AN8008 US $19, 9999count, 1uV, 0.01uA, 0.01Ohm, 1pF resolution meter

[Titan 4380]

I received one of these meters about a week ago and feel rather indifferent about it. The first thing that hits you is that it is tiny. If you are used to holding a chunky Fluke in your hand, this thing feels minuscule .Dave's video on this meter is bang on. For the money it is a great meter with surprisingly accurate DC measurements, but I definitely won't be sticking any of its probes into a wall socket. The square wave generator setting does a great job and produces very nice patterns on my 2000 series scope right up to 5000khz with no overshoot or distortion. At the moment Aldi is selling a 6000 count meter that does a ton of things. Even for $25, I haven't done any rudimentary tests on the thing, but I can't see it being too impressive.

[Fungus]

I received one of these meters about a week ago and feel rather indifferent about it. The first thing that hits you is that it is tiny. If you are used to holding a chunky Fluke in your hand, this thing feels minuscule .Dave's video on this meter is bang on.

Take a look at the ANENG DT850B+. It's as big as a Fluke and has all the features that are missing on the AN8008 for only $20.


(but not real input protection...)

[evava]

AN8008 problem -  bias current on mV range is too big:

When meter in mV range, connected 10.1MOhm resistor causes 0.101mV reading - non of my other meters does that to such an extent - it means input bias current is about 10pA!
Smaller resistors cause proportionally smaller value on display and vice versa.
Can be used for measuring resistors up to 9999MOhms on 99.99mV range, thats nice  

Can someone check, whether it is just my meter this bad/nice?

[electronic_eel]

Temperature - I enabled in EEPROM on the mV function, works °C/°F and factory CJC CAL is already set at 25°C fixed.

Cold Junction Temperature fixed to 25°C?

Does that mean there is no thermistor or other means in the ic to measure the temperature inside the multimeter to have a somewhat proper cold junction temp? Or is it just that you calibrate the proper cjc at 25°C?

[floobydust]

AN8008 problem -  bias current on mV range is too big:

When meter in mV range, connected 10.1MOhm resistor causes 0.101mV reading - non of my other meters does that to such an extent - it means input bias current is about 10pA!
Smaller resistors cause proportionally smaller value on display and vice versa.
Can be used for measuring resistors up to 9999MOhms on 99.99mV range, thats nice  

Can someone check, whether it is just my meter this bad/nice?

I also measure 0.100mV across a 10MEG resistor, for 10pA or 0.01nA leakage current. Second try it hones in on -0.135mV. Using no leads and Faraday shield.
Johnson-Nyquist noise of a 10MEG resistor is significant here. Try add a disc cap in parallel

[floobydust]

Temperature - I enabled in EEPROM on the mV function, works °C/°F and factory CJC CAL is already set at 25°C fixed.

Cold Junction Temperature fixed to 25°C?

Does that mean there is no thermistor or other means in the ic to measure the temperature inside the multimeter to have a somewhat proper cold junction temp? Or is it just that you calibrate the proper cjc at 25°C?

The DMM IC has an internal temperature sensor (diode?), but needs to be calibrated.
The EEPROM comes with default values and I'm not sure how accurate they are:
0BH, 0CH: 00FAH: Ambient temperature (25.0°C)
0DH~0FH: 03BE00H: Ambient temperature Default ADC value

A one-point CAL is supported (offset?), you basically tell the meter what the ambient temperature is (25C).

[Kalvin]

AN8008 problem -  bias current on mV range is too big:

When meter in mV range, connected 10.1MOhm resistor causes 0.101mV reading - non of my other meters does that to such an extent - it means input bias current is about 10pA!
Smaller resistors cause proportionally smaller value on display and vice versa.
Can be used for measuring resistors up to 9999MOhms on 99.99mV range, thats nice  

Can someone check, whether it is just my meter this bad/nice?

I tried with my two AN8008 meters with 10 Mohm and 1 Mohm resistors in mVDC range. The meters were too sensitive to how I held the probes in order to make any definitive conclusions.

However, when I first short-circuit the probes (display shows 0uV) and then let the probes hang in the air freely, the display shows increasing negative offset voltage. With 10 Mohm resistor I can get fluctuating readings around -125 uV ... -150 uV and with 1 Mohm around -10 uV ... -18 uV. I repeated the tests with my other AN8008 and obtained about similar results: Open leads increasing negative offset voltage, 10 Mohm resistor giving around -130 uV, 1 Mohm resistors giving around -10uV ... -15 uV.

So, I can repeat your findings. Whether or not this is an issue - it depends on what you are measuring. If you are measuring a circuit with an impedance of 100 kohm or less, this will not be an issue. However, if you are measuring very high impedance circuits (with impedance > 100 kohm) and want to get accurate uV measurements this may be an issue.

