Another fnirsi fail

[Doctorandus_P]

@ 11:10 the fnirsi DSO-TC3 measures an ESR of 5 Ohm with an elco with a 33 resistor in series.



Yet another example how bad these things are. On itself, a single measurement like this does not mean much, but fnirsi repeatedly makes equipment that delivers unreliable results. They are also appealing to beginners, who are likely to have difficulty in recognizing it is their "test equipment" that does not work properly instead of whatever project they are working on.

Don't be tempted by the reviews that show that fnirsi stuff is capable on showing plausible results in some cases. The problem is that it fails too often, which makes it unreliable and you're always second guessing this garbage.

[modoran]

Don't be tempted by the reviews that show that fnirsi stuff is capable on showing plausible results in some cases. The problem is that it fails too often, which makes it unreliable and you're always second guessing this garbage.

Yes, there are some "influencers" on youtube that don't know how to use the test equipment they get for free.   


Fnirsi has a long track record of lying to customers, why even bother with them ?

[pcprogrammer]

Eleshop sells it, so it must be good. 

Even the reviewer of elektormagazine thinks it is good value for money. 

[ArdWar]

Fnirsi has a long track record of lying to customers, why even bother with them ?

Lack of affordable alternative. As cheap as those Fnirsi junks are, they often still cost a week worth of average salary where I come from.
That's also why DT830 clones are extremely popular here.

[ataradov]

Both devices use guessing for ESR, since they don't do the actual testing at frequency. So, I would not say it is specifically Fnrsi fail, but fail of that widely used firmware and testing approach.

[KungFuJosh]

None of these devices should be used in place of a real LCR or proper DMM.

[nctnico]

Both devices use guessing for ESR, since they don't do the actual testing at frequency. So, I would not say it is specifically Fnrsi fail, but fail of that widely used firmware and testing approach.

I agree. The measurement is only wrong if it shows wrong values with an actual capacitor attached. Also, the 5.0 Ohms shown may be a limit of what the Fnirsi can measure. The number being exactly 5.0 is kind of a giveaway here. I don't think there are general purpose / commonly used capacitors with an ESR greater than 5 Ohms when they are new. So how realistic is it to put a 33 Ohm resistor in series to begin with? Putting a smaller value like 1 Ohm or 2.2 Ohm in series is much more realistic.

[flash2b]

Why did this guy put his DSO in roll mode?

He could have captured the start-up in normal acquisition model assuming he has some memory in his DSO. On roll mode a DSO is very, very, very slow and frequency pulse will not be shown.

[nctnico]

Why did this guy put his DSO in roll mode?

He could have captured the start-up in normal acquisition model assuming he has some memory in his DSO. On roll mode a DSO is very, very, very slow and frequency pulse will not be shown.

Unless you enable peak-detect (assuming this Fnirsi has this feature).

[ebourg]

Besides the mismeasured ESR, is this product good for testing components?

[flash2b]

Why did this guy put his DSO in roll mode?

He could have captured the start-up in normal acquisition model assuming he has some memory in his DSO. On roll mode a DSO is very, very, very slow and frequency pulse will not be shown.

Unless you enable peak-detect (assuming this Fnirsi has this feature).

In the video he uses a Siglent DSO.....

[ataradov]

Besides the mismeasured ESR, is this product good for testing components?

Yeah, it works fine. It is just a nice GUI and package over the same firmware. If you need any component tester, I'd pick this one over the old one where the battery is flopping unless you come up with a case of your own.

[ledtester]

The device on the right looks like a "LCR-TC4" which is an implementation of the open-source TransistorTester project:

https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/
https://github.com/Mikrocontroller-net/transistortester

This project has a long history (10+ years) and many implementations have appeared on Aliexpress and Ebay. It is quite safe to say that any "LCR" meter that features a ZIF socket is based on this project. A recent list of implementations offered for sale may be found in this post:

https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg4733444/#msg4733444

The design document for the project may be found here:

https://github.com/Mikrocontroller-net/transistortester/blob/master/Doku/trunk/pdftex/english/ttester.pdf

The explanation of how ESR is measured by these devices may be found starting on page 98. There are actually two proposed methods - the description of the second method starting on page 101.

Neither of these methods are the traditional way of measuring ESR and the document admits that.

Another thing to keep in mind is that the TT project was designed for the AVR microcontroller - in particular 5V operation. Recent implementations are starting to use alternative MCUs - like the LGT8F328 clone of the atmega328. A couple of years ago FNIRSI released the DSO-TC2 -- their first mash-up of a DSO138-like oscilloscope and a TransistorTester:

https://www.eevblog.com/forum/testgear/new-(toy)-dsolcr-meter-fnirsi-tc2-(june-2022)/

The DSO-TC2 they actually used two MCUs -- a MM32F103RET for the TransistorTester and a WCH CH32F103C8T6 presumably for the oscilloscope. According to the above EEVblog thread the DSO-TC2 didn't report ESR.

In any case, the TransistorTester on the right in the video is very likely using an AVR MCU -- like an atmega328 -- and the original TT code whereas the DSO-TC3 is using a different MCU and different code base.

[aeg]

Also, the 5.0 Ohms shown may be a limit of what the Fnirsi can measure. The number being exactly 5.0 is kind of a giveaway here. I don't think there are general purpose / commonly used capacitors with an ESR greater than 5 Ohms when they are new.

A new "low ESR" Nichicon UPW2E2R2MPD 2.2uF/250V capacitor measures 9.8 ohms on my Bob Parker meter.

[Paul T]

The FNIRSI seems to be way out of line.  In general, these devices aren't very accurate testers, they are convenient component identifiers & basic function checkers.  Great for binning lots of components quickly, or the first step for determining what a poorly marked component might be.  Later, when I want to use one of those binned parts, I go directly to a "real" LCR meter or DMM to accurately check the performance before making use of it.

On use, it's best to press the transistor test button three times to see how repeatable the result is.  The results I get are usually not reassuring, I often document the median value.  Sometimes inductors are identified as diodes, etc. which goes into a "test on something else" bin.  Also try putting the component in the receptacles different orientation and see what you get.

When it's working correctly, give 10% lee-way on resistance (accuracy range 10Ω-10MΩ) and capacitance (range 100pF-20MF) and 20% error on inductance (range 20uH-20H).  This is based on two LCR-T7's I have that I've evaluated across an array of reference components I have collected.

On these units I've added a foot pedal switch connection option and also a recessed slider power switch.  The only purpose of that power switch is to disconnect the battery when it freezes on testing.  This happens from time to time testing mosfet-looking components that might stump the microprocessor.

I also have several different external termination adapters: hooks, grabbers, spring terminal board, SMD plate made from an old SIM card with the little bond wires cut. 

One final recommendation: Use these devices off of the battery, not plugged in.  Keep away from sources of EMF during use, including things like LED bulbs and fans.