2N3055 performance problems

[maijaz99]

rca 2n3055

I have made a 12V DC to 280V AC inverter using 2N3055 in push pull amplifier design and stepup transformer. But performance is not upto expectations. with three transistors attached to both sides of center taped 12V winding, the circuit calims to give 500W output but in actual it gives around 160W ouput, while the output of single 2N3055 is around 115W according to datasheet. The transistors are mounted on fairly large heatsink and get hot when load is increased from 160W. Can any body suggest what is the problem with this?

Thanks

 

[FvM]

2n3055 specs

From a rough hand calculation, I see that the 2N3055 would be stressed up to it's nominal current limit (or above, considering a less optimistic efficiency). Did you design your circuit for the 2N3055 current gain expectable at high currents? It's rather low.

P.S.: Please avoid cross postings. There's a very clear statement regarding this in rules and points.

 

[maijaz99]

2n3055 circuits

The circuit is meant for inverter and a dedicated design for a load of less than or equal to 500W out put

 

[FvM]

2n3055 parallel

The circuit is meant for inverter and a dedicated design for a load of less than or equal to 500W out put

If you know that it's working by design, there's nothing to discuss here farther. I hope, you manage to convince your hardware, too.

 

[maijaz99]

inverter 2n3055

I discuss this cz there are no reasons that circuit should behave this way. I think there is no reason to be annoyed of discussion.

regards

 

[IanP]

2n3055 600w

What is the VA rating of your transformer?
It is best to choose a transformer which has a va (W) rating higher than your load ..
This way your load will never be starved of power when working hard ..

Rgds,
IanP

 

[FvM]

2n3055 darlington pair

Im absolutely not annoyed of discussions. The problem is, that you don't give any circuit information but want to know, why the cicuit doesn't work. I tried to discuss a possible reason, and you just tell, it's designed correct.

This may be the case, possibly. However, you stopped the discussion, I just noticed this fact.

 

[Borber]

modern than 2n3055

Current gain of your 2N3055 at high collector currents is too low or base driver is not capable to deliver needed base current. Because of that transistors does not saturate at higher loads and they get overheated. I expect that transformer has at least 25cm^2 cross-section.
This is what FvM meant by "manage to convince your hardvare".

 

[maijaz99]

2n3055 inverter circuit

I admit my mistake. I could not attach circuit diagram cz I had it at home and I posted first message from office. Ok its there in attachment for you. Some one told me that the transistor I used might not be manufactured in accordance with original specifications thats why it reflected performance problems. Is it true? Is this may be the cause? The transistor is labelled with following data

ST 2N3055
9M104 VW
9M 642
MYS

I just need to mention that I replaced transistor SK100 with BD140 cz SK100 is not available.

I use the transformer rated for 650VA

Thanks

 

[Borber]

how to use the 2n3055

I heard too that today's 2N3055 could be a fake but it does not matter. Measure transistor current gain before any other action!
As it follows from your posts you modified push-pull circuit by paralleling transistors. Are you shure you did it right?

 

[FvM]

2n3055 transistor specifications

The schematic gives much clarity. I understand, that you connected three 2N3055 in parallel as mentioned in the text. Paralleling of BJT without emitter resistances isn't guaranteed to give balanced current distribution normally.However, in this case, I think it's O.K., cause the CE saturation resistance achieves a balance.

I don't see any power or current notice in the schematic, and I basically doubt that it's capable to deliver 500 W in the present dimensioning. Calculating with typical 2N3055 current gains, you can easily see, why.

The pre-driver is supplied by 5 V and may source around 20 mA (3V/150 ohm) to 2N3055 driver. It has e. g. a current gain of 75 and sources 1.5 A, 0.5 A to each output 2N3055. It achieves e. g. current gain of 15 and 7.5 A output current. All three together carry 22.5 A. Neglecting transformer magnetizing current and any switching losses, and assuming only 2 V voltage drop for the darlington, you get 10V * 22.5 A = 225 W total output power. A bit more, than the reported 160 W, but actually, there are switching and transformer losses and a magnetizing current, too.

Furthermore, the 2N3055 may have a much lower current gain without being a fake product, completely within manufacturer specification. To be operational by design, a circuit must consider specified minimal current gain.

 

[maijaz99]

2n3771

Thanks FvM for such elaborated analysis. It appears I have to focus on BD140 to adjust its ouput between 50-60 mA so that it may drive the 2N3055 for appropriate power.

I have observed the circuit continuously draws 1.5 - 1.7 A from battery when no load is applied. :?:

Regards

 

[FvM]

oscillator am with 2n3055

Increasing the drive current is a way to get more output power from the circuit. I see, that SK100 has a 500 mA IC rating, cause it's operated as a switch, reducing the 150 ohm resistor to 33 ohm or even less should be possible without changing the driver. The SK100 3.3k base resistor should be also reduced to e. g. 470 ohm.