[amspire]

AN8008 problem -  bias current on mV range is too big:

When meter in mV range, connected 10.1MOhm resistor causes 0.101mV reading - non of my other meters does that to such an extent - it means input bias current is about 10pA!
Smaller resistors cause proportionally smaller value on display and vice versa.
Can be used for measuring resistors up to 9999MOhms on 99.99mV range, thats nice  

Can someone check, whether it is just my meter this bad/nice?

I also measure 0.100mV across a 10MEG resistor, for 10pA or 0.01nA leakage current. Second try it hones in on -0.135mV. Using no leads and Faraday shield.
Johnson-Nyquist noise of a 10MEG resistor is significant here. Try add a disc cap in parallel

There can also be the issue with a big resistor that the AC pickup is saturating the DC amplifier causing an apparent voltage. A better way to test for offset current is probably to use a capacitor such as a 10nF capacitor and measure the time constant. At 10pA, you should get -1mV/second voltage.

I tested my Richmeter 101 (Aneng AN8001) and got -18mV over 60 seconds with a 360pF capacitor. That corresponds to 360pF x -0.018/60 = -0.1pA offset current on the mV range.  I put the capacitor directly in the multimeter socket to eliminate any additional effect of the leads and to minimize AC pickup.

Edit: looks like there may be a capacitance inside the meter and 360pF is not enough. Definitely need a much bigger capacitor. Repeated the test with a good quality 2.68uF capacitor. The meter shows 0000 until the count gets to 0005 - I think that quirk has been mentioned before. After 60 seconds, the voltage was -0.18mV. That is a current of 2.68uF x 0.00018 / 60 = 8pA.

[amspire]

Repeated my test with an AN8001 type meter over 5 minutes and the result was -5.7pA. Looks like the mV range input impedence is pretty high - well over 1G ohm. Probably hard to measure with the offset current present.

[floobydust]

You have to be careful measuring uV at high-Z (10MEG)
Many effects you would not normally notice come to surface, such as contact potential and thermal EMF from dissimilar metals, AC hum and RFI.
You need a shielded, grounded work area. You need symmetric metallurgy on the leads and plugs etc. otherwise you get a net thermocouple.
The Metrology boyz surely have more guidance 

Connect your scope (10:1 10MEG) to the multimeter with both on lowest mV scale.
Notice you can see on the scope the DMM analog input circuit pulses, I about 1mVpp at 3.3kHz
The DMM IC has CAZ chopper before the input buffer op-amps, so I guess this is making some noise that shows up at 10MEG ohm. Can't see it at 1MEG though.
Adding 820-1,000pF eliminates most of it; with 10nF it is all gone. The readings shift maybe 10uV adding the cap.

So some of the displayed uV offset is resistor noise, DMM IC leakage current and CAZ noise, with high source resistance.
It is not entirely picoamps of leakage.

[evava]

Yes, I almost forgot: when you dont understand whats going on, look at it with scope 

Thank you all for explanation!

P.S. no such behavior with my UT61e - maybe it has something with high input impedance (>3GOhm supposedly on 220mV range, so 10MOhm is negligible probably)

[JohnPen]

I tried the 10M test with my 8008 with the following results.

10 M carbon film resistor on mV range and resistor connected directly between the sockets on the 8008. The meter was freestanding well away from any other equipment.

Initially reading cycled between -108 mV and -111mV after a few minutes it cycled around -114/116 mV.

Interestingly if one held the resistor's connection to the COM socket between finger and thumb the reading dropped to ~74 mV.

Heating just the resistor a few degrees, as expected, increased the mV reading quite rapidly 200mV plus.

None of this really matters too much for normal everyday use of course.  Still a very useful, versatile meter for low voltage/current electronics work.  Especially for the price.

John

[coromonadalix]

There is a new model look a like : SURPEER AV4     it is a 20000 counts model ? maybe worth checking ? sure it is not a 20$ meter ...

on Ebay: http://www.ebay.com/itm/digital-multimeter-true-rms-4-1-2-voltmeter-with-auto-range-multi-meter/112444656128?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D555018%26algo%3DPL.SIM%26ao%3D2%26asc%3D41376%26meid%3Dbdba934ac7de4c11bf5d7231a7c8164c%26pid%3D100005%26rk%3D4%26rkt%3D6%26sd%3D152355178736

[ebastler]

There is a new model look a like : SURPEER AV4     it is a 20000 counts model ? maybe worth checking ? sure it is not a 20$ meter ...

Not an AN8008 competitor, due to the price and the larger form factor. But it looks somewhat promising -- might even have proper fuses?
Although the 600V CAT IV rating and the claim that it is "FDA certified" do not instill confidence... 

[bitseeker]

Although the 600V CAT IV rating and the claim that it is "FDA certified" do not instill confidence... 

FDA certified! Never saw that before. Apparently, True RMS by Amazon. Excellent. Covering as many bases in one product as possible.

