There are a variety of Arduino (or similar) curve tracer projects on the Web. If you have some nice ADC and DAC modules for Arduino you can learn quite a bit by putting one together. It is a substantial project.
But a question would be: What specific parameter are you trying to validate? Some transistors it will be important to validate noise, others breakdown voltage, other die size/power/current capability, others linearity.
It becomes quite a project.
If you simply need to validate a handful of Ebay or AliExpress BC546 then maybe it is easier and less expensive to buy from a Digikey or Mouser level distributor. Or take genuine transistors out of wreck chassis.
But a question would be: What specific parameter are you trying to validate? Some transistors it will be important to validate noise, others breakdown voltage, other die size/power/current capability, others linearity.
It becomes quite a project.
If you simply need to validate a handful of Ebay or AliExpress BC546 then maybe it is easier and less expensive to buy from a Digikey or Mouser level distributor. Or take genuine transistors out of wreck chassis.
Learn how to do basic static tests using a meter on ohms and diode ranges.
The voltages and currents on a transistor are always unique to the application although the base to emitter voltage is a very important figure to look at when looking for faults. The base will normally be approximately 0.6 to 0.7 volts higher than the emitter for an NPN and vice versa for a PNP.
The voltages and currents on a transistor are always unique to the application although the base to emitter voltage is a very important figure to look at when looking for faults. The base will normally be approximately 0.6 to 0.7 volts higher than the emitter for an NPN and vice versa for a PNP.
back in de 80-ies I used to buy bc547´s from a surplus parts graveyard for production of mixing desks for home use. I tested them in a small high gain stage and looked at the HP339 meter showing the noise. bad ones had much 1/f noise, showing the needle moving randomly. 1/f noise is often the result of process contamination. japanese brand transistors were always good. In those days fakes did not exist.
There are different test setups to measure various parameters. Many of the early data sheets do show the test circuits and you can probably find them in the app notes from Motorola and later On-Semi (same company). Noise would be one parameter you'd want to know for small signal development.
Some instruments have been developed to make this job easier. You have the old class of equipment, curve tracers. Relatively new would be SMUs (source - measure units). We measure bandwidth at different current levels, same for capacitance between electrodes.
Most important is education to know what to do to measure and what is important for what the use of the part is. We can't teach this on the internet. Really good techs and engineers are highly trained individuals. It ain't easy or quick. Your best bet is to read the application notes. They are written by true experts and how I gained a lot of knowledge.
Some instruments have been developed to make this job easier. You have the old class of equipment, curve tracers. Relatively new would be SMUs (source - measure units). We measure bandwidth at different current levels, same for capacitance between electrodes.
Most important is education to know what to do to measure and what is important for what the use of the part is. We can't teach this on the internet. Really good techs and engineers are highly trained individuals. It ain't easy or quick. Your best bet is to read the application notes. They are written by true experts and how I gained a lot of knowledge.
Hello, I came across a DN2540 from a trusted seller on ebay...as soon as I put it on the test circuit I realized that something was wrong. In fact from a subsequent test it seems to be a simple mosfet. Probable fake component?
PS: d-s has no values in ohm-meter and diode range.
PS: d-s has no values in ohm-meter and diode range.
Attachments
I dont know if it's OT, but here is a very good website about transistor dies:
https://www.richis-lab.de/
https://www.richis-lab.de/
Too bad it wasn't in English ...
Anyway, not very many of us will bother to crack the case to examine the die. In a sealed epoxy or composite case things are rather more difficult.
So, buy from accepted authorized distributors only if you can. If you can't, then you better know enough to be able to (properly) measure hFE, leakages and capacitance on the junctions. Then you need to be able to test the device in at least on configuration appropriate for that device. It would greatly help if you have a current limited HV supply so you can measure breakdown voltage. I use a 6186C HP DC current source (300VDVDC compliance voltage) for that purpose.
I'll be honest, it isn't worth the risk and work to buy from unknown sources for me. Most parts I get today are fakes, even special purpose ICs! Why would anyone buy an op amp for high performance from an unknown source ????? !
It comes down to using your head and do not give in to wishful thinking. The odds are against you, face it. But, it's up to you. May as well spend the money on lottery tickets, at least you know the odds.
-Chris
Anyway, not very many of us will bother to crack the case to examine the die. In a sealed epoxy or composite case things are rather more difficult.
