i have constructed the circuit that is attached but instead of bd908 & bd907
i use tip2955 & tip3055.
first problem
as soon as i get the power to the board the tda2030a in the twitter stage will becoming very very hot.
second problem tip2955 & tip3055 also will becoming very hot although i use heatsink for them . can you tell me what should i do and what test can i perform to find the problem????????
(without getting any signal to the board)
please help me
i use tip2955 & tip3055.
first problem
as soon as i get the power to the board the tda2030a in the twitter stage will becoming very very hot.
second problem tip2955 & tip3055 also will becoming very hot although i use heatsink for them . can you tell me what should i do and what test can i perform to find the problem????????
(without getting any signal to the board)
please help me
Attachments
Hi,
It sounds like your amplifer is unstable. I.e it is oscillating at some frequency by itself.
You circuit diagram does show power supply decoupling. If you do not have this then add this first. Power supply decoupling needs to be very close to the PSU pins or it will not work. Looking at your PCB layout it doesn't look like it is close enough. Try soldering the parts to the back of the PCB on the pins of the amplifer.
If this doesn't work, disconnect the power supplies from two of the amplifers and concentrate on solving the problem for one of the amplifiers first. Start with the tweeter amplifer. They may be affecting with each other so this may solve the problem. If it does then you need to isolate the ground returns from each amplifer back to your power supply. Looking at your layout this is quite likely to be a problem. Tyr reading up on star grounding as your ground return paths look very poorly controlled.
If this doesn't work then one approach to amplifer instability is to try to create a dominant pole in the response. This is a brute force method but it may work. Try about 100pF across the 2.2K feedback resistor. Be warned this could make things much worse so be ready to turn it off quickly if it starts heating up.
Many people put filament light bulbs in the power supply when first running up amplifers to limit the current. If the bulb lights then you have a fault. Look for a bulb suitable for your power rails a 40V or 50V bulb would probably be ok. (I have not used this method as I have a current controllable PSU so if someone else can advise you better on bulb choice take their advice)
Regards,
Andrew
It sounds like your amplifer is unstable. I.e it is oscillating at some frequency by itself.
You circuit diagram does show power supply decoupling. If you do not have this then add this first. Power supply decoupling needs to be very close to the PSU pins or it will not work. Looking at your PCB layout it doesn't look like it is close enough. Try soldering the parts to the back of the PCB on the pins of the amplifer.
If this doesn't work, disconnect the power supplies from two of the amplifers and concentrate on solving the problem for one of the amplifiers first. Start with the tweeter amplifer. They may be affecting with each other so this may solve the problem. If it does then you need to isolate the ground returns from each amplifer back to your power supply. Looking at your layout this is quite likely to be a problem. Tyr reading up on star grounding as your ground return paths look very poorly controlled.
If this doesn't work then one approach to amplifer instability is to try to create a dominant pole in the response. This is a brute force method but it may work. Try about 100pF across the 2.2K feedback resistor. Be warned this could make things much worse so be ready to turn it off quickly if it starts heating up.
Many people put filament light bulbs in the power supply when first running up amplifers to limit the current. If the bulb lights then you have a fault. Look for a bulb suitable for your power rails a 40V or 50V bulb would probably be ok. (I have not used this method as I have a current controllable PSU so if someone else can advise you better on bulb choice take their advice)
Regards,
Andrew
Last edited:
- Status
- Not open for further replies.