EEVblog 121GW Multimeter Issues

[1anX]

Will instructions be provided for replacing the 4007 diodes with the TVS for current 121gw owners in the future?

Good question, which you could also add and a suitable supplier for 1 of TVS components.
There is also a resistor that Dr. Frank has identified that may need replacing to enable a higher current for stability reasons. Its probably a bit early for Dave to do a video on a 121GW hack/modification/repair just yet as more hardware may need changing.

[EEVblog]

There is also a resistor that Dr. Frank has identified that may need replacing to enable a higher current for stability reasons.

Sorry, I have not been following ever post, link to post?

[Dr. Frank]

The 50 Mohm range uses 20nA reference current by its internal R9 = 10MOhm reference resistor, which causes a susceptibility to leakage currents (I will demonstrate that later). Maybe it's possible to increase the Vref (by firmware) from about 0.2V to 2V, to get 200nA test current, but probably that's not possible due to supply/reference voltage  limitation, as in total >4V would be required. See HY3131 Configuration document, page 31.
 
The 50 Ohm configuration also follows  the HY3131 document, see page 33, so R26 = 1k is used also for 50 Ohm, giving 470µA test current only, and a sensitivity of 470nV on the last digit, causing the noise trouble. The analog multiplex bus of the HY3131 is completely occupied, so it's not possible to add another 100 Ohm reference resistor for 5mA test current.

Frank

[1anX]

Ground, Earth, Static Field... whatever.

I propose simply to calculate the equivalent voltage on the last digit in 50mOhm range, that is about 0.4 mA times 1mOhm = 400nV.

That's simply extremely sensitive, so shielding is also difficult in a handheld DMM.

The reference current should have been 5mA, to get a stable reading as in 500 Ohm range.

Instead both ranges are derived from the same 1k reference resistor, but the 50 Ohm range uses an amplified AD range, which is the culprit
Frank

This is the post I'm referring to!

[JS]

The 50 Mohm range uses 20nA reference current by its internal R9 = 10MOhm reference resistor, which causes a susceptibility to leakage currents (I will demonstrate that later). Maybe it's possible to increase the Vref (by firmware) from about 0.2V to 2V, to get 200nA test current, but probably that's not possible due to supply/reference voltage  limitation, as in total >4V would be required. See HY3131 Confirguration document, page 31.
 
The 50 Ohm configuration also follows  the HY3131 document, see page 33, so R26 = 1k is used also for 50 Ohm, giving 470µA test current only, and a sensitivity of 470nV on the last digit, causing the noise trouble. The analog multiplex bus of the HY3131 is completely occupied, so it's not possible to add another 100 Ohm reference resistor for 5mA test current.

Frank

  There is the 15V generator for the diode test, with that and a ratiometric measurement with the 10M reference something could be worked out. I have to take a second look at the datasheet to see what else could be done in FW.

For the 500Ω range I should take another look as well, isn't there any other resistor that can be paralleled to the existing one? Or again using the 15V zener source to do measurement against the 1k reference. I haven't looked enough to this, I will take a look at the datasheet and come back tomorrow.

JS

PS: two more ideas, there is the 100Ω resistor I the μA range, the resistor might need to be connected to that input for measuring small resistors but who cares our connected by u11 but I guess that last wouldn't work. The second once, couldn't we spare C30 to get the extra resistor there.

[Dr. Frank]

Following Daves hint, I removed the over voltage protection diodes D7 and D8.



By means of an Impedance Synthesizer circuit, I measured different resistors from 5MOhm to 50MOhm.



The gain calibration of the 121GW is now -6% low, that means, that the nominal test current of about 20nA is decreased by that amount also, that is -1.2nA.



These 1.2nA were leaked by D7, or by the difference in leakage currents between D7 and D8.
At 6V reverse voltage, I have determined leakage currents of about 500..800pA for the one diode, but about 3nA for the other, so this difference causes this effect, at least.

Without the diodes, I re-calibrated the 50MOhm and 5MOhm ranges, and measured the change of reading over a 3°C temperature change.



It's now -0.04% only, giving a virtual T.C. of about -133ppm/°C, instead of +6400ppm/°C, as with the diodes assembled.

