• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

How to reduce 120Hz frequency at amp output?

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
1,264
Likes
1,788
Location
North-East
#1
A few recent measurements of older amplifiers, including one that I did of Forte 1a, show a fairly high level of 120Hz or 2x the mains frequency. This appears to be the output frequency of a full-wave rectifier that's not completely eliminated by the filter capacitors.

1578194413877.png


How would I go about determining if this due to:
  1. Undervalued capacitors, need to replace with larger ones
  2. Old capacitors that just need to be replaced with the same sized ones
  3. Poor PS design
  4. Something else?
 
Last edited:

Hayabusa

Active Member
Joined
Oct 12, 2019
Messages
136
Likes
64
Location
Netherlands
#2
-98dB @ 120Hz will most likely be inaudible as the hearing threshold at 120Hz is around 25dB.
I just did a test with my setup: I stop hearing 120Hz below -85dB (which should be 20dB at the listening position)
 

Doodski

Major Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
1,224
Likes
807
Location
Edmonton , Canada
#3
Use the RC time constant ( resister rating in Ohms times the capacitance value in farads) to calculate the time required for 1 tau using a 1K Ohm resister to charge the capacitor to ~63% of charge. If the time required for 1 tau is too long then decrease the value of the resister you use. If you increase the value of the capacitors you may need a larger rated full-wave bridge rectifier.
https://en.wikipedia.org/wiki/RC_time_constant
 

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
1,264
Likes
1,788
Location
North-East
#4
-98dB @ 120Hz will most likely be inaudible as the hearing threshold at 120Hz is around 25dB.
I just did a test with my setup: I stop hearing 120Hz below -85dB (which should be 20dB at the listening position)
Agreed. I'm not bothered by it being audible, just the fact that it's there. Considering the age of the amplifier, I want to know if this is a possible symptom of the capacitors aging and if I should recap before they go completely bad.
 

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
1,264
Likes
1,788
Location
North-East
#5
Use the RC time constant ( resister rating in Ohms times the capacitance value in farads) to calculate the time required for 1 tau using a 1K Ohm resister to charge the capacitor to ~63% of charge. If the time required for 1 tau is too long then decrease the value of the resister you use. If you increase the value of the capacitors you may need a larger rated full-wave bridge rectifier.
https://en.wikipedia.org/wiki/RC_time_constant
I can see how this would help if I were building a PS, but how does this help me to determine if my caps are going bad in an existing amp and PS?
 

Doodski

Major Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
1,224
Likes
807
Location
Edmonton , Canada
#6
I can see how this would help if I were building a PS, but how does this help me to determine if my caps are going bad in an existing amp and PS?
By testing the charge and discharge rate you can test a capacitor. These are large caps and so I doubt you will find a handheld meter with the range to test that way. Otherwise you could just use a capacitance meter and cross your fingers and hope it's good.
EDIT>
If the capacitors where small value inexpensive we would obviously just replace them but they are going to be costly and there are 4 of them I see from pictures I googled. After you are organised and perform the first test the other 3 tests should go pretty efficiently.
 

Doodski

Major Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
1,224
Likes
807
Location
Edmonton , Canada
#9
Oh and need the voltage rating of the caps too. Do you have a DC power supply and a DC volts meter. That is if you want to test them.
 

Doodski

Major Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
1,224
Likes
807
Location
Edmonton , Canada
#11
Rated at 100v, I believe rail voltage is 39v for this amp. As long as I don’t need to take it all apart, I’m willing to try testing. I have various DC power supplies, but not a bench one.
To test them you will need good safe access to the terminals on the caps. That will require some disassembly. Due to the high capacitance and voltages involved you need to be very organised for safety reasons and that means good access to the caps terminals with nothing in the way. It really is best if you make a test jig and pull the caps and do the test on a workbench. If you take a charge from these size of caps it could cause damage to yourself. So safety is critical. Then there is the voltage used for the test. It's better to have a higher DC voltage for the test. What DC voltage supply do you have?
 

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
1,264
Likes
1,788
Location
North-East
#12
To test them you will need good safe access to the terminals on the caps. That will require some disassembly. Due to the high capacitance and voltages involved you need to be very organised for safety reasons and that means good access to the caps terminals with nothing in the way. It really is best if you make a test jig and pull the caps and do the test on a workbench. If you take a charge from these size of caps it could cause damage to yourself. So safety is critical. Then there is the voltage used for the test. It's better to have a higher DC voltage for the test. What DC voltage supply do you have?
I’m used to working with power supply capacitors, know how to discharge them safely and not to short them with test leads :)

I can probably get clean access to the capacitor terminals, as they are the screw-on type and the connectors face up in the amp.

I have a regulated 12v/5A supply that can be adjusted +/-4v or so.
 

