Rigol DM3058e Bug - Relay Banging in DC Current Measurement

[kandrey89]

Hi,

So I've had Rigol DM3058e for just a little while, purchased it to measure photodiode current from 10uA to 10mA.

I discovered a bug that would cause the range switching relay to bang back and forth incessantly and contacted Rigol. After making a video of the issue, sending in my unit on their dime, and RMA guy verifying that yes he sees this same issue on their office unit, he talked to engineering and they said it's not a BUG, it's meant to work that way.

Well here's how to reproduce it, please try it on DM30XX and report if you also see it.

Set Rigol DMM into DC Current measurement mode, Auto Range, Filter ON.
Provide constant current of 2.0mA
Increase current by 0.1mA
At 2.4mA you should start hearing relay banging and the current measurement will fail 
Continue to increase current by 0.1mA
At 3.8mA the relay banging will stop and the DMM will read the value properly again.
This cannot be reproduced if your Filter is set to OFF or if you're not using Auto Ranging!
What this means is that 2.4-3.8mA cannot be measured in this configuration, making this equipment unusable when you need to use both 2mA and 20mA ranges in filtered and auto ranging mode because you don't know what the current is going to be and require averaged precision measurements.

So, would you buy a DMM that had a broken Auto Ranging and the manufacturer was telling you it's not a bug? 

Link to video: https://youtu.be/6Uf7gT3Mml0

[2N3055]

It is not a bug..

How do you provide current? You use constant current source or maybe voltage source with resistor in series?

[kandrey89]

It is not a bug..

How do you provide current? You use constant current source or maybe voltage source with resistor in series?

Watch the video, constant current source Keithley unit.

Oh really, so I should just be OK with not being able to measure part of the range? How would you like an instrument that advertises 0.001V-300V range, except 2-5V doesn't work.

[2N3055]

OK video is not very clear as to what are you doing exactly..
Auto-ranging is always annoying when measuring something right smack in the middle of switchover area between the ranges..
Filter introduces time delay that creates oscillation..

You are right, Rigol obviously chose not enough hysteresis in switchover so comparator oscillates, and didn't account for filter behavior..
They could fix this with firmware update,  I agree, it can be considered bug.

But to say you can't measure with it, or that you are missing measurement range is funny. I would do that measurement with manually fixed range anyways, to avoid any range switching...

[kandrey89]

If you want to measure using your arse, that's up to you, but I prefer to measure properly and as painlessly as possible.
My photodiode range of values does not allow me to choose one range or the other, 2mA vs 20mA. The DMM was bought for automatic measurements. The way it works now is unusable and not practical to use manual ranging. I need the filter to smooth out measurement values for precision.

Gosh, you're like a baby, spoon feed you everything, and you keep your mouth closed not eating, how about your read the video description or the post and see the how to reproduce steps and assume that's what I am doing.

Rigol RMA guy said it's hardware, cannot be fixed in firmware.

Auto ranging is annoying? Maybe, when doing automatic measurements and some measurements take 3 seconds instead of 1 second to perform, but since it's automatic I don't much care as long as I get valid data. Keithley 6485 picoammeter has these things implemented very well, so it's not impossible for Rigol to do the same.

[2N3055]

Ok now let's start with "Yo' mama..." jokes...

What's with all the rage.. 
What's next: You are so pissed at them, that you'll put commando knife in your teeth and parachute on Rigol factory in China, Rambo style, and seek your revenge "with them Commies engineers" ...  LOL

I said you're right and it's a bug..
Rigol might as well made auto-ranging with the hardware and not software and in that case it is serious engineering oversight, and cannot be fixed with software, but I doubt that.. It probably can be.

And auto-ranging IS annoying sometimes, and it creates visible artifacts in measurement data... Also, there is difference in burden voltage, resolution, offset and gain calibration between ranges... That complicates interpreting data.....

But that all depends what and how are you doing...
If instrument in current state is not good for you and can't be fixed than return it and get one that does work for your purpose...

[Dwaine]

I have the Rigol DM3068 and it does not have that filter option on DC current.  Only AC for voltage and current.

I blame the Kiethley 2400.  It does not have any blackfin magic. 

