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

[floobydust]

My meter came in from AliExpress ANENG store. Price has gone up since I ordered it now USD $24.18 and supposedly 29% off the regular $USD 34.

Took it apart and did the usual ritual IPA + Q-tip clean of flux, slag, solder balls. I resoldered the fuse clips, they had no solder on the top layer. Overall the build quality is good and seems a new level for china. Next I'll run it.

Replacement fuses in North America... I have 10A for current and 200mA for input.
Littelfuse 3.6x10mm fuses and Bussmann/Eaton, Schurter SPT 3.67x10; This size is 250VAC interrupt-rated...
But all have pigtails 
So either never blow the fuse, or I'll try the 20mm mod. instead of waiting for fuses from china.

[deflicted]

Well, after watching Daves video I just.... bought another one 

I'd be interested to see how well they compare. With everyone buying these and putting them through various tests, it's still somewhat of an open question whether the results are repeatable from one unit to the next.

Are you saying Dave just got lucky? On all ranges?

Not at all. I'm saying maybe Dave just got a generally high quality unit. You wouldn't expect that a meter that's dead on at 5V would be WAY off at 5mV, although I suppose it's possible. But with such a cheap meter, I wouldn't be at all surprised if there's some significant variation in quality from one unit to the next. Maybe he gets one AN8008 that measures 5.000V with his voltage reference set to 5V, but then he orders a second AN8008 and it reads 4.986V. Would that really surprise you for a $19 meter? Even Fluke, Agilent, etc, don't have perfect production processes, QC, etc. So maybe one out of every 10,000 Fluke units is not up to snuff, or something like that. What's your guess as to what Aneng's percentage is? That's all I was getting at. A sample size greater than N=1 (using the same set of references, verified with the same high quality bench meter, etc) would be interesting to see.

Edit: Given that one horribly bad solder joint that Dave showed in the video, I don't see why one would expect consistent high quality on every AN8008 unit.

[sleemanj]

But all have pigtails 
So either never blow the fuse, or I'll try the 20mm mod. instead of waiting for fuses from china.

I bought replacements from RS with pigtails, flushcutters fixed that.

[Mark Hennessy]

Does anyone know anything about the internal switch? On the AN8002, it's range.

it was tried in a utube review, it is fully funtional manual range override.
more interesting question is why it's there - is it used in the calibration process??

Yes, it's a range button on the AN8002 - as I say. I tried it myself months ago.

However, what about the AN8008? Nothing happens when you short it. It's also connected to one of the test pads (PT1.1, which is the same in the AN8002) - so yes, perhaps it is part of the calibration procedure? Looking at the DTM0660 data sheet, you need 3 keys to operate the calibration procedure, and I guess it seems likely that the IC is in the AN8008 works in a similar way...

The datasheet states that "PT1.1" is pin 23 of the IC, which is described as "Data input/output, SPI communication port (SCE), OTP read/write port (PSDI), interrupt source (INITI)".

[fish2eggs]

Greetings all,
has anybody tested the burden voltage on the low current ranges?

[EEVblog]

However, what about the AN8008? Nothing happens when you short it.

Not that I could see.

[ocw]

has anybody tested the burden voltage on the low current ranges?

The burden voltage is dependent on the current. 

The shunt resistance of the AN8008 was 100.561 ohms when the meter was in the uA position and 0.0301 ohms when the meter was in the mA/A position.

[ocw]

As somewhat referred to previously, the AN8008 is a poor meter to use between 1 - 100 mA.  I just completed my detailed accuracy verification of the lower current readings.  For an approximately 1 mA current reading I had the choice of 989.8 uA (1.01% low due to the 100.561 meter insertion resistance) or 1.0 mA.  Yes, that's two digits, 10% error represented by the least significant digit.  That's the best resolution possible just above 1 mA.  The -1% uA error continued down to around 200 uA.  When the reading was corrected for the added resistance, the uA accuracy was typically below 0.1%.

