Got some problems and questions about my recently built BrianGT lm3886 chipamp, didn't want to hijack several threads, so sorry for the long post right away...
1. Transformer humming - when I power up my amp I hear this relatively loud humming from the toroid which wasn't there a few days back. It hums for like 5sec, then quiets down for a second, and repeats that cycle over and over. Sometimes during playback it goes silent for some time, but will start humming all over again.
It's a 220V / 2x28V toroid, though my mains is 230V so the amp sees 2x40V or even a little more. Working on regs to bring it down to 2x35V or less.
2. When building carlosFM snubbered regulators - do I have to change the 100nf cap before the regs to 470nf as in his scheme or its not crucial? Also, is there a big difference in replacing the C1/C3 100uF PanasonicFCs with 1000-2200uF as he states or can I keep these?
3. Dual-mono/monoblocks or keep it stereo? How much of a difference is there going from stereo amp to monoblocks? I guess shorting the length of speaker cables should be beneficial by itself, but is it worth the cost of another psu and case, or is it better to spend that money on better caps, say nichicon FG, Elna Silmic II..? Though I'm not saying I won't do both...
1. Transformer humming - when I power up my amp I hear this relatively loud humming from the toroid which wasn't there a few days back. It hums for like 5sec, then quiets down for a second, and repeats that cycle over and over. Sometimes during playback it goes silent for some time, but will start humming all over again.
It's a 220V / 2x28V toroid, though my mains is 230V so the amp sees 2x40V or even a little more. Working on regs to bring it down to 2x35V or less.
2. When building carlosFM snubbered regulators - do I have to change the 100nf cap before the regs to 470nf as in his scheme or its not crucial? Also, is there a big difference in replacing the C1/C3 100uF PanasonicFCs with 1000-2200uF as he states or can I keep these?
3. Dual-mono/monoblocks or keep it stereo? How much of a difference is there going from stereo amp to monoblocks? I guess shorting the length of speaker cables should be beneficial by itself, but is it worth the cost of another psu and case, or is it better to spend that money on better caps, say nichicon FG, Elna Silmic II..? Though I'm not saying I won't do both...
I would stop using the amplifier and properly recheck all the transformer connections and ensure the transformer is working properly !
I checked all my connections and everything is fine. I can still hear the humming when the trafo is out of the amp and connected to the mains only.
A friend of mine heard it and said its not uncommon, that it could be a cheaper trafo made from metal plates and not a solid core. Any comments on this?
A friend of mine heard it and said its not uncommon, that it could be a cheaper trafo made from metal plates and not a solid core. Any comments on this?
If this is an indication of your safety awareness when working with mains electricity, then I suggest you stop Playing at amp building.
OK, Now that we have your safety awareness clarified.
Here is something I do. But be aware it does leave mains exposed and does require DMM probes being used to measure Mains Voltage.
Run the mains cable from the plug top to an insulated terminal strip. The type with the protective towers surrounding the securing screws.
Attach a 1r0 resistor across two spare terminals of that 5way strip. Insert a shorting link from the Neutral to one side of the 1r0.
Take the three wire flex from the L terminal and the Earth terminal and the other end terminal of the 1r0 to your equipment.
Wrap the whole terminal strip in insulating tape.
Screw the terminal strip to a backing board using the spare fixing hole/s between the terminals.
Now you can power up the equipment with the 1r0 in the Neutral lead.
The Probes of a DMM can be used to "punch through" the insulating tape down onto the securing screw head. You can measure Mains voltage. You can measure Volts Drop across the 1r0.
From these two measurements you can derive other consumption information.
When you withdraw the probes, you are left with small holes in the insulating tape and these are mm removed from the top of the securing screws. You cannot accidentally electrocute yourself if you are careless. Nor can anyone else who wanders in supervised.
But don't leave it plugged in and leave it unattended, just in case an unsupervised person decides to investigate.
DO NOT use any meter set to DCV to measure Mains voltage.
Here is something I do. But be aware it does leave mains exposed and does require DMM probes being used to measure Mains Voltage.
Run the mains cable from the plug top to an insulated terminal strip. The type with the protective towers surrounding the securing screws.
Attach a 1r0 resistor across two spare terminals of that 5way strip. Insert a shorting link from the Neutral to one side of the 1r0.
Take the three wire flex from the L terminal and the Earth terminal and the other end terminal of the 1r0 to your equipment.
Wrap the whole terminal strip in insulating tape.
Screw the terminal strip to a backing board using the spare fixing hole/s between the terminals.
Now you can power up the equipment with the 1r0 in the Neutral lead.
The Probes of a DMM can be used to "punch through" the insulating tape down onto the securing screw head. You can measure Mains voltage. You can measure Volts Drop across the 1r0.
From these two measurements you can derive other consumption information.
When you withdraw the probes, you are left with small holes in the insulating tape and these are mm removed from the top of the securing screws. You cannot accidentally electrocute yourself if you are careless. Nor can anyone else who wanders in supervised.
