Fnirsi SG-003A & SG-004A Signal Generators

[black6host]

@Phil1977

That's what I thought, industrial control.  Not having worked in that environment I'm not familiar with it.  The youtube video in reply #12 was great for showing what it did and accuracy but not what it might be used for.  Thank you for answering!  Cheers!

[LastManStanding]

Is it this behavior?
(Attachment Link)
(Attachment Link)

I never cared to much about this. Around 4mV P2P at 5V output voltage is less than 0.1% relative and less than 0.02% over the output voltage range.

No, I don't mean noise. It's a slow voltage drop of 2mV (within ~23 seconds) followed by a rapid increase of 2mV (1 second). This results in a sawtooth profile in the recording. Additionally, the device heats up quite quickly, causing an additional deviation of +1.5mV/+1°C. This makes it hardly usable as a reference, and the values are outside the specifications (0.02% +0.003). The manufacturer did not respond to my inquiries, which is unfortunately the usual approach for many Far Eastern manufacturers. Produce, sell, forget... But this is also a reason for the low prices.

I returned the device and received a refund. I assumed that this phenomenon was a defect in my SG-004a. For that reason, the money wasn't worth it to me.

[Phil1977]

I think I know this "feature": Once I tried to calibrate the SG04 itself - as far as I remember in current mode. The current was slowly *increasing* and then jumped back a few uA. I think the SG04 is drifting quite a lot with temperature but it has an internal compensation based on its NTC. It seems like the readout of the NTC isn't smooth, so that the temperature compensation is done in steps and leads to this "sawtooth" profile.

In my case this behaviour was still in the tolerance range. Anyhow, such a €100,- device should never be seen as a rock stable absolute reference. For precise calibrations I´d always additionally connect a trustable DMM. But for checking what happens if e.g. if a sensor runs away the SG04 is really nice.