First time on this site. I just registered. If I make some mistakes, please forgive me.
I have a board set up with 16 - 2N3055's. I hoped to use them as switches to power some coils. I have designed a TTL logic circuit that works well and gives me the logical highs and lows to turn on and off the transistors. My problem, I know very little about transistors. I'm not sure what I need to interface them from the TTLs. The 3055's will be switching 12VDC coils at about 2.5 amps.

 

Do you have a schematic of what you're planning?

It's always good to post your most current schematic when you start a thread; that way it saves a lot of questions back and fourth.

2N3055's have a very low hFE (gain). In order to get the transistor pretty well saturated with an Ic of 2.5A, you will need to supply the base with around 80mA, or an hFE of around 30.

If you used TIP120 transistors instead, you could drive them directly from TTL.

 

Do you have a schematic of what you're planning?

It's always good to post your most current schematic when you start a thread; that way it saves a lot of questions back and fourth.

Not sure how to insert a schematic.

 

Click on the "Go Advanced" button below the text reply box.
Then click on the "Manage Attachment" button near the bottom.
It's best to use .png format for things such as schematics; they will be compact and not distorted.

 

I have the schematic on AutoCad LT. I'll have to got to it and save it as a different extension.

 

Not familiar with Autocad. You might be able to do an export as a .png type.
If not, you might be able to do a screen print by pressing Ctrl+PrintScreen, then pasting it into Windows Paint.

 

I pasted it into Paint successfully and tried to attach. It keeps telling me that the upload failed. In Paint it is only about a 2 KB file. I'll try re-drawing it in another format and get back on here later.
Basically, I have a set of photo switches that count slots on a rotating cam. Approx. 400 inputs per second. The Photo switches feed directly into a 74LS04. The outputs of the 74LS04 feed another 74LS04and also to a terminal. The outputs of the other 74LS04 also go to a terminal. That is where I'm stumped. I don't know how to interface these to the base of the transistor.

 

Anyway, here's a couple of options for you.

One is using 2N3904 NPN transistors as drivers for the 2N3055 transistors; basically building a Darlington transistor from two discrete transistors.

The other is using TIP120 Darlington transistors.

The 74LS04 is just an example gate. To the left of the example gate is a simulated signal generator.

In both cases, R1 limits the output current from the TTL device to a safe level.
In the 2n3055, R2 limits the collector current of Q2 to a safe level.

 

The first example of the 2N3904 looks like it should work. I don't have any but should be able to buy locally. I'll try it.
Before I registered on this site, I viewed several of your replies to other questions and hoped you might respond to mine. hanks a lot!!

 

I went back over our conversation and noticed something I missed. Your first reply could have solved my problem.
"If you used TIP120 transistors instead, you could drive them directly from TTL."
I may try them.
Again Thanks

 

Well, as a last resort you can get them at Radio Shack. They have a couple of 15-transistor assortments; one NPN and one PNP. Might as well get one of each; they're handy to have around - still a bit on the pricey side, but one of their better buys.

The exact driver transistor to use isn't important, just as long as it's NPN, capable of an Ic of 100mA, and has an hFE of 100 or more. 2N2222, 2N3904, 2N4401, and many others would work fine.

You can buy transistors very inexpensively from places like Mouser.com, Digikey.com - even Fairchildsemi.com Direct has decent prices. Buying from authorized distributors or direct from the factory ensures that you're getting the genuine item. There are lots of "fake" semiconductors on the Web nowadays; if it sounds too good to be true, it probably is.

Let us know how you make out.

 

If you're going to be doing much fiddling around with transistors, it's a good idea to at least have a DMM (digital multimeter) that has a transistor test feature.

I bought a DVM810 somewhere years back for $10; very compact and surprisingly accurate. Don't know where you might find one nowadays.

Harbor Freight Tools has a couple of really inexpensive meters; you can frequently find them on sale for as low as $2.99. They're also amazingly accurate for the low price; and also have the hFE transistor test built in. Really comes in handy when you're categorizing random transistors. Just looking at a datasheet isn't enough.

