Need help with a push pull amplifier problem

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Okay I built a push pull amplifier, there are other various filters before this part, but this is were the problem lies. The problem I get is a large current flowing through the two transistors, from 56+ to 56-. I have checked everything and it setup correctly. I tested the collector - emitter resistance with no voltage applied to the base and it reads at 0.5ohms. I no this is not good, it means a large (VERY large, enough to make pcb tracks flash and break) and once I figured this it became clear that the transistors where getting internal shorted, the questions is how. I used a new set of transistors and grounded the bases. Everything seamed to be working and the capacitors in the power supply where holding a voltage around 56+ and 56- whereas last time they were pretty much zero, due to the transistors shorting out. When I switched the power off there was a flash, a large current had flown and broken one of the tracks. I tested the resistance of the transistors and it was zero. I will make some other things clear. Each transistor is on its own and the cases of the transistors are not in contact with each other. In the attached diagram the tip42 is a npn despite the slight mistake in the emiter arrow direction. PS the circuit worked fine with 12v rails and 15K resistors instead of the 20k biasing resistors. Is it my biasing technique? The biasing resistors? No thermal runaway resistors at the output of the emitters?
All suggestions welcome thank you
 

Attachments

  • screenshot.png
    screenshot.png
    5.2 KB · Views: 417
Could it be that your transistors are counterfeit? I don't see how anything in the diagram could cause your problem. Where did you order them from? You could always break them open and see, though I heard there were chemicals you want to avoid in doing so.

Also: The TIP142 is rated for 100V. This means that at full output, your transistor has 112V across it. I don't know if this is enough to cause damage.

Is it a problem with the power supply? If your power supply suddenly dumps a lot of power into the circuit when it shuts off, that may be the problem.

- keantoken
 
The way this circuit is set up, the high rail voltages do no good at all. It can't and won't swing beyond the op-amp's maximum output voltage. If it's working on +/-12 (or 15) leave it there. If you want more output than that threre's a way to do it but the +/-56v power supply will need to be re-configured.

As to what blew the transistors, they're underrated, you got fakes, or both.
 
this circuit

and type of topology might work for a rail voltage of tops 24+24 volts ....

design with this type of transistors has a lot of troubles and especially at this rail voltage ....

you actually use a circuit that is designed for 20w while rails push it to 120 w

not working

search for a post called a darlington story you will get some information from there

smoky regards :hot: :hot:
 
thanks for the reply. I bought the transistors from digikey and wouldn't have thought you would get fakes from there. As far as my circuit design goes the +-56v rails are useless then if it will only swing from +15 to -15. This is the first time I have tried it on a pcb, I will try it again today but with 12v rails. If I get the same results then the problem must lie in the pcb because it worked on breadboard with 12v. Maby the base emitter voltage is getting to high. I will try it with 12v rails and If it works with then I will just go out and buy a 12v toroidal, a much safer voltage to deal with. If anyone has any other theory's please post them, as it could save on un-necessary spending. I didn't realise the transistors outputs would only swing from +-15, I thought they were amplifiers, although I know they will increase the available current to the driver? Is it the negative feedback design across the whole amplifier opamp, is that limiting it to +-15v swing? Without the higher voltage unless I use a lower impedance driver I wont get the power output I was looking for. The driver is a 12" sub woofer rated at 100w rms and it doesn't have a brand, the coil measures about 4ohms. So max power is about 45w not bad but not the best.

The real questions is why are both transistors shorting and a huge current flowing through them from +56 to -56. It didn't happen with the lower voltage on breadboard, so its either fakes or something else is wrong.
 
keantoken I just read your part on the 112v, yes you are right but its not 112 volts over each transistor. I just dawned on me that this might be the problem. I think I have figured it out. Asuming that push-pull operation is working correctly then when it swings to positive there will be a voltage at the emitter junction that could get to 56+ which would mean there would be 112 volts over the pnp transistor and it would blow. this would cause an emxtremly large current to flow through the pnp and when the npn is conducting a large current would flow through it, possibly causing it to blow. I will try using a lower voltage and see what happens.
 
yeap !!!!!

you need to get familiar with the basic of AB amplifiers ....
your approach and estimation of opinion from others is completely wrong ...

all of it .... swing, rails, and so on with the voltage you use the bias that is created for the outputs is astronimical ...this is what blows your transistor neither the quality nor the pcb ....

but really ...do you understand anything about bias ??? if not you need to read a bit ...or ask for more help ....

one way or another the particular circuit if ever manage to work will sound terible compaired to simplest amplifier existing in the forum ...

good luck
 
furlow said:
I didn't realise the transistors outputs would only swing from +-15, I thought they were amplifiers, although I know they will increase the available current to the driver?

