XJW01 Auto LCR meter review ($120 bench top LCR meter)

[indman]

That looks rather good, well done. I think it would be better if they had also included the push-in terminals like the DE-5000 so you didn't need to have those leads laying across the bench.

Thank you, I agree with you. I would also like to note the very low quality of Chinese BNC connectors. Once I got this kit, my first desire was to get rid of these substandard BNCs and replace them with Canare BNCs.

I was looking at picking up the Hantek LCR meter. Both have 0.3% accuracy. The Hantek has a lot of frequency ranges up to 100K, so is there any advantage to buying this instead?

This meter has not very good reviews from the owners. A raw product, I do not recommend it for purchase.
If 100 kHz is important for you for measurements, then I advise you to pay attention to the time-tested DE-5000 or the new device from UNI-T UT622C

[floobydust]

I bought an XJW01 last year "PCB Rev. 1.1", "Circuit Rev. 3.1", "Date: 2017/03/28" the price is reasonable.
Calibration instructions took a while to translate, a lot of pain and actually the hardest aspect of the kit.

They come with a cheap op-amp U108 is either TL081 or NE5532. The meter does OK, not great, I cannot read relay coil inductance or low value WW resistors (inductance) for some reason. It's as if the noise floor is poor. It struggles at 7.8kHz too, the MCU is slow.

There are several enhancements to F/W and H/W that the West does not get. They went STM32 and AD9850 DDS but the designer got tired of his work being ripped off so I think he closed off the design. The Russian's competed with their "RLC Meter Digital" design, all using HP's original design.

[ResistorRob]

I was looking at picking up the Hantek LCR meter. Both have 0.3% accuracy. The Hantek has a lot of frequency ranges up to 100K, so is there any advantage to buying this instead?

This meter has not very good reviews from the owners. A raw product, I do not recommend it for purchase.
If 100 kHz is important for you for measurements, then I advise you to pay attention to the time-tested DE-5000 or the new device from UNI-T UT622C

I watched a couple reviews on Youtube covering the Hantek and they gave it really high marks. I looked into pretty heavy before, but don't recall what the pros and cons were. Thanks for the heads up on the UNI-T.  Somehow I missed that one. Seems new and almost no mention of it online. Checks all the boxes of what I'm looking for. That's definitely the one I'm going for. They charge a lot more for the "e" model which adds DSR, so will probably just order the "c" version.

[precaud]

There appears to be a new version of the XJW01, though it's not labeled as such, it is branded "Rod rain". Black case, 2-layer pcb with SMD components, built-in battery charger, and the BNC's are along the front edge.

https://www.ebay.com/itm/Digital-LCR-Bridge-Tester-Resistance-Inductance-Capacitance-Meter-ESR-w-Battery/284098649802

[indman]

https://www.ebay.com/itm/Digital-LCR-Bridge-Tester-Resistance-Inductance-Capacitance-Meter-ESR-w-Battery/284098649802

No, I don't like this model. A 2-line display(or small text on LED) instead of a 4-line display significantly impairs the usability of the device. In addition, the location of the BNC connectors assumes that the device will be in a horizontal position and a person will have to look into the display from above. It also complicates a comfortable measurement environment.

[exe]

https://www.ebay.com/itm/Digital-LCR-Bridge-Tester-Resistance-Inductance-Capacitance-Meter-ESR-w-Battery/284098649802

It goes only to 7.8KHZ

[precaud]

It goes only to 7.8KHZ

It's the same for them all!

No, I don't like this model. A 2-line display(or small text on LED) instead of a 4-line display significantly impairs the usability of the device. In addition, the location of the BNC connectors assumes that the device will be in a horizontal position and a person will have to look into the display from above. It also complicates a comfortable measurement environment.

Agreed on the display lines, but I like the BNC position, you can easily attach fixtures to it. And having batteries inside with charger is a plus.

My post is not to condone, only to inform...

[pizzigri]

I bought an XJW01 last year "PCB Rev. 1.1", "Circuit Rev. 3.1", "Date: 2017/03/28" the price is reasonable.
Calibration instructions took a while to translate, a lot of pain and actually the hardest aspect of the kit.

They come with a cheap op-amp U108 is either TL081 or NE5532. The meter does OK, not great, I cannot read relay coil inductance or low value WW resistors (inductance) for some reason. It's as if the noise floor is poor. It struggles at 7.8kHz too, the MCU is slow.

There are several enhancements to F/W and H/W that the West does not get. They went STM32 and AD9850 DDS but the designer got tired of his work being ripped off so I think he closed off the design. The Russian's competed with their "RLC Meter Digital" design, all using HP's original design.

Hello  Floobydust,
For some coincidence I have two of these meters, one I got as a gamble on Amazon as it was listed for something totally different for like 20 US dollars.... well in any case, i wanted to ask you if you think the device could be improved, both by changing the admittedly poor bnc connectors and maybe replace the op amp you mention above with a better part (which one?).
Also, I plan to change the Kelvin clips with something better- would the GWinstek type clips be an improvement to the performance of the LCR meter? I would have to modify them to use BNC rather that Banana clips though.
Best,
Franco

Eta mod the name I mistook a fellow member for another...

