nanoranger: portable 8.5 decade ammeter with display micro-review

[exe]

Hello everyone!

TL;DR: link to the device along with specs, etc https://www.altonovus.com/nanoranger . Manual: https://irp-cdn.multiscreensite.com/1b24b2b0/files/uploaded/20210220-NR_Instruction_Manual-V2_NdwgCNJnTiica1f6MpY1.8-FINAL.pdf .



Since a long time ago I was eyeballing nanoranger, an 8.5 decade ammeter. It was showing up regularly when I googled for a (reasonably) cheap sensitive ammeter, but, to my surprise, I didn't find any reviews. Because of lacking reviews I hesitated to buy this one. But after a year of hesitation I made my mind

Look and feel. It feels pretty solid, A bit bigger that I expected, but not too big. It's as wide as my smartphone (oneplus 8pro) but shorter, so it's not gonna take too much desk space. It has rubber grips on the side and comes with rubber feet. My only complain is that battery compartment is very tight (it fits two AA batteries).

Teardown. There is warranty sticker that I didn't want to destroy, so... I asked the company to give me pics of internals . To my surprise they said that's possible. I'll post them as they give it to me. Update: see the post below for internals.

A tiny strange detail, I'd expect positive terminal to be on the right, and negative on the left, but it's the opposite. Although, now that I'm thinking about it, this way it might be easier and more intuitive to plug it into the circuit on high side (i.e. high-side current sensing).

Some specs and features. It has 9 ranges, 20na to 800mA. According to specs, it switches ranges within 30us, has a 12bit ADC and 0.6ksps which, with averaging, gives 16bit resolution internally. But it only displays 2500 counts max to suppress noise and 50/60Hz hum. It can be connected to PC, has an isolated usb port, supports up to 150 sps in fast sampling mode, supports SCPI commands,
Battery consumption (from specs): 15mA with backlight, 10mA without, but it's not very readable without backlight, imho. It has 10pA resolution at the lowest range. Comes with factory calibration.

Burden voltage: below 50mV. I guess that explains low resolution.

Documentation. I'd it's pretty comprehensive, check this out: https://irp.cdn-website.com/1b24b2b0/files/uploaded/NR_Instruction_Manual-V2.9-FINAL.pdf

So far I'd say it's nice little device. There are only two four things that I'd like to be different:
- it's uni-polar, it doesn't work with "negative" currents. Bipolar operation requires two nanorangers
- only 3.5 counts (ranges switch automatically at above 2150 counts and below 150 counts, over-range 2500 counts). But then again, burden voltage of 50mV probably is the limit here.
- statistics (min/max values, etc): can only be reset by power cycling or by SCPI. Not a big deal, but would nice to have Enter button clears stats
- one more physical button to navigate menu. I think power button could be used for that (long press would power off, short press would work as "menu up" or reset stats, or, may be double tap would do that).
- may be a trend chart, but I think PC could be used for that. I haven't checked the PC program yet.

Those "issues" are not a show-stopper for me, and probably a design choice. I still would like to see one more digit of resolution, even if it is noisy.

Price. It's a bit pricey for a hobbyist (175 pounds/200euro before taxes and shipment), but also understandable, considering how much effort it was put into it.

My conclusion after 5 mins: the device does essentially one function and does it well. It feels solid, it has professional firmware, there is a lot attention to detail. In a sense, the device feels a bit like a transistor tester: it's handy, convenient to use, doesn't take much space. I think I'll use it a lot.

It's impossible to capture everything in such a short review, so please check the manual and follow the thread if interested.

PS I'm going on vacation next week, so, unfortunately, I won't have much time to add to this review, but may be other owners of this little device will something here.



EDIT: typos, moar typos, removed irrelevant information
EDIT2: enter button clears stats, didn't know that

[exe]

I just got back from the company, and I was given a few pics, posting the sharpest one.

At first glance looks reasonably straightforward:
- top right corner: likely usb isolation (U11) and esd protection (U10)
- middle below terminals: range switching.  Three big mosfets and six smaller ones for range switching (nine total, matches number of ranges)
- mcu: ATSAML21G18B . I suspect they use ADC built-in into the mcu.
- U12 is probably a mux to control range switching.

