Multimeter for matching transistors

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

Hi there,

Im having some trouble finding a multimeter for matching transistors and diodes. I dont really want to spend too much, as little as possible really...

Im getting a bit confused about display counts, like what the difference between 3999 or 4 1/2. i have found a few items from farnell, but i dont know exactly what would be suitable. I really want to get a bulk of the building done next week...

Is there anything from farnell that someone could reccomend.

many thanks

ali

Uncle Krunkus · Posted: Fri Oct 23, 2009 1:34 am Post subject:

As long as it's an auto-ranging DMM it would do fine. We sell one at work for AU$40 which would do the job.
AFAIK:-
For diodes, you're checking the voltage drop across them with a similar current as the circuit they're for. So if the application will put 15mA through it, set it up with a series resistor to have that (nominal) current passing through the diode, then check the voltage drop across the diode. Swap the diode, repeat.
For transistors, you want to check that with a similar Base-Emitter current to the application (fixed) the transistors produce the same Collector-Emitter current. I found two DMMs handy for this one.
In both situations, make sure you swap the components out using tweezers (maybe padded with heatshrink/rubber tubing) so your fingers don't heat the components.

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

thanks for the reply, I thought the multimeter had to be fairly accurate to measure 1mV, and able to display 4 1/2 or more digits or 4000pts. I didn't think a cheaper multimeter would do it.

This seems to fit the discription
http://uk.farnell.com/extech-instruments/ex330/multimeter-dmm-v-detect/dp/1271508

Or do I need something more like this
http://uk.farnell.com/tenma/72-7730/multimeter-usb/dp/1283642

thanks

Uncle Krunkus · Posted: Fri Oct 23, 2009 3:18 am Post subject:

The first one would be fine.
When they say:-
Voltage Measuring Ranges, DC:0.1mV to 600V
Voltage Measuring Ranges, AC:0.1mV to 600V
No. of Digits:3999
Series:EX300
AC Current Measuring Ranges:0.01µA to 10A

It means that the lowest DC Voltage range is .1mV
That means full scale deflection it will show up as "1000"uV on a four digit display.
The other ranges are a similar situation. 3 decimal places of accuracy, which is way more than you need. (a cat farting in the room next door is gonna make the current through a Collector-Emitter junction float around at 3 decimal places! Laughing

)
Remember too, that when matching transistors etc. you're looking at how each one differs from the others. Actual accuracy of the reading is not anywhere near as important as difference between readings. (There is a lot more technical way of saying that, but I think you know what I mean. Cool

)

Uncle Krunkus · Posted: Fri Oct 23, 2009 3:27 am Post subject:

BTW
The second one has 1 extra decimal place of accuracy.
When matching transistors, that extra decimal place will be bouncing around so much that it will be almost irrelevant.
They are both nice DMMs.
The first though will do the job just as well.

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

Thanks for the reply!

Do you think what Im going for is a bit overkill? Or do you think its generally better to have a decent meter.

thanks again, you have been a gr8 help!

hopefully gonna get most of it built this week!

ali

Uncle Krunkus · Posted: Sat Oct 24, 2009 3:42 am Post subject:

If you can afford the extra $20 then yeah, go for it. A nice multimeter is just that. An even better one is too! Laughing

I'm just saying that you don't need it to match transistors. Smile

What is it that you're actually building?
Most of the really specky VCOs etc. would use something like the MAT01 or LM394, so matching transistors wouldn't be an issue.

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

hehe, well umm, im building pretty much every module.
Probably biting off much more than I can chew...but I have all week to do it.

It modules like the VCA and Wavefolder that I am going to have to do the matching for. And the Moog filter.

Im not sure which multimeter to go for now tho. I would rather spend less than more.

Uncle Krunkus · Posted: Mon Oct 26, 2009 1:39 am Post subject:

go with the cheaper one then. You won't regret it.

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

Hi Mr krunkus!

I got my new multimeter, i went for the cheaper one. Loving it! much better than anything else I have used. Really nice feel to it. thanks for the advice.

A quick question tho;

I meassure diodes using the diode test function, when I first connect the diode it starts at a figure of about 0.568v. Then it quickly decreases, normally to around 0.562v.
Then if I keep the multimeter connected it will decrease, slowly, to about 0.558. And every diode seems to go this low if I keep it connected long enough.
Im not sure which value to take to match the diodes for the LFO and Waveshaper. The first 0.562 seems right to me, because it seems to stablize around there. But now I am a little confused.....

Any help would be great.
I have finished soldering most of the components apart from the matched and heat sensitive components and im well on the way to some modular madness!

kind regards

ali m

Uncle Krunkus · Posted: Sat Oct 31, 2009 4:18 am Post subject:

As far as I know, the diode test cannot help but slowly warm the diode up due to the current passed through it. So I'd say that is what is making the drop fall away a bit.
So,
I'd connect them, then let them settle, then take the reading. Again, make sure you don't touch them too much, or that will make them wander around a bit as well.

Ali M · Joined: Sep 01, 2009 Posts: 96 Location: uk

Yeah I think that would be the most sensible option, I was just thinking that the value might keep desending.
I have been sticking 50 or so into polystyrene foam. then leaving them for a few hours before testing them.

thanks