#239

[@Stephen_Heathcote]

Yep, that's the only issue with voltage references that that one, they need an accurate voltage to power them. If that ones off, then the output from it is off also.

[@TonyAlbus]

Yes i think that exacly the prob;em with this one..

[@jaymclemore9497]

Sure like that Siglent! Looking forward to the video on comparing different voltage references. Do you have any videos or suggestions on entry level bench meters? I watched your video on OWEN, but it seemed disappointing.

[@TonyAlbus]

The owon is indeed entry level, you can also try to find a good HP Agilent or Keithley second hand.

[@craxd1]

The Racal-Dana 5000 series DMMs have two LM399 ref sources in them. I use a 5004, which can be used as a 4-1/2 or 5-1/2 digit. One ref is the digitizer voltage ref, and the other is the ohms and current source ref. One ref is a 399H, with the other a 399Z.

[@TonyAlbus]

Thanks for sharing, interesting

[@markjones9180]

Thanks for this very timely video. Have been looking at these to compare meters and I’m going to give it a miss. Looking forward to the V ref roundup! Just subscribed. Thanks for the great content!

[@TonyAlbus]

Thank you Mark! Welcome!

[@juergenhussels9591]

Hello Tony, I am having a very similar reference since a few days - which also seems to be off a bit to the written values. But when looking into the specs of my multimeters - I am unsure what the real reason is. At least is my reference - after heating up for a while - amazingly stable. Checking your SDM3065x specs, it says 0.0040 + 0.0004 Accuracy ± (% of reading + % of range) after 1 year. This would add up to +-280 micro Volts with 5 V in the 20V range - now comes the question what is really off, the meter, the reference ? We need to check our references against some high precision instrument with a lower guaranteed tolerance, but I have none available. Even worse, my 6.5 Digit Schlumberger is 35+ years. I am off 0.006% - which is still good, I believe 🙂. Another point is, we don't know what instrument the Chinese vendor used - what were its tolerances what can we expect? best regards, Jürgen

[@TonyAlbus]

Hi Juergen, yes with more digits comes more resonsability :)) Ambient temp, heat-up time etc.. little test here: kzbin.info/www/bejne/aJu4knqlZs6LoZo

[@juergenhussels9591]

@@TonyAlbus Hi Tony, I just checked the power supply voltage "rejection". For me I found, I can tune my device with the power supply voltage to the "right" value (according my old instruments). Even though not 100% perfectly for all output voltages. When 20 Volts instead of 15 Volts are applied, I have a change of 500 uV in the 7V Zener voltage. The datasheet would allow for a change from 0.5mA to 10 mA (delta9.5 mA ) Zener current a max of 12mV ( = 1.26mV/mA) - that would be for the 5V change a delta of 0.67mA (7.5k resistor) resulting in a theroretical reference voltage change of 630 microV. All according to spec, but. I believe the biggest mistake of the Chinese Vendors is omitting a decent 15V Voltage regulator e.g. like 78l15. And they probably use just an adjustable DC power supply and don't care about the supply voltage. For me it seems they used 20V. I will implement a LM317 adjusted to 20V. - With that I am on the save side and power up with a 24V power supply.

[@TonyAlbus]

@@juergenhussels9591 Thank you for this feedback, very nice

[@zumzethuhurez5331]

Could the value written on the calibration certificate be actually 9.96375? It would be consistent with the other low values.

[@TonyAlbus]

You could have a very good point there!, thanks! i will check that

[@allthegearnoidea6752]

That’s a nice and useful little device thanks for sharing. Regards chris

[@TonyAlbus]

Thanks Chris!

[@ivolol]

LM399 is not designed to be accurate at all, but it is designed to be extremely extremely stable. If you feed it a good quality voltage and exact current into the zener it will give you stable voltage all day all year. You just have to know what that voltage is :') I think it is reference used in some near-high end DMMs, like 5 & 1/2 digit. The 10v it puts out and others, the absolute accuracy is completely reliant on the resistor dividers values being accurate and/or trimmed. So you are completely reliant on chinese calibration sheet actually being real to get any use out of it. If you don't have other calibration standard, then AD584 gives way better absolute accuracy. Best AD584L initial accuracy is ~0.05%, LM399 is 2%! But AD584L tempco is 5ppm, and LM399 tends be

[@TonyAlbus]

Thanks for the info, i do have a few other standards if you are a frequent viewer.