That clearly indicates that the diodes were the root cause for the bad temperature behavior and out-of-spec of the 50MOhm range.

The replacement of the 1N4007s by a TVS (SM6T22CA) probably is much better for ESD / overvoltage protection, as these TVS react much faster than ordinary power diodes.
A TVS can't be used in that application, though, because they have leakage currents of up to 200nA @ 25°C, which is 10 times the reference current of the 50M range.
In other words, a TVS in place of D8 would simply short both the 5MOhm and 50MOhm ranges.

So a solution with two JFETs would be the correct measure, although I doubt that they'd withstand Joes ultra-violent HV pulse testing 

[chronos42]

The voltage reference is connected to AGND, not AVSS which is the battery negative. The full battery voltage is not across the reference, as AGND is a referenced voltage rail lifted above ground.
If you actually measure the voltage across the reference I think you'll find it's fine.

Doh...

i was sure I made some mistake.
I have mixed up AVSS with AGND, sorry.
I must subract 1.8V from B+ of course.

So it is clear that the 121GW can be used with lithium batteries.
In case that the lithium cells are brandnew the voltage at the reference chip is above the operation voltage for a short time, but never above the absolute maximum voltage.

Thanks for clarifying.

[IanB]

The 50 Ohm configuration also follows  the HY3131 document, see page 33, so R26 = 1k is used also for 50 Ohm, giving 470µA test current only, and a sensitivity of 470nV on the last digit, causing the noise trouble. The analog multiplex bus of the HY3131 is completely occupied, so it's not possible to add another 100 Ohm reference resistor for 5mA test current.

I can get stable voltage readings when I independently apply a 1 mA source current and have 1 µV sensitivity on the last digit. This is the same order of magnitude as 470 µA. So I think there is more to the noise than just the magnitude of the test current.

[dcac]

The voltage reference is connected to AGND, not AVSS which is the battery negative. The full battery voltage is not across the reference, as AGND is a referenced voltage rail lifted above ground.
If you actually measure the voltage across the reference I think you'll find it's fine.

Doh...

i was sure I made some mistake.
I have mixed up AVSS with AGND, sorry.
I must subract 1.8V from B+ of course.

So it is clear that the 121GW can be used with lithium batteries.
In case that the lithium cells are brandnew the voltage at the reference chip is above the operation voltage for a short time, but never above the absolute maximum voltage.

Thanks for clarifying.

Beware that the AGND voltage seems to change depending on measurement mode. It's i.e. 1.09V in Ohms mode and 0.38V in Diode mode - and that seems to put more voltage over the ADR3412 than it perhaps was designed for - all depending on how fresh the batteries are.

Or am I also missing something? I measured the voltage between the batteries (-) which = VSS and the COM terminal which = AGND.

[Digital Corpus]

Is there a specific reason why the thermocouple update rate is 1 Hz? Its as such in 1.22, 1.25, and 1.26. I realize the time constants for thermocouples is usually fairly high, but I have probes with TC's less than a second that I was hoping to not need to spin a custom solution for.

[EEVblog]

The replacement of the 1N4007s by a TVS (SM6T22CA) probably is much better for ESD / overvoltage protection, as these TVS react much faster than ordinary power diodes.
A TVS can't be used in that application, though, because they have leakage currents of up to 200nA @ 25C

That's not correct.
Not only is the 200nA a maximum figure (not typical), it's at the rated voltage.
It drops drastically with lower voltage.

[chronos42]

The voltage reference is connected to AGND, not AVSS which is the battery negative. The full battery voltage is not across the reference, as AGND is a referenced voltage rail lifted above ground.
If you actually measure the voltage across the reference I think you'll find it's fine.

Doh...

i was sure I made some mistake.
I have mixed up AVSS with AGND, sorry.
I must subract 1.8V from B+ of course.

So it is clear that the 121GW can be used with lithium batteries.
In case that the lithium cells are brandnew the voltage at the reference chip is above the operation voltage for a short time, but never above the absolute maximum voltage.

Thanks for clarifying.

Beware that the AGND voltage seems to change depending on measurement mode. It's i.e. 1.09V in Ohms mode and 0.38V in Diode mode - and that seems to put more voltage over the ADR3412 than it perhaps was designed for - all depending on how fresh the batteries are.