So, buy from accepted authorized distributors only if you can. If you can't, then you better know enough to be able to (properly) measure hFE, leakages and capacitance on the junctions. Then you need to be able to test the device in at least on configuration appropriate for that device. It would greatly help if you have a current limited HV supply so you can measure breakdown voltage. I use a 6186C HP DC current source (300VDVDC compliance voltage) for that purpose.
I'll be honest, it isn't worth the risk and work to buy from unknown sources for me. Most parts I get today are fakes, even special purpose ICs! Why would anyone buy an op amp for high performance from an unknown source ????? !
It comes down to using your head and do not give in to wishful thinking. The odds are against you, face it. But, it's up to you. May as well spend the money on lottery tickets, at least you know the odds.
-Chris
Is there a reliable way (without acid) to remove the case so that the top of the die can be examined under a microscope? When I have crushed a few transistors I have not had much luck examining the top surface of the die afterwards. I can often see the size of the die but not much luck in accessing the top surface in one piece. Interested primarily in looking at some TO-92, TO-220 and TO-3P. More often than not I can see where the back of the die was bonded to the lead frame.
I am interested in knowing if the layout of the generic parts (made by one of the very largest manufacturers on the planet) resemble the layouts of the "original" historical parts. I also want to compare them to a couple of name brands because I am curious if the production is subcontracted.
I do indeed measure a great deal about the parts including a homemade curve tracer and then a 60 dB LNA (with 3x parallel 2N4403) to get an idea of the noise performance.
The homemade curve tracer is not like the HP 4142 SMU systems that I used years ago with Electroglas probers but it is ok. I miss being able to wander down to the failure analysis department for a decap.
Another part of the reason is that more and more of the "real parts" are going out of production so I am curious about the generics or copies from one particularly large manufacturer.
I do indeed measure a great deal about the parts including a homemade curve tracer and then a 60 dB LNA (with 3x parallel 2N4403) to get an idea of the noise performance.
The homemade curve tracer is not like the HP 4142 SMU systems that I used years ago with Electroglas probers but it is ok. I miss being able to wander down to the failure analysis department for a decap.
Another part of the reason is that more and more of the "real parts" are going out of production so I am curious about the generics or copies from one particularly large manufacturer.
Also some people use discontinued parts to repair stuff, so it's good to know if the parts are genuine...
Most parts even from Mouser etc....?
Most parts even from Mouser etc....?
Hi jpk73,
I do, and it is critical they are real. that's why I generally find a current substitute if possible and go that route. Today's transistors are simply better in every way. Since my name goes on each repair, I do not fool around.
I wouldn't think so. But NTE and replacement brands are remarks. Who knows what they really are?
What they really are is what NTE can get. If originals are available, they used to get them and stick their ECG number on them. The problem is when the original is NLA (which is most of them these days). Then the NTE you get is either a second source (which may or may not be up to spec), something “close” but generic, or an outright fake.
The best course of action is to find a current production version intended for the same service as the original (which will be “simply better in every way”) and use it instead of something which might be dubious.
The best course of action is to find a current production version intended for the same service as the original (which will be “simply better in every way”) and use it instead of something which might be dubious.
Hi,
Exactly.
My personal experience is that I used to have to sell ECG and use them in service at a place I was working. I got to a point where I stuck the leads through the bag and tested first. Most were nowhere near what they had to be. Once I started using "real" parts I substituted, or real originals, my return rate dropped to very near zero. The take-away is that NTE and ECG were not up to snuff. Sometimes even the cross reference was incorrect as well.
That's how the cross-reference manuals earned the name "the book of lies". The real hacks would cross reference a part to ECG, then something they thought would substitute (crossing in and out) which most times ended up with a transistor nowhere near the ratings it had to be. These people couldn't understand they were doing something wrong even after it is explained to them. I even used single syllable words spoken slowly and softly. lol!
Exactly.
My personal experience is that I used to have to sell ECG and use them in service at a place I was working. I got to a point where I stuck the leads through the bag and tested first. Most were nowhere near what they had to be. Once I started using "real" parts I substituted, or real originals, my return rate dropped to very near zero. The take-away is that NTE and ECG were not up to snuff. Sometimes even the cross reference was incorrect as well.
That's how the cross-reference manuals earned the name "the book of lies". The real hacks would cross reference a part to ECG, then something they thought would substitute (crossing in and out) which most times ended up with a transistor nowhere near the ratings it had to be. These people couldn't understand they were doing something wrong even after it is explained to them. I even used single syllable words spoken slowly and softly. lol!