The TIP42 in bridged class ab amp is hot.

Thread Starter

mike_canada

Joined Feb 21, 2020
239
But that's when they're turned on?

I read on another post here that a transistor is considered hot if you can't keep your finger on it. Well I can keep my fingers on mine.
I don't notice any smell.

I would think the prime current limiter are the 82 ohm resistors?
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
guess ill have to add 4.7 ohm emitter resistors.

My last design had heat but not as much as this design but it didn't have emitter resistors but was using 7.2V. then again the resistors in the old design were 330 ohms and 56 ohms but if I use values too high then the sound output wont be loud
 

MrSalts

Joined Apr 2, 2020
2,640
guess ill have to add 4.7 ohm emitter resistors.

My last design had heat but not as much as this design but it didn't have emitter resistors but was using 7.2V. then again the resistors in the old design were 330 ohms and 56 ohms but if I use values too high then the sound output wont be loud
That's not going to help much. You need a few more things. What is your simulator saying the current flow is through Q1 and Q3 (collector current?
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
That's not going to help much. You need a few more things. What is your simulator saying the current flow is through Q1 and Q3 (collector current?
The emitter resistors made the waveforms worse.

When I disconnected the input, the PNP (tip42) collector currents are both the same: -381mA

For the NPN (tip41) collectors, its 381.4mA

I then did curent measurements of both transistor collectors with a 40Hz input waveform of 0.37V amplitude and the results are this:

c.png
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
You know what I think might be going on... I think there's a resistance in each TIP transistor that is going in parallel with the external resistances because the external resistances alone will give me worst case 7.2/82 = 87mA. 87mA x 7.2 = 0.632W

But based on simulator data, measurements with same values is 381mA and that x 7.2 = 2.74W which means TIP transistor needs a heatsink to be OK.

since I'm low on heatsinks and all TIP transistor holes connect to the same place electrically, would I be able to connect them all together using thick wire and solder as a giant heatsink and then connect that to the same signal on the PCB?
 

MrSalts

Joined Apr 2, 2020
2,640
You know what I think might be going on... I think there's a resistance in each TIP transistor that is going in parallel with the external resistances because the external resistances alone will give me worst case 7.2/82 = 87mA. 87mA x 7.2 = 0.632W

But based on simulator data, measurements with same values is 381mA and that x 7.2 = 2.74W which means TIP transistor needs a heatsink to be OK.

since I'm low on heatsinks and all TIP transistor holes connect to the same place electrically, would I be able to connect them all together using thick wire and solder as a giant heatsink and then connect that to the same signal on the PCB?
The tab on those transistors is connected (electrically) to the collector pin (middle pin). You would be shorting across the load if you connect all four. You can connect the left NPN to the left PNP but, your circuit may not be optimal.

Is this a class D or just a bridged Linear amplifier?
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
its a bridged class ab amp.

I was thinking only connecting the two TIP41's tabs together then that to VCC then the two TIP42's tabs together to ground. But then again would I get a bad ground loop or no?
 

Ian0

Joined Aug 7, 2020
6,739
The PNP Tip42 will be off while the TIP41 NPNs are on. But...

The R8 and R6 are pulling Q1 base up and turning on current flow.
Once current flows through Q1, the base of Q7 is pulled high to turn it on.
theat pulls the base of Q3 low (a PNP) and it turns on.
nowmyou have Q1 and Q3 on to connect 7.5v to ground (essentially).

this is the fragment that could be improved. Caps removed to show DC steady state current flow.
View attachment 277366
You‘ve not copied that correctly. The cathode of the lower diode goes to the base of Q3 - it’s a standard push-pull output stage.
 

Ian0

Joined Aug 7, 2020
6,739
View attachment 277363

Why would the TIP42 transistors become hot in a matter of seconds as soon as I start it up?
Why don‘t you read my post when you asked the same question in your previous thread.
https://forum.allaboutcircuits.com/threads/simplest-bridge-amplifier-with-transistors.189365/page-2
it’s post #29.
It’s hot because it’s over-biased. The voltage across the two diodes sets the Vbe of the TIP41, And for the Vbe that is it setting, the collector current is quite large. It therefore gets hot.
 

Martin_R

Joined Aug 28, 2019
132
It may help with long term thermal management if the bias diodes are in thermal contact with the output transistors.
Also try emiter resistors of 0.47ohm, not 4.7ohm as you previously stated.
 

Martin_R

Joined Aug 28, 2019
132
Just noticed a glaring problem with the circuit, the mid voltage point for the load is not defined, meaning a standing current can flow through the load. It's made worse by the fact that it's a bridged amp, and neither side is designed with tight DC performance.
 

Ian0

Joined Aug 7, 2020
6,739
The TS is determined to make a rubbish amplifier. SeverAl experts contributed excellent circuits to his previous thread, so he closed it and started a new thread with his same old circuit that doesn’t work.
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
Just noticed a glaring problem with the circuit, the mid voltage point for the load is not defined, meaning a standing current can flow through the load. It's made worse by the fact that it's a bridged amp, and neither side is designed with tight DC performance.
Can you elaborate more?

and if its because I have no capacitor attached to the load then I'm trying to get more power from my design and someone suggested making a bridged amplifier.
 

Audioguru again

Joined Oct 21, 2019
5,437
I think the TS wrongly assumes that the amplifier he builds with mismatched transistors will work exactly the same as the simulated circuit that has perfectly matched transistors. Also, he does not know that the simulation does not measure the temperature of anything.

The idle (no signal) DC voltages at the outputs of each amplifier on the bridged amplifier should be almost identical then a series output capacitor is not needed.
 

Martin_R

Joined Aug 28, 2019
132
As in post #17, the voltage gain transistors Q5 and Q7 will ultimately determine the no signal operating point at the emitters of the output transistors which should be 1/2 supply rail. The 2N2222 specified have a gain spread from 50 to 300, and just the 22k resistor to bias them. The operating point will be all over the place, and also the gain alters with temperature too. A simple amplifier like this will never have stable operating points to make bridging possible.
 

Audioguru again

Joined Oct 21, 2019
5,437
As in post #17, the voltage gain transistors Q5 and Q7 will ultimately determine the no signal operating point at the emitters of the output transistors which should be 1/2 supply rail. The 2N2222 specified have a gain spread from 50 to 300, and just the 22k resistor to bias them. The operating point will be all over the place, and also the gain alters with temperature too. A simple amplifier like this will never have stable operating points to make bridging possible.
Bootstrapping increases the positive-going output swing. Therefore the idle voltage should be 0.4V higher than half the supply voltage.
 

Attachments

Last edited:

Thread Starter

mike_canada

Joined Feb 21, 2020
239
The idle (no signal) DC voltages at the outputs of each amplifier on the bridged amplifier should be almost identical then a series output capacitor is not needed.
When I ran such tests with and without the speaker (600uH inductor and 6 ohm resistor in series) the voltages at each output point is exactly the same. I also use the exact same components for both amps that I bridge together.

I am aware that parts have what are called a tolerance meaning that the values for each part aren't necessarily dead-on and I think its possible that one transistor can behave somewhat a bit different than another transistor of the exact same part number, but don't they have some standard (like a tolerance) to comply to?

And yes I make an effort to use bootstrapping in my amp circuit.

I should probably look at increasing the pull-up resistors from 82 ohms and suffer with slightly lower volume because my batteries might not handle the job with such low resistor values. I'm running the unit for 90 minutes at a time and my batteries are 7.2V and have capacities of 6800mAh and 4000mAh.
 
Top