I need help eliminating a buzz, and a radio station.

[Hextejas]

I just finished this amp and though I paid a small of attention to the wire routines, apparently not enough.
There is a buzz from both speakers, the right more than the left. Plus I can hear faintly some radio station from both.
From what I read, I only have the amp and phono plugged into the same mains. The phono stage and pre amp are both battery powered.
The attached pictures show my rats nest of wiring. The other pictures show the designers layout in the case as well as his actual wiring.
In spite of the buzz and radio it still sounds terrific.
Oh yeah, I don't have the thiele thing in there.

Thanks

 

[nigelwright7557]

Radio pickup is usually due to lack of screening and/or no RF filter on input.

 

[Hextejas]

Radio pickup is usually due to lack of screening and/or no RF filter on input.

Thank you Nigel
Looking at the schematic and supposedly there are 2 RF filters that filter all except
FM: 87.5 to 108 MHz
AM: 535 to 1605 kHz
The filters are at the input pin3, of the LM3886 so is it possible that the radio is getting in after the amplifier?
The sound is very feint and it takes a struggle to hear it.

 

[tauro0221]

Hi,
One change I would do it is to change the input ground wire from the PS board to the board ground.

 

[stocktrader200]

it will be an AM station, a 68pf across the input(+) and feedback(-) pins on the 3887 should eliminate the radio. the hum is likely 120hz from power supply.
I would attach signal ground to the input ground then to main ground.

 

[Mark Whitney]

Hex, I am guessing your amp isn't the one shown in the third attachment. That is what it should look like with tightly twisted leads and a ground loop breaker.

 

[MarcelvdG]

You can make it easier for forum members to help you by posting or linking to the full schematic, assuming there are no copyright or other issues preventing that.

There are some experiments you can try to help locate the problem:

1. Disconnect the source, connect shorted RCA plugs straight at the input connectors and check if you still have a buzz and if you still hear a radio station.

2. Connect the loudspeakers with a much shorter cable than normal and check if you still have a buzz and if you still hear a radio station.

By the way, the Thiele networks are meant to keep the amplifier stable independent of the length of the loudspeaker cables and to suppress undesired radio reception via the loudspeaker cables.

 

[Hextejas]

Hex, I am guessing your amp isn't the one shown in the third attachment. That is what it should look like with tightly twisted leads and a ground loop breaker.

Thank you Mark and yes my amp is the messy wired one. I will beg some forgiveness as the original designer, when he called out the specs for the case and heat sinks, really steered me wrong. The case was way too small and the heat sinks were way too large. I eventually made them fit but it was a struggle. I have toyed with the idea of re wiring and re drilling it but that might mean moving the input and speaker connectors.
I have had such a hard time with this that I was very pleased with myself that I was able to get any kind of amplification out. It was an expensive and frustrating lesson but given what I know now, I Just might attempt it.
All the grief that I had was of my own fault.

Oh yeah, I think that I have included a ground loop breaker. It's that odd looking small circuit board that appears to be suspended in the air. I wanted to take a crack at point-to-point wiring.

 

[Hextejas]

You can make it easier for forum members to help you by posting or linking to the full schematic, assuming there are no copyright or other issues preventing that.

There are some experiments you can try to help locate the problem:

1. Disconnect the source, connect shorted RCA plugs straight at the input connectors and check if you still have a buzz and if you still hear a radio station.

2. Connect the loudspeakers with a much shorter cable than normal and check if you still have a buzz and if you still hear a radio station.

By the way, the Thiele networks are meant to keep the amplifier stable independent of the length of the loudspeaker cables and to suppress undesired radio reception via the loudspeaker cables.

Power supply and amp schematics attached.

 

[Hextejas]

Speaker wire routing ?

I started working on a small setup in order to fix the buzz. Then I noticed that as I moved the speaker wire, the buzz increased and decreased as I moved the wire. Hmmmm says I, so where should I route them ?
I took a picture of my speaker wires. It is lamp cord with banana plugs on the end. I remember reading somewhere that lamp cord made an excellent speaker wire. Now I am not so sure.
The speaker wire also is routed on top of a UPS so that might have something to do with it.
Ok then, what is an ideal speaker wire and how should it be routed ?



Thanks

 

[Aridace]

According to the data sheet, the LM3886 has a high CMRR so the hum is unlikely to be caused by the power supply. Sensitivity to wire placement would suggest a cause not mentioned yet: the amp is oscillating. Thus it can be rather sensitive to RF pickup that (via the usual non-linear behavior) results in (unwanted) radio reception. Check the amp with a scope, and when it oscillates, adjust the feedback network: start with R8=1k which might cure it.

