Hello all,
I've built my proto type lm3886 on a bread board and after a testing it out I've come across an issue. When I turn up the volume the sound cuts off and on. Now I've built the basic schematic from the datasheet and haven't included the "optional components" as I'm still testing. So in regards to the issue my thoughts are that my amp may be clipping upon increased volume or it might have to do with my unregulated power supply. The reason my power supply isn't regulated is that I haven't found a way to regulate 45 volts as of yet.
My transformer is a toroidal 250va 30 + 30 with about 10,000 uf per rail.
Any thoughts or advice would be appreciated.
I've built my proto type lm3886 on a bread board and after a testing it out I've come across an issue. When I turn up the volume the sound cuts off and on. Now I've built the basic schematic from the datasheet and haven't included the "optional components" as I'm still testing. So in regards to the issue my thoughts are that my amp may be clipping upon increased volume or it might have to do with my unregulated power supply. The reason my power supply isn't regulated is that I haven't found a way to regulate 45 volts as of yet.
My transformer is a toroidal 250va 30 + 30 with about 10,000 uf per rail.
Any thoughts or advice would be appreciated.
Here's a quick guess. You don't have a big enough heatsink for the 3886's. So, when you drive them hard, they warm up, and the thermal cutout triggers, stopping the sound.
Update My Dynaco
Akitika GT-101 Audio Power Amplifier Kit
Update My Dynaco
Akitika GT-101 Audio Power Amplifier Kit
Interesting guess as you might be correct. I do have an undersized heat sink that I had to use because its on the bread board. I have a fan blowing on it as well to reduce the heat, and it did reduce the cutoffs but they still occur sometimes. I also made a small speaker burn as I was testing the cutoffs at high volume...
How long are the typical cutoffs? Mine are for a second or two before playing again and then repeating the process.
Last edited:
The power supply voltage is too high, the "optional components" in most cases are not truly optional, but necessary. I recommend you revisit the data sheet, read the whole thing through (maybe a couple of times), then come back with your questions. TomChr, a forum member here, has a nice web page dedicated to this very subject, I highly recommend you read it also. http://www.neurochrome.com/audio/?page_id=941#ThermalDesign
Mike
Mike
I've read the datasheet. I didn't expect the voltage to jump from 30v to 45 when making my power supply. The site I was reading said a transformer between 25-28 volts should be used, 3v isn't that much higher. The datasheet says you don't need a regulated supply and the chip can handle up to 84 volts as well.
I 30V-0-30V transformer will make 42.4264 Vdc.
30VACrms*sqr2 or 30V * 1.414 =42.4264Vdc out of the supply.
This exceeds the 84Vdc for the chip for a Bipolor Supply configuration (2x).
Even though there is some voltage drop across the rectifier diodes (typically .3 to .6 volts) it may not be enough of a drop.
Also what is the impedance that you are trying to drive as it sounds like spike is kicking in as mentioned.
When you are running them at such a high voltage it is advisable that you should use a Paralleled configuration (PA100).
FWIW
jer
30VACrms*sqr2 or 30V * 1.414 =42.4264Vdc out of the supply.
This exceeds the 84Vdc for the chip for a Bipolor Supply configuration (2x).
Even though there is some voltage drop across the rectifier diodes (typically .3 to .6 volts) it may not be enough of a drop.
Also what is the impedance that you are trying to drive as it sounds like spike is kicking in as mentioned.
When you are running them at such a high voltage it is advisable that you should use a Paralleled configuration (PA100).
FWIW
jer
I 30V-0-30V transformer will make 42.4264 Vdc.
30VACrms*sqr2 or 30V * 1.414 =42.4264Vdc out of the supply.
This exceeds the 84Vdc for the chip for a Bipolor Supply configuration (2x).
Even though there is some voltage drop across the rectifier diodes (typically .3 to .6 volts) it may not be enough of a drop.
Also what is the impedance that you are trying to drive as it sounds like spike is kicking in as mentioned.
When you are running them at such a high voltage it is advisable that you should use a Paralleled configuration (PA100).