I'm not sure what U7, U9 and U13 and I don't see the IC numbers. So, may be I'll crack my unit open at the end to see that

I was told there is not much on the underside, just a display and membrane keyboard.

PS serial number was redacted by me just in case.

[exe]

Small update: I unglued the warranty sticker to check the part numbers on ICs U7, U9 and U13. So:

- U7: markings 1A98 F26a , tsop-8 package. Idk what that is.
- U9: ADUM1286ARZ digital isolator
- U11 ft230xs FTDI usb-to-uart converter
- U12 TSZ124 quad auto-zero micropower opamp
- U13 TSZ122IST same, but dual opamp

Switching mosfets: marking is 1C125L, which googles as Nexperia's PSMN1R1-25YLC fet. It's a 1.15 mΩ 100 beast, feels a bit overkill for the device .

[Chance92]

Thanks for your review. I'm thinking about getting one for myself. I suppose you have been using it for more than a year now. What do you think of it?

[slugrustle]

My guess is that this is a feedback ammeter, hence the low burden voltage.

This looks really cool.  Once I can get the funds together, I plan to buy one of these, primarily for measuring leakage current in transistors.  I can't imagine anything else capable of doing that job at such a low price.

[Kleinstein]

Ideally this should be mixed: normal shunts for the high currents maybe down to a 1 mA or 100 µA range and than feedback (TIA) typ for the lowest currents.
The TIA type for the low currents keeps the voltage low and this way leakage from the switches (and maybe protection diodes - they may use the FETs for this too).
The higher current ranges need a low burden alone for keeping the heat down.

From the pictures it looks like some of the shunts do not use seprate sense and drive.  If the FETs are part of the measurement resistance it needs very low R_on fets to keep the temperature drift withint limits. It would still be an accuracy issue, so that more resolution would be of limited use. More ADC resolution could still help at the low current end to get away with less ranges / switching.

[exe]

Thanks for your review. I'm thinking about getting one for myself. I suppose you have been using it for more than a year now. What do you think of it?

I rather like it. I was building a power supply, and for me it was small helper to fine-tune constant-current mode. I didn't use it much, but when I needed it, it was convenient. To me it's a little helper like LCR tweezers.


Some random notes:
- 50mV burden voltage is max at full-scale current of the range. It's proportionally lower if current is lower
- maximum current it registers is 920mA (that exceeds 0.8A spec). It doesn't indicate when overloaded
- it's unipolar, and I think there is built-in diode that protects from reverse polarity. It will show zeros when connected this way, even if current flows. Voltage drop is 0.47-0.78V when changing reverse current from 1mA to 800mA.
- when measuring low currents care should be taken. Like, measurements are screwed if I use an SMPS. Or, in other words, I'm still figuring out how to measure currents below 1uA and what are low-current capabilities of the device.

[TimFox]

Ideally this should be mixed: normal shunts for the high currents maybe down to a 1 mA or 100 µA range and than feedback (TIA) typ for the lowest currents.
The TIA type for the low currents keeps the voltage low and this way leakage from the switches (and maybe protection diodes - they may use the FETs for this too).
The higher current ranges need a low burden alone for keeping the heat down.

From the pictures it looks like some of the shunts do not use seprate sense and drive.  If the FETs are part of the measurement resistance it needs very low R_on fets to keep the temperature drift withint limits. It would still be an accuracy issue, so that more resolution would be of limited use. More ADC resolution could still help at the low current end to get away with less ranges / switching.

Yes:  traditional Keithley analog electrometers had "fast" and "normal" modes for current measurement, where "normal" measured the voltage across a burden resistor and "fast" placed the resistor in the feedback path to obtain a much lower burden resistance.
They often forgot to mention on the front panel that "fast" mode could be used only for currents below 1 x 10^-5 A (for model 610C).

[Tangent_Tracker]

Sorry to reserect, is there a chinesium version of this yet? I know there is an auto-ranging version of Daves ucurrent measuring device, but they are all a little outside my budget atm.... I suppose I could make one though!