 

[FauxFrench]

Thick lamp cord (normally thicker than ordinary lamp cord - some 3-5mm2) makes fine speaker wire. With the speaker impedance of 4-8 Ohm in one end and the amplifier dynamic output impedance (order of 0.1 Ohm) in the other end, coupling directly onto the speaker wire is unlikely.
The other gents, who have already given you advice, are much more experienced with using LM3886 amplifiers than I. The generic manner I use in case of humm is:
1) Short circuit the inputs of the power amplifier preferably through 1K resistors.
2) Disconnect any pre-amp or the like.
3) Turn all unnecessary electrical gear or appliances near the power amplifier off.
4) Turn the power amplifier ON with speakers connected. Does humm appear in the speakers? If yes, it is the amplifier itself that generates the problems. If no, the problems arrive from the external.
5) If no humm from the amplifier itself, start turning electrical appliances ON (not the pre-amp) and hear if any appliance radiates magnetic fields to be heard in the speakers as humm. I see some vacuum-tube gear - hmmm.
6) If no external appliances seem to radiate humm into the power amplifier, a first hurdle is passed. Remove the short circuits of the power amplifier inputs. Connect the pre-amplifier with short-circuited inputs (of the pre-amplifier) and turn it ON. Is the humm back? If yes, it is either the pre-amplifier itself (noise radiation into the pre-amplifier or interference with the pre-amp power supply) or interaction between the pre-amp power supply and the power amplifier.

This way you work your way backwards until you run into trouble and can more precisely identify the source of problems.
Humm and radio interference are difficult issues to handle because not only is it a question of which circuit you have implemented but also a matter of which exact components you use and exactly how you have implemented the electrical connections.
It must be a bit like predicting if you can send radio signals through a thunderstorm.

 

[Hextejas]

I
It must be a bit like predicting if you can send radio signals through a thunderstorm.

Hello again Faux. Do I need to worry about the pre-amp since it is battery powered?
I was under the impression that if I avoided AC power that I eliminated problems dealing with 60 cycle power humm .

 

[FauxFrench]

Hello again Faux. Do I need to worry about the pre-amp since it is battery powered?
I was under the impression that if I avoided AC power that I eliminated problems dealing with 60 cycle power humm .

Battery operation reduces the risk of hum a lot. Often humm arise from more power supplies in the chain. Humm may still be radiated into the pre-amplifier circuits.

 

[Hextejas]

it will be an AM station, a 68pf across the input(+) and feedback(-) pins on the 3887 should eliminate the radio. the hum is likely 120hz from power supply.
I would attach signal ground to the input ground then to main ground.

Thank you stocktrader .
I am a newbie to this and I confess that I cannot translate English into pin numbers. Call me embarrassed. I think that the input is pin 10, and the feedback is pin 9 ????

An interesting aside is that when I grab the speaker wire, the buzz diminishes considerably.
What does that tell you ?

 

[FauxFrench]

An interesting aside is that when I grab the speaker wire, the buzz diminishes considerably.
What does that tell you ?

Very often successful laying on of hands is an indication that not all metallic structures are well grounded. Make sure first that they are not connected to a different potential but then make sure the chassis and the heatsinks are connected to signal ground. Preferably in a star-structure. I guess you also have one point that connects signal ground to power ground?

 

[Hextejas]

Hi,
One change I would do it is to change the input ground wire from the PS board to the board ground.

Thank you taro, I am about to try this and am unsure of what exactly I need to do.
I see the input ground wire, ( actually I see 2 of them, L and R, ), on the PS board.
I am thinking that I would need to move these 2 wires (which originate at the RCA jack), and move their termination to the star ground, which is a bolt drilled into the chassis and the 3rd wire from the mains comes to it.

These grounding connections and discussions are a very confusing topic for me.
>really, would it be optimal to run all the ground wires to the star connection at the chassis ?

Can anybody help me here ?

 

[Mark Whitney]

With this design the star has been placed at the end of the PSU board.
You have posted photos of two amps. One with and one without hum. Rebuild the one to look like the other.

 

[tomchr]

Radio pickup is usually due to lack of screening and/or no RF filter on input.

+1

If you clean up the rat's nest so it looks more like the second picture you posted, your amp will have a fighting chance.

Looking at the schematic and supposedly there are 2 RF filters that filter all except
FM: 87.5 to 108 MHz
AM: 535 to 1605 kHz

Not sure what your point is there. All you need is a lowpass filter on the input. Assuming the signal gets in via the input, that is.

The filters are at the input pin3, of the LM3886 so is it possible that the radio is getting in after the amplifier?

Pin 3 is the output. I'm guessing you're referring to the Zobel and Thiele networks. Those aren't for RF filtering. They're there to ensure stability of the amplifier. I do recommend that you put the Thiele network in.

I'd short the input. I.e. connect RCA centre to RCA shell. If you still have the problem, the issue is within the amp.

You also mention a phono stage. Those tend to have high gain, thus be better "radio receivers". Have you eliminated that as the RF injection point?

Tom

 

[Hextejas]

+1

If you clean up the rat's nest so it looks more like the second picture you posted, your amp will have a fighting chance.

Unfortunately, other than starting all over, I will probably need to live with the wiring mess. It is getting better in that I did some rearranging. I learned a powerful lesson in planning ahead as to where to mount the external connectors.

+1
I'd short the input. I.e. connect RCA centre to RCA shell. If you still have the problem, the issue is within the amp.

ok, I will try this

+1
You also mention a phono stage. Those tend to have high gain, thus be better "radio receivers". Have you eliminated that as the RF injection point?

Ok, I will give it a try.
The hum has gotten a lot better and seems to be related to the position of the speaker cables. I might be able to move some of them so that they avoid the power cables. Part of my problem is that I have a subwoofer in the scheme and it is located about 10 feet from the amp.