FWIW
jer
Ohh I thought it meant 84v per rail...crap. I'm getting 44.38 volts per rail exactly. I plan on using 2 chips plus a few other circuits for my entire project and didn't want to cut the voltage to close. I haven't found a suitable solution to lower the voltage to a decent level either.
There are a few options but not many and they may or may not apply to your nreeds.
1#~ You can make a regulator or get the proper voltage transformer.
2#~ You can add some diodes in series with the rails but this would take a lot of them to drop at least 3-4 volts and they would have to be high current types, Not very cost effective but it works.
3#~ You can use 2 or 4 chips per channel in a bridged amp (BA100 or BPA200) configuration and run it on your 44.38Vdc supply in as a single ended supply configuration.
This would be great providing your transformer can supply enough current, else stick with the 2 chip design such as the BA100.
4#~ You can also use a single chip in a single ended configuration on your 44.32Vdc supply providing you use a output capacitor on the amp as well.
jer
1#~ You can make a regulator or get the proper voltage transformer.
2#~ You can add some diodes in series with the rails but this would take a lot of them to drop at least 3-4 volts and they would have to be high current types, Not very cost effective but it works.
3#~ You can use 2 or 4 chips per channel in a bridged amp (BA100 or BPA200) configuration and run it on your 44.38Vdc supply in as a single ended supply configuration.
This would be great providing your transformer can supply enough current, else stick with the 2 chip design such as the BA100.
4#~ You can also use a single chip in a single ended configuration on your 44.32Vdc supply providing you use a output capacitor on the amp as well.
jer
There are a few options but not many and they may or may not apply to your nreeds.
1#~ You can make a regulator or get the proper voltage transformer.
2#~ You can add some diodes in series with the rails but this would take a lot of them to drop at least 3-4 volts and they would have to be high current types, Not very cost effective but it works.
3#~ You can use 2 or 4 chips per channel in a bridged amp (BA100 or BPA200) configuration and run it on your 44.38Vdc supply in as a single ended supply configuration.
This would be great providing your transformer can supply enough current, else stick with the 2 chip design such as the BA100.
4#~ You can also use a single chip in a single ended configuration on your 44.32Vdc supply providing you use a output capacitor on the amp as well.
jer
I thought about the diodes but before I go that route I wanted to explore my options. How would I go about making a regulator?
One other option I didn't mention is that if you are using a toroid type power transformer it is possible to unwind a few turns from the secondary to drop your output voltage as well.
Typically the secondary winding is wound on to of the HV primary winding, But don't do this if you are not familiar with the construction and risks involved with transformers and the HV side.
I have done this and it is not that hard to do, but should be done only as a last resort as I wouldn't want anyone to get hurt trying it if one is not familiar about such things.
As far as Chip amps go you could use a TDA7293 instead as they are good for +/- 50Vdc (60Vmax).
http://www.st.com/web/en/resource/technical/document/datasheet/CD00001887.pdf
The TDA7294 is rated at 50Vmax as well but the data sheet advises only a +/- 40Vdc supply should be used.
http://www.st.com/web/en/resource/technical/document/datasheet/CD00000017.pdf
But, I think either would work on your supply as is.
jer
Typically the secondary winding is wound on to of the HV primary winding, But don't do this if you are not familiar with the construction and risks involved with transformers and the HV side.
I have done this and it is not that hard to do, but should be done only as a last resort as I wouldn't want anyone to get hurt trying it if one is not familiar about such things.
As far as Chip amps go you could use a TDA7293 instead as they are good for +/- 50Vdc (60Vmax).
http://www.st.com/web/en/resource/technical/document/datasheet/CD00001887.pdf
The TDA7294 is rated at 50Vmax as well but the data sheet advises only a +/- 40Vdc supply should be used.
http://www.st.com/web/en/resource/technical/document/datasheet/CD00000017.pdf
But, I think either would work on your supply as is.
jer
Check out some of these threads,
https://www.google.com/search?num=1...1.0....0...1c.1.52.hp..12.28.1296.PC67ZVAy5kk
Here is one such example,
Building an LM3886 Gainclone chip amp with a regulated power supply (PSU).