Doodski

Major Contributor
Forum Donor
Joined
Dec 9, 2019
Messages
1,224
Likes
807
Location
Edmonton , Canada
#13
Hmmz... it's better if you have more voltage. 12 volts can be used but to go through all the effort a higher voltage is the way to go. I suggest using more voltage. You would need to make this circuit. The value of R is to be determined and then we will know the charge time(seconds).
 

DonH56

Technical Expert
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
3,892
Likes
5,711
Location
Monument, CO
#14
I used to pull the lead off a cap, insert a 1 k-ohm, 10 W resistor in series (value may vary but make sure it has high enough power rating), then measure the voltage across the resistor with a 'scope to get the RC time constant, then calculate C, using the old i = C*dV/dt equation or measure a couple of voltages (and their corresponding time stamps) on the curve and use V2 = V1 * [1 - e^(-t/RC)].
 
Last edited:

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,798
Likes
7,499
Location
Gold Coast, Queensland, Australia
#15
It's unlikely to be deteriorating main filter capacitors, in my opinion. Being as it has twin supplies, one for each channel and it behaves the same way in each channel (doesn't it?), I'd be looking at common causes.

Look at the design: (identical PCBs, flipped for each channel)

1578210776093.png


With the input RCAs sitting above the line fuse, there's some of the 60Hz.

Look at the positioning of those main filter caps. Big fat capacitors with lots of current (class A) sitting right next to the front end (VAS). It's amazing it doesn't make more noise.

Depending on how the chassis and the two 0V (CT of the transformer secondaries) are earthed/grounded and whether it's a common point or not, there could be chassis eddy currents too.

We also have no idea of the front end VAS stage V+/- rail decoupling as it appears to use the main rails (red and blue wires in the middle of the PCB) and I cannot see any medium value electrolytics on the PCB. Probably just supplied via a few resistors by the looks. That's where I would starting with my scope because if the supply is full of 120Hz ripple there, it's going to end up at the speaker terminals.

I don't think it's a design that is worth trying to look for problems- I think they may have always been there...
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,321
Likes
8,018
Location
The Neverlands
#16
^ Yep, exactly my view ^

Reducing it may be a difficult task because it could mean raising the transformer or getting it out of the cabinet for tests and changing ground/common points and RCA inputs etc. Not an easy task.
Another test could be to use an extra set of smoothing caps + a resistor (or inductor) fed from the main power supply and then feeding the amp or use a lab power supply instead of the transformer.
 

pma

Addicted to Fun and Learning
Joined
Feb 23, 2019
Messages
631
Likes
850
Location
Prague
#18
A few recent measurements of older amplifiers, including one that I did of Forte 1a, show a fairly high level of 120Hz or 2x the mains frequency. This appears to be the output frequency of a full-wave rectifier that's not completely eliminated by the filter capacitors.

How would I go about determining if this due to:
  1. Undervalued capacitors, need to replace with larger ones
  2. Old capacitors that just need to be replaced with the same sized ones
  3. Poor PS design
  4. Something else?
First, please always specify the level to which dB plot is related. dB is a relative number and depends what is the 0dB voltage level. 1V? 7V? Anything else?

Then I would start with (4), possible mains frequency throughput into measuring system. To check, please try a setup as below.
1) disconnect any ground loop.
2) turn the amp on. Connect measuring cable to amp speaker binding post as shown, both live and gnd of the measuring cable to the amp binding post that is at signal ground. Measure the spectrum with the soundcard and this will tell the residual measurement error. DUT amplifier input to be shorted.

test_hum.png


Even some linear amps may have very low hum voltage, like this Parasound
1578213961462.png
 

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,798
Likes
7,499
Location
Gold Coast, Queensland, Australia
#19
Strange that they placed the RCA inputs all the way in the middle rather than next to each amplifier board.
I agree. It certainly is strange, but a lot of people back then used twin RCA leads and if you were like me, you didn't want to pull them apart. :facepalm:

Now, everyone pretty much uses individual RCA L/R cables and we like our inputs to be separated by a decent distance on "dual mono" gear.

Certainly, it would would have removed a bunch of screened cable and improved the front end if they'd just used a PCB mount RCA jack and popped it through the rear panel right at board level. I guess it's easy to be wise after the event, but I'd completely redesign the whole thing from the bottom up. :)
 

JohnYang1997

Major Contributor
Joined
Dec 28, 2018
Messages
2,559
Likes
2,315
Location
China
#20
It's probably the interference from the magnetic field from the transformer. If you are able to place the transformer further from the circuit, there is high chance that you will see it being reduced.
Btw 65uV hum is huge. 2uV or lower is good. And i only see 200nV or lower as high performance.
 

Similar threads

Top Bottom