I've just bought a whole bunch of MJ15003 from what I thought was a good supplier.
Now the element of doubt has been installed in my mind as they are marked Motorola.
I'm contemplating pulse testing one of them.
I've got an 80V transformer so I can supply it with ~115VDC
I've got 3 x 47R 300W resistors so can load it to about 1A.
Any thoughts anyone.
Now the element of doubt has been installed in my mind as they are marked Motorola.
I'm contemplating pulse testing one of them.
I've got an 80V transformer so I can supply it with ~115VDC
I've got 3 x 47R 300W resistors so can load it to about 1A.
Any thoughts anyone.
Attachments
Last edited:
Thoughts... well Motorola do make them, but if you have doubts beyond that then you will always wonder. A test like this proves nothing tbh.
I thought all MJ15003 were made by On-SEMI since 2000 ish.
Testing at 115V would prove they weren't re-badged 2N3055s. Beyond that I am looking for suggestions.
Testing at 115V would prove they weren't re-badged 2N3055s. Beyond that I am looking for suggestions.
Usually fakes will fail if tested well into their SOAR area ?
I chose 115V to separate it from the 2N3055.
Maybe I could go the other way and try it at 50V and 5A. I've got 4 x 47R 300W resistors. I've got 5 actually.
50V at 5A might not be so easy, I've got a couple of 35-0-35V 250VA transformers that hover around that voltage.
I chose 115V to separate it from the 2N3055.
Maybe I could go the other way and try it at 50V and 5A. I've got 4 x 47R 300W resistors. I've got 5 actually.
50V at 5A might not be so easy, I've got a couple of 35-0-35V 250VA transformers that hover around that voltage.
Last edited:
Could be NOS if that is the case. It's not me you have to convince 🙂
Just checking hfe and the fact it doesn't break down at 115 volts proves nothing. There are dozens, hundreds, of devies that could pass these simple tests.
Does the "good supplier" have a name or is it an ebay purchase ?
Just checking hfe and the fact it doesn't break down at 115 volts proves nothing. There are dozens, hundreds, of devies that could pass these simple tests.
Does the "good supplier" have a name or is it an ebay purchase ?
It was E-Bay but he's sold hundreds with no complaints.
Personally I've never had a fake from E-Bay by simply doing a bit of homework on the supplier, this guy is UK based.
http://www.ebay.co.uk/itm/191138460126?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
Personally I've never had a fake from E-Bay by simply doing a bit of homework on the supplier, this guy is UK based.
http://www.ebay.co.uk/itm/191138460126?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
Last edited:
at that price they should be perfect.
If they had been half that price I would be looking elsewhere.
You need to de-rate for Tc
If passing 100W through the device raises the Tc to 100°C then de-rate to
175-100 / 175-25 * Pmax = 75/150*250 = 125W
The test to 100W is using up 100/125 of it's guaranteed capability, i.e. 80% of maximum rating.
You must measure Tc and adjust Ptest accordingly.
If they had been half that price I would be looking elsewhere.
You need to de-rate for Tc
If passing 100W through the device raises the Tc to 100°C then de-rate to
175-100 / 175-25 * Pmax = 75/150*250 = 125W
The test to 100W is using up 100/125 of it's guaranteed capability, i.e. 80% of maximum rating.
You must measure Tc and adjust Ptest accordingly.
Thanks. I agree with Andrew, not cheap. CPC have ST branded devices (and as part of Farnell these should be 100% genuine) at under £2 including free delivery.
As to testing, I still say that you can't verify the devices provenance using home test methods.
Testing >100V at a few mA and in saturation at >10A will weed out the most common die substitutions. These were not super fast exotic diffusion devices
I think the question of authenticity can be just as easily answered by the age. Are these 14+ years old or the more usual sort, looking brand, spanking new? Is the batch number/date code consistent with this? Did you actually test the marks, case features etc?
For the price of the genuine article from a reputable selling agent, I can't see why you don't go directly to US distributors such as Mouser and Digi-key. It doesn't take much to add a few items and thus spread out the delivery charges for an order.
MJ15003G ON Semiconductor | Mouser
With due respect for their service to your local market, even distributors like Farnell and RS have been caught out with fake Motorola/ON, Toshiba and Sanken product here over many years here at least. I don't think they have been any less keen than unknown, small companies to profit from cheap, undocumented or unverified sources, when the opportunity arose. They haven't been so dimwitted as to still offer Motorola product to the service industry, though.
