a good suggestion because it would have prevented the damage seen by the OP. The SDM3055 isn't a particulary cheap instrument to begin with anyway so some protection against user errors would be nice.

[tautech]

RTFM !

[HKJ]

Perhaps it is also a good idea to put two big diodes anti-parallel to the shunt. That is the recommended way to prevent an overcurrent (surge) from damaging a shunt resistor.

You mean something like this (Multimeters often use a bridge rectifier to get 2 or 3 diodes in series):

[nctnico]

Perhaps it is also a good idea to put two big diodes anti-parallel to the shunt. That is the recommended way to prevent an overcurrent (surge) from damaging a shunt resistor.

You mean something like this (Multimeters often use a bridge rectifier to get 2 or 3 diodes in series):

This looks more like a circuit to protect the input circuitry against an open current shunt and not to protect the shunt itself. But if the shunt resistor is already operating very near it's design limits then it won't work ofcourse.

[HKJ]

You mean something like this (Multimeters often use a bridge rectifier to get 2 or 3 diodes in series):

This looks more like a circuit to protect the input circuitry against an open current shunt and not to protect the shunt itself. But if the shunt resistor is already operating very near it's design limits then it won't work ofcourse.

Just about any multimeter uses a diode bridge across the uA and mA shunt, usual there is only fuse protection for the 10A shunt.

This meter also has another trick to protect the uA and mA ranges: It will auto range to higher range when it is overloaded, also when auto range is disabled.

[nctnico]

Just about any multimeter uses a diode bridge across the uA and mA shunt, usual there is only fuse protection for the 10A shunt.

This meter also has another trick to protect the uA and mA ranges: It will auto range to higher range when it is overloaded, also when auto range is disabled.

Either way the current sensing circuit shouldn't be damaged by an overcurrent. Or put differently: the current sensing circuit should survive until the fuse blows (which in the case of the Siglent 3045X DMM means the resistor is underrated / insufficiently protected).

[rf-loop]

Another thingvthat bothers me: why did the internal fuse blow and not the external? and why didi it not protect the internals?

Can you explain bit more this situation when this overload happen when Fuse blow. Also was it first time when any fuse blow in its history or have you previously changed back panel fuse. I mean, is is so that both fuses was original fuses after factory.

Then can you also explain exactly what happen in this case when this overload happen. What kind of current source. Its voltage and what kind of very short time peak current it can produce or some info for somehow estimate it.

Back panel Fuse is 10A and internal fuse is 12A.  (I do not know what kind of speed class they are and need be, but my opinion is that least back panel Fuse need be fast(F) or very fast (FF))

Because it is weird that internal blow first, really weird. It must not happen. Si it is good to know bit more about it.



Also I want bit criticize towards Siglent.

10A fuse in place what can not last continuous 10A load. Newer seen fuses characteristics (I2t, arcing time, clearing time)?
For enough protection it need be very special 10A fuse if it need open before damage in all circumstances including also small long time over current or fast very high short circuit peak. Also even if it is aka "ultra fast"  fuse (or what ever factory have installed (right or accidentally wrong)  and designed)  normal user without enough education and shared information may change wrong type fuse. Ampere and dimensions is not enough information.

For example, if a current of 20A was flowing through a fuse rated at 10A, a very fast acting fuse might open in 0.1 seconds, a fast acting fuse may open in one second, whilst a slow acting fuse may take tens of seconds to open.

FF   Very Fast Acting Fuse (Flink Flink)
F   Fast Acting Fuse (Flink)
M   Medium Acting Fuse (Mitteltrage)
T   Slow Acting Fuse  (Trage)
TT   Very Slow Acting Fuse  (Trage Trage)

[BillB]

RTFM !

To be fair, I looked over both the Quick Start Guides and the User's Manuals for the SDM3045X and SDM3055 and nowhere is the 7A/30sec recommendation to be found.  Only in the Datasheets are there tiny little footnotes describing this.  Something this important should probably be mentioned in those documents at least, and since no one really reads manuals any more, it might have been nice to have that on the front or back panel.

Honestly, knowing that there was a 10A fuse protecting the 10A range, I assumed 10A measurement was probably not a good idea.  But I certainly didn't remember the 7A/30sec recommendation until I went back through and found the footnote.

[ahope]

A little background on the blown fuse.

The instrument was almost unused, bought new a couple of montgs before.
I was probing inside a chassis and accidentally touched the 220v and poof! it blew. quite a bang actually, those fuses dont go quietly,

The internal fuse is a 12A slow ceramic, the rear fuse a fast 10A. makes absolutley no sense that the internal should blow before the rear fuse.

This is the one and only fuse I have had to replace on this instrument.

[tautech]

The internal fuse is a 12A slow ceramic, the rear fuse a fast 10A. makes absolutley no sense that the internal should blow before the rear fuse.