[Damianos]

Even if the DM3058 is a stupidly implemented instrument, it seems that you have not done a proper device choice for the task!

You stated that you want to make precise measurements, BUT:
- You are changing ranges and expecting what? They have different specifications!
- You are expecting "unlimited" dynamic range, by using auto-range in a 5 digit instrument? Consider of using one with 7 or 8 digits!
- The two ranges, that you try, use different methods to measure the current. The lower uses a current to voltage converter and the higher uses the usual shunt resistor. This has consequences ...
- The mentioned filter creates additional errors and does not improve the precision...
- For DC measurements, the general advisement is to keep the signal clean from any disturbances. If the signal is fluctuating by itself consider of using a digitizer.

The instrument that you use is not an exception, it has a datasheet and something that is called "User Guide". In them, even in "rigolish" language, you can find some information of how this thing work. Try it!

However if you play more with this instrument you will discover that it has a lot of annoying behavior as have other instruments, even if they have more than twice the cost of this...

[kandrey89]

- Unlike some, I read the specification and have chosen the instrument appropriately.
Out of all the models available for <$1800 for current measurement instruments, DM3058 is the only one that can measure 1nA of current, which is as low as I need to go for my purpose, because it has 200uA range with 5 digits, all the others have 5 or 6 digits but have higher minimum range thus actually not being able to measure lower currents or cost a lot more. Even Keithley DMM has a resolution of something like 100nA even though it's 6 digits because it's min range is like 200mA.
- Why do you assume things and then put them in my mouth? Where do I say I expect unlimited dynamic range? I checked the instrument burden voltage before ordering, and it's acceptable to me. Calibrated instrument should measure the same value in either range provided it has sufficient precision, but I agree, the same value to a point, I'm not expecting 10A range to give me current down to nA that's what the ranges are for, but if I'm close to the range transition of the 2 neighboring ranges it's OK. I did have dedicated instruments for photodiode measurement before that went out of calibration, were old and probably not very good and had ranges with different offsets, resulting in jumps when ranges would switch.
- Where did you find such information about which range uses which method to measure current? The air moves, the water flows, it has consequences...
- I tested the filter with noisy input, it does it's job fine for me.
- Thank you, I know how to RTFM and I did before buying!

Even if the DM3058 is a stupidly implemented instrument, it seems that you have not done a proper device choice for the task!

You stated that you want to make precise measurements, BUT:
- You are changing ranges and expecting what? They have different specifications!
- You are expecting "unlimited" dynamic range, by using auto-range in a 5 digit instrument? Consider of using one with 7 or 8 digits!
- The two ranges, that you try, use different methods to measure the current. The lower uses a current to voltage converter and the higher uses the usual shunt resistor. This has consequences ...
- The mentioned filter creates additional errors and does not improve the precision...
- For DC measurements, the general advisement is to keep the signal clean from any disturbances. If the signal is fluctuating by itself consider of using a digitizer.

The instrument that you use is not an exception, it has a datasheet and something that is called "User Guide". In them, even in "rigolish" language, you can find some information of how this thing work. Try it!

However if you play more with this instrument you will discover that it has a lot of annoying behavior as have other instruments, even if they have more than twice the cost of this...

[Damianos]

...
- Why do you assume things and then put them in my mouth? Where do I say I expect unlimited dynamic range?

Read the entire phrase and not cut it in independent parts... I am saying that you are trying that by changing ranges.

- Where did you find such information about which range uses which method to measure current? The air moves, the water flows, it has consequences...

RTFM. For the higher ranges they say what shunt resistors are used, except for the two lower that they refer the input offset voltage. There was a document describing it, but I don't know where it is buried!

- I tested the filter with noisy input, it does it's job fine for me.

This is the problem! Have in mind that the filter may reset the measurement and, because it is at the edge of the range, it starts ... dancing! Auto-range and stupid filters, in cheap instruments, are not ... friends!

The solution is to not using this filter, clean up the signal before it arrives to the instrument...
OR
Make your software and use the instrument as the "detector" (I did that about two years ago with the help of the "command expert" from Agilent and a spreadsheet program). Doing this you will discover another stupid thing: they don't round but truncate the last digit!!!