They'll amplify current, yes. But in that configuration (emitter follower), not voltage. To get voltage gain, you need to take the signal off the collector, not emitter.

The real questions is why are both transistors shorting and a huge current flowing through them from +56 to -56. It didn't happen with the lower voltage on breadboard, so its either fakes or something else is wrong. [/B]


Disconnect the outputs entirely, and measure the voltage that would be applied from the bias network between the two bases. If that's more than about 2.2V, those transistors will both go into hard conduction. Emitter resistors (.2 to .5 ohms) will help keep the bias from running away as the outputs heat up, but if it's too high to begin with it's too high. The way you have it built, the higher the rail voltage, the more bias. That's probably why it didn't go poof at +/-12. Reducing the 51 ohm resistors will bring the bias down. Those trannies shouldn't be run at over +/-40V anyway, even with the bias where it should be. Fix *that*, and you're still stuck with "how do I get voltage gain?".
 
sakis, trust me it does not sound terrible, the audio quality was brilliant when I was using 12v and breadboard. Yes I'm making lots of assumptions. I think its my calculations, which were lazyly done using a piece of electronics modeling software. I should have done them my self as well. The real problem was when I bought my transformer, which was awhile ago I didn't understand at the time the dc voltage would be ruffly 1.4*AC. I "thought" I would be able to use but over looked many things. You learn from your mistakes. I will get a lower voltage transformer and look into ways of amplifing the voltage. Back to some reading. My many problem is I do A levels in the uk and don't have very much spare time to do electronics and therefore i forget things.

Thanks to everyone for all your help
 
furlow said:
You learn from your mistakes. I will get a lower voltage transformer and look into ways of amplifing the voltage. Back to some reading.


If your main power supply (was +/-56 but lower is prolly better) is floating - completely separate from the op-amp's supply with a center tap that can be isolated from ground, you can use the basic QSC circuit topology to add voltage gain to the output stage without adding anything to the circuit. If you search here you can find schematics of the basic concept - it's not hard to do at all.
 
sakis, trust me it does not sound terrible, the audio quality was brilliant when I was using 12v and breadboard.

Yes , a boosted op-amp can sound good ,but for a reliable,
stable piece of audio equiptment ,a little more support
circuitry is needed.I have breadboarded Syn's VSOP clone
and it works off of +-75.. same basic idea as the QSC but
diamond powered :D . (attached).
I run the Ic (lm353) off of its own lm317/337 supply.



wg_ski - there would have been pieces of opamp embedded in the walls/ceiling.
Or be breathing it in. :(
OS
 

Attachments

  • vsop1.gif
    vsop1.gif
    23.6 KB · Views: 179
ostripper said:


Yes , a boosted op-amp can sound good ,but for a reliable,
stable piece of audio equiptment ,a little more support
circuitry is needed.I have breadboarded Syn's VSOP clone
and it works off of +-75.. same basic idea as the QSC but
diamond powered :D . (attached)...
I think that's going a bit too far for a novice! A CFP output stage with a little gain is a nice and simple way to increase the usable rail voltage. Or a high-voltage op-amp like OPA445 (expensive, but really useful in general).
 
Mr. evil -I think that's going a bit too far for a novice!

+-56 v can "bite" pretty hard as well :bigeyes: , can also
turn a screwdriver into red hot balls of metal !

A novice should start at a LM3886 with +-32V rails or a
DX amp (+- 35v). Just a little "bite" :bawling:
In between is the lm4702 at +-55v with a pair of big darlingtons.
I do agree , +-70 or more is scary.. I tell family to stay away
from "deadly" power supply (led power indicators and bleeders
to discharge "death").

Easy and novice are subjective terms.
OS
 
one way or another the particular circuit if ever manage to work will sound terible compaired to simplest amplifier existing in the forum ...

.
If we are to suggest new circuit, I believe this is a good statement to focus on.

Try focusing on the fundamental reasons why the circuit will sound terible.:rolleyes:

There are many..........:smash:



BTW, could it be instability that is responsible for the smoke being let out of these cheap outputs? I seem to remember these sometimes were prone to this problem, particularly if they are fakes.
 
uh, guys..... i hate to point this out........ look at the original post, and carefully at the drawing....... uh, EF output, with the PNP collector connected to the positive rail, and NPN collector connected to the negative rail........ doesn't that look a bit odd? that might explain why the magic blue smoke refuses to stay inside the transistors............

not just a bias problem, but it's a reverse bias problem.........
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.