This is the one and only fuse I have had to replace on this instrument.

No sense at all !

As mentioned earlier I've seen this before in SDM3065X.
Questions to the factory have been asked.

[nctnico]

A little background on the blown fuse.

The instrument was almost unused, bought new a couple of montgs before.
I was probing inside a chassis and accidentally touched the 220v and poof! it blew. quite a bang actually, those fuses dont go quietly,

The internal fuse is a 12A slow ceramic, the rear fuse a fast 10A. makes absolutley no sense that the internal should blow before the rear fuse.

This is the one and only fuse I have had to replace on this instrument.

If you compare the graphs of the fuses you'll probably see the time to open is identical so it is not a surprise they both blew. In mains installations the rule is to go two sizes up when putting fuses in series to get proper selectivity. Over here common values are 4,6,10,16,20,25,30 and 35A. If you have a 10A fuse then the upstream fuse should be at least 20A to make sure only the 10A fuse blows. Siglent probably put the 12A fuse on the board to make sure there is some safety when the user replaced the fuse with a piece of wire. Unfortunately 12A is way to close to 10A for this trick to work. A better way would be to install a user replaceable 7A fuse to make sure the internal 12A fuse doesn't blow. This likely doesn't solve the problem with the current sense resistor becoming damaged though. A fuse isn't the answer to protecting components. A fuse is there to prevent fire.

For example: the VC8145 bench DMMs have a 13A fuse even though they are rated for 20A for 15s (the latter is also printed on the front of the VC8145). Remember a fuse is designed to carry the rated current forever so it will only blow when there is a significantly larger current flowing through it.

[tautech]

Any observant hobbyist has witnessed the different characteristics between FB and SB fuses even of the same A rating and IME 12A SB (on PCB) and 10A FB (rear panel) would indeed be sufficient to protect an instrument and give correct cascading of fuse failure/opening.

But here this has not happened (2x IME) so tests must be made to discover why.

Could the fuses be installed incorrectly or the fuse rating characteristics be called into question ?

Careful tests need be made to discover if fuse behavior is 'as designed' and if it isn't, could this have contributed to shunt damage.

Has the factory changed fuse supplier.....there are many things yet to be discovered.

[nctnico]

Any observant hobbyist has witnessed the different characteristics between FB and SB fuses even of the same A rating and IME 12A SB (on PCB) and 10A FB (rear panel) would indeed be sufficient to protect an instrument and give correct cascading of fuse failure/opening.

That depends entirely on the current and voltage. Compare Littlefuse 216 and 218 series for example. The fast acting opens within a second at 25A and the slow one at 30A. All in all the difference is much smaller than you'd expect. And at mains voltages there is also the effect of arcing which doesn't interrupt the circuit right away. There is a good reason why mains installations require a large margin between downstream/upstream fuses! There is no sense in denying that.

[rf-loop]

A little background on the blown fuse.

The instrument was almost unused, bought new a couple of montgs before.
I was probing inside a chassis and accidentally touched the 220v and poof! it blew. quite a bang actually, those fuses dont go quietly,

The internal fuse is a 12A slow ceramic, the rear fuse a fast 10A. makes absolutley no sense that the internal should blow before the rear fuse.

This is the one and only fuse I have had to replace on this instrument.

Do you have still available data how these fuses was marked...  All markings what can find in one fuse end or how ever they have marked voltage, current and speed class. Example (F 10A 250V) or how ever they are.

Also, it need note that some glass fuses 5x20mm have really very low interrupt rating what may reach quite easy in total short circuit (depending source impedande)

Then over current - time - break curves can be really unpredictable, there may be really wide tolerances. In one manufacture lot one 10A fuse may break after 10s 15A but and 1s 25A and 0.1s 60A and 0.01s perhaps 200A and 0.001s perhaps 1000A. And then, min and max curves may differ really lot.
Of course these depends fuse type. But fast or ultra fast and slow fuses have not very big difference in some circumstances in practice even if I2t is many fold.
Some 5x20 glass fuse may have as low as 35A - 100A interrupt rating. (means that if current is more big it can not break in worst cases, example Littelfuse type 218   <4A and example 10A  have 100A interrupt rating and some other type have 10kA)

If really want build, using fuses, even some kind of circuit (passive) component protection it really need do lot of study and tests and very extremely carefully and clever select fuses. Mostly no one do it seriously. And it can understand because result is still poor or just borderline And still not for usual semiconductors.
Fuse is, as nctnico told, more like preventing fire etc.  (as told, of course also some amount circuit protection. But as can see if really look fuses true data... 10A fuse make still over 1000A short peak possible.

