Substituting a transistor

I am not seeing your specs on the data sheet:

But to get back to your question, many substitutes might be possible, but to decide what specs must be well matched and which may simple be exceeded, you need to understand what the transistor is limited by in the circuit in question. If you want a substitute that will work in any application, all the specs must be equaled (at least). Besides collector voltage and current ratings, you might need to consider DC gain at low current, the current that has the peak gain and the gain at the current where the gain falls to something like half of the peak gain. You might have to be concerned with saturation voltage (if being used in a switching application), and collector to base capacitance and fT, if speed is a concern. And you might compare noise figure if that is important (and the two data sheets are at all similar in their definitions and test conditions).

Mr. Popelish has given good advice. But I'd like to know why you didn't provide the exact part number of the transistor you're replacing.

I'd assume you would probably be able to replace a TSxxxx with a BCxxxx. ;-)

If you want an almost certainly good substitute, you need to know what you're substituting. Make sure you equal or exceed all the relevant specs. And there are many other specs besides working voltage and maximum current.

But if there are no exact matches, then you have to get a circuit schematic and work out exactly what the circuit is doing. Figure out what's important and what's not. For instance, if you're using a transistor as a slow speed switching relay driver, f(t) probably isn't going to be too important. Nor are various capacitances. You'd be more concerned with h(fe), Ic(max), and V(ceo).

Many times, especially with power transistors and high frequency circuits, there is no good substitute.

Good luck Chris

** Nothing close to a BC547 here - even the pin outs are different. ** Transistor type number substitution is not a simple exercise.

The very first thing is to get the polarity, package and pin outs the same.

Vce, Ic and Hfe (beta) need to be the same or somewhat higher.

Then check the Ft (gain-bandwidth product) is similar.

Then the power dissipation figure, which can vary greatly even with the same package.

Some devices are specialised for radio frequency work or switching, so look at the maker's description to.

........ Phil

"Chris"

** Huh ???????

The OP mentioned it was a BC547 no less than three times.

........ Phil

Thanks, Mr. Allison. I read "I am going to trade a BC547 transistor for a TSxxxx transistor (Taiwan Semiconductors)." and somehow interpreted that to mean he was replacing a Taiwan Semiconductors something with a BC547, instead of the other way around.

My mistake. Thanks again for the spot.

Cheers Chris

This is very interesting for me. I have a datasheet from "Vishay Semiconductors" It shows:

50V,50V,6V,100mA

while your specifications from "Fairchild" show

45V;50V,6V,500mA

and I also found the specifications of "Philips"

45V;50V;??;500mA

I guess I was wrong when I thought I could just download circuits from the internet, buy the parts and assemble them. (When even a partnumber can have different meanings) But I'll go with the 45,50,6,500 as you showed me.

Thank you for your reply Chris.

TS13007CZ: Vceo = 400V Vcbo = 700V Vebo = 9V Ic = 800mA

I guess this could then replace the

BC547: Vceo = 45V Vcbo = 50V Vebo = 6V Ic = 500mA

Would this work? (By the way it will probably be used in applications like a PIC-microcontroller programmer and a USBtoSerialport Converter)

Thank you in advance BramGo

** A TO220 pack device - you f****ng moron.

** NO - you half witted baboon.

......... Phil

Hi, BramGo. I'm not sure why you're focused on Taiwan Semiconductor (unless you work there). There are a couple of problems with your possible substitution. Both transistors are NPN. However, the replacement you're suggesting is a much larger TO-220 package. While it exceeds the BC547 on voltage specs by an order of magnitude, it's a relatively slow transistor (f(t) = 4MHz). That's going to be an issue if you're trying to switch voltages at microsecond speeds. It won't switch as quickly. Also, the pinouts are wrong. Like a lot of power transistors, the TS13007 has a BCE pinout. That isn't close to the BC547 -- if you wanted to force it into an existing board, you'd have to twist around some leads. Also, if you've got an existing board, the physical board holes for a TO-92 transistor are too small for a TO-220 lead -- it may not fit in the holes.

NPN transistors of the BC547 type are called gumball parts -- they're commonly manufactured by many sources, and are relatively inexpensive.

The BC547 is a gumball-type part. If you look at the data sheet:

a couple of things come to mind. First, the NPN transistor part pinout is a little odd -- the leads on the TO-92 package are backwards from the standard EBC configuration (looking at the flat part on the front of the package, with the leads down, reading from left to right). The BC547 is CBE. If you have a circuit board you're working with, you can solve this problem by replacing with a more standard transistor, and putting it in backwards (like the NTE123A mentioned above). If you're doing a prototype, though, you don't have to worry about it.

The second thing that comes to mind looking at the datasheet is that if you're using this as a voltage switch at microsecond speeds, you might be able to do better. This Fairchild data sheet says maximum current is 100mA, with maximum power dissipation 500mW. Many TO-92 transistors can switch more current.

On the plus side, the BC547 is optimized for higher voltage gain and speed at that current.

If I were looking at this, I'd see if your application is using voltages over 40V (from your description, probably not). I'd then grab an NPN transistor called the 2N3904 (Vceo = 40V, Ic(max) = 200mA, all other specs similar) and just put the transistor in the board backwards if this is a replacement application.

Look at the similarities on the datasheets, and see for yourself. By the way, if you're going to post again, please mention if you're repairing an existing board, or building something new. Also please describe more about what your transistor is doing.

Good luck Chris

Foley questioned your apparantly arbitrary exclusion of all but a single source. I do as well.

$$Search?OpenForm

Greg Szekeres apparantly has a cross-reference page, but, having run it thru the WorldWide Web Consortium's HTML checker, I have to wonder what horrible Horrible HORRIBLE tool he used to construct it:

Here's an old dead-tree variant: ISBN 0-672-21335-4

The word "gumball" (aka "jellybean") was also used. That could be apt. The more details you provide, the better the advice you are likely to get.

Hi again,

Thank you for all your comments. And thank you for those hyperlinks.

The sequence of pins aren't really that important. I am using a prototyping board so it's easy to switch some wires.

Electronics is totally new to me to be honest. In september I started studying Industrial Engineering in the field of Electronics. Normally it would take 4 year to graduate in it. But I received an adapted scheme due to my previous studies in the field of Computer Science. So I am doing the first 3 years in just 1 year. The problem is: most of my teachers assume I allready had over 2 year experience with the basics of analog electronics. But I'm actually a total beginner. Next year I'll have to write a big paper and do a training period in order to graduate. So in preperation of that I am trying to make up the experience I don't have.

Choosing transistors is pretty complicated for me. Maybe it would be easier if I would better understand the circuits. I don't get the exact meaning of it in certain situations. (I do have a basic knowledge of what a transistor is: I know it can be used as a kind of amplifier depending on the current on the base.) But there are certain situations where I do not understand the importance of it. For example in the diagram on the following webpage:

It would be nice if somebody could give a little bit of background

It would probably "work", depending on your application, but you don't need anywhere that high of a voltage:

Try an MPSA06: Vcbo = 90V Vce0 = 80V Vebo = 4V