It does not matter what you are repairing. There are times you "may" want to check transistor DC gain (hFE). One thing that is missing from the discussion is how conditional hFE can be. I am going to take 2SD718-R as an example to setup test while mitigating the condition that must meet minimum requirement. Going by Toshiba, Yamaha, Pioneer, Panasonic and Mitsubishi training from good old days, it goes like this. The first step in testing a device is to obtain a reference data.
Then interpret the test graph to obtain data. In the case of 2SD718 and all high power transistors you know the Base voltage often is given at least 700mV (most are over 1.0V) or higher on schematics diagrams for power amplifiers. Just sticking a device into a test instrument of any kind without satisfying the condition could result a disappointment such as low gain, or unnecessarily noisy.
You might NOT be able to test the device at a collector current of 1A but from the graph you can see that the hFE value reaches closer to minimum of 60 when collector current is 0.01A (10mA) at 25 degree C. You don't want to use a low collector current as you can not be sure of device temperature which is not a room temperature. Since it is a none linear function, scalability is tricky but quick approximation is possible. Leader LTC905 and Tektronix 7CT1N recommend using 0.1mA or 0.2mA for base current for power transistors. You can always increase or decrease base current according to the trace and wave characteristics. See below.
Then interpret the test graph to obtain data. In the case of 2SD718 and all high power transistors you know the Base voltage often is given at least 700mV (most are over 1.0V) or higher on schematics diagrams for power amplifiers. Just sticking a device into a test instrument of any kind without satisfying the condition could result a disappointment such as low gain, or unnecessarily noisy.
You might NOT be able to test the device at a collector current of 1A but from the graph you can see that the hFE value reaches closer to minimum of 60 when collector current is 0.01A (10mA) at 25 degree C. You don't want to use a low collector current as you can not be sure of device temperature which is not a room temperature. Since it is a none linear function, scalability is tricky but quick approximation is possible. Leader LTC905 and Tektronix 7CT1N recommend using 0.1mA or 0.2mA for base current for power transistors. You can always increase or decrease base current according to the trace and wave characteristics. See below.
This DC current gain graph (right side) is very important: it the device should maintain its gain over the intended application range. The flatter and wider, the better (but wishfull thinking). Usable up to 8A max I'd say, above that the base current explodes (left graph).
As @Citizen124032 described it correctly, the following data points are useful for readily available curve tracers such as Leader LTC-905 and Tektronix 7CT1N. Following the manual carefully, for power transistors on average a base current (IB) of 0.1mA or 0.2mA is needed at minimum. Condition considered for two points is 25°C. The Leader LTC-905 is even unique such that it actually places a good bit of load that will warm-up the device under test for the same setting conditions as Tektronix 7CT1N.
Attachments
Has the Leader LTC-905 not been cloned? @STOXX not that I know of. It is quite possible.
I did have a typo on the previous post #3 JPG file. The table header was missing the current symbol (A). In this case the 2SD718 base current can be set at 0.1mA or 0.2mA depending on the transistor suffix R or O.
I did have a typo on the previous post #3 JPG file. The table header was missing the current symbol (A). In this case the 2SD718 base current can be set at 0.1mA or 0.2mA depending on the transistor suffix R or O.
Thank you for your service manual. Since I own the Leader LTC-905 I did have the manual. I also have my own copy made using NI Multisim drafting app.
The LTC905 has a fixed number of stair steps. I worked on it to change that so the number of steps are adjustable from step 1~7. But I went against it for now since parts for this unit are very scarce. I have asked my former coworker (Onkyo USA) in Japan Osaka to check on the parts.
The LTC905 has a fixed number of stair steps. I worked on it to change that so the number of steps are adjustable from step 1~7. But I went against it for now since parts for this unit are very scarce. I have asked my former coworker (Onkyo USA) in Japan Osaka to check on the parts.
I have never heard of a HD2529P logic chip or micro controller.
What the heck is it?
Hey, nice Multisim!
What the heck is it?
Hey, nice Multisim!
HD2529P J-K flip flop. probably a single flop in a package. They also used a discrete multivibrator. Nothing complex except the transformer and the switches. Exactly why it would not be easy to duplicate.
Would it not be easier and cost effective to use a small nickle ninety eight micro controller instead?
I would paste in a cmos jk flip flop. $.15?
Using a micro to square up the clock seems way over the top.
Using a micro to square up the clock seems way over the top.
When it comes to test instrument cost is not the question. The availability of the part in Japan was kind of interesting.
Micro controller? That requires a design from grounds up and I know where we are at this point given that UK and China are peddling inexpensive digital micro processor controlled transistor testers. They are good to some extent for small transistors and they do good in identifying pins.
Let me ask you something.
If it was as simple and inexpensive to make transistor curve tracers as the peddlers think, how come well known digital voltmeter (DVM) manufacturers did not make them. Example; Fluke which most of use for bench-top and portable DVMs don't exactly make curve tracers, but do so for solar panels. The next brands are Keysight Technologies and Tektronix which are recognized leaders in this field.
A proper design of transistor curve tracer shall have a means to enter transistor parameter from data sheet given by respective companies. Or at least a select switch to match test conditions. Another example of a good design is Sencore LC101, LC102, LC103 capacitor testers. They provide data entry keypads which the tester will use to enter data for comparison. Based on the entry, it will return GOOD or BAD.
Transistors don't have a fixed specification, rather a range specified by some sort of interval. For tens of thousands of transistors having their own independent data, the cheapest micro processor controlled design is to have a means for data entry just like Sencore did. Otherwise the instrument should have a massive memory to store transistors data to recall upon need.
At the end for most of US, the old test instrument such as Tektronix 7CT1N or Leader LTC-905 are still useful and do professional test. They allow a user to select test conditions matching data as we have so far discussed.
Micro controller? That requires a design from grounds up and I know where we are at this point given that UK and China are peddling inexpensive digital micro processor controlled transistor testers. They are good to some extent for small transistors and they do good in identifying pins.
Let me ask you something.
If it was as simple and inexpensive to make transistor curve tracers as the peddlers think, how come well known digital voltmeter (DVM) manufacturers did not make them. Example; Fluke which most of use for bench-top and portable DVMs don't exactly make curve tracers, but do so for solar panels. The next brands are Keysight Technologies and Tektronix which are recognized leaders in this field.
A proper design of transistor curve tracer shall have a means to enter transistor parameter from data sheet given by respective companies. Or at least a select switch to match test conditions. Another example of a good design is Sencore LC101, LC102, LC103 capacitor testers. They provide data entry keypads which the tester will use to enter data for comparison. Based on the entry, it will return GOOD or BAD.
Transistors don't have a fixed specification, rather a range specified by some sort of interval. For tens of thousands of transistors having their own independent data, the cheapest micro processor controlled design is to have a means for data entry just like Sencore did. Otherwise the instrument should have a massive memory to store transistors data to recall upon need.
At the end for most of US, the old test instrument such as Tektronix 7CT1N or Leader LTC-905 are still useful and do professional test. They allow a user to select test conditions matching data as we have so far discussed.