Why internal 12A blow and not external 10A.  There may be several and also natural reason. One reason may be fuses tolerances in over current time curves. It can be very possible that if short circuit current is example 50A first blow 12A and after it blow 10A is still ok condition if both are not in they tolerances middle ones.. if other is from other side and other is from other side of normal distribution it is very possible even if external 10A is F(ast) and internal 12A fuse is T(race)

Even if there is example under 10A fuse (yes it can and 10A not continuous measurement is possible)
There can be  still very high maximum short circuit current (if source can feed it). In this game it do not mean lot if short maximum  is reduced by fuse to 200A instead of 250A or how ever it is)

Here one tiny example about time current curve (average - and it mean that every fuse have min and max curve and area between these are big.)

But then, here is just ONE example about one type min and max curves. All fuses have different these curves. This may give some imagine how tolerances may affect and give one possible explanation also for this case. As can see differences in min and max is huge.

[tautech]

SDM3065X
Rear panel fuse: 5x20 glass LF (Littlefuse) marking: T10AL250VP 

Internal PCB fuse: 5x20 HRC LF (Littlefuse) marking: T12A250VP

[rf-loop]

SDM3065X
Rear panel fuse: 5x20 glass LF (Littlefuse) marking: 10AL250VP  As per datasheet spec this should be FB fuse.
Internal PCB fuse: Later when more time available............ but already we know it's ceramic.

"10AL250VP"

Is it possible there is installed wrong fuses?
(Need ask because China and fuses - I do not wonder anything. I have seen there with my own eyes so much...)

Where is speed identifier? It need be before current.

Here is Littelfuse 5x20 cap marking

Littelfuse product catalog and design guide, small fuse "bible"

[rf-loop]

A little background on the blown fuse.

The instrument was almost unused, bought new a couple of montgs before.
I was probing inside a chassis and accidentally touched the 220v and poof! it blew. quite a bang actually, those fuses dont go quietly,

The internal fuse is a 12A slow ceramic, the rear fuse a fast 10A. makes absolutley no sense that the internal should blow before the rear fuse.

This is the one and only fuse I have had to replace on this instrument.

Do you still have this original rear panel 10A (not blown) fuse what was there installed at this time when internal 12A fuse blow.

[ahope]

Do you still have this original rear panel 10A (not blown) fuse what was there installed at this time when internal 12A fuse blow.

Yes. I have checked it and it is indeed a 10A fast glass fuse

[tautech]

SDM3065X
Rear panel fuse: 5x20 glass LF (Littlefuse) marking: 10AL250VP  As per datasheet spec this should be FB fuse.
Internal PCB fuse: Later when more time available............ but already we know it's ceramic.

"10AL250VP"

Is it possible there is installed wrong fuses?
(Need ask because China and fuses - I do not wonder anything. I have seen there with my own eyes so much...)

Where is speed identifier? It need be before current.

Here is Littelfuse 5x20 cap marking

Littelfuse product catalog and design guide, small fuse "bible"

Rear panel fuse # corrected, it does indeed have T but maybe F or FF would be best.

[tautech]

Do you still have this original rear panel 10A (not blown) fuse what was there installed at this time when internal 12A fuse blow.

Yes. I have checked it and it is indeed a 10A fast glass fuse

Can you please get the codes for the internal fuse ? .........I'll check my 3065X

See reply #45

[rf-loop]

Do you still have this original rear panel 10A (not blown) fuse what was there installed at this time when internal 12A fuse blow.

Yes. I have checked it and it is indeed a 10A fast glass fuse

You have checked. How?
Do you  believe it is fast or do it have marking what it IS fast and how fast.
If there is mark for speed class, what it is. (ref Tautech next msgs)

[rf-loop]

SDM3065X
Rear panel fuse: 5x20 glass LF (Littlefuse) marking: 10AL250VP  As per datasheet spec this should be FB fuse.
Internal PCB fuse: Later when more time available............ but already we know it's ceramic.

"10AL250VP"

Is it possible there is installed wrong fuses?
(Need ask because China and fuses - I do not wonder anything. I have seen there with my own eyes so much...)

Where is speed identifier? It need be before current.

Here is Littelfuse 5x20 cap marking

Littelfuse product catalog and design guide, small fuse "bible"

Rear panel fuse # corrected, it does indeed have T but maybe F or FF would be best.

Rear panel fuse really have speed class marking T ?

Oh what a total mess.

Normally T = Slow Acting (Trage)

Not TT, not T, not M.... Only acceptable is F or FF and really from trusted manufacturer.

[Qw3rtzuiop]

At my SDM3045X the internal Fuse is a ceramic LF T12AH250VP. The rear one is marked LF T10L250V.

[HKJ]

I have just published a review of this meter: https://lygte-info.dk/review/DMMSiglent%20SDM3045X%20UK.html
Including a tear down.

[rf-loop]

At my SDM3045X the internal Fuse is a ceramic LF T12AH250VP. The rear one is marked LF T10L250V.

