You should download the datasheet each time you buy and look for the specific package code. Newark sold me some TIP147 from On Semi that were in TO220 instead of TO247. They may handle the watts on a liquid cooled heatsink but not on my aluminum one. !@#)$*(&^
Power transistors are loosing gain when current increases.
Genuine TIP transistors are showing gain 60-80 on the same meter.
All other BJT transistor I'm testing with this meter also show gains matching their datasheets (at small currents)
I literally tested 100s of BJT power transistors with it.
If these fake ones show gain 12 with a small current, with bigger current gonna show even less.
All in all, I'm more concerned about the pinout of these transistors, not the gain.
Wrong pinout makes them unusable no matter what.
Have you tried them in circuit? Just to see what they do? I’d at least put them in and put voltage to them.
Dan
Dan
And lose gain at low currents. That may be the problem here. I agree that it looks suspicious, and I have already mentioned that the typical gain on a Bourns datasheet was 100 over the 1-100mA range, so we agree your devices are not tpyical. But that does not mean they are out of spec.
It is not true to imply that a gain of 12 at the current your meter measures the gain at means it will be less at higher currents. That depends on whether the current your meter is measuring is below or above the peak gain. If your transistors have a peak gain at 500mA it could well be that the gain at 1mA is low.
Do you know what current your meter uses to measure the gain?
All I was suggesting is that you make some other measurements to verify your suspicion as others have also said. And the collector-to tab connectivity.
But it IS the point, that´s why we suggest established measuring methods, instead of relying on an unknown meter programmer´s decisions.
I already mentioned at least twice, that I do not use this meter to measure anything.
This meter was only used to visualize the problem with pinout for the purpose of this thread; low beta was just an icing on the cake.
After I emailed Mouser executives, as Rayma suggested, I got a promise of refund,
and they investigating the issue to find out what exactly happened.
So at least will get my $40 back.
Kudos to emily.gay@mouser.com, their Service Excellence Manager.
Now the question - where should I buy genuine TIP2955s ?
and they investigating the issue to find out what exactly happened.
So at least will get my $40 back.
Kudos to emily.gay@mouser.com, their Service Excellence Manager.
Now the question - where should I buy genuine TIP2955s ?
LCSC... 50 cents each for JSMicro. https://www.lcsc.com/product-detail/Bipolar-Transistors-BJT_JSMSEMI-TIP2955_C2900598.html
https://datasheet.lcsc.com/lcsc/2109301730_JSMSEMI-TIP2955_C2900598.pdf
Or 1.50$ for STMicro: https://www.lcsc.com/product-detail/Bipolar-Transistors-BJT_STMicroelectronics-TIP2955_C262991.html
Or 3$ for onsemi... https://www.lcsc.com/product-detail/Bipolar-Transistors-BJT_onsemi-TIP2955G_C906636.html
Personally, I'd go with the JSMicro.
https://datasheet.lcsc.com/lcsc/2109301730_JSMSEMI-TIP2955_C2900598.pdf
Or 1.50$ for STMicro: https://www.lcsc.com/product-detail/Bipolar-Transistors-BJT_STMicroelectronics-TIP2955_C262991.html
Or 3$ for onsemi... https://www.lcsc.com/product-detail/Bipolar-Transistors-BJT_onsemi-TIP2955G_C906636.html
Personally, I'd go with the JSMicro.
Last edited:
Suppose one of their employees switches good parts for bad ones ...
You aren't supposed to withdraw fake money from an ATM, but there are cases reported...
It would be a major issue for their ISO 9001 as they would risk losing their trusted supply chain status
Not credible. There are much easier ways to make a lot more money.
Hold on to the old parts in case they want to examiner them.
I am glad that Mouser is now working with you.
If Mouser does come up with an explanation I would like to hear it.
It is not possible to detect pinout on "any" BJT for certain. Some Germanium "choppers" were truly reversible C for E. When there is a difference it shows in B-E breakdown or forward/reverse hFE at normal current. The ancestral 2N3055 was a BIG device with huge low-end fall-off. The modern "3055" and complement 2955 does better but still may be a worst-case that the tester-maker had not seen.
In industry, you buy a $1000 crate of parts, your Incoming Stock Inspector spot-checks even before they go into the warehouse. Trust But Verify. Yes, yes, I know, I have a bunch of DIY parts which turned out to be not what I thought I was getting....
If Mouser does come up with an explanation I would like to hear it.
...except pin-out connections!
Did you measure the connectivity between the centre lead and tab?
If the transistor tester on a cheap DMM is used with proper connection, how useful is the reading?
Well, that rather depends. I'm guessing that a cheap DMM puts a low current (e.g. 10uA) into the base and measures the collector current and that becomes the gain with suitable manipulation of the decimal point for a power-of-10 current. If that collector current coincides with the typical operating current of a transistor then a high gain would suggest the transistor is fine unless over spec. when it might indicate a leaky or shorted device. The readings I have made with small signal transistors have beenOK, but that is as far as I would go.
For power transistors we are talking typical currents of 1A or more which is well above what a cheap DMM could operate at. As we see here, if it is necessary to check a transistor then a measurement against the specifications is required, so that the conditions can be ascertained more accurately. In this case a low cost meter is not a reliable guide, I would suggest. If you get a high or believable reading that may be fine. If it says otherwise, it is wiser to check.
I would not use one for testing devices myself, as it is somewhat undefined. Usually, if I need the parameters, I need to know them at specific current /voltage.
Last edited:
So you are saying a tester with a proper current is the right unit for checking transistors.
Useful, possibly of great use when many transistors are to be tested.
But for $40 worth of transistors, I don't think making one is a good idea.
Useful, possibly of great use when many transistors are to be tested.
But for $40 worth of transistors, I don't think making one is a good idea.
I generally test all transistors I am going to use in an amplifier, and have built fairly simple test jigs for Vbe matching but sometimes just use my scope as a curve tracer for power devices. Ideally this needs to be a pulsed measurement for power transistors because the rate of temperature rise makes static Vbe measurements unreliable. So I feed the devices with a square wave low duty cycle driven base current. Even then the power devices are bolted directly onto a heatsink. Building test jigs for various measurements also enables Spice model parameters to be determined, if you need a Spice model for a device for which a model is not available or dubious. That includes using a high impedance FET input voltage amplifier/buffer for low current Vbe measurements.