EEVblog #972 - Operating Chips Outside Their Spec

[EEVblog]

Dave tests the AD8436 True RMS converter chips inside the 121GW multimeter to see if it works well below it's voltage rail specification.
Bonus thermal testing in the chamber.
http://www.analog.com/media/en/technical-documentation/data-sheets/AD8436.pdf

[jfasoc]

The whole video I was waiting for you to test how low it will work.

[nidlaX]

Can't wait to get one of these. I passed on the $200 U1252Bs so I can save my multimeter budget for a 121GW!

[KerryW]

Have you thought about using Batterizers with this meter?   

[EEVblog]

The whole video I was waiting for you to test how low it will work.

Sorry, didn't occur to me to do that! 

[JS]

Dave, we all want a batteroo inside our meters instead of the low voltage operation!! 

JS

[integritetus]

I am curious as to whether Analog Devices will guarantee that the AD8436 will operate to its specifications at a supply voltage of 3.6 V, or whether they are saying that it works but the design user is on his own to characterize operation at voltages below the minimum 4.8 V.

Dave mentioned this his DMM design partner is doing testing.  I would expect that they could characterize some number of parts across voltage, temperature, and frequency, but I wonder how they would characterize the part across process variation (that is, the variation in process parameters for the IC process in which the AD8436 is fabricated).  I would expect that only the part designer/manufacturer could characterize the part across process (unless the design user were supplied with a significant number of parts binned by process parameter).

For a sufficiently small production of units, a design user could actually in theory characterize each part received.  But I suspect that is not practical for this project.

[kandrey89]

Is Dave making an EEVBLOG branded multimeter?
Is it branded only, or did Dave help design it?
Is there a pre-sale, spec page or something, can't find anything anywhere about it?
ETA and other details?

[tszaboo]

Some time ago: We were testing a FET driver. Specification said maximum voltage 18V. Sure we can go higher than that! They give large amount of headroom on these specs, and it will work fine at 20, if we want to.
Of so we started cranking up the voltage on the lab supply. At 18.3V the smoke flew out of the part 

[Supercharged]

AM i the only one who saw a Wifi or Bluetooth module at 4:18?
Wireless Confirmed!

[NivagSwerdna]

Is Dave making an EEVBLOG branded multimeter?
Is it branded only, or did Dave help design it?
ETA and other details?

Firstly, I'm pretty sure those questions are answered somewhere on the forum.
Secondly, you just need to think about it for one second... just try and quantify the engineering required to have a decent product.  It's clearly going to be either a rebadged existing model or as has been suggested a rebadged exisiting model with maybe a tweak or two.  Personally I don't have any issue with that as its better to have a tool that works than a flakey v0.00001 of something.
But... I've been madly googling images and for the life of me I cant work out the manufacturer.  I'm was going on the basis that its got to be a reasonable manufacturer so that eliminates Hung Lo brands.  My line of search is working on the DMM glass.  It's likely that this model will have the same glass as an existing or planned model since an EEVBlog run will only form part of a larger commercial run..

Am I warm?  Any suggestions?

(If the price isn't outrageous I'll be getting one)

[testian]

AM i the only one who saw a Wifi or Bluetooth module at 4:18?

Definitely looks like one

[Len]

Is Dave making an EEVBLOG branded multimeter?
Is it branded only, or did Dave help design it?
Is there a pre-sale, spec page or something, can't find anything anywhere about it?
ETA and other details?

https://www.eevblog.com/forum/chat/new-eevblog-branded-multimeter-coming/

[DavidDLC]

AM i the only one who saw a Wifi or Bluetooth module at 4:18?
Wireless Confirmed!

No you are not

[Brumby]

Is Dave making an EEVBLOG branded multimeter?
Is it branded only, or did Dave help design it?
ETA and other details?

Firstly, I'm pretty sure those questions are answered somewhere on the forum.
Secondly, you just need to think about it for one second... just try and quantify the engineering required to have a decent product.  It's clearly going to be either a rebadged existing model or as has been suggested a rebadged exisiting model with maybe a tweak or two.  Personally I don't have any issue with that as its better to have a tool that works than a flakey v0.00001 of something.
But... I've been madly googling images and for the life of me I cant work out the manufacturer.  I'm was going on the basis that its got to be a reasonable manufacturer so that eliminates Hung Lo brands.  My line of search is working on the DMM glass.  It's likely that this model will have the same glass as an existing or planned model since an EEVBlog run will only form part of a larger commercial run..