But please notice, that with 2N3055 absolute maximum current of 16 A, you still won't achieve the intended 600 W. Cause 2N3055 current gain is unspecifed above 10 A, a profound design would never calculate more than 10 A per output transistor, resulting in about 30A * 10 V = 300 W maximum output power. You need more paralleled or stronger (e.g. 2N3771) transistors.

Regarding the observed no load current. One drawback of the circuit is it's 50% duty cycle without any means to prevent simultaneous current flow in both output branches, e. g. delayed switch on. This may be tolerable for a low frequency inverter, but also the output transistor have no parallel base resistors, thus switch-off is rather slow. You should try, if 50 or 100 ohm in place of the 1N5402 reduces the idle current.

 

[maijaz99]

2n3055 dc-analysis

Since I am using BD140 in place of SK100, a 30 ohm resistance comes in my calculation too for 100mA output. Firstly I shall change the 150 ohm resistance with available value 27, 30 or 33 ohm and if there is any improvement, I shall add one or two pairs of 2N3055 transistors. I agree with you that I should not try to operate power transistor near its upper limits (15A). Increasing the number of transistors is a better way for more reliable and smooth performance.

Thanks and regards

 

[Audioguru]

simple inverter 2n3055

You posted the schematic of an extremely old circuit for a simple square-wave inverter.
TTL oscillators have not been used for 40 years.

Inverters today are inexpensive and use Cmos ICs and Mosfets.

The maximum useable output current of a 2N3055 treansistor is about 8A when its base current is 2.6A.
So two pairs will make an inverter with an output of only 154W.
But the very old circuit does not provide each output transistor with that much base current so it won't work.

 

[FvM]

calculate heatsink 2n3055

This sounds like the circuit would be bad, cause it's old. I guess, the circuit simply isn't designed for the itended power rating. As you experience at EDAboard every day, using modern parts as MOSFET doesn't necessarily assure better designs.

 

[Audioguru]

2n3055 control schematic

An old 2N3055 transistor has a max saturation voltage loss of 3.0V when its collector current is 10A and its base current is 3.3A. That is a lot of loss so the design must account for it so the output voltage will not be too low. The voltage loss times the current equals heating of up to 30W. The power output of the inverter will be very low.

A modern Mosfet like an IRFZ44 (there are better ones) has a max saturation voltage loss of only 0.28V at 10A and heats with only 2.8W. The Mosfet doesn't need any base current. The IRFZ44 can conduct with up to 30A.

Many electronic products incuding power tools don't work properly on the square-wave output of a simple inverter because they rely on the much higher peak voltage of a sine-wave from the mains.
You cannot simply increase the voltage of a square-wave inverter because then incandescent lights and heaters will draw too much current.

A "modified-sine-wave" inverter has a rectangular waveform which has the average and peak voltages the same as a sine-wave and most products work fine from it. Its circuit is a little more complicated than a square-wave inverter.

A pure-sine-wave inverter operates without making much heat and has a very complicated circuit. Special small ferrite high frequency transformers are used.

Many inexpensive Chinese inverters use a pure-sine-wave circuit and they have voltage regulation so the output voltage does not change when the load changes or when the battery voltage runs down.

I have never used an inverter because my electricity is very reliable.

 

[FvM]

step up mit 2n3055

I don't want to doubt anything you said about advantages of modern inverters. Personally, I also don't use inverters normally, except for some outdoor industrial projects.

Regardig 2N3055 data, I already mentioned previously, that the minimum specified current gain results in a very poor circuit behaviour. However, I've used these transistors at times, when few better devices have been available and of course, you can make profound designs with them. They once have been called work horse of electronics.

I don't have the original RCA data book at hand now, but I think the actual Onsemi datasheet is almost identical. It shows typical (not worst case) saturation voltage below 0.8V with Ic=10A and Ib=1A, which isn't that bad, I think. If you don't want to accept the worst case data, your free to select devices respectively to put some exemplars to the electronic scrap.

As another point, I notice that some EDAboard members don't have the rich choice of electronic parts we are used too.

 

[Audioguru]

2n3055 history

Maybe you have the luxury of ordering hundreds of 2N3055 transistors so that you can test, sort and throw away the weak ones then your circuit will work better.
But a manufacturer might have ordered all the typical and better ones so you might be left with all weak ones.

I design circuits using worst case specs so that every circuit works properly.

 

[FvM]

2n3055 driver

I design circuits using worst case specs so that every circuit works properly.

I'm under the impression, that you try hard to misunderstand any serious argument.

The above sentence sounds good, and it should be applicable for many of the almost simple DIY and beginner level projects that are discussed at EDboard all day. It should be applicable also for the present inverter circuit, at the expense of using some more transistors. Unfortunately it's not completely applicable for many more sophisticated designs, for various reasons. Thus I think, it's also, excuse me, a little - naive.