But don't leave it plugged in and leave it unattended, just in case an unsupervised person decides to investigate.
DO NOT use any meter set to DCV to measure Mains voltage.
Last edited:
DC offset help?
After reading about DC offset and some Peter Daniel's posts, I need some clarification about it. First of all - is DC offset measured per channel, or is it the difference between L and R channels?
I measured the offset with and without Ci. With Ci installed the offset on left channel is 0.4mV and on right channel it's -0.3mV (yes, negative, checked it a few times). So does this mean I have 0.7mV (difference) DC offset?
Without Ci the offset is -91mV on left and -68mV on right.
In both cases I took the measurements after leaving the amp on for about 15 mins, no input and no speakers connected, and there is no attenuator connected, so the amp is at full gain.
Can somebody comment on this? I'm guessing that the offset is way too much without Ci, but how much lower would you expect it to be when the volume is at say 1/4?
After reading about DC offset and some Peter Daniel's posts, I need some clarification about it. First of all - is DC offset measured per channel, or is it the difference between L and R channels?
I measured the offset with and without Ci. With Ci installed the offset on left channel is 0.4mV and on right channel it's -0.3mV (yes, negative, checked it a few times). So does this mean I have 0.7mV (difference) DC offset?
Without Ci the offset is -91mV on left and -68mV on right.
In both cases I took the measurements after leaving the amp on for about 15 mins, no input and no speakers connected, and there is no attenuator connected, so the amp is at full gain.
Can somebody comment on this? I'm guessing that the offset is way too much without Ci, but how much lower would you expect it to be when the volume is at say 1/4?
test output offset for ONE channel with the input shorted and the output open.
Then connect your Source and check again.
If you do not have a DC blocking capacitor on the input then you will find that the Source and different Sources will change the output offset.
You will also find that an unbuffered vol pot will also change the output offset.
You should be trying to find the worst case set of circumstances for all your connections that lead to worst case output offset.
If this is high then you have to decide if omitting the DC blocking capacitor is a risk you want to accept !!!!!
Another unhelpful post?
Then connect your Source and check again.
If you do not have a DC blocking capacitor on the input then you will find that the Source and different Sources will change the output offset.
You will also find that an unbuffered vol pot will also change the output offset.
You should be trying to find the worst case set of circumstances for all your connections that lead to worst case output offset.
If this is high then you have to decide if omitting the DC blocking capacitor is a risk you want to accept !!!!!
Another unhelpful post?
@AndrewT - I will keep the DC blocking caps since I will be connecting the amp to a lot of different sources and don't know what to expect. Also, you say to short the input, I read a few articles and posts on it and noone mentions shorting the input? Do you mean "not attenuated" or really shorted from hot to barrel?
***
So, after a few listening tests it became obvious that Ci cap is really restricting the amp's dynamics. The DC offset is in the 60mV range without it, but the amp plays incomparably better.
On the other hand, the 3,3uF MKT cap between V+ and V- also restricts some of the dynamics, but I wonder if it's the way I put them in place - I have those big green Vishay 100V MKT caps so I had to extend the leads to about 2cm to solder them - do I need to shorten those leads? The general advice is to solder them directly on the pins, so could this be a source of problems in dynamics?
These are not some high quality caps, does anyone hear a difference in sound when using higher quality compared to standard cheap ones?
***
So, after a few listening tests it became obvious that Ci cap is really restricting the amp's dynamics. The DC offset is in the 60mV range without it, but the amp plays incomparably better.
On the other hand, the 3,3uF MKT cap between V+ and V- also restricts some of the dynamics, but I wonder if it's the way I put them in place - I have those big green Vishay 100V MKT caps so I had to extend the leads to about 2cm to solder them - do I need to shorten those leads? The general advice is to solder them directly on the pins, so could this be a source of problems in dynamics?
These are not some high quality caps, does anyone hear a difference in sound when using higher quality compared to standard cheap ones?
Ceramic capacitors are designed for decoupling duty. X7R are periodically recommended, but any of the NOT np0 and NOT c0g will do the job.
Don't use plastic film caps for decoupling. They are too big and that ensures long lead and/or trace lengths.
Now go back and find out why Ci is not working properly.
When you have the correct type and the correct value you will find that it has virtually no effect on the sound quality.
Don't use plastic film caps for decoupling. They are too big and that ensures long lead and/or trace lengths.
Now go back and find out why Ci is not working properly.
When you have the correct type and the correct value you will find that it has virtually no effect on the sound quality.
Hmm, I put the MKT cap there because in the carlosfm scheme it particularly states so. What value would you suggest to use across V+ / V- then? Don't have any 100V ceramics at hand.
As for the Ci, it is Panasonic FC 47uF 50V, standard from BrianGT kit, didn't try any other value/type. I will try some more combinations with Ci, input caps and this V+- cap and see what I can hear
Cheers!
As for the Ci, it is Panasonic FC 47uF 50V, standard from BrianGT kit, didn't try any other value/type. I will try some more combinations with Ci, input caps and this V+- cap and see what I can hear
Cheers!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.