The DC voltage reading accuracy was typically 0.06%  The resistance reading accuracy was typically below 0.3%.  Most of my resistors used for accuracy verification were rated at 0.01% accuracy.

More tests tomorrow.

[floobydust]

Experimenting with DCV using a homemade precision DC reference, I find the readings are a bit quirky.
-ve input always reads a few counts less than flipping the leads.

I see the (long-term) averaging algorithm has hysteresis of at least +/- one count, and you can get it to "stick" either way. (If you've ever written code to average A/D readings, you know how hard it is to react to fast changes and also filter out noise).

The meter seems to jump to a reading based on the first (fast) voltage sample, then after 2-4 seconds the long-term average adjusts the reading. Since the long-term average has hysteresis, it may or may not correct. It seems to depend on the difference between the samples, or if the sign changed.

What I am saying is if the meter reads 5.000V right away, it stays there.
If the meter reads 4.997V initially, it may slowly move to 4.998V or 4.999V.
If the meter reads 4.964V initially, after 2 seconds it corrects to 4.998V

You can try making glitches- connect/disconnect the leads quickly and notice the readings are inconsistent by about 3 counts. Not a big deal, I 'm used to getting the same number regardless.

[matura713]

what I don't understand - considering any DMM based on DTM0660L chipset can be hacked just with changing few bytes in the EEPROM from 6000 to 9999 counts and that is widely known, is there any real difference in accuracy between AN8008 (ZT109) and for example AN8002 (ZT102) hacked to 9999 counts? or hacked HoldPeak HP890CN (the OEM revision of HoldPeak HP-770HC)?

[sleemanj]

what I don't understand - considering any DMM based on DTM0660L chipset can be hacked just with changing few bytes in the EEPROM from 6000 to 9999 counts and that is widely known, is there any real difference in accuracy between AN8008 (ZT109) and for example AN8002 (ZT102) hacked to 9999 counts? or hacked HoldPeak HP890CN (the OEM revision of HoldPeak HP-770HC)?

For those not aware of the hack that matura713 is referring to, which is for the AN8002 ZT102, and it's various other BM0660L based meters, the translated datasheet here http://www.kerrywong.com/2016/04/03/dtm0660-datasheet-translated/ details the EEPROM data amongst other stuff, which happens to include the # of counts.  This post http://kazus.ru/forums/showpost.php?p=1130725&postcount=1874 has a link to a (java) program which supposedly allows changing the eeprom settings in an easy manner, I have not tried it (err, run random executable linked from Russian website... might do that in a virtual machine) so I have no idea how it works.

Note that this is for the DTM0660 based meters, this new AN8008 is supposedly DM1106EN based I think, Google hasn't revealed a datasheet for it.

[matura713]

Note that this is for the DTM0660 based meters, this new AN8008 is supposedly DM1106EN based I think, Google hasn't revealed a datasheet for it.

yes, but isn't the chipset/hardware difference only that "square wave output" feature on AN8008 (ZT109)? and everything else is exactly the same? I mean (AN8002 + EEPROM hack to 9999 counts) is equal to (AN8008 - digital square wave feature) or there is real difference in accuracy? (BTW, on first glance even both PCBs of AN8008 and AN8002 looks very very similar).

in any way, I am sure soon people will release EEPROM dumps from AN8008 and there are such from AN8002 for a long time, as well everyone having any of those DMMs can make EEPROM dumps very easy.

in any way, what I would be interested to see is comparison with any of the previous DTM0660-based meters hacked to 9999 counts and AN8008, because for example for only several USD more you can get UT139C, hack it to 9999 counts and that has adequate quality and input protection as well.

This post http://kazus.ru/forums/showpost.php?p=1130725&postcount=1874 has a link to a (java) program which supposedly allows changing the eeprom settings in an easy manner, I have not tried it (err, run random executable linked from Russian website... might do that in a virtual machine) so I have no idea how it works.

you can always use hex-editor and modify your EEPROM dump manually, but in general I guess for most people here it's ultra simple hack: dump the EEPROM of your DTM0660 based DMM, modify it, write it back to the EEPROM chip in your DMM.