I think that reason why there was a T speed class 10A fuse, it need investigate and I believe it is now under investigation.
Personally, I also do not understand why a 10A fuse was used there. I think the fuse selection is doubtful. Including speed and its current rate.

[nctnico]

If Siglent is smart they fit a 7A fuse and fix current sensing circuitry so it cannot be damaged before the fuse blows.

[ahope]

Hi all,

issue is now resolved.  Siglent provided new boards and everything now checks out fine and within specs again.  Verified against accurate voltage sources and the original test circuit where i discovered the problem in the first place.

One last thing: I ordered brand new 10A FAST fuses for the rear fuse panel! 
thanks to siglent for great service.

So to all you out there - I am still a big fan of Siglent. based on this experience i will continue to buy more instruments from them.  I am very happy with my 3045X and my 1202xe.


/andreas

[Kleinstein]

If Siglent is smart they fit a 7A fuse and fix current sensing circuitry so it cannot be damaged before the fuse blows.

7 A fuses are not a common size, but there are 6.3 A and 8 A available.

So maybe downgrade the specs to 6 A and use a 6.3 A fuse in the back.

[al777]

SDM3065X
Rear panel fuse: 5x20 glass LF (Littlefuse) marking: 10AL250VP  As per datasheet spec this should be FB fuse.
Internal PCB fuse: Later when more time available............ but already we know it's ceramic.

"10AL250VP"

Is it possible there is installed wrong fuses?
(Need ask because China and fuses - I do not wonder anything. I have seen there with my own eyes so much...)

Where is speed identifier? It need be before current.

Here is Littelfuse 5x20 cap marking

Littelfuse product catalog and design guide, small fuse "bible"

Rear panel fuse # corrected, it does indeed have T but maybe F or FF would be best.

Sorry folks... I know it's a 5-year old topic... I'm making a zombie I guess... :-)... But just so that the new info gets indexed by Google and someone like myself will save time on this topic in the future, here is what I've found.

First of all, on the new SDM3065X (non-SC, factory calibration is March 2023) the rear panel fuse is a glass-body one, marked as "T10AL250VP", with LittelFuse ("LF") logo, it's resistance (if anyone cares) is 4.575 milliohms, as measured with the RC3563 milliohm meter. There are 5 agency approval logos on the other cap.  My best guess is: it's the LittelFuse Series 218 5x20mm glass body time-lag fuse.

Second, I wanted to buy spares, just in case as incidents do happen - and I thought Fast Blow series would potentially be better (that would be LittelFuse Series 217). What I've found by reading datasheets is some new knowledge I wanted to share here.

The Time-Lag Series 218 outpaces the Fast Blow Series 217 for overload currents of 210% and only gets behind above that overload level.

218 series (slow) @ 210%: 120 seconds maximum; while 217 series (fast) @ 210%: 30 minutes maximum.

So, for DMM overload protection use, especially if there is an internal Time-Lag fuse for 12A - use of the external Time-Lag series fuse is well justified. Otherwise the "fast blowing" external fuse may allow the internal fuse to open first, at currents between 10A and 20A. Certainly, if you try to measure voltage of a low internal resistance source with DCA input :-) - then yeah, fast blow will be faster, but at 100's of amperes the damage may still happen... to both fuses and the DMM itself... Besides, slow characteristics may help in real life work, when your load (DUT) is capacitive for example... may save you few bucks on fuse replacements due to inrush current spikes...

So, in my opinion - the right replacement external fuse for SDM3065X is LittelFuse Series 218, glass body Time-Lag T10AL250V, here is the datasheet link:
https://www.littelfuse.com/~/media/electronics/datasheets/fuses/littelfuse_fuse_218_datasheet.pdf.pdf  (double .pdf at the end is not a typo).
They now go to my next Digikey order's list, $2.63CAD qty 1.

BTW, I've checked LittelFuse's ceramic body Time-Lag fuses as well - interestingly, while they do have way higher interruption capacity (not often a necessity for a bench meter) - they do what the label says, they lag behind, in a "mild overload" zone as well. So, Siglent engineers did choose the right fuse - that's my view of course :-). I'm not associated with Siglent in any way :-)!

Hope this will help someone to save time looking for the right replacement fuse for SDM30x5X DMMs. I'll now leave this old topic alone :-). All the best, folks!

[KungFuJosh]

I think the model number for the 10A fuse you're referring to from LittleFuse is 0218010.HXP.

[al777]

I think the model number for the 10A fuse you're referring to from LittleFuse is 0218010.HXP.

Thanks, KungFuJosh!

...or 0218010.TXP or 0218010.MXP... I think those are packaging variations, same fuse. Few cents difference between them, I usually add to the basket those which have the highest stock at the moment, to make sure my near-future order will not be delayed :-).