Or am I also missing something? I measured the voltage between the batteries (-) which = VSS and the COM terminal which = AGND.

I was not aware that the virtual ground is changing with the measurement mode. With 0.38V in Diode mode my doubts would be true, event though I made a mistake with AGND. That needs further investigations. But I think like you, we maybe overlook something. i cannot believe that UEI has overlooked such a problem.

Update:
Meanwhile I made the same measurement, the voltage between com terminal and - battery. It is -1.8V as expected in all modes, except Diode test. In 15V Diode test mode there are indeed only 0.38V. I assume because of the high 15V level they had to shift the virtual gnd level in Diode mode, prevent the adc to overrange.
And now the problem is back again, with lithium batteries the voltage at the reference element is over 6 V in 15V Diode test mode.

Update 2:

Actually there is no problem.
With alkaline batteries the voltage at the reference chip will never exceed 6 V, even in the 15V Diode mode. According to the manual there are only alkaline batteries mentioned.

This meter cannot be used with lithium batteries. In the 15V Diode test mode is a high risk to damage it.

So: Do not use the meter with lithium batteries!

That is the point we have overlooked.
It should be a warning in the manual.

[Dr. Frank]

The replacement of the 1N4007s by a TVS (SM6T22CA) probably is much better for ESD / overvoltage protection, as these TVS react much faster than ordinary power diodes.
A TVS can't be used in that application, though, because they have leakage currents of up to 200nA @ 25C

That's not correct.
Not only is the 200nA a maximum figure (not typical), it's at the rated voltage.
It drops drastically with lower voltage.

Hello Dave,
you're right, that these 200nA are a maximum value at Vrm ~ 18.8V, but the term I have used: 'up to' is an equivalent expression to 'maximum', I think.

There is no specification or diagram of the behavior of this leakage current at lower voltages, so your statement that this current drops 'drastically' has to to be quantified / measured on an appropriate  TVS, probably with Vr = 18V, Vrm ~15V.

As these are sort of zener diodes, I expect the leakage current being much higher at 1..6V than  these 1N4007, several 10nA maybe, therefore still being unsuitable for this application.

Frank

[chronos42]

Hi,
I noticed that the meter is locked in an auto-ranging loop in the DC+AC mode in some conditions (Ver. 1.26).

For example: Generate this signal with a function generator:
6.4Vss, sinus, 800Hz together with a DC offset of +6.6V.

With this values a stable auto-ranging loop occures.
No other of my meters (Agilent U1252B, Brymen BM869S, Gossen Metrahit Energy) has this problem.

In manual mode the 121GW shows the correct value, the reading is very close to the other meters.

[emece67]

.

[Bratster]

The voltage reference is connected to AGND, not AVSS which is the battery negative. The full battery voltage is not across the reference, as AGND is a referenced voltage rail lifted above ground.
If you actually measure the voltage across the reference I think you'll find it's fine.

Doh...

i was sure I made some mistake.
I have mixed up AVSS with AGND, sorry.
I must subract 1.8V from B+ of course.

So it is clear that the 121GW can be used with lithium batteries.
In case that the lithium cells are brandnew the voltage at the reference chip is above the operation voltage for a short time, but never above the absolute maximum voltage.

Thanks for clarifying.

Beware that the AGND voltage seems to change depending on measurement mode. It's i.e. 1.09V in Ohms mode and 0.38V in Diode mode - and that seems to put more voltage over the ADR3412 than it perhaps was designed for - all depending on how fresh the batteries are.

Or am I also missing something? I measured the voltage between the batteries (-) which = VSS and the COM terminal which = AGND.

I was not aware that the virtual ground is changing with the measurement mode. With 0.38V in Diode mode my doubts would be true, event though I made a mistake with AGND. That needs further investigations. But I think like you, we maybe overlook something. i cannot believe that UEI has overlooked such a problem.

Update:
Meanwhile I made the same measurement, the voltage between com terminal and - battery. It is -1.8V as expected in all modes, except Diode test. In 15V Diode test mode there are indeed only 0.38V. I assume because of the high 15V level they had to shift the virtual gnd level in Diode mode, prevent the adc to overrange.
And now the problem is back again, with lithium batteries the voltage at the reference element is over 6 V in 15V Diode test mode.

