I am reading through some material on output transistor selection. One of the suggestions is to get a bunch of a particular transistor and handpick the ones that measure an Hfe of at least 20 and have the closest to eachother at 15A. A nice simple task for a newbie/novice equipped with a soldering iron, a meter and a case of beer?
Greetings nonsub. Know the feeling. Very frustrating. There is a good design for a high current hfe meter to be found on the site of "Elliot Sound Products". That is a local diyaudio site here in Australia. Punch that into Google and it should pop up. It will be found in the "test instruments" section or something similar. Greetings from Australia on this 26th of Jan which is in fact AUSTRALIA DAY! Cheers, Jonathan
Edit. it is Project No. 106 in the Test Equipment section. Good luck.
Edit. it is Project No. 106 in the Test Equipment section. Good luck.
Last edited:
The device shown on the Elliot's site (Nice resource - thanks Jonathan, and happy Australia Day!) seems more complex than you need and only goes to 3A.
This is how I would rig a test stand. I ran a setup like this recently, but only @ 5A.
I have to say right off, this is my suggestion only, and I make no claims as to it's safety. At 15A there is plenty of power to burn things up including, but not limited to all the devices involved, jumper wires, and yourself. In other words, test at your own risk, and don't try it unless your are comfortable and confident of your ability.
If you really need to test at 15A, You can use a 12V source - like a car battery or charger (with filter caps) or a heavy duty bench supply. Use a pot as a voltage divider to bias the base, with a pushbutton between the wiper and transistor base. Rig up some power resistors (or make one from steel wire) that ends up a tad under .8 ohms (12V/15A = 0.8R) .5R should give you some range. You'll need to be able to handle 180W for your test duration, which probably won't be too long (15A X 12V = 180W (worst case)). Make a chart based on your resistor, and determine your target voltage across the transistor. Connect a milliammeter in the base circuit. Start with the bias pot all the way down, turn it up a bit and push the button, creep up on the target voltage that will give you the 15A collector current. Read the base current, and do the math. You'll probably have to heatsink the transistor under test. There's a lot of energy potential here, and the chance of burns, smoke etc. is very real, so I'd save the beer for a post grading celebration.
This is how I would rig a test stand. I ran a setup like this recently, but only @ 5A.
I have to say right off, this is my suggestion only, and I make no claims as to it's safety. At 15A there is plenty of power to burn things up including, but not limited to all the devices involved, jumper wires, and yourself. In other words, test at your own risk, and don't try it unless your are comfortable and confident of your ability.
If you really need to test at 15A, You can use a 12V source - like a car battery or charger (with filter caps) or a heavy duty bench supply. Use a pot as a voltage divider to bias the base, with a pushbutton between the wiper and transistor base. Rig up some power resistors (or make one from steel wire) that ends up a tad under .8 ohms (12V/15A = 0.8R) .5R should give you some range. You'll need to be able to handle 180W for your test duration, which probably won't be too long (15A X 12V = 180W (worst case)). Make a chart based on your resistor, and determine your target voltage across the transistor. Connect a milliammeter in the base circuit. Start with the bias pot all the way down, turn it up a bit and push the button, creep up on the target voltage that will give you the 15A collector current. Read the base current, and do the math. You'll probably have to heatsink the transistor under test. There's a lot of energy potential here, and the chance of burns, smoke etc. is very real, so I'd save the beer for a post grading celebration.
and how do you keep the current pulse @ <100us and keep the duty cycle @ ~5%? and still get meaningful readings?
And for good measure, how do you keep Tj & Tc = 25degC at the start of each test run?
And for good measure, how do you keep Tj & Tc = 25degC at the start of each test run?
Hi
If you have a pulse generator available and a oscilloscope.
You can solder a 1 ohm resistor on the collector and one on the base connect a power supply feed a pulse with very low duty cycle and measure across the 1 ohm at the collector until you have a 15v pulse then measure the 1 ohm on the base and you know the hfe.
Rob
If you have a pulse generator available and a oscilloscope.
You can solder a 1 ohm resistor on the collector and one on the base connect a power supply feed a pulse with very low duty cycle and measure across the 1 ohm at the collector until you have a 15v pulse then measure the 1 ohm on the base and you know the hfe.
Rob
Now that's a strange advice for output bjt selection. What's the point in matching hfe at 15A per bjt?
That amount of current will straightly burn out the degeneration resistors that are necessary for current sharing in paralleled output stages. I can't see how 15A could be a reasonable operating current for which it makes sense to match parts for.
I would get high hfe output transistors (MJW0281A and similar) and don't worry about matching. That is not really needed for bjts, mostly necessary for mosfet outputs.
Have fun, Hannes
That amount of current will straightly burn out the degeneration resistors that are necessary for current sharing in paralleled output stages. I can't see how 15A could be a reasonable operating current for which it makes sense to match parts for.
I would get high hfe output transistors (MJW0281A and similar) and don't worry about matching. That is not really needed for bjts, mostly necessary for mosfet outputs.
Have fun, Hannes
Hi nonsub. Btw what was the actual project you were looking at that required the output trannies?
Hi guys. The project is replacing NA001 transistors on NAP250s and also considering changing the BDY58s as well on the even older units. As far as I can make out BUV20 is possibly = NA001 and I want to stick as close as possible to the original spec. The specific advise I was given when I asked about matching was "... you do not need to match them as long as they have hi-current HFE close to each other (15 to 20 at 15A)..." so I am confused! Thanks!
Hi nonsub. I am not an expert on power tranny specs but I would have thought there was some predictable correlation between the Hfe at 15 amps and that at a lower current (say 3amps) that wouyld allow you to move forward on this.
A search around power tranny characteristics might save you some time (and grief!). I gather some recent ONSEMI units are marketed as having a fairly flat gain across the current spread. But I am getting out of my comfortable knowledge zone at this point......
And I guess some of us would be interested in the findings too.
cheers, Jonathan
A search around power tranny characteristics might save you some time (and grief!). I gather some recent ONSEMI units are marketed as having a fairly flat gain across the current spread. But I am getting out of my comfortable knowledge zone at this point......
And I guess some of us would be interested in the findings too.
cheers, Jonathan
Yeah theres no point in electrocuting myself on my first DIY attempts! OK I did check the specs for that transistor and they are min 20 max 60 Hfe. So I am still unclear what is meant by the suggestion like what is there for me to even check?
the datasheet often shows a graph of typical hFE vs Ic. This graph usually/always drops at very high current.
The data sheet also shows typical and min/max values for hFE @ particular Ic, sometimes quite a variety of Ic.
Go look at ONsemi MJ15003 cf MJL3281.
One sheet only tells us what to expect @ 5A and there is no graph. The other sheet gives min data for 0.1A, 1A, 3A, 5A, 7A, 8A, 15A and shows a graph for Ic from 0.1A to 24A. and for 3 Tj and for 2 Vce.
Note how low the guaranteed min hFE is, compared to typical hFE.
The data sheet also shows typical and min/max values for hFE @ particular Ic, sometimes quite a variety of Ic.
Go look at ONsemi MJ15003 cf MJL3281.
One sheet only tells us what to expect @ 5A and there is no graph. The other sheet gives min data for 0.1A, 1A, 3A, 5A, 7A, 8A, 15A and shows a graph for Ic from 0.1A to 24A. and for 3 Tj and for 2 Vce.
Note how low the guaranteed min hFE is, compared to typical hFE.
- Status
- Not open for further replies.