Nice board but where are the bridge rectifiers? The amp is good for 55 A! Are those mounted externally?
A hint Mayday, and others:
1 Make the heatsink ready for drill holes etc.
2 Mount all LM3886 to the heatsink.
3 When all seems to fit and all LM's are vertical and nice mounted solder them in two pins
4 Remove the pcb from the heatsink and solder the rest of the LM pins.
This method will get you LM's in a straight line.
A hint Mayday, and others:
1 Make the heatsink ready for drill holes etc.
2 Mount all LM3886 to the heatsink.
3 When all seems to fit and all LM's are vertical and nice mounted solder them in two pins
4 Remove the pcb from the heatsink and solder the rest of the LM pins.
This method will get you LM's in a straight line.
Is the design debugged properly or do you use the "big bang" method? Valid schematic? How about the grounding, any development aorund that? When you have so much currents available you can't just consider the groundplane as "ground". You must lead the currents where they won't hurt anything.
It's good if it can be confirmed. Saves you lot's of trouble.Mayday said:
About the 7915, soldermask is to regard as "uninsulated", mind that. The regulator must not be in contact with the groundplane. I usually removed the groundplane under such parts just because I don't want any trouble in the future.
Drilled holes, soldered in bindingposts.
Vincent I have a question for you. On each bindingpost there are two connections, one leading to ground.
The inner (closest to the chips) does not lead to ground.
When I hook this board up Loudspeakerterminals, how do I connect the wires? And please explain like you would to a five year old.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Vincent I have a question for you. On each bindingpost there are two connections, one leading to ground.
The inner (closest to the chips) does not lead to ground.
When I hook this board up Loudspeakerterminals, how do I connect the wires? And please explain like you would to a five year old.
Onboard capacitance seemed a bit low, so I decided to go for external bridgerectifiers and an external electrolytbank.
In from rectifiers:
Negative/GND/Positive:
The whole thing(I hope I remembered correctly how to face the caps):
And my worldclass soldering
In from rectifiers:
An externally hosted image should be here but it was not working when we last tested it.
Negative/GND/Positive:
An externally hosted image should be here but it was not working when we last tested it.
The whole thing(I hope I remembered correctly how to face the caps):
An externally hosted image should be here but it was not working when we last tested it.
And my worldclass soldering
An externally hosted image should be here but it was not working when we last tested it.
Looking good! The connections that are NOT connected to ground should be connected to the output terminals. There should be two posts per channel connected to the output. The + and - doesn't matter for either channel but make sure that if you connect the rightmost post to the output of a channel's +, that the leftmost is connected to the - of the other. In other words, DO NOT MIRROR + and -! It needs to have the same configuration on either side; otherwise, you will get stereo phase and you won't be able yo hear where the sound is coming from. It will come from inside your head! 
Just kidding, but It will sound weird and it will not be a good thing to let that happen unless that's what you re looking for.
The main difference between this amplifier and a normal parallel amplifier is that a bridged amplifier can have half of the amp working in phase with the other, and then connected to the output so then it can maximize the efficiency of the entire circuit and is more optimized for power. This technique is known as bridging.
Maybe not for a kindergardener but definitely explains volumes of information over here!
Hope this how-to helps!
-Vince
Just kidding, but It will sound weird and it will not be a good thing to let that happen unless that's what you re looking for.The main difference between this amplifier and a normal parallel amplifier is that a bridged amplifier can have half of the amp working in phase with the other, and then connected to the output so then it can maximize the efficiency of the entire circuit and is more optimized for power. This technique is known as bridging.
Maybe not for a kindergardener but definitely explains volumes of information over here!
Hope this how-to helps!
-Vince
Today the popyprops came in the mail. Haven't soldered them in yet, just trying out the layout.
I see that I need to move the left one a bit forward for easier access to inputs. Reason I've laid them down like this to get to bindingposts and other components better. And there's a lot of free space on the board.
I see that I need to move the left one a bit forward for easier access to inputs. Reason I've laid them down like this to get to bindingposts and other components better. And there's a lot of free space on the board.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Is it just me, or does the schema for zeroing out the DC offset seem to be a bit...useless? By looking at the DC condition, how does varying the 50K resistor do anything? All it would do is limit current, but since it's just a voltage reference it should have no effect. It might have some effect on HF cutoff frequency in conjunction with the 220pf cap, but DC? I don't see how that would work.
The diagram is here:
http://www.shine7.com/audio/bpa300.htm
I can confirm real world results after building this circuit - the variable resistor does nothing. Maybe a .1mv movement (on a 20mv offset) , but that could just be the result of drift as the circuit or psu settles/heats.
I think that this circuit would give a better bias correction. With the indicated values, a single turn of a 15 turn pot would give about a 25mv swing.
The diagram is here:
http://www.shine7.com/audio/bpa300.htm
I can confirm real world results after building this circuit - the variable resistor does nothing. Maybe a .1mv movement (on a 20mv offset) , but that could just be the result of drift as the circuit or psu settles/heats.
I think that this circuit would give a better bias correction. With the indicated values, a single turn of a 15 turn pot would give about a 25mv swing.
Attachments
No need to alarm. The used method of offset adjustment will work in many cases but it may not cover all cases. I wouldn't have chosen this solution because of this reason. What happens if you'll need more than 50k pot? When you'll chose this solution you must calculate for worst case plus some more and I fear the designer hasn't done that but still it's only to change the pot or some of the feedback resistors.
Wait: I'll notice that the pulldown resistor will share input bias current from 5 LM3886! The resulting current may be rather much => 50 k is the wrong value. I'll guess this was a typical debugging issue
Wait: I'll notice that the pulldown resistor will share input bias current from 5 LM3886! The resulting current may be rather much => 50 k is the wrong value. I'll guess this was a typical debugging issue
Well, I'll admit that I didn't use exactly the same values as the original schematic. To reduce Johnson noise, I lowered the value of the feedback and input resistors to about 47K and 2.5K respectively. I also raised the value of the feedback capacitor appropriately. The variable resistor I used was a 15K one. All parts but the big caps and the trimpot are SMD. I'm pretty sure my board layout and soldering is fine since this isn't my first time (about 70+ chipamps built) and all four of the amps work just fine individually, I just can't adjust the DC offsets.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.