Help!!! I keep blowing transistors!

[zking786]

It seems I've gone through 6 Transistors (should have used fuses!) with my circuit. Can someone explain to me why it's not working.

 

Here's the diagram:

http://www.i4at.org/lib2/inverter.htm

 

Here's what I've used in the circuit:

I've used resistors with 5% tolerance.

The transformer is 300W

The diodes are FR155 (should be equivalent to HEP 155)

I have used heat sink to secure the transistors (so as to avoid burning them out) and confirmed they work before running the circuit

 

I've observed:

A small spark at the battery's mains when I connect it, and thereafter it no longer sparks (something's blown thereafter).

The current voltmeter on the output end showed 10VAC briefly (like for half a second) the first time and then nothing.

The Base-Emitter junction doesn't pass on either transistor (therefore it must be blown)

The Base-Collector passes on both transistors, however, one yields a diode test value of around 500 and the other 001. I wonder why.

 

 

Can someone please help me by explaining my obervations?

[YT2095]

impossible to determine without looking at YOUR circuit, there could be a million and one things going on.

[zking786]

Well, assuming I hooked it up correctly, what could be wrong?

 

Also, is there a problem with the schematic, or should it work?

 

Lastly, could the tantalum capacitors be blown (due to the heat of soldering)? What could be blown, aside from the transistors?

[YT2095]

hang on a second, a thought just occured to me, are both transistors mounted on the same metal heatsink?

 

edit: rather than just leave my question like this for a yes/No answer, I ant you to perfom a continuity check between the metal mounting plate and the transistor case, do that for both of them and let me know if it shows continuity, as well as my 1st question

 

as for the tantalums, they do take some punishment, but will NOT tollerate reverse bias! not even a little bit compared to other types.

[zking786]

As for question 1, I don't think both transistors are not mounted on the same metal heatsink. I have them mounted on a single breadboard, though. They are not touching eachother.

 

As for question 2, when I do a diode test using the multimeter, I get a low value (001), so the transistor's case and the metal mounting seem to be joined.

 

Also, on the tantalum it has + and - in small print. I believe I've plugged them in correctly, but was wondering if my poor soldering skills could have blown them (that is, the extensive exposure to heat).

 

Should I post pics of my actual circuit? Would these help?

[YT2095]

yes it would help tremendously!

close-ups as well if possible.

[zking786]

I'm very sorry for the delay. I couldn't find a memory stick and had to use a phone's camera. Will work on getting better pictures if these are inadequete.

 

 

 

 

 

 

[zking786]

Also, it might be important to note that a website I've seen is telling me that the circuit, itself, is flawed and won't work. I've even been told that the capacitors should be switched in direction to work. Interestingly, I've blown transistors only (I didn't hear any 'pop' from the transistors or the diodes, as a matter of fact, not even the transistors).

 

Anyway, any analysis you can provide will be helpful.

 

Thanks!

[YT2095]

the cap polarity is indeed backwards in that cct diag, I actualy built a small version of this cct this morning and it worked just fine for me, the difference is placed my caps the correct way (without thinking, as I can build an astable almost blindfolded), the cct diag presented is flawed in that respect.

I didn`t use the diodes either (naughty me), but it was only low powered version with little back EMF.

take a look here: http://www.play-hookey.com/digital/experiments/rtl_astable.html

you`ll note that the only difference (other than the values) is that the 2x 1k resistors are simlply replaced by the transformer winding(s).

 

I was actualy going to suggest that you tested your existing cct with resistors instead of the transformer, when that`s working (use a little speaker to test it with) Then put your transformer in place.

 

keep me updated of your progress

[zking786]

Hopefully this is the right thread

 

Wouldn't it have blown the capacitor if the polarities were reversed? Earlier, you mentioned that tantalums are really sensitive when it comes to polarity.

 

Also, I was reading your explanation about how the astable multivibrator works and didn't completely understand it. Firstly, why is it that when one transistor is off the other turns on. Secondly, why aren't both capacitors charged at the same time?

[YT2095]

well HOW it works is well beyond the scope of this thread and would take a Rather Large post that would itself generate alot more questions, so i`ll not even attemt to explain it, other than that a transistor is simply a Switch that uses less power to turn the switch On than the switch is capable of Carrying.

 

and yes, it`s quite possible (to likely) that these caps have been compromised.

 

I want you to measure the resistance of the transformer input that your using, I want the measurements taken fro the center tap to either of the outer.

then I want you to look at the cct diag I gave you last and build that with your existing parts on that board but replace only the 1k resistors with ones that match your transformer resistance reading.

let me know when you`ve done that

[zking786]

The resistance is 1.3 ohms. I definitely won't be able to find a resistor with a low enough resistance.

 

Also, what do you mean by the 1k resistors? The resistors of the circuit are for 10 ohms and 180 ohms. Which ones should I replace and with what resistors?