[sleemanj]

in any way, I am sure soon people will release EEPROM dumps from AN8008

http://kazus.ru/forums/showpost.php?p=1126917&postcount=1827

[matura713]

in any way, I am sure soon people will release EEPROM dumps from AN8008

http://kazus.ru/forums/showpost.php?p=1126917&postcount=1827

haha, yeah, the Russians already did it and Google translates the comment in Russain there to "that this is a complete analog DTM0660, well or some advanced version."

which is exactly my point and suspicion, either the same or maybe the same + square wave generator. So, for me the only question - is there any real difference in accuracy compared to DTM0660 hacked to 9999 or not. Because if not, then there are much better DMM for basically the same money than AN8008.

[EDIT] I Google translate more from the Russian forum: what I got is adding temperature hack to AN8008 is the same as for DTM0660, i.e. the same bytes in the EEPROM.

also, a note was made there that EEPROM of DMM model "BSIDE ADM08" has in its EEPROM data for the mode setting that are not consistent with and described in the datasheet, but i believe "BSIDE ADM08" is DTM0660-based and they mentioned that to make a point DTM0660 datasheet that is floating around is simply not complete.

[kalel]

I'm also interested in any differences between a modded AN8002 and this. I also don't know if the lower models that look the same on the outside perhaps also use the chip (e.g. AN8000/8004).

in any way, I am sure soon people will release EEPROM dumps from AN8008 and there are such from AN8002 for a long time, as well everyone having any of those DMMs can make EEPROM dumps very easy.

But, I'd also like to add that not everyone (as in 100%) here can easily dump and flash EEPROMs. Maybe it's almost everyone, but perhaps someone who is not so experienced with flashing also bought the multimeter?

I don't even have a dumping and flashing tool (I guess those might be cheap and USB based these days), and if I did, wouldn't know the procedure (e..g. if I have to desolder/resolder the chip, or attach some jumper wires, or even do it in circuit without disconnecting things). Someone did mention a bit about it here, and there might be tutorials either for the specific multimeter or just general procedure, so it might be possible, but anyhow still seems like a risky prospect unless you're experienced.

That said, I'm carefully reading and considering all of the info here. To me, most of the AN models seem pretty great, even the 6000 count ones (with temperature -8002).

[sleemanj]

Because if not, then there are much better DMM for basically the same money than AN8008.

Again for the curious, the first post of that russian thread ( http://kazus.ru/forums/showthread.php?t=112135 ) has a fairly comprehensive list of the meters, and various reverse engineering and hacks of the 0660 based meters.

[matura713]

But, I'd also like to add that not everyone (as in 100%) here can easily dump and flash EEPROMs. Maybe it's almost everyone, but perhaps someone who is not so experienced with flashing also bought the multimeter?

so it might be possible, but anyhow still seems like a risky prospect unless you're experienced.

if you look at some of my posts in this forum, you will see i am complete amateur, my knowledge of electronics is below what i would consider a very basic level. however, dump/modify/flash EEPROMs I consider very easy, but maybe because I can program in several programing languages and thus reading and writing memory feels like something more natural to me than other common electronic tasks.

anyway, if you have no experience, I believe the easiest is to get cheap USB EEPROM programmer like CH341A, it's about 3 USD delivered and 10 x EEPROM chips - they cost less than 1 USD delivered - 24C02N is EEPROM chips compatible with DMMs we're discussing here like AN8002 and AN8008.

so, you can even prepare few 24C02N with EEPROM dumps available online, then desolder from the DMM you're hacking and keep your original EEPROM chip and its content as backup and just solder the new EEPROM chip you prepared with your programmer to your DMM - that makes the risk to minimum or basically no risk, assuming you can solder.

[matura713]

Because if not, then there are much better DMM for basically the same money than AN8008.

