I am getting a very mild, buzzy distortion of the bass end of the spectrum from my LM1875 chip amp. Many times I have checked the circuit and my wiring etc. over the past few days and have found no answers to my problem. I am using the circuit from here, Single Chip 25W Amplifier (Project 72), which doesn't seem to have any of the problems people have mentioned in other threads.
I am using a small piece of old, crappy veroboard, which I have checked multiple times for continuity, and everything seems alright. To be sure, I coated each of the tracks with solder, and rechecked the continuity, and nothing seems to have changed.
I am convinced it is the amp causing the distortion, although it is noticeably reduced when I connect it to two speakers in series, or use a transformer on the output to have a higher output impedance.
Thanks in advance for any help
I am using a small piece of old, crappy veroboard, which I have checked multiple times for continuity, and everything seems alright. To be sure, I coated each of the tracks with solder, and rechecked the continuity, and nothing seems to have changed.
I am convinced it is the amp causing the distortion, although it is noticeably reduced when I connect it to two speakers in series, or use a transformer on the output to have a higher output impedance.
Thanks in advance for any help
yes, 1.1 c/w (Heatsink to suit SY4084/SY4086 - Jaycar Electronics), mounted with a silicone rubber insulator, no thermal paste (TO-220 Silicon Rubber Transistor Insulator - Pk.20 - Jaycar Electronics). I'm not using thermal paste, because the insulators supposedly do the same job. Even so, the chip is barely getting warm
Hi,
if the noise is present all the time (i.e it is there when no music is playing) it is almost certainly power supply noise.
If it is present only when music is playing then check the speaker on another amplifier first, if the speaker is OK you may have instability.
This can occur even if the circuit is wire up correctly. The most common causes are the decoupling capacitors are too far from the amplifier or if you have crossed the input and output signals over causing parasitic capacitive feedback. So look at your circuit topology to see if this is the case.
If you can borrow a scope this would identify this problem with a signal generator as you would be able to see the fuzz of instability on top of the signal wave form.
Regards,
Andrew
if the noise is present all the time (i.e it is there when no music is playing) it is almost certainly power supply noise.
If it is present only when music is playing then check the speaker on another amplifier first, if the speaker is OK you may have instability.
This can occur even if the circuit is wire up correctly. The most common causes are the decoupling capacitors are too far from the amplifier or if you have crossed the input and output signals over causing parasitic capacitive feedback. So look at your circuit topology to see if this is the case.
If you can borrow a scope this would identify this problem with a signal generator as you would be able to see the fuzz of instability on top of the signal wave form.
Regards,
Andrew
That absolutely can cause it. Decoupling capacitors needs to be as close to the chip as physically possible. If you are dead set on using the veroboard, you need to use a star grounding scheme, and have independent ground returns for the decoupling caps, the load, and the IC itself, all back to a center point. This will help stop ground currents from making their way into other parts of the circuit.
If you have a DMM, you should check the AC voltage on the output with a load and without. Do this with the input shorted. This can help give you an idea if it's a load effect. An oscilliscope is invaluable, but yes they can be higher cost. Keep your eyes peeled for an older CRT oscilliscope, they can be had pretty reasonable sometimes.
Also, that part is designed to run at a gain of 10 or higher. Not doing so will give insufficient phase margin and it will oscillate.
Lastly, no part with a lot of bandwidth and gain likes purely capacitive loads. They form another pole in the feedback transfer function giving you another 90 degrees of phase shift, usually on top of the dominant pole used for compensation(if the capacitor is sizeable). Depending on the capacitor and the open loop output impedance of the part, this extra pole might degrade your phase margin enough to start oscillations. If your load is largely capacitve, add some series resistance between the output and the load to form a partial zero to correct for some of the phase shift.
If you have a DMM, you should check the AC voltage on the output with a load and without. Do this with the input shorted. This can help give you an idea if it's a load effect. An oscilliscope is invaluable, but yes they can be higher cost. Keep your eyes peeled for an older CRT oscilliscope, they can be had pretty reasonable sometimes.
Also, that part is designed to run at a gain of 10 or higher. Not doing so will give insufficient phase margin and it will oscillate.
Lastly, no part with a lot of bandwidth and gain likes purely capacitive loads. They form another pole in the feedback transfer function giving you another 90 degrees of phase shift, usually on top of the dominant pole used for compensation(if the capacitor is sizeable). Depending on the capacitor and the open loop output impedance of the part, this extra pole might degrade your phase margin enough to start oscillations. If your load is largely capacitve, add some series resistance between the output and the load to form a partial zero to correct for some of the phase shift.
Here is a post that covers a lot of things that need to be done in certain ways, to avoid problems:
http://www.diyaudio.com/forums/chip...-between-rf-ground-loop-buzz.html#post3290979
http://www.diyaudio.com/forums/chip...-between-rf-ground-loop-buzz.html#post3290979
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.