Anybody ever build an amp that oscillated between 750 to 1.2 kHz?
The first time I built this amp, 2 weeks ago, it worked perfect.
This time I did a little reworking and now it oscillates when its just sitting there with the input grounded. The only difference is that I have no heatsink compound between the transistors and the heatsink. Each transistor is on there own heatsink so there electrically isolated. I did this only because I didn't want to fiddle with the goop and the amp last time barely got warmer than room temperature and never really warmed up much this time.
The first time I built this amp, 2 weeks ago, it worked perfect.
This time I did a little reworking and now it oscillates when its just sitting there with the input grounded. The only difference is that I have no heatsink compound between the transistors and the heatsink. Each transistor is on there own heatsink so there electrically isolated. I did this only because I didn't want to fiddle with the goop and the amp last time barely got warmer than room temperature and never really warmed up much this time.
Thanks for all the responses.
Yes, I figured it would oscillate at a lot higher frequencys as well. When I had it hooked up the first time, it oscillated around 3 MHz, so I put a cap across the output, which took that away.
I am going to come in and build it again this weekend, making sure I take my time at each step.
I think I will put the proper heatsink goop on to isolate the transistors and try grounding the heatsinks.
Thanks again.
Yes, I figured it would oscillate at a lot higher frequencys as well. When I had it hooked up the first time, it oscillated around 3 MHz, so I put a cap across the output, which took that away.
I am going to come in and build it again this weekend, making sure I take my time at each step.
I think I will put the proper heatsink goop on to isolate the transistors and try grounding the heatsinks.
Thanks again.
Hi davidallancole,
Motorboating is normally a "plop .. plop .. plop" sound at a low frequency. Depending on wiring this can be higher in frequency. There are times that high frequency oscillation can be modulated by a lower frequency (blocking). This may make a similar sound.
So use you 'scope. Look at the power supply lines (B+ and B-) for similar waveforms to what's coming out the speaker terminal. Also check for a high frequency or bursts of high frequency on the supplies and output. Once you know what's going on you can track it down.
Jaycee has already pointed out that ungrounded heatsinks can cause these problems. It's happened to me before. That was easy to fix (thank goodness!). You will need to use insulators and grease on your outputs before grounding the heatsinks (sorry if I'm pointing out the obvious).
-Chris
Motorboating is normally a "plop .. plop .. plop" sound at a low frequency. Depending on wiring this can be higher in frequency. There are times that high frequency oscillation can be modulated by a lower frequency (blocking). This may make a similar sound.
So use you 'scope. Look at the power supply lines (B+ and B-) for similar waveforms to what's coming out the speaker terminal. Also check for a high frequency or bursts of high frequency on the supplies and output. Once you know what's going on you can track it down.
Jaycee has already pointed out that ungrounded heatsinks can cause these problems. It's happened to me before. That was easy to fix (thank goodness!). You will need to use insulators and grease on your outputs before grounding the heatsinks (sorry if I'm pointing out the obvious).
-Chris
The waveform of the oscillation looks just like how you describe the sound anatech. I put the scope every where and I see this signal every where. When the op-amp section is seperated from the output devices and the top transistors base is grounded, I don't see any of this happening. The power supply lines are pretty clean. I am going to put the resistor from the op-amp output to the base of the transistor.
Poobah, you were saying that even if the heatsinks are not hot, the devices might still be burning up. I put my finger on the case everytime I put my hand on the heatsink. Is the case a better measure of the temperature?
I am going to take all your peoples suggestions and use some careful planning and layout to build it this weekend.
Thank you again.
Poobah, you were saying that even if the heatsinks are not hot, the devices might still be burning up. I put my finger on the case everytime I put my hand on the heatsink. Is the case a better measure of the temperature?
I am going to take all your peoples suggestions and use some careful planning and layout to build it this weekend.
Thank you again.
Hi davidallancole,
Can you describe exactly what you see? HF bursts or not?
I don't think you should run the op amp off the same supply as the output. The negative supply terminal of the op amp is more sensitive to noise than the positive, but you need to pay attention to all of it. Pay close attentoin to the speaker current return path as well.
-Chris
Can you describe exactly what you see? HF bursts or not?
I don't think you should run the op amp off the same supply as the output. The negative supply terminal of the op amp is more sensitive to noise than the positive, but you need to pay attention to all of it. Pay close attentoin to the speaker current return path as well.
-Chris
Might want to check out your grounds. Make sure all the ground points of the amp are connected and you did not forget a connection.
I built an amp where the NPN and PNP output pairs had their own heatsinks, no insulator, just good thermal grease, and I don't get oscillation. I don't think your non insulated heatsinks are the problem. If you still are uncertain, you could connect a zobel (4.7 ohm resistor with 0.1 or 0.2 uf cap) to each heatsink and to ground.
I built an amp where the NPN and PNP output pairs had their own heatsinks, no insulator, just good thermal grease, and I don't get oscillation. I don't think your non insulated heatsinks are the problem. If you still are uncertain, you could connect a zobel (4.7 ohm resistor with 0.1 or 0.2 uf cap) to each heatsink and to ground.
Dave,
All other things being equal... your heatsink will run at the same temperature above ambient, with or without grease. But your devices temperature (not power dissipation) is function how well it is connected thermally to heat sink... the better the connection, the lower the temp of the device. A transistors life goes in half for every 11 deg C rise in temp. Always use grease or a silpad and don't get carried away when you tighten the screw.
All other things being equal... your heatsink will run at the same temperature above ambient, with or without grease. But your devices temperature (not power dissipation) is function how well it is connected thermally to heat sink... the better the connection, the lower the temp of the device. A transistors life goes in half for every 11 deg C rise in temp. Always use grease or a silpad and don't get carried away when you tighten the screw.
Well, the first time I built it, it worked perfect. I think its just an issue with the physical layout because I tried rebuilding it a bit and thats when all hell broke loose. Like I said a few times before, i will rebuild it this weekend and see what happens.
Thank you again for your suggestions.
Thank you again for your suggestions.
Well, I solved my problem. It looks like I had a case of dumbassness. I have a digital oscilloscope that displays the frequency in the bottom corner. I haven't been wearing my glasses lately (yes, an excuse) so I didn't notice the k in front of Hz.
So I took a .1uF cap from the output to ground and it totaly eliminated my problem.
Thank you all for your kind input.
So I took a .1uF cap from the output to ground and it totaly eliminated my problem.
Thank you all for your kind input.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.