Again for the curious, the first post of that russian thread ( http://kazus.ru/forums/showthread.php?t=112135 ) has a fairly comprehensive list of the meters, and various reverse engineering and hacks of the 0660 based meters.

and just for completeness as that is good 0660 list, the known DMMs with the "new" DM1106EN, which so far looks like slightly modified version of 0660 chipset is:

* BSIDE ADM08
* PeekMeter PM18
* AN8008 (ZT109) --> discussed in the forum thread here

oh and BTW, in case someone don't know 0660 chipset is cheap Chinese clone of HYCON HY12P65 and I believe HY12P65 is in very reputable DMMs including the one that is EEVBlog branded.

[plazma]

if you look at some of my posts in this forum, you will see i am complete amateur, my knowledge of electronics is below what i would consider a very basic level. however, dump/modify/flash EEPROMs I consider very easy, but maybe because I can program in several programing languages and thus reading and writing memory feels like something more natural to me than other common electronic tasks.

anyway, if you have no experience, I believe the easiest is to get cheap USB EEPROM programmer like CH341A, it's about 3 USD delivered and 10 x EEPROM chips - they cost less than 1 USD delivered - 24C02N is EEPROM chips compatible with DMMs we're discussing here like AN8002 and AN8008.

so, you can even prepare few 24C02N with EEPROM dumps available online, then desolder from the DMM you're hacking and keep your original EEPROM chip and its content as backup and just solder the new EEPROM chip you prepared with your programmer to your DMM - that makes the risk to minimum or basically no risk, assuming you can solder.

IIRC the calibration parameters are in the EEPROM. So you need the meters original dump for modifications.

[plazma]

I used a modified PicKit 3 clone to read and write the EEPROM in the AN8002.

[matura713]

IIRC the calibration parameters are in the EEPROM. So you need the meters original dump for modifications.

good reminder, it's correct, but from my point of view it's another advantage - if you want to calibrate it better by yourself.

[Fungus]

My meter came in from AliExpress ANENG store. Price has gone up since I ordered it now USD $24.18 and supposedly 29% off the regular $USD 34.

Ooooh, they make a cute miniature multimeter as well.  (with teardown pic in the store!)

Edit1: Reading the small print, that meter needs a weird 12V battery. And ... what the hell is this? I never saw that on a meter before.


Edit2: I also found/ordered something else they make that's very interesting but I'm keeping it secret for a while seeing as how everybody else gets their stuff a week ahead of me.  (no my AN8008 didn't arrive yet )

Replacement fuses in North America... I have 10A for current and 200mA for input.
Littelfuse 3.6x10mm fuses and Bussmann/Eaton, Schurter SPT 3.67x10; This size is 250VAC interrupt-rated...
But all have pigtails 

Can't you just cut them off? 

[Fungus]

For an approximately 1 mA current reading I had the choice of 989.8 uA (1.01% low due to the 100.561 meter insertion resistance) or 1.0 mA.  Yes, that's two digits, 10% error represented by the least significant digit.  That's the best resolution possible just above 1 mA.  The -1% uA error continued down to around 200 uA.  When the reading was corrected for the added resistance, the uA accuracy was typically below 0.1%.

I can live with that. I don't think I need to measure 10mA with two decimal places very often.

Besides, nobody around here owns just one multimeter. I'm sure you own other meters if you need that particular reading.

(or use a shunt)

[ocw]

I don't think I need to measure 10mA with two decimal places very often.

You set a low standard to meet.  My accuracy verification tests showed that the AN8008 current reading accuracy ranged from 2.1 - 6.6% for readings between 2 - 10 mA.  Many economy analog meters can do better than that!  While you can't expect perfection from a $20 meter, one which can provide an accurate 20 uA reading should be able to read 2 mA accurately.  Being required to get a different meter when a reading which you are peaking goes above 1 mA doesn't make sense.

I'm glad that I purchased the meter.  It's a handy small meter for certain tasks.

[JohnPen]

Fungus you are not alone!  I have also been waiting patiently,enviously and increasingly jealously for this electronic dream since ordering on 30th June!  Hopefully both ours will arrive very soon!