AC mains noise filter for audio amp power supply

dudwity

Member
2006-12-13 9:40 pm
Hello,
im assembling a power audio amplifier based on commercial available PCBs.
toroidal power transformer, 2x NAIM NAP140 pcb's, rectifier and capacitor bank 2x15000uf, output protection PCB with delay relay for speaker outputs.

here's the question:
i have an AC line filter that i disassembled from a heavy duty xerox machine.
should i install it? pros? cons?
IMG_2952.jpg



thank you for your help.
 
Last edited:
That TDK thing on the right looks like just the thing. I can't read the writing except I see inductors in series with two lines coming in. Those are good. The toroid around one line coming in is good also, inhibits 1100 v spikes from large motor shut off and distant lightning strikes. Make sure the wires from the IEC socket to that are very short, if you can't salvage an IEC socket with that stuff built in.
I have a 5 lb EI transformer preceded by a 330 vac MOS supressor right by the power switch in the ST120 and I still get big pops if a lightning strike is nearby. But the CS800s switcher supply with no transformer pops louder. How much do you want to know about the weather from your audio amp? I want to know nothing. I see you're in a metal chassis, that is a good thing. Also a 11 turn coil around a china marker form (1 cm diameter) series the speaker outputs helps inhibit RF intrusion that way, also. Those are usually parallel a 10 ohm 3 watt wirewound (inductive) resistor. The fourier transform of a lightning pop has a lot of high frequency stuff I don't need to hear. That RF filter needs to be right near the speaker jacks to avoid making antennas of the wires from OT's to jacks.
Having the OT heat sinks inside the metal chassis is a little weird. You need to cut a rectangular hole & stick those out, or 2 round holes with a fan in one and a metal mesh and filter (to keep out spiders) in the other. Try to get airflow over the heatsinks.
I cut rectangular holes by drilling the 4 corners with 3/8" bit, then sticking a Stanley carbide saw through and sawing the straight parts by hand. The reason a 3/8" hole, the tangs on the saw blade are that wide. Use safety glasses with power tools. If you use a regular hacksaw blade, the holes have to be 1/2", which calls for a $99 1/2" chuck drill motor instead of a $69 3/8" chuck drill motor. Also bigger drills can warp sheet metal as they punch though. Drill big holes left handed to avoid spraining your wrist as the bit grabs when it punches through.
 
Last edited:

dudwity

Member
2006-12-13 9:40 pm
I bought in some cheap Chinese output transistors in my naivety.
They ran fine for 10 minutes until I unplugged my soldering iron.
The glitch on the mains got straight into the amp and popped the output transistors.
I guess the trick is to keep glitches below breakdown voltage of transistors in the amp.


my old NIKKO STA-6060 started popping. at first i thought that it needed a capacitor replacment. i checked all the capacitors with an ESR meter and to my surprise it was all fine after all these years.
maybe its a power line issue. some air conditioner or other motor noise. or lightning. but we don't have a lot of lightning in Israel.
 
Popping usually is either bad solder joints or semiconductors. See whether you can't narrow down the area of the fault - e.g. whether it affects both channels, is it volume pot dependent, etc. Also go through the types of transistors and compare to a list of known problematic ones.

Note that ESR is just one relevant capacitor performance parameter. You may still have leakage issues that would have to be checked independently.
 

phase

Member
2004-10-04 11:59 pm
The pros for using that is obviously potentially less line noise, is in a nice self contained housing. It looks like along with the common-mode (common - neutral) filtering that it provides, it also uses the mains earth to couple, via small capacitors.
In my experience, other’s may vary, having any connection in noise filters like that from mains earth to the common and/or neutral lines can introduce noise.

My devices have the safety earth connection to the steel case of the amplifier only, fused both at the mains inlet, and on the DC rails. This also requires that the rca jacks be isolated from the case, as well as any potentiometers. All A/C connections are crimped, not soldered, output DC protection on amp. I use an X rated capacitor across the incoming A/C, and a ferrite (as you now have) inside the amplifier case.
So you may want to try it with, and without the TDK filter, and determine if there’s any difference, good/bad. It’s a very tidy looking amp, and that filter does fit in nicely in there.

And if you can get ahold of a non-contact thermometer to check the temperature of the output transistors, I would recommend that, if you haven’t already.
 
my old NIKKO STA-6060 started popping. at first i thought that it needed a capacitor replacment. i checked all the capacitors with an ESR meter and to my surprise it was all fine after all these years.
Leakage current at rated voltage is more to the point. Just because a capacitor is fine at 2 v (test voltage) doesn't mean it is fine at 50 or 80.
Also if watching with a DVM, realize dual slope digital meters average signals over 2 - 4 seconds. Pops are microseconds to milliseconds. Scope or analog current meter are a lot faster.
Beside sgskloss suggestions, old mica insulators under power transistors can short and pop.
 
Last edited:

dudwity

Member
2006-12-13 9:40 pm
The pros for using that is obviously potentially less line noise, is in a nice self contained housing. It looks like along with the common-mode (common - neutral) filtering that it provides, it also uses the mains earth to couple, via small capacitors.
In my experience, other’s may vary, having any connection in noise filters like that from mains earth to the common and/or neutral lines can introduce noise.

My devices have the safety earth connection to the steel case of the amplifier only, fused both at the mains inlet, and on the DC rails. This also requires that the rca jacks be isolated from the case, as well as any potentiometers. All A/C connections are crimped, not soldered, output DC protection on amp. I use an X rated capacitor across the incoming A/C, and a ferrite (as you now have) inside the amplifier case.
So you may want to try it with, and without the TDK filter, and determine if there’s any difference, good/bad. It’s a very tidy looking amp, and that filter does fit in nicely in there.

And if you can get ahold of a non-contact thermometer to check the temperature of the output transistors, I would recommend that, if you haven’t already.

i am using crimped terminal for all AC connections and also DC connections.
the TDK filter does show 2 capacitors to gnd but since there are only 4 external connections i guess the capacitor connects to the filter metal chassis. isolating it from the enclosure chassis will render the connection to gnd idle. i will try with and without. i wonder if i will hear the difference.
i have an infra red thermometer. the black heatsinks are a good mach for it.
i haven't tried it with a speaker yet. just hooked it to dual bench power supply +-30vdc ' injected sine wave and tested the freq response with no load.
2v max input before clipping.
it went over 40 khz in BW.
i will add phones output.(maybe i can disable the relay outputs with a nice looking button)
the plan is to use XLR for the inputs and SPKON for the outputs.


as for the ESR meter(DE-5000) you are all correct. i didn't wanted to desolder and take out each and every capacitor and check it out of circuit.
i might just recap the whole thing one day. surprisingly the popping stopped so no recapping anytime soon.
 

phase

Member
2004-10-04 11:59 pm
And be really careful anytime you’re dealing with mains voltage, helps to keep one hand in your pocket and have clip on probe leads onto the higher voltage areas, and use a dim bulb tester as well. Usually I only think of the bulb tester for saving amplifier parts, however it could have some safety benefits as well.
You may already be aware of the potential current path when one hand is on something grounded and the other contacts something hot...