As I don't have the time or inclination to test every purchase and then dicker over returns, I simply don't buy items that can't be easily verified either from original packaging, marks or in service. All that means, is not buying semis, electrolytics, ferromagnetics, switches, relays or electrolytics on Ebay or from spare parts stockists whose lines are typically generic. Doing the odd repair still, I have to stand by the components and it only takes one goof-up to blow what little reputation I may have. The same issues apply to selling gear I'm no longer interested in.
For the price of the genuine article from a reputable selling agent, I can't see why you don't go directly to US distributors such as Mouser and Digi-key. It doesn't take much to add a few items and thus spread out the delivery charges for an order.
MJ15003G ON Semiconductor | Mouser
With due respect for their service to your local market, even distributors like Farnell and RS have been caught out with fake Motorola/ON, Toshiba and Sanken product here over many years here at least. I don't think they have been any less keen than unknown, small companies to profit from cheap, undocumented or unverified sources, when the opportunity arose. They haven't been so dimwitted as to still offer Motorola product to the service industry, though.
As I don't have the time or inclination to test every purchase and then dicker over returns, I simply don't buy items that can't be easily verified either from original packaging, marks or in service. All that means, is not buying semis, electrolytics, ferromagnetics, switches, relays or electrolytics on Ebay or from spare parts stockists whose lines are typically generic. Doing the odd repair still, I have to stand by the components and it only takes one goof-up to blow what little reputation I may have. The same issues apply to selling gear I'm no longer interested in.
Last edited:
I'd have thought that someone might have come up with a test circuit that only a genuine device would work as expected with a measureable method of failure ?
There is far more to it than that. What about switch off times, junction capacitance, reverse breakdown parameters for all the junctions, the list goes on and on.
Just because a device passes "hard" tests related to breakdown and SOAR doesn't make it the device you think it is.
You are obviously not happy using the devices and so the only thing you can do is put it down to experience and buy devices from a certified source.
Open one and check the size of the chip. There are photos here with which you can compare for at leat the size.
Gajanan Phadte
Gajanan Phadte
Looks reasonable.
The main point of using , say, MJ15003 instead of, say, 2N3055 is their improved power handling and SOAR.
Other parameters are important, but SOAR separates Men from Kids.
So far none of your tests checked this.
1) go to the datasheet and look for the "official" SOAR spec, which in this case is:
2) build a test jig which applies either 50V or 100V CE while making it pass 5 or 1A , with a monostable (555?) which applies a single 1 second pulse to base.
Mount the transistor to any heatsink with some grease, no mica , even an aluminum bar, you want just to add some thermal mass to the case, not dissipate continuously.
If the transistor survives, it´s the good one.
If in doubt, wait , say, 30 minutes for it to cool down *completely* and repeat.
3 or 4 times if you wish, just to prove it was not a happy accident, but always with 30 minutes between tests.
We have an old Gaucho proverb: "en la cancha se ven los pingos" meaning "you are only certain about a horse when you test it in the racetrack" .
Which is what we are doing here 😉
By the way, you don´t meed to test all of them, just 1 or 2 samples should be enough, if they all come from the same batch.
EDIT: remember to fuse the 50/100V line accordingly.
The main point of using , say, MJ15003 instead of, say, 2N3055 is their improved power handling and SOAR.
Other parameters are important, but SOAR separates Men from Kids.
So far none of your tests checked this.
If you want to:
1) go to the datasheet and look for the "official" SOAR spec, which in this case is:
2) build a test jig which applies either 50V or 100V CE while making it pass 5 or 1A , with a monostable (555?) which applies a single 1 second pulse to base.
Mount the transistor to any heatsink with some grease, no mica , even an aluminum bar, you want just to add some thermal mass to the case, not dissipate continuously.
If the transistor survives, it´s the good one.
If in doubt, wait , say, 30 minutes for it to cool down *completely* and repeat.
3 or 4 times if you wish, just to prove it was not a happy accident, but always with 30 minutes between tests.
We have an old Gaucho proverb: "en la cancha se ven los pingos" meaning "you are only certain about a horse when you test it in the racetrack" .
Which is what we are doing here 😉
By the way, you don´t meed to test all of them, just 1 or 2 samples should be enough, if they all come from the same batch.
EDIT: remember to fuse the 50/100V line accordingly.
Last edited:
- Status
- Not open for further replies.