Am I warm?  Any suggestions?

(If the price isn't outrageous I'll be getting one)

I think you'll find Dave's involvement is a bit more than that.

Dave has already done a (sort of) rebadge with the BM235 ... but the 121GW is something special. 

[NiHaoMike]

That part is rated to well above double the battery voltage and uses less than 1mA. Perhaps you could use a spare PWM on the microcontroller to drive a charge pump.

Some time ago: We were testing a FET driver. Specification said maximum voltage 18V. Sure we can go higher than that! They give large amount of headroom on these specs, and it will work fine at 20, if we want to.
Of so we started cranking up the voltage on the lab supply. At 18.3V the smoke flew out of the part 

Switcher driver chips for mains applications often have an internal zener clamp to allow a simple resistor to supply startup power.

[obiwanjacobi]

Hee Dave, is this what you were working on in #900 STM32 Development Board?

[oldway]

Operating chips outside their spec is a bad practice, I don't agree with this.
Ask to the manufacturer to change their specifications....

The statement " the manufacturer (Analog Device) says it can work at lower voltage than specified" is not a profissional statement....It means that good working in this condition is not garanteed by the manufacturer of the chip.

You can't specify that your multimeter will work with 3.6V if the chip is only specified to work with 4.8V minimum voltage, that's not acceptable.

I worked in quality control and I would reject your multimeter....

Sorry Dave....

[EEVblog]

Secondly, you just need to think about it for one second... just try and quantify the engineering required to have a decent product.  It's clearly going to be either a rebadged existing model or as has been suggested a rebadged exisiting model with maybe a tweak or two.

Nope, and nope.

It's likely that this model will have the same glass as an existing or planned model since an EEVBlog run will only form part of a larger commercial run..

Nope again.

[EEVblog]

I worked in quality control and I would reject your multimeter....

You don't get a say in it
Apart from not buying it of course.

[oldway]

And what about using 1.2V rechargeable batteries ?
No way !

The only solution is to use 5 batteries instead of 4 or a 9V battery.

NB: 1.5V batteries often leaks, I have a Fluke 87 IV badly damaged by leaking 1.5V batteries....I never had such problem with 9V battery.
So I prefer using a 9V battery in a multimeter...
So I prefer a multimeter with a 9V battery....

[f4eru]

Question: why use a rms calculator chip?  I would rather use two adc channels and sample both I and U at the same time, and calculate the result. This way you can also measure power.
Also note that you don't need absolute add precision for your 6000 counts, you can use 12 bits, for example, the massive oversampling gives you the needed precision.

Concerning chips out site specs, it can be  OK,  if the risks have been properly rated, and the outside spec validated as thoroughly as possible. A eol test is a good bonus. I seen it done at a  big Corp for million automotive products. It was done properly.

[oldway]

Question: why use a rms calculator chip?  I would rather use two adc channels and sample both I and U at the same time, and calculate the result. This way you can also measure power.
Also note that you don't need absolute add precision for your 6000 counts, you can use 12 bits, for example, the massive oversampling gives you the needed precision.

Concerning chips out site specs, it can be  OK,  if the risks have been properly rated, and the outside spec validated as thoroughly as possible. A eol test is a good bonus. I seen it done at a  big Corp for million automotive products. It was done properly.

I don't think Dave intend to make a multimeter as the Metrawatt Gossen Energy....
https://www.gossenmetrawatt.com/resources/tt/metrahit_energy/db_gb.pdf
It is far too expansive !
Even this multimeter seems to use two rms converter and not your solution, as max bandwith is only 100Khz.

[EEVblog]

Question: why use a rms calculator chip?

Because most multimeters use front end chipsets, and these integrate well with external True RMS chips.

I would rather use two adc channels and sample both I and U at the same time, and calculate the result. This way you can also measure power.
Also note that you don't need absolute add precision for your 6000 counts, you can use 12 bits, for example, the massive oversampling gives you the needed precision.

It's not that cheap or easy to get fast decent 12bit converters and integrate them into the design.
It's much easier to use a proper True RMS converter chip that integrates into the measurement system nicely, can handle high crest factor waveforms, and can use the existing chipset converter and firmware.
And the meter is 50000 counts, not 6000 counts.

[f4eru]

Expensive? Noooo, it's not! There are plenty of us today with integrated adc 12 bits or higher. Take one that is fast enough. The cost of the Micro is already included