No more 60 Hz hum in a PL2000 preamp.

[George S.]

I never liked how the factory wired these units with the power switch between the cord and fuse. Decided to do some testing and wired the primaries directly to the IEC socket. Found I got appx 10 dBFS less 60hz noise by wiring the striped primaries to the line and solid primaries to the neutral.
I then made red the line color, and green the neutral color.
Line runs from the IEC, to fuse tip and from base, to switch, then to striped primary. Green runs from IEC to switch, then to solid primary. The resistor and caps across the switch are removed for testing. I'll probaly replace the caps later, but not the resistor.
Here's result of testing outside. Thats with a 142 mV RMS 1kHz sine wave injected. Theoretically, the sound card base line #s gets subtracted from the unit being tested #s for final results. I'm going to do more work on the 60hz noise once I get the Quasimodo Test Jigs built.

 

[mlucitt]

I wonder what grounding the chassis to the mains would do to the 60Hz noise?
Did you remove your bridge rectifier bypass capacitors?

 

[Gepetto]

Here are some pictures with my probe grounded (way too noisy when not grounded):
Outputs of the Transformer Secondary Windings (38V Peak) Don't know why the sine waves are clipped, this is "in circuit" measuring.
View attachment 53382

Outputs of the Bridge Rectifier (DC with .4V Peak AC Sawtooth) .01uF snubber caps across the bridge rectifier diodes will likely fix this.
View attachment 53385

The secondary outputs are clipped because of the filter capacitors which are being charged at the peak of each sine wave. The resistance of the transformer primary and secondary cause this effect. The caps present a very low impedance to the transformer once the diode bridge turns on; it is like the transformer is driving into a short circuit once the secondary winding voltage exceeds the rectified DC voltage plus the 2 diode drops of the bridge.

 

[George S.]

Removed the ceramic caps across the switch terminals that PL says are to limit arcing of the switch contacts and accessory power sockets.
Hum is non-existent in the system with the volume turned up.
I think this is as good as it gets in "stock" form. But, I am going to do further work on it once I can ring that transformer with a square wave using the Quasimodo jig.
The bridge rectifier caps are new Nichicon UFG swapped in for older Muse

 

[mlucitt]

The secondary outputs are clipped because of the filter capacitors which are being charged at the peak of each sine wave.

Joe, is this an undesirable condition, or just the way it works? Is there a work around to get a better looking sine wave from the transformer secondary? Do the diodes care if the sine wave is clipped?

 

[George S.]

Mark, post 16, ran the IEC ground to a transformer screw, higher noise. Used simple wire. May experiment with a safety cap in the future. Incidentally, if I pull the RCA cable from the preamps output jack (this is the soundcards input cable)and hold it out like a antenna, I get a huge 60hz spike on the screen. Then, if I short the jack, the spike disappears. This is with the laptop running on battery. The equipment detects a huge amount of 60hz anywhere close to the house.
Tomorrow I'm going to try a different sound card. Been using a Asus U7 MK2 USB card. Going to try a Focusrite Scarlett 2i4 USB sound interface.
Have to wonder how the labs keep test signals clean from external factors like 60hz. House was built in 77 with all copper wiring, breakers, ground fault in breaker box, 3 conductor Romex, good ground rod bonding to soil best I can tell. No electrical issues in the house. Here's a photo of test setup outside

 

[mlucitt]

Removed the ceramic caps across the switch terminals that PL says are to limit arcing of the switch contacts and accessory power sockets.

I would think the ceramic caps are out of the circuit as soon as the switch contacts close...

 

[mlucitt]

Also, did you see this picture? What are those sneaky resistors doing in the picture?

 

[mlucitt]

No electrical issues in the house.

Could be a bad florescent light fixture ballast somewhere. Somewhat off topic, but as a HAM radio guy, I occasionally walk around the house with an old transistor AM radio (with a ferrite rod antenna), and using headphones, listen for the telltale 60Hz static. I find it in the weirdest places. Bad refrigerator compressor (still under warranty!). Just an idea.

 

[Gepetto]

Joe, is this an undesirable condition, or just the way it works? Is there a work around to get a better looking sine wave from the transformer secondary? Do the diodes care if the sine wave is clipped?

Not undesirable Mark. It is the way it works. That tiny transformer has fine (small) wire in the primary and somewhat larger wire in the secondary. This causes winding resistance and poorer load regulation in the transformer. The way you solve that is a monster transformer with large gauge wire windings. Seriously unnecessary for a low power preamp application like this. The diodes do not care at all.

 

[George S.]