[psycho0815]

Not sure if anyone cares but fyi i just received a Tougs M103 from amazon germany, from a german seller for 23€ delivered and it appears to be the exact same meter. Just in case you live in the eu and don't wan't to go through all the china hassle.

[plazma]

Not sure if anyone cares but fyi i just received a Tougs M103 from amazon germany, from a german seller for 23€ delivered and it appears to be the exact same meter. Just in case you live in the eu and don't wan't to go through all the china hassle.

With a black case. Nice

[Fungus]

My meter finally turned up. 31 days after purchase.

[MacMeter]

I ordered one on Amazon, knowing it would be slow from China. But this was a 1st for me, kind of a time buying scam, I assume they lied about it being in stock. I'm pasting the sellers response when after tracking the package from China, I received a tiny envelope with a "free" shrink wrap sleeve for a remote control, worth about 10 cents. Along with a business card saying the real package would still be coming. I then put in for a refund, and Amazon lets the seller contact me via email. See if this response is logical!

"We sent it out on Jul 17, 2017, normally it would take 10-25 workdays from China to US,
As the item is electric product,  it would take more days to complete the customs clearance.
we are afraid it could cause a delay, so a free case for remote controls goes to you first and the real package is after. Normally you can receive the item you oredered within 12 working days, would you please wait for it?"

So I then said, just give me a tracking number for the REAL meter, a few days later, they send it, only shows as "label created". So I'm assuming the meter was NOT in stock, so they send something to buy time, and then ship it when it's in stock.

Just thought I'd share this odd story, another reason I mostly avoid buying directly from China. Hopefully, in another 2-3 weeks it will arrive.

[floobydust]

I wonder if that 8uV gap around zero (meter doesnt show values +-1uV +-2uV +-3uV +-4uV, instead of it only shows 0 or +-5uV or more) could be eliminated or at least reduced?
Is that set in EEPROM, or is it invariably written in firmware?
Is it possible to find out?

Otherwise it seems to be stable enough, I see no need to hide these small values (+-1uV +-2uV +-3uV +-4uV).

I see that deadband on the uV range centered around zero. Readings start at ±5uV up. Input from -4 to +4uV reads zero.
This would be done in firmware to mask noise.  With this meter, to read uV that low, I think a preamp is needed.  In other words, AN8008 gives 1uV resolution at 4uV up.
DMM's don't have specs for their reading deadbands and hysteresis.

The other "stickiness" people are seeing in readings I think are due to the averaging and filtering of A/D readings done by firmware - within the product spec of  ±3 counts.
This got better after I did the capacitor mods on Vss and Vref to lower H/W noise.

Other uV effects are thermocouple EMF with test leads and probably AC hum and noise.

[JXL]

I ordered one on Amazon, knowing it would be slow from China. But this was a 1st for me, kind of a time buying scam, I assume they lied about it being in stock. I'm pasting the sellers response when after tracking the package from China, I received a tiny envelope with a "free" shrink wrap sleeve for a remote control, worth about 10 cents. Along with a business card saying the real package would still be coming. I then put in for a refund, and Amazon lets the seller contact me via email. See if this response is logical!

I ordered mine from eBay and got the same shrink wrap sleeve and a similar message.  My guess is that they are collecting orders so they can do their volume buy to get volume discounts.  Anyway, I only bought the meter out of curiosity and to reduce wear and tear on my "better" meters. 

[RedSky]

As far as the AN8008 readings around µV are concerned...

I have a Uni-T UT70D (80,000 count) and a Brymen 867s (up to 500,000 count) both can read down to a resolution of 1 µV
Both have trouble at around the 0 mark at either staying on 0 when leads are shorted or reading accurately in the -10 to +10 µV range, no matter the source of the measurement.   When the reading goes up substantially they read very close to one another.
Someone will probably get their Brymen out now and try it and it will show a perfect 0 when leads are shorted and yes that happens to me a fair amount but not always, sometimes it might just count up or down very slowly as well or show an off 0 measurement.  Both meters do the same thing and they use different chipsets.

Basically if two several hundred dollar meters have trouble at near 0 µV this AN8008 is still doing very well..   Not that i have mine yet, ordered a month ago now, still waiting doh!
As others have already noted, it's probably either the averaging algorithm used or an attempt to hide some noise near zero in firmware.   Virtually no chance this thing has a 13+bit ADC in it... yet it's trying to do 10,000 resolution has to be a fair amount of averaging going on.

[amspire]

It sounds like many multimeter customers want the meter to read 0.000mV if the leads are shorted. The manufacturers do not want to have to explain offset voltages and thermal voltages to complaining customers I guess.

It would be great if there is a way to turn this dead zone off.

[plazma]

I added RANGE and REL buttons to my AN8002. Let see if button modification is possible with the AN8008.

[tronde]

I added RANGE and REL buttons to my AN8002. Let see if button modification is possible with the AN8008.

What kind of switches have used? Isn't the rubber buttons to soft for normal push buttons?