[parasole]

Quite an old topic, but let me answer you, who knows, maybe you still need it

Since I killed my DE5000, I was in an intense investigation for a DYI meter, not because it would be cheaper, but because I want full control and independence from a vendor who told me the repair will be more than a new meter...
Now thanks to this forum I think I found most of the available DIY meters, and this one seems to be simplest in the way it is executed, although all of them are more or less equal and use the same bridge architecture, just for this one I have everything in my hands including source code.

Possible improvements I see here are not that many, considering it is an entry-level unit, changing to a much better OPA has no reason, I am intending to use LF347 simple since I have many of them and they are a little bit improved version of TL082/84. Basically, for the front end, you need a high impedance input (FET input) and low voltage offset, so keeping stock TL082/84 or a little bit better OPA is OK. This meter has 7k8 highest test frequency hence even for G=9 as one stage is having, there is no problem with bandwidth. However, one point to look for is current sense ref resistors, having those at list 0.1% tolerance and low TCR (planning for 25ppm in my case) would help to increase the stability and calibration. One possible source of noise tests sine signal generation, maybe it would benefit from a better filter, but that requires more verification if that indeed is the case.

[mawyatt]

How did you destroy your DE-5000?

Did you attempt a repair, or was it a total loss?

Best,

[exe]

How did you destroy your DE-5000?

My guess is by connecting a charged capacitor.

[precaud]

FWIW, a couple weeks ago I borked my XJW01 by accidentally connecting a charged 'lytic. Fortunately, the input protection diodes did their job admirably. Replaced a couple shorted 1N4148's (easily accessible) and it was back in business.

As for the stability, at first I thought it may be a problem, but quickly realized that my expectations and use habits were the the issue. When I use my HP meters, I always give them time to warm up. But with the XJW01 (and other handheld meters) I tend to use them immediately after power-up, unrealistically expecting them to be fully accurate and stable. Fact is, it should be no surprise that there is significant drift (esp when measuring low impedances) until it has reached stable temps inside.

[mawyatt]

Agree, warmup is always required for accurate results, whatever the instrument is. However, the DE-5000 seems good right from Power On, even with the previous Open/Short Cal values, assuming of course that the same measurement adapter is utilized.

Think that most of the performance for the DE-5000 is attributed to the Cyrustek Chip set utilized, seems like a sophisticated mixed-signal chip set, altho this link doesn't show much about what's inside!!

http://www.cyrustek.com.tw/wp-content/uploads/ES51920.pdf

Best,

[coromonadalix]

if you respect the DE-5000  open and close calibrations,  never had bad surprises with it

[parasole]

How did you destroy your DE-5000?

My guess is by connecting a charged capacitor.

You guess it right
Stupid mistake after almost 10 years of use...

Did you attempt a repair, or was it a total loss?

Yes, wrote to both, Cyrustek and DER, first ignored my mail, and the second requested to send it for repair in exchange for 70USD plus postage... new one for me (as I am now in Japan) is about 90, so not a big deal to get one...DIY, in fact, will be probable twice more expensive as I plan to build at least two or even more variants and compare them...  The turning point was DER answer, I don't like the lack of control over the unit and no chance for repair by replacing the analog front-end chip, digital is still ok.
Embarked now on a DIY version and have a lot of fun learning about digital bridges and the math behind them 

Think that most of the performance for the DE-5000 is attributed to the Cyrustek Chip set utilized, seems like a sophisticated mixed-signal chipset, altho this link doesn't show much about what's inside!!

http://www.cyrustek.com.tw/wp-content/uploads/ES51920.pdf

After inspecting DE5000 front-end PCB circuitry, I am sure that the front-end chip is nothing more than a bunch of analog switches and OPA as you may find in any digital bridges out there, even I-sense ref resistors have the same values. The second chip is handling the display and all the math, again the same as any other DYI, one like Elector is more sophisticated while xjw01 or LCR2 Go is probably on the lower side.
Another aspect is they seem to have factory calibration as there is a serial memory chip installed, and that is a pro point for DE5000.

FWIW, a couple weeks ago I borked my XJW01 by accidentally connecting a charged 'lytic. Fortunately, the input protection diodes did their job admirably. Replaced a couple shorted 1N4148's (easily accessible) and it was back in business.

Exactly, and unfortunately, DE5000 luck any even this rudimentary type of protection...

if you respect the DE-5000  open and close calibrations,  never had bad surprises with it

I fully agree, it was a great meter, bad enough I destroyed it by stupid accident... On another side, if that would not happen, I would never learn about these devices' inner work 

[parasole]

Overall, I think I like the STM32 version better. I do wish it had the AFE of the 8051 version.

I suspect the STM32 version is newer.

Hi
not sure Danny will read this post so questioning the public
Regarding AFE, what would be the major benefit of using a dedicated diff-amp for V and I? To me, it looks like sharing a single diff-amp for both channels would be better as the same offset/bias will be applied... do I miss something? 

[sonpul]

I fully agree, it was a great meter, bad enough I destroyed it by stupid accident... On another side, if that would not happen, I would never learn about these devices' inner work 

Maybe it will help
https://www.ittsb.eu/forum/index.php?topic=714.0

[parasole]

Maybe it will help

Thanks, but in my case the stored energy was too high as AFE chip got fully shorted. I desoldered it and the meter may start, but then complain of an error and shutdown. I think the chipset datasheet is confidential for a reason, I am pretty confident by now that AFE side consists of a number of analog switches and diff PGA and probably would not be that difficult to substitute should we have a good chipset description...