This is my first DIY project . I am happy with the sound . Except, I have a constant si--si--si sound if I put my ear near to the speaker unit . The background noise do not go up with the volume ,but I still think it could be a problem when listening . Any suggestion to kill the background noise ? Please .
Attachments
I used LM3875TF and followed Greg's NIGC design . The 50K pot is omitted as I use pre-amp to control music volume . I built it with cheap componets and hardwire them together . all carbon film resistors, NCC 1000uF and philips MKT box capacitors . I am using single Airlink 300VA dual 25V AC toroidal for both channels .
Attachments
I am using alumium enclosure for heatsink as well . The surface is warm to touch.
An externally hosted image should be here but it was not working when we last tested it.
I hardwired everythings together with the shortest length as possible . The red/black wire is +/- power . I used copper wires to connect groundings .
An externally hosted image should be here but it was not working when we last tested it.
Felix said:
Felix,
Looks very nice. I hope when you tryed the amp you used a heatsink .
First check if the noise is there if you disconect the amp from the preamp, but connect a resistor of 47k or so paralel to the input (you can connect the input to gnd too). It could be that the preamp is noisy and this capable amp amplifies everithing that shows at its input.
I didn't get that kind of noise level.
After you are sure it's not the preamp, make sure it's not the amp oscilating. Do you have the RC network paralel to the speaker output?
Last:
Try 10k/220k or 4.7k/100k FB resistors. It may decrease the noise.
Let's see if it's the preamp first that's making most of the noise.
My guess is for the preamp.
Greg
Re: Re: My First Gainclone have background noise
I disconnected the pre-amp and short the input to ground with the resristor . The noise is still here , but level is slightly lower .
I used the enclosure as heat sink . The heat generated is not that much . It is so nice to touch it in winter .GregGC said:
GregGC said:
I disconnected the pre-amp and short the input to ground with the resristor . The noise is still here , but level is slightly lower .
Yes, I have a resistor of 1 ohm in series with a .22uF box cap . I have tried not having the RC network . The chip become unstable . It sometimes have humm sound when power up. When I measure the DC . It has 1.2V at output !GregGC said:
I think that lower the gain would be benefit to my problem . Many thanks for your hints .GregGC said:
Two tricks for reducing noise:
1. Smaller resistors. Resistors have thermal noise which is proportional to resistance. Noise created in the gain-setting resistors does not change with the volume control. Reduce the two gain-setting resitors by the same proportion so that the gain remains the same. Also be aware of possible effects on the input impedance.
2. Add a small cap (around 0.1 uF) across the two power supply pins (V+ and V-) very close to the IC. I have found this to reduce hiss. Make sure that the cap is rated for the total voltage!
Also, if you disconnect the preamp and short the input to ground, does the hiss disappear or remain? If it disappears, then the pre-amp may be the source of the noise. If it remains, then the gainclone is the source.
1. Smaller resistors. Resistors have thermal noise which is proportional to resistance. Noise created in the gain-setting resistors does not change with the volume control. Reduce the two gain-setting resitors by the same proportion so that the gain remains the same. Also be aware of possible effects on the input impedance.
2. Add a small cap (around 0.1 uF) across the two power supply pins (V+ and V-) very close to the IC. I have found this to reduce hiss. Make sure that the cap is rated for the total voltage!
Also, if you disconnect the preamp and short the input to ground, does the hiss disappear or remain? If it disappears, then the pre-amp may be the source of the noise. If it remains, then the gainclone is the source.
Could you take a picture of the solderside too?Felix said:
This IC is a high current device and care must be taken how you design the current paths. I suspect that you have a ground/decoupling problem or problems with the output filter (oscillations). Have you carefully read the datasheet and the AN-1192 ?
Don't forget to check these too.
http://www.national.com/appinfo/audio/files/Overture_Design_Guide13.xls
http://www.national.com/appinfo/audio/files/Using_Overture_Design_Guide.pdf
Looking at the downside, where do you connect the power ground from the single(?) transformer? have you tried to test one channel at the time? As we have pointed out earlier you must or at least ought to, add 10-100 nF close to the power pins. One from V+ to ground and one from V- to ground, close to the IC. Try to implement the "star ground concept"
I hooked up one channel after another . Both channels have background noise. I will get some .1uF box cap for power rails and hope it will solve the problem . I am not sure if I understand 'ground star concept '. I am going to make a PCB that connect all ground to one point . Is this correct ?
Attachments
GregGC said:
Opps ! It seemed I have done somthing pretty wrong .
Okay. Let's try again 
Greg, I copied your schematic , but omitted 50K pot as I do not need it for power amp . There are resistor of 680Ohm and 470P ceramic capacitor at input . Should I omit them as well ? Can you suggest a schematic for power amp without volume pot .
Macboy, thanks for pointing out the ground route . It is very useful .
Thanks everybody for your support . You are so friendly .
Felix said:
Opps ! It seemed I have done somthing pretty wrong .
Greg, I copied your schematic , but omitted 50K pot as I do not need it for power amp . There are resistor of 680Ohm and 470P ceramic capacitor at input . Should I omit them as well ? Can you suggest a schematic for power amp without volume pot .
Macboy, thanks for pointing out the ground route . It is very useful .
Thanks everybody for your support . You are so friendly .
Depends a lot on what kind of preamp you have.
1. The gain of the GC you need?
The best to be between 20 and 30. In the setup below I used 22 (10k/220k NFB resistors R6/R4). You can use diferent res values here: 1k/22k 2k/47k; 4.7k/100k, 6.8k/150k. Pic the lowest res values with reasonable DC offset at the output. I think 10k/220k should be the optimum, but if you want to go with lower noise level, try the other ones too (4.7k/100k), though I don't think you'll get any less noise.
2. Does your preamp have a DC present at the output?
Measure with the DC voltmeter. If you build it and you know you have a cap at the output of the preamp, then don't use the input cap (4.7uF in my case).
3. HF noise.
If your GC has a separate case from the preamp I would add the input RC (in my sch. it's 15k/47pF in your case should be 680/470pF) just in case the the interconnects pickup some RF. The RC network should be as close as possible to the RCA, so that it kills the RF before it enters the housing of the GC.
If the housing of the preamp and the GC are the same one, then ommit the 680/470pF.
I'm posting 3 dif sch. for each case.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.