 

I do use Mouser. It's suprising how cheap these componets are. The TIP120 sells for 46 cents and 38 cents if you buy 10 or more. Shipping is the bigest part of the expense.
I will let you know how it comes out.

 

I do have three Fluke DMM's and one Fluke Scopemeter. They are very handy. I find that when troubleshooting a circuit, it is very handy to have more than one meter to watch multiple points.

 

One thing I didn't really explain - if you're going to be driving inductive loads (eg:coils) then you really need to have EMF protection diodes (sometimesm called "flywheel diodes") to keep the transistors from getting fried. When you suddenly cut off current to an inductive load, the inductor's magnetic field collapses; the current tries to keep going, but it can't. What happens is that the voltage polarity across the coil reverses, and the voltage peaks can be extremely high if they are not clamped or snubbed. These peak voltages will kill semiconductors in an instant.

See the attached revision. The 1N5401 is a 100v 3A diode; they're even available at Radio Shack. It's attached so that it will only conduct when the polarity across the load is reversed.

Now C1 requires some explaining. Power rectifiers are slow to turn on. Without C1, the peak voltage could build up quite high before the diode had a chance to conduct. C1 takes some time to charge, so it basically slows the rising edge of the reverse EMF spike, giving the diode more time to conduct. Somewhere between 330pF and 1nF ceramic disc capacitor is usually good enough. But, if you happen to have 470pF handy, use that.

The diode becomes forward biased, and gives the inductive load a current path. This prevents the high voltage peaks.

It LOOKS like there is a diode in the TIP120 - but actually, the model I used is a TIP141, not a TIP120. The TIP120 doesn't have an integral diode - but I didn't have one in my library.

 

SgtWookie,
Yesterday you helped me with my problem with my transistor problems.
Since I can't buy parts until Tuesday I thought I would try playing with this configuration. The idea is:
When a logic high one set of transistors allows current through the load in one direction.
When a logic low the other set of transistors allows current through the load in the other direction direction.
The load is a coil creating a magnetic field that I need reversed at given intervals.
Could you take look and see if you think it will work using those TIP120's?
Thanks

 

I can't read your schematic.

There is far too much "white space" around it, and the characters are so small that they are all blended together.

What you really want is an H-bridge. You can build H-bridges with transistors, but power MOSFETS are preferred due to their extremely low ON resistance. They are far more vulnerable to ESD than standard BJT (bipolar junction transistors) are, which is their proverbial Achilles' heel. However, if proper precautions are observed, you can wind up with much more power transferred to the load, and much less being wasted in the drivers.

Try to re-post your schematic with less "white space". .png format is preferred over .gif and .jpg, as the former will result much more clear, and the latter will be quite fuzzy/blurry due to the "lossy" compression algorythm/format.

 

Here it is again. Hopefully it will work.

 

Even though I can't read the values on your schematic, I can make out the logic symbols.

All logic gates have what is called a "propagation delay". This delay may be anywhere from the nanosecond to the millisecond range. The consequence of this delay is that for a fraction of a second when the input signal level is changed, the output will not reflect the change in the input signal.

This can have disastrous implications in power control circuitry. If both the high side and low side of an H-bridge are on simultaneously, the result is commonly known as "shoot-through", or a dead short across the power supply. This usually results in blown transistors/MOSFETS, but can also result in the primary supply self-destructing due to the overload.

Dedicated H-bridge controllers have built-in circuits to cause what is known as a "dead time" delay; that means that both the upper and lower sides of the H-bridge are turned completely off for a brief period before one side or the other side is turned on. This effectively prevents the "shoot-through" condition.

MOSFETs have a huge advantage in their low Rds(on) statistics. This does not translate to common BJTs, which have "gain" and "saturation" statistics instead. Usually, a transistor requires 1/20 to 1/30 of the collector current to be in saturation. All MOSFETS need is a voltage level on their gate.

 

Here it is again. Hopefully it will work.

The only text that I can read is the 12VDC for the supply.