diy voltage source

[p.larner]

i have an old nixi tube advance dmm2 multimeter,It needs calibrating firstly with a 1.99v feed.any idea how i cn get this from a battery either minh or lion?,it cost £3 from a radio rally and gives garbage readings,all electrolytics test ok so i assume it needs calibrating.

[tooki]

You need a whole slew of voltages to calibrate all the ranges, all the way up to 1kV. This is well outside of what’s easily done with simple DIY methods. Even the 2V range requires 1.999V ±0.05%.

Sure, you can use a bench power supply (linear, please!) to adjust the ranges up to 20V into the general ballpark. That’ll be good enough to test whether it’s actually working at all. But to make it useful you need another meter with at least 4.5 digits with which to precisely measure the actual voltage you’re applying.

For example, if you set your bench supply to 1.99V, connect that in parallel to both the Advance and your 5.5 digit meter, and the latter shows 1.98363V, then you’d twiddle R54 so the Advance shows 1.984V. That’s close enough to full-scale to be useful.

If truly all you have is batteries, then get some nice fresh alkaline batteries (voltage when brand new, without load: 1.6-1.7V) and measure a single one in the 2V range, using the method above. It’s not quite full-scale, so not ideal, but close enough.

But this will only calibrate the 2V range, for the higher ranges you’d need lots and lots and lots of batteries in series to get the necessary voltages.

[Solder_Junkie]

You can make a 1.8V 0.05% Voltage reference using Texas Instruments REF35180QDBVR

They are cheap and stocked by Mouser.

Check the part number, but I think it’s the 1.8V one.

SJ

[dobsonr741]

gives garbage readings

Define garbage readings. Is it 10% off or not even in the range?

[p.larner]

the readings are like 50+%off,the best i can do for the 1v voltage source is a c cell in series with a germainum doide so about 1.2v will that be ok,or if i do do the first call with that will t slew the rest?.

[coromonadalix]

for an 0.1% 0.2%  no need to go crazy,  stables voltage is all it needs

[Kim Christensen]

i have an old nixi tube advance dmm2 multimeter,It needs calibrating firstly with a 1.99v feed.any idea how i cn get this from a battery either minh or lion?,it cost £3 from a radio rally and gives garbage readings,all electrolytics test ok so i assume it needs calibrating.

the readings are like 50+%off,the best i can do for the 1v voltage source is a c cell in series with a germainum doide so about 1.2v will that be ok,or if i do do the first call with that will t slew the rest?.

Sound to me like it needs repair first. Not calibrating. Trying to calibrate broken equipment will create more trouble than it solves.
ie: Say you do the cal procedure when it's broken. Now the adjustments are cranked to some extreme. Then you go and fix some fault but it still looks wonky? Hmmm... Now you must try and cal it again... OOOppps! That didn't work... Find another fault... Cal again... etc.
If you find the faults, it's going to be within 5-10% once those repairs are done.

[p.larner]

The thing is i dont know if it is faulty as its how i brought it,its just the readings are nonsenseicle so assume its been screwdrivered.

[Kim Christensen]

You could mark the trimpot positions with a sharpie. Then it's easy to put things back, close to the way they were. Later, you can wipe the markings off with a tissue moistened with alcohol.

[p.larner]

its on the tuit pile lol,at least all the nixies+ drivers are ok,guess the main ic is ok too as it does measure but way way off,all caps+ power rails are ok i guess i better look for drifted resitors then try to cal it.

[kevin.gibbs]

You can start by calibrating at any voltage by comparing the readings with a working multimeter. If you can establish the same multimeter readings, no repair is necessary.

[inse]

A general advice for vintage equipment: watch out for carbon composition resistors, especially the higher ohmic ones.
Those might drift up to +50%, though not used in the signal path but can drive auxiliary circuits out of operation range.
In a Schlumberger multimeter I swapped all of them.
Do some random test on >100k types to find evidence.
P.S. forget about it, no carbon composition resistors used there