one of my multimeters has a hfe testing feature. so I decided to run a few 2n5551 by it. Well, I did about twenty of them, hfe came to be anywhere between 95 - 102. surprisingly close. the manual said the hfe is measured at 1ma so that's pretty realistic.
has anyone tried this? if this is true, do we really need transistor matching?
has anyone tried this? if this is true, do we really need transistor matching?
i buy these in the "ammo box" style -- 2k at a crack -- you would think that the hfe would be tighter (ie. a lower std deviation) when you buy them this way or on a reel -- many of the parameters will change just by handling the devices !
when you batch the transistors together, you shouldn't be surprised that hfe and Vbe are often lumped together -- just as with HEXFETs or Lateral MOSFETS in which Vgs and transconductance wind up overlapping on a histogram. As I said in the previous paragraph, however, temperature does have an effect on the measurements.
when you batch the transistors together, you shouldn't be surprised that hfe and Vbe are often lumped together -- just as with HEXFETs or Lateral MOSFETS in which Vgs and transconductance wind up overlapping on a histogram. As I said in the previous paragraph, however, temperature does have an effect on the measurements.
Same experience -- sort of.
If I buy a dozen or so from Digikey the lot numbers printed on them are often the same and the hfe measurement range is pretty small. BUT, if I go down to the local surplus shop where one literally scoops them out of a bin (I'm not making this up), you need two dozen to find a single match.
Conclusion, yes you need to match them because you CANNOT assume that a bunch of them will always measure in a narrow range. However, if you know you are going to use them in current mirrors or differential amps, buy a bunch of them from some source where they are likely to all come from the same lot. Then match up the hfe, that may also help narrow the range of parameters you can't match.
If I buy a dozen or so from Digikey the lot numbers printed on them are often the same and the hfe measurement range is pretty small. BUT, if I go down to the local surplus shop where one literally scoops them out of a bin (I'm not making this up), you need two dozen to find a single match.
Conclusion, yes you need to match them because you CANNOT assume that a bunch of them will always measure in a narrow range. However, if you know you are going to use them in current mirrors or differential amps, buy a bunch of them from some source where they are likely to all come from the same lot. Then match up the hfe, that may also help narrow the range of parameters you can't match.
millwood,
yes, we do need matching but it depends on the circuit.
Last I had to replace 4 Transistors of 2 differential amps, one pnp, the other npn.
These dual diff amp took besides other things, both care of the feedback, hence the output offset of the amplifier.
Selecting the transistors with my multimeter HFE didn't work out:
i had selected 4 transistors with the same hfe, but in the circuit it didn't work, large output offset.
So i made a test jig, a copy of the diff amp, and selected again for equal collector voltage(=equal collector current with the same resistor load).
This turned out to work fine.
Dick.
yes, we do need matching but it depends on the circuit.
Last I had to replace 4 Transistors of 2 differential amps, one pnp, the other npn.
These dual diff amp took besides other things, both care of the feedback, hence the output offset of the amplifier.
Selecting the transistors with my multimeter HFE didn't work out:
i had selected 4 transistors with the same hfe, but in the circuit it didn't work, large output offset.
So i made a test jig, a copy of the diff amp, and selected again for equal collector voltage(=equal collector current with the same resistor load).
This turned out to work fine.
Dick.
Matching
Hello Dick,
Not my experience. I matched all four input transistors (two differential amps with PNP and NPN’s) of my Ampzilla and it worked out great. Sound was MUCH better. There was also an article in the Audio Amateur looong ago about the matching issue. Yes they builded a rig. If you don't have a full complementary design I believe it is easier and better using a dual monolithic pair like 2SC3381, MAT02, LM394 to name a few. Right Hugh?
The differential pair works by virtue of the equality of the two halves. See also:
http://db.audioasylum.com/cgi/m.pl?forum=tweaks&n=50223&highlight=elso+john&session=
djmiddelkoop said:
Hello Dick,
Not my experience. I matched all four input transistors (two differential amps with PNP and NPN’s) of my Ampzilla and it worked out great. Sound was MUCH better. There was also an article in the Audio Amateur looong ago about the matching issue. Yes they builded a rig. If you don't have a full complementary design I believe it is easier and better using a dual monolithic pair like 2SC3381, MAT02, LM394 to name a few. Right Hugh?
The differential pair works by virtue of the equality of the two halves. See also:
http://db.audioasylum.com/cgi/m.pl?forum=tweaks&n=50223&highlight=elso+john&session=
Goedemiddag Elso,
mayby there is something wrong with my multimeter HFE measurement ?
Yes, it was a Ampzilla circuit look-a-like, a Metaxas amp.
Using a dual monolithic pair is indeed much better, as you say :
the differential pair works by virtue of the equality of the two halves.
thanks for your reply,
Dick.
mayby there is something wrong with my multimeter HFE measurement ?
Yes, it was a Ampzilla circuit look-a-like, a Metaxas amp.
Using a dual monolithic pair is indeed much better, as you say :
the differential pair works by virtue of the equality of the two halves.
thanks for your reply,
Dick.
Matching
Back in time when I did the matching I did it with an analog Pantec transistor tester. Later I got that fancy multimeter with hFE tester. I got a totally different value on the latter meter but two transitors that proved equal on the Pantec were very close on the multimeter too. Beware the reading depends on the battery voltage. A fresh battery gives higher readings.
BTW do you have a copy of the Metaxas circuit? (I hardly dare to ask)
Hallo/hello Dick,djmiddelkoop said:
Back in time when I did the matching I did it with an analog Pantec transistor tester. Later I got that fancy multimeter with hFE tester. I got a totally different value on the latter meter but two transitors that proved equal on the Pantec were very close on the multimeter too. Beware the reading depends on the battery voltage. A fresh battery gives higher readings.
BTW do you have a copy of the Metaxas circuit? (I hardly dare to ask)
Vbe matching is often more important than Hfe in diff pairs and current mirrors, although emitter degeneration can reduce sensitivity to mismatch
Pease warns that digital noise from multimeters can confuse these measurements due to noise rectification, he put C across his meter and a used series R to reduce noise rectification in his diode slope measurements at the end of his book
any idea on how much Early Voltage varies? - again circuit techniques such as cascode/super-pair can help
Pease warns that digital noise from multimeters can confuse these measurements due to noise rectification, he put C across his meter and a used series R to reduce noise rectification in his diode slope measurements at the end of his book
any idea on how much Early Voltage varies? - again circuit techniques such as cascode/super-pair can help
> why Vbe matchin more important than Hfe?
Depends entirely on the application.
If the impedance the Bases see is very low, hFE does not matter, but Vbe may. This is true for some diff-pairs and mirrors, the cases JCX was talking about.
If the impedance the Bases see is very high, hFE matters a lot. This can be seen in the diff-pairs of some op- and audio amps which use high value bias resistors.
Basic circuit theory, down to Ohms Law, is a good guide.
Depends entirely on the application.
If the impedance the Bases see is very low, hFE does not matter, but Vbe may. This is true for some diff-pairs and mirrors, the cases JCX was talking about.
If the impedance the Bases see is very high, hFE matters a lot. This can be seen in the diff-pairs of some op- and audio amps which use high value bias resistors.
Basic circuit theory, down to Ohms Law, is a good guide.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.