Multimeter selection...

[Jim Johnson]

Funny we should hit the multimeter subject today - last night I realized my
current DVM won't do what I need...

I'm in the process of calibrating a Roland SH-5. The service manual calls
for the +-14 v PS to be calibrated +- 10 mV, and the 10v precision
reference to be calibrated +- 1 mV. My 15 year old Fluke 8020 won't do this
- it's only 3 1/2 digits, and I have no idea what its inherent accuracy is.
Plus when I short the inputs together, I get a reading of -30 mV on the 20V
scale - I don't even want to try to calibrate the synth with this.

Am I going to have to spend $300 for a meter that can be used for
calibrating synths? Is there anything affordable out there that can provide
this kind of resolution? I can live without transistor checking, etc -

TIA

Jim Johnson 
Metaphoric Software
-------------------
Makers of Techno Toys
Software for Electronic Music
http://www.technotoys.com
info at technotoys.com



*********** REPLY SEPARATOR ***********

On 4/6/98, at 6:09 PM, Paul Schreiber  wrote: 

>The DVM in question is no longer made. The RS meters over $49 all use the
ASIC that Doug Curtis and I
>designed. If you buy the $69-$99 meters with true RMS, then any one will
work. I like the $89 hand held that has
>the cap meter and the transistor checker. This is a good way to match beta
in transistors, and it will
>automatically identify the leads (E B C) if you don't have a data sheet.
RS likes to change the cosmetics every year,
>but I still call them by price!
>
>They are not as accurate as a Fluke 87 ($280 meter) but certainly good
enough for 99% of use. The only real difference in the RS and the Fluke is
absolute accuracy, and the RS meters are generally 0.2%, which is 30mv
>out of a 15V signal. A Fluke is like 0.05% or 7mv error in a 15V signal.
The main reason is the cost of the 14
>precision resistors used in the divider network. The 0.05% resistors,
which are in a laser-trimmered array
>(you trim the ratio of them, not the exact value) cost about $2ea (a
whopping $28) and the 0.2% resistors
>in the RS meter are around 11 cents each. In fact, I have my RS meter
'hopped-up' with 0.05% discrete resistors.
>
>The other limit is the temperature compensation of the bandgap voltage
reference. The RS meter uses a statistical
>averaging (using the on-board uP, a screaming 4bit Samsung part) with 2
trims. The Fluke uses a 'Zener zapped'
>(laser trimmed) reference that has about the same overall % drift as mine,
but the initial accuracy is much
>tighter: so over the 0-50C range, it drifts 'closer to home'.
>
>Voltmeters are almost as fun to design as synths. Don't miss the UL
testing, though ("Let's put it on the OHM scale
>and apply 1000 VAC for 1 minute!!!")
>
>Paul Schreiber
>Synthesis Technology
>www.synthtech.com
>
>
>----------
>From: 	Haible Juergen[SMTP:Juergen.Haible at nbgm.siemens.de]
>Sent: 	Monday, April 06, 1998 12:02 PM
>To: 	synth-diy list; dr john tuffen
>Subject: 	AW: Multimeter selection...
>
>
>
>	>I remember a few months back, there was a discussion about which
>tools
>	>should be in a basic toolbox - and multimeters were among the
>things
>	>discussed. In particular there was a Radio Shack meter that seemed
>to be
>	>recommended..... anyone tell me which it was? (It was (co-)designed
>by a
>	>listmember IIRC)
>
>Be sure you have a frequency counter, a capacitance meter, and a
>transistor beta tester built in.
>
>JH.