P.S. I must have two separate 30V winding's that are not connected together in order for the above example to work for you!
jer
https://www.google.com/search?num=1...1.0....0...1c.1.52.hp..12.28.1296.PC67ZVAy5kk
Here is one such example,
Building an LM3886 Gainclone chip amp with a regulated power supply (PSU).
P.S. I must have two separate 30V winding's that are not connected together in order for the above example to work for you!
jer
Last edited:
Crap. That's why I was wondering If there was a practical way to drop the voltage a bit. Most of the regulators including the ones in the above links have a max voltage of between 30 and 35 volts, and I'm pushing 44 for each rail.
see "There is regularly a need to reduce the mains voltage. In some cases, it's because where you live it's just too high and causes problems with electronic equipment. Sometimes, you might have a great transformer for a project, but the voltage is just a bit higher than recommended."
from Bucking Xfmrs
from Bucking Xfmrs
Crap as in I have to much voltage. I have looked at multiple sites but haven't found a way to reduce 44 volts. The highest example was 26Vac and mine is 30.
The LM117/317 HV series is capable of a higher input voltage of 60V max.
LM317HV | Single Channel LDO | Linear Regulator (LDO) | Description & parametrics
It could be used in the Decibel Dungeon design, which is the second link I posted.
As they are only rated at 1.5 amps you can either parallel a few of them or add a extra pass transistor in order to handle more current.
I have done this with the LM317's and it works well.
Since you only need to drop a few volts the power dissipation in the regulator won't be too high.
LT doesn't have any regulators that can handle anything over a rating of 45V and their models for that voltage range can only handle 500ma or less so a series pass transistor would still be required as well.
These links can give you an idea of regulator design and many examples of how to add a series pass transistor.
https://www.google.com/search?num=1....10.0....0...1c.1.52.hp..0.17.712.9yedqO4oynk
https://www.google.com/search?q=vol...9Ho2wggTXuILADg&ved=0CB8QsAQ&biw=1024&bih=678
Typically you can just use a 2N2955 transistor to any LM317 or Lm78xx circuit but it is a PNP type and may be getting harder to find.
A common NPN power type can be easily adapted instead with the addition of a small signal pnp transistor as well.
Regulators aren't really that hard to design but it does add to cost and complexity of your final circuit.
I have searched this heavily myself for a similar/same exact application but there just aren't many ready made chips out there to handle such an application, mainly not for the current rating required anyhow.
I have 10 transformers that are 35V !!
jer
LM317HV | Single Channel LDO | Linear Regulator (LDO) | Description & parametrics
It could be used in the Decibel Dungeon design, which is the second link I posted.
As they are only rated at 1.5 amps you can either parallel a few of them or add a extra pass transistor in order to handle more current.
I have done this with the LM317's and it works well.
Since you only need to drop a few volts the power dissipation in the regulator won't be too high.
LT doesn't have any regulators that can handle anything over a rating of 45V and their models for that voltage range can only handle 500ma or less so a series pass transistor would still be required as well.
These links can give you an idea of regulator design and many examples of how to add a series pass transistor.
https://www.google.com/search?num=1....10.0....0...1c.1.52.hp..0.17.712.9yedqO4oynk
https://www.google.com/search?q=vol...9Ho2wggTXuILADg&ved=0CB8QsAQ&biw=1024&bih=678
Typically you can just use a 2N2955 transistor to any LM317 or Lm78xx circuit but it is a PNP type and may be getting harder to find.
A common NPN power type can be easily adapted instead with the addition of a small signal pnp transistor as well.
Regulators aren't really that hard to design but it does add to cost and complexity of your final circuit.
I have searched this heavily myself for a similar/same exact application but there just aren't many ready made chips out there to handle such an application, mainly not for the current rating required anyhow.
I have 10 transformers that are 35V !!
jer
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.