Update 2:

Actually there is no problem.
With alkaline batteries the voltage at the reference chip will never exceed 6 V, even in the 15V Diode mode. According to the manual there are only alkaline batteries mentioned.

This meter cannot be used with lithium batteries. In the 15V Diode test mode is a high risk to damage it.

So: Do not use the meter with lithium batteries!

That is the point we have overlooked.
It should be a warning in the manual.

Can Dave maybe weigh in on this? It would be nice to know if there is a legitimate cause for concern using lithium batteries in this or not.

Seems like this would be a fairly serious issue...

Sent from my Fi Moto x4 using Tapatalk

[gdewitte]

I agree that an input from Dave would help. Specifically, can one use Lithium or LSD NiMH batteries? Maybe pros and cons of each? I've had enough things ruined or compromised by leaking alkaline batteries and don't want to sacrifice my 121GW.

[Mr.B]

I put Lithium in mine the first day I got it - wanting to avoid any potential leakage problems with the usual chemistries.
Now I am worried about the Lithiums...

[Bratster]

I put Lithium in mine the first day I got it - wanting to avoid any potential leakage problems with the usual chemistries.
Now I am worried about the Lithiums...

Same as me...

Sent from my Fi Moto x4 using Tapatalk

[JS]

Hi,
I noticed that the meter is locked in a auto-ranging loop in the DC+AC mode in some conditions (Ver. 1.26).

For example: Genrate this signal with a function generator:
6.4Vss, sinus, 800Hz together with a DC offset of +6.6V.

I've seen this very same auto-ranging loop in AC+DC simply measuring the mains 50 Hz, 230 V.

Regards.

I've seen it too, unable to repeat as I turned off and on again and was good.

JS

[Zucca]

can one use Lithium or LSD NiMH batteries?

Looks like the problem is when the batteries voltage goes above 6V. Full Lithium batteries are at 1,7V and they stay above 1,5V until they are done:

http://data.energizer.com/pdfs/l92.pdf

So it's 1,7x4=6,8V

LSD NiMH are more relaxed:

https://www.kronium.cz/uploads/BK-3MCCE.pdf

they start at 1,5V and they drop to 1,2V.

So LSD NiMH are just fine, Lithium are dangerous, see posts above.

PS: In my future projects the MAX AA Battery Voltage MUST be 1,8V 2V (EDIT: see post below) to be safe.

[chronos42]

can one use Lithium or LSD NiMH batteries?

Looks like the problem is when the batteries voltage goes above 6V. Full Lithium batteries are at 1,7V and they stay above 1,5V until they are done:

http://data.energizer.com/pdfs/l92.pdf

So it's 1,7x4=6,8V

LSD NiMH are more relaxed:

https://www.kronium.cz/uploads/BK-3MCCE.pdf

they start at 1,5V and they drop to 1,2V.

So LSD NiMH are just fine, Lithium are dangerous, see posts above.

PS: In my future projects the MAX AA Battery Voltage MUST be 1,8V to be safe.

Actually 6.38V and above is the critical value.
My 121GW showed 7.7V (!!) with brandnew lithium cells. That value settled down to 6.7V after some time. Still too much for the reference chip.

[Zucca]

My 121GW showed 7.7V (!!) with brandnew lithium cells...

wow, so in the future Vmax of a AA must be considered 2V.

Okay, idiots will always win:



 

[JS]

Use a bateroo!

If designed for four 1.5V to 2V cells two 3.7 cells could do but you need to change the arrangement (or put one backwards ), but Li could go over 4V when fully charged, so we need to provide a way to work with batteries between 0.8V to 4.2V, bit streetchy without a switching regulator...

About leaking batteries, just replace them every a year or two, that's all it takes, even ig they are not fully depleated, and use them till they die in the mouse, tv remote or torch.

JS

[darik]

About leaking batteries, just replace them every a year or two, that's all it takes, even ig they are not fully depleated, and use them till they die in the mouse, tv remote or torch.

That's not all it takes, unfortunately. I've had batteries well under a year old that weren't depleted pop on me. I think they aren't building them as heat resistant as they used to. I have become pathological about checking on my alkalines now because they seem to pop on me all the times these days and they never used to do that.