Also, did you see this picture? What are those sneaky resistors doing in the picture?

View attachment 53708

I remember that photo. Think Lee asked same question to Joe. To possibly eliminate popping when adding or removing phono plugs if I remember correct.
Yes, that REW setup makes a fine RFI sniffer. 60hz is everywhere here and most likely everywhere. The second, unmodified PL2000 I aquired as a "control" also exhibits the same 60hz spikes .
Going to look at a few things on the scope today. Those Quasimodo Jig cards are in the mail and headed my way. Should have sent the Gerber files to a board maker and had a dozen made. I just didn't know any better at the time. Oh well, I learn best by making mistakes.

 

[George S.]

Mark, here's a photo of my modded unit with the probes connected to the pcb trace where the secondaries are soldered. Same as yours. Been thinking about what Joe said. Have considered using a regulated power supply add-on board, but don't really want to. Having too much fun learning about this! As far as I'm concerned the unit works perfect as is, just want to see if I can get the inaudible 60hz lower.
And, best of of all, I don't see the Browndogs misbehaving, the main reason I pursued getting the spectrum analyzer up and running.
Sitting here listening to Black Sabbath "Snow Blind" on a old Akai integrated amp and Pioneer CS-99 speakers. Damn this old gear sounds good!

 

[J!m]

If your ungrounded interconnect is picking up RF, you may consider an interconnect upgrade.

Another idea, which may have side effects, is to remove the transformer from the chassis (external power supply) and bring only DC up there. I know a preamp is not as dependent on a massive power supply as an amplifier is, but a simple phono pre (Aragon 47k) has been transformed to another league by upgrading the external power supply to “ludicrous” grade, even though it draws nothing. It’s on my short list, although I have boards for another new design to try someday…

 

[mlucitt]

Having too much fun learning about this!

George, I'm with you. I never considered poking a probe into my preamp power supply. Then I see the truncated sine wave, but Joe says it's OK. I always thought the PL2000 transformer was too small, but it seems to work just fine. And why fix it if it's not broke. Will installing a regulated power supply make the PL2000 sound any better, probably not. Certainly not to my ears.

 

[George S.]

Mark, my modded unit sounds amazing compared to the stock unit. Night and day difference, especially the phono section. I want to thank you all for your help so far. Once we get the snubber figured out, I think it's a done deal and it'll be time to write it up. Should have the Quasimodo Jig boards this weekend, then I'll order components, including a wide variety of caps and resistors to configure a tuned CRC between each secondary lead and center tap to critically damp that transformer. It'll be interesting to see the results on the scope and REW spectrum analyzer.
Also want to test some slow recovery diodes for the rectifier. Have some new On Semi in there now, just standard 4004s.
Then hopefully someone smarter than me will build one with better improvements and share the info.
The price of PL2000s may be rising!

 

[George S.]

Testing the phono section RIAA curve with a Motronix reverse RIAA filter. First photo shows a 1 kHz square wave with RIAA equalization applied. Second photo shows same square wave with the Motronix (the small circuit board) offsetting the RIAA equalization. The Motronix is commonly used for such testing or turning a unused phone stage into a normal line level input.
I did a frequency sweep from 20 Hz to 20 kHz and square wave maintains form showing RIAA curve of modded phono section is correct.
Phono section uses Panasonic polypropylene films, Nichicon UFG where electrolytics must be used, and CDE silver mica where cost effective ( thanks for that tip Joe!). And Browndog RC4739 with OPA2134 opamp. No ceramics and all resistors are 1% metal fillm.

 

[laatsch55]

Good work George!!

 

[George S.]

Ordered the components today to build the Quasimodo's. Maybe next weekend I can get that transformer snubbed.

 

[George S.]

Bad turn on thump if wired much different from factory. The power switch has two sets of contacts. Use the top switch lugs to make and break line to turn the unit on. Bottom lugs make and break line for the switched accessory power jacks.
Definitely worthwhile getting the fuse holder wired to line before the switch. This is configuration I'm going to keep. Keep the neutral unbroken.

 

[George S.]

Doh, sometimes I can't see the forest because the trees are in the way! They wired the fuse for the preamp after the switch because they are switching the accessories directly at the power switch!
So, who needs 2 wire accessory plugs these days, and why stress that irreplaceable power/volume switch? Let's not wire those accessory plugs and eliminate the wiring on the bottom lugs of the switch. This will allow wiring the fuse properly between the IEC socket and switch. And if one set of contacts in the power switch fails for whatever reason in the far future, there is the other set. Will test this coming weekend.