[@redmist4963]

The resistor are also quite expensive with 2 $ a piece for 2ppm stability and going up to 10$ for 0.2ppm

[@wernerpeter9706]

The difference from 240 µV (2.5 V) over 300 µV (5 V) to 530 µV (7.5 V) to the exact measured value at 10 V is very strange. If you have no way of checking the calibration data, then you have to be able to rely on these numbers. But if you can check these values, then you don't need to buy the calibrated version. So your video shows very nicely that you cannot blindly trust the calibration data. Then the only question that remains is, how large can the deviations in the calibration data be if, for example, two stickers have been swapped over?

[@TonyAlbus]

Usualy even with the cheap once, is was sort of correct, never this far off..so maybe indeed wrong sticker.

[@wernerpeter9706]

@@TonyAlbus Yes, as I said, the purchase and comparison is no longer necessary if you cannot rely on it. I have only a sdm3055, but the calibrated reference should be at least a decade more accurate in order to work meaningfully with it.

[@manuelcv9475]

Hello Tony and Followers: As I said earlier that reference of tension does not look like the originals, it gives me the impression that it is falsified. I do not know that Chinese gossip, but to accurately measure the multimeter has to be high impedance, I measure with the HP3456A, Fluke 8505A and with the Agilent / HP 34401A, the latter has the option in the menu of selecting high entry impedance of more than 10 gohm, and for the photos that you put seems to measure pretty well assuming that the reference of the photo is well calibrated, but it is a multimeter that has a good review and ground / mass connections are always correct It will measure well and without fluctuations. LM399H references (good and original) have to be connected enough hours, days, weeks or months to age and are even more stable, this is noted much since at first the multimeter will measure well from ~ 1.5 hours And as the months pass, it will stabilize faster, that can be seen maintaining the temperature of the laboratory. You have to understand that resistive dividers also have a DRIFT with temperature and is always tended to rise inside the equipment itself, because of the power supply or transformer, and to reach its stabilization, many hours are needed. I hope you make a video of the Agilent 34401A. Greetings.

[@TonyAlbus]

Thank you, yes correct. The video of the 34401A will come.

[@manuelcv9475]

Hola: Lamento que este sitio donde hay tanta información y tan valiosa sea dirigido por mentes obtusas y sean tan dependientes de otras políticas que hacen más daño que bien a la humanidad. MI mensaje no va dirigido a Tony Albus, si no a los dueños del sitio, ya que sin explicaciones eliminan mi propiedad intelectual, y asi nunca sabemos el porqué de ello pecando de nuevo por no tener una explicación clara. Por este motivo me daré de baja en todos los canales e iré eliminado mis videos pues considero que no he hecho daño a nadie y si es asi me merezco una explicación. Entretanto me despido de este estupendo canal. Saludos. Hello: I regret that this site where there is so much information and so valuable is directed by obtuse minds and be as dependent on other policies that do more harm than good to humanity. My message is not addressed to Tony Albus, if not to the owners of the site, since without explanations they eliminate my intellectual property, and so we never know why sinning again for not having a clear explanation. For this reason I will give myself down on all the channels and I will be eliminated my videos because I consider that I have not hurt anyone and if it is so I deserve an explanation. Meanwhile I say goodbye to this great channel. Greetings and even always

[@TonyAlbus]

Sorry to see you go, you also added to the information for others.

[@manuelcv9475]

@@TonyAlbus Hello: I'm leaving all the channels, more than 40 I am subscribed, thanks to the KZbin Policy, I have disappointed me, and I am from those who react logically before these injustices, besides trading with our information we They punish without reasons. I am not willing to follow or google or youtube while using injustice. Greetings

[@dennisqwertyuiop]

something I dont have,,all my flukes and bench dvm are good for me,,this is not nasaworking on space shuttle HA HA i

[@TonyAlbus]

No indeed :)