"Better" values for an RG Keen style transistor tester?

[mzy12]

Hey all.

I'm currently building a transistor tester to evaluate, well, BJT transistors. I think I saw a post here about changing the resistor values to better work with Si BE and CE voltage drop, but I cannot for the life of me find it again. I could probably do the maths on it myself, but I'm worried I'd overlook something. I'm just going to put them on a switch so I can toggle between Ge and Si.

Also, about the resistor values in general, would it make sense to adjust them to make I_c reflect what datasheets typically show when measuring HFE? Or are all the recommendations in the pedal based on the values people got from RG Keen's test circuit.

I've also read on this forum that compared to a DCA55/75, the leakage values can be vastly different from the RG Keen test circuit. There wasn't really any conclusion on what to do with that information. I'm wondering if anyone has any more input on this matter?

Thanks in advance for any help!

[mozz]

 My opinion, I would recommend in this order, DCA55/75, MK328, RG tester. The cheap MK328 also test other components R,L,C, besides transistors and other semis. I did buy the DCA55 first, but had built the RG tester so i had something to get me started testing the germanium i had picked up. I also use a few benchtop meters with hfe tests and i can say between them testing at different currents and voltages, i get different readings. I would say get the MK328 or similar for ~$20 and build the RG tester while saving your dollars for the DCA55.

[Rob Strand]

I'm currently building a transistor tester to evaluate, well, BJT transistors. I think I saw a post here about changing the resistor values to better work with Si BE and CE voltage drop, but I cannot for the life of me find it again. I could probably do the maths on it myself, but I'm worried I'd overlook something. I'm just going to put them on a switch so I can toggle between Ge and Si.

Silicon Calculation:
It's only a minor tweak to account for the different VBE, which I might add isn't constant across different devices, collector current or temperature
https://www.diystompboxes.com/smfforum/index.php?topic=107358.msg1249075#msg1249075

Which you can compare against RG's article,
http://www.geofex.com/Article_Folders/ffselect.htm

On RG's tester the battery voltage will vary say from 7.4V to 10V and that will introduce errors because the base current is no longer the base current assumed in the calculations. (Yes you can use a regulator but it needs to lower than 9V, 6V wouldn't be a bad choice.)

I use test set-ups like RG's tester all the time.  However I measure everything and put the numbers in a spreadsheet.  I even account for the DMM loading in the measurements.  That's because I'm looking for accurate values for a specific transistor.

For general work you don't need such accurate values.  All the secondary points which affect the hFE can be swept under the rug:
- battery voltage
- hfE vs current
- hfE vs temperature
- effect of VCE on measure IC ; Early effect
- Vbe differences
- Vbe variation with temperature
- In some cases even the measuring circuit itself adds specific forms of error.

There's a trade off between convenient measurements, which give ball park values, and precise measurements characterizing a particular transistor.

The DMM hFE testers and some of those components testers fall in the convenience category.  I like RG's setup because you know what it's doing, there's very little to go wrong.  However, I do play with VCE and IC so I don't have fixed base resistor in there.

Because these testers are non-specific you can't complain if one doesn't agree with another.  Even if the devices are precisely calibrated (which they often aren't) the secondary effects will cause different measurements.

Things like leakage can be very hard to get a precise value.  You might find the leakage wanders due to temperature, or it simply wanders with a mind of it's own - do the test as well as you can today do it again tomorrow and you will see quite different values.

For general use you can clearly tell a hFE=100 device from a hFE=400 device.  There are issues where hFE varies with current.    If the test current in the tester is IC=1mA and the circuit is operating at 10uA then you can't rely on the hFE from the tester agreeing with the hFE in-circuit.  Here's some sample curves,



If you do actually measure some hFE at different currents you can see they the current variation does *not* agree with these plots or with the plots in the datasheet.  I've found some brands produce much more variation of hFE with collector current even though it's the same part value, eg 2N3904.   I've measured plenty of parts which have very flat hFE down to low currents even though the manufacturer's data shows otherwise.

So the bottom line is:
- You can't always believe the numbers spat out by testers,
- Convenient testers are good enough for ball-park values.
- For the reasons given it doesn't pay to get caught up with minor variations
  (say 20%) from tester to tester.  The circuit's don't care.
- if your circuit is IC=10uA and and your tester is IC=1mA, the hFE will only be roughly like the in-circuit value.
- If you want to do precise work you need to set-up a tester like RG's with all the
  parameters tune-up to suit the conditions you want, then take the measurement.
  That measurement only has precise meaning under all those specific conditions.

[antonis]

Out of curiosity, you are interested for Si leakage current, beta or both..??

[mac]

I think I saw a post here about changing the resistor values to better work with Si BE and CE voltage drop, but I cannot for the life of me find it again

This one?

https://www.diystompboxes.com/smfforum/index.php?topic=102014.20

mac

[mzy12]

Sorry, I didn't communicate what I'm actually trying to do here very well! I'm building a version of the RG Keen's transistor tester on vero-board for permanent installation in an aluminium enclosure, powered by a constant dc supply voltage (i.e. standard pedal power supply with nominal 9v). I'll have switches for changing between NPN and PNP and for changing the I_C from RG Keen's 4ua to a current that you'd see on a datasheet (pretty easy to calculate), for calculating HFE.

build the RG tester while saving your dollars for the DCA55

This makes sense to me, but I do want to build the best version I can while I'm at it

It's only a minor tweak to account for the different VBE, which I might add isn't constant across different devices, collector current or temperature

I think the best course of action is for me to build an excel graph and let that deal with the calculations. Thanks for the link.

Out of curiosity, you are interested for Si leakage current, beta or both..??

Just beta for Si haha. This is just for audio amplification after all, not trying to beat moore's law with measuring pA of leakage

This one?

https://www.diystompboxes.com/smfforum/index.php?topic=102014.20

mac

I think that was the one, maybe! Hahah my memory's not so great, thanks for the help  I think I'll just deal with stuff in an excel table.

EDIT:  Forgot to say I will be measuring all resistors and adding trimmers to get the exact values I want.

[mozz]

I think the DCA75 is what you want, does curve tracing and a lot more. Right now, i would never be putting values into a excel graph, much easier to have the software import right into windows for you.

[mzy12]

I think the DCA75 is what you want, does curve tracing and a lot more. Right now, i would never be putting values into a excel graph, much easier to have the software import right into windows for you.

I do have my eye on one. I think I can justify it based on what my college course will have me do hahah.

much easier to have the software import right into windows for you.

In my case, it would have to be imported into Wine running the Windows program, because I am not only an insufferable pedal nerd, I am also an insufferable Linux nerd!

[antonis]

Out of curiosity, you are interested for Si leakage current, beta or both..??

Just beta for Si haha. This is just for audio amplification after all,

Don't get me wrong but an audio amplifier (actually, almost any circuit) based on particular item beta isn't a good design..

[mzy12]

Very true, but some people did something back in the 60's, and now here we are measuring DC current gain on mostly obsolete parts