[YT2095]

see the cct here: http://www.play-hookey.com/digital/experiments/rtl_astable.html

take your 10 ohms out (and the diodes), and use your 10 ohms in place of the 1k in that cct keep you 180s in place as the 15ks.

so basicly you build the same cct with the parts you already have minus the transformer and diodes.

 

do that and get back to me

[zking786]

Wow! It works!!!! When I don't have it hooked to the transformer (without the diodes, as you indicated), it puts out ~23V when I test one of the terminals to the center tap using a voltmeter. It must have been the capacitors.

 

Why did you have me remove the diodes? Should I put them back in? Assuming my transformer can handle 300W, can I use this circuit without the diodes to generate 300W?

[zking786]

Before I get too excited, actually I already did , it seems something is amiss. When using a voltmeter to test each of the output ends to the center tap, I initially found about ~26VAC. The right bank worked perfectly and sustained a constant voltage. The left, however, didin't. It gradually declined from 26V to 15V, when I heard a slight pop (I'm not sure if it was from the circuit, though). It's important to note, that the circuit continues to output 15VAC consistantly, now, from that bank (hasn't stopped outputing current). The capacitors were only a bit warm, the transistors weren't even a bit warm, but one of the resistors was BURNING hot. The rubber casing seemed to be melting again. This resistor was the one diagonally opposite the terminal which was giving a lower voltage.

 

What does this mean? Should I replace the capacitor or the resistor? What's going on?

[YT2095]

personaly I`de rebuild the whole thing again with the caps in the correct way, making sure you check and double your wiring.

also, if a tantalum cap gets warm, it`s blown, it`s bin food.

a circuit like this has no built in redundancy, and so if One part fails then you have a runaway problem that`ll cascade and take down much of the others until the resistance becomes high enough to stop the failing.

as for testing it when your done, don`t start off giving it full power, half power will still give a result as the circuit`s so simple that it will tollerate a wide operating voltage below spec

[zking786]

I used all new parts to set it up (and put in the diodes) with TIP35C transistors instead of the 2N3005 ones. It won't work (no AC voltage generated). Interestingly, this time, one of the lower power resistors (180 ohms) got a bit warm this time.

 

This circuit is really confusing me! Is there a cheaper alternative to the tantalums, that can still work in a 300W circuit? The tantalums are starting to get expensive.

[YT2095]

sure, use electrolytics

 

infact that`s exactly what I used in my demo cct when I tested that diag of yours (I also used TIP3055s instead) and a 20-0-20 center tapped transformer, as I said earlier, it was only a low power thing, but electolytics will operate Just Fine

[woelen]

The transformer may also cause troubles. I have made circuits for driving motors. The inductive peaks, when the motors were switched on could be so powerful that semiconductor devices are blown out.

 

The voltage across an inductor is proportional to the time derivative of the current through the inductor. If the circuit is switched on and a large current is running through it at once, then you could obtain very high peak voltages.

[YT2095]

that`s the function of the reversed polarised diodes across the coil(s), to preclude this occurance.

[zking786]

Are the diodes placed correctly in the described circuit?

 

Interestingly, now I won't get any AC voltage (at either end). The resistors (10 ohm, 5W) seem to take turns heating up (when I plug in the circuit one of the resistors gets really hot, then when I disconnect it and plug it in again, the other might get hot). This time I used all new components. Any ideas why it's no longer working? I've tested the transistors and they aren't blown, the resistors show the correct resistance values, the capacitors are new. What could be wrong?

[YT2095]

the protection diodes in the cct diag you gave in the OP are indeed the right way around.

a TIP35c, I`m unfamiliar with that one personaly, I use TIP29 and TIP30`s of that series.

 

the biasing may be entirely out of range for that tranny, the HFE`s probably greater too, and thus would require less "switch on" current.

the fact that they take turns getting hot could actualy be a Good thing though, meaning that you`re Not far off, as an Astable takes turns in which tranny is On or Off.

 

it`s just So limiting having to try and explain in words/text smething that I could get up and running in 5 mins if it were on the bench here!

[zking786]

Isn't the TIP30 series PNP and not NPN?

 

I redid the whole circuit and it still doesn't work. Prior to installing each component, I checked its functionality. The transistors work (BE and BC diodes work), the capacitors charge when I plug them into a low voltage power source (1.5V AA battery), and the resistors all show the necessary voltages when hooked to a multimeter. What could be wrong? This time I even measured wire lengths of the wires I soldered to make connections.

 

Three things I was concerned about include:

1. The diode tests I conducted to ascertain the functionality of the transistors seemed to pass for both diodes, though I got differing diode test values for the transistors at the BC and BE junctions. For the BC junctions I get diode values of 521 and 596; for the BE junctions I get diode values of 475 and 576.

2. I got both transistors from the same place, though, they seem to be labeled differently. One says 2N3055E and the other says CDIL and then 2N3055. Are these fundamentally different? I searched the net and can't find much info on the 2N3055E.

3. Though the capacitors both charge to the same voltage when connected to a AA battery, is this really a good test for functionality? Could they still be blown?