Author |
Message |
Wobuse
Joined: Dec 07, 2011 Posts: 34 Location: Lyons,il
|
Posted: Tue May 14, 2013 5:54 pm Post subject:
Measureing Micro volts? ? Subject description: Matching NPNs for the T.Henry VCO 1 |
|
|
I built Ian Fritz's transistor matching circuit ,and it seems to work great 'cept my meters only go down to .1 mV....
I think I need to go down to uV,to really get tight pairs .
What are your recommendations for meters that can go so low.
Thanks; in advance. _________________ "It's nice 2 B important; but it's important 2 B nice"
Visit my studio ; LimeStoneLabs at: www.Facebook.com/LSlabs
Hear some of my sound track work ,on the web series ,, www.cafegangster.com/16
https://soundcloud.com/james-phillips-75 |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Thu May 23, 2013 2:30 am Post subject:
|
|
|
AFAIK matching below 2mV is good enough!? i read it somewhere, i just can't recall... _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Thu May 23, 2013 2:48 am Post subject:
|
|
|
okay, i found it: in the minimoog service manual robert moog recommended +/-2mV for the VCO's NPN pair.
however, the 1st pair in a CA3086 is matched to less than 1mV, the pair in a LM394 is matched to 50μV. _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
Wobuse
Joined: Dec 07, 2011 Posts: 34 Location: Lyons,il
|
|
Back to top
|
|
|
marvkaye
Joined: Mar 14, 2011 Posts: 225 Location: Fla
|
Posted: Fri May 24, 2013 7:57 am Post subject:
|
|
|
I've been matching transistors using the Ian Fritz circuit to +/- 50 uV using a 50,000 count BK Precision bench meter, but recently added to my meter collection with a Keithly 177 digital microvolt meter which is accurate to +/- 2uV. Picked it up on Ebay for under $100 including shipping, and have seen them there since my purchase for considerably less (of course... ) Lots of meters out there today, though, that are accurate to the microvolt, just look around.
Most of the matching articles suggest that trannies are matched when they're +/- 2 mV Vbe, though, which is an order of magnitude less precise than what I've been trying to hold. Like most folks have mentioned, you can get within 1-2mV just by picking trannies from the same batch. Lots has changed in semi manufacturing since those early Moog days, that's for sure.
FWIW.........
<marv> |
|
Back to top
|
|
|
Wobuse
Joined: Dec 07, 2011 Posts: 34 Location: Lyons,il
|
Posted: Fri May 24, 2013 11:13 pm Post subject:
|
|
|
marvkaye wrote: | I've been matching transistors using the Ian Fritz circuit to +/- 50 uV using a 50,000 count BK Precision bench meter, but recently added to my meter collection with a Keithly 177 digital microvolt meter which is accurate to +/- 2uV. Picked it up on Ebay for under $100 including shipping, and have seen them there since my purchase for considerably less (of course... ) Lots of meters out there today, though, that are accurate to the microvolt, just look around.
Most of the matching articles suggest that trannies are matched when they're +/- 2 mV Vbe, though, which is an order of magnitude less precise than what I've been trying to hold. Like most folks have mentioned, you can get within 1-2mV just by picking trannies from the same batch. Lots has changed in semi manufacturing since those early Moog days, that's for sure.
FWIW.........
<marv> |
I did pick up a better meter... Check the link in my reply to Fonik.
I think I need to rework that test circuit though ,because , at the time I only had a 100k trim ,so I used that for one side & a resistor on the other.I can match the ohmage out to .000, but that's not good enough..
I need to use two resistors and a low ohm multi turn trim ,(like Ian suggested)so I can completely null the circuit ,and eliminate all the error, which I haven't been able to do,in my current set up.
Any tips for finding those super matched pairs?Is there a particular:manufacture you favor? _________________ "It's nice 2 B important; but it's important 2 B nice"
Visit my studio ; LimeStoneLabs at: www.Facebook.com/LSlabs
Hear some of my sound track work ,on the web series ,, www.cafegangster.com/16
https://soundcloud.com/james-phillips-75 |
|
Back to top
|
|
|
marvkaye
Joined: Mar 14, 2011 Posts: 225 Location: Fla
|
Posted: Sat May 25, 2013 8:11 pm Post subject:
|
|
|
I bought a couple of these Vishay/Dale 0.01% 15ppm resistors for the matched pair in a wheatstone bridge circuit to use for matching other resistors for R/2R ladders. A 50 ohm trimmer in series with the smaller valued one got them as close to identical as I could measure them using a uCurrent box. This, BTW, is way overkill for matching transistors, IMHO. There's a better way to use Ian's circuit that is much simpler. Put any transistor of the type you're trying to match in one of the sockets and just leave it there as a standard. Put your next transistor to match in the other socket (call it the DUT socket), let it settle down and note the difference in Vbe. Using a plastic parts box with 36 compartments as a sorter put that first DUT transistor into one of the central bins and note its Vbe difference on a bin chart stuck to the inside of the lid. Put another transistor into the DUT socket and repeat the test... note its difference and assign it to a bin relative to the first one. Keep going until you have multiple transistors in various bins... trannies with the same difference in Vbe from the standard match each other. This is much quicker than siwtching emitters, working out math problems, or even worrying about whether or not your resistor pair is close enough. Try it, I think you'll like it.
<marv> |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Sun May 26, 2013 2:52 pm Post subject:
|
|
|
good point, marv. simplicity is beautiful _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
|