No more 60 Hz hum in a PL2000 preamp.

[Gepetto]

I can also be instrumentation pickup from your measurement point to the instrument doing the measuring and not be apparent in the signal itself. In other words, are you sure it is real?

It appears from the spectrum analysis that you were better off at 60 Hz before you separated the primary and secondary bundle. By about 10dB

 

[George S.]

I can also be instrumentation pickup from your measurement point to the instrument doing the measuring and not be apparent in the signal itself. In other words, are you sure it is real?

It appears from the spectrum analysis that you were better off at 60 Hz before you separated the primary and secondary bundle. By about 10dB

Your correct, pretty sure the test set up is picking up significant 60 Hz from the rooms environment. Need to get set up outside and test. That's next weekend's goal.

 

[Gepetto]

Hi George
Even when you move outdoors and get rid of some of the ambient room noise, you will still need to be rigorous regarding cabling to the PL2000 because the line cord will be radiating as will the transformer and internal AC wiring. The levels you are measuring are vanishingly small so almost anything can be the source.

 

[George S.]

What sort of patch cords would they use in a laboratory setting to connect to the PL2000. Truth be told I'm using old skinny Radio Shack ones and where they lay does make a difference in results. I'm very willing to upgrade as I want to pursue this further, especially once I retire.

 

[Gepetto]

What sort of patch cords would they use in a laboratory setting to connect to the PL2000. Truth be told I'm using old skinny Radio Shack ones and where they lay does make a difference in results. I'm very willing to upgrade as I want to pursue this further, especially once I retire.

Hard to say George. Maybe post some pictures of your setup and we can make some suggestions. Instrumentation cables I use are male BNC to male BNC type with RG 316 cable.

 

[George S.]

So your using BNC to RCA adapters or using installed BNC jacks for product development that later get swapped for RCA jacks? I can easily make Belden 83284, RG-316 cables, have probably 30' on hand. Just need to get the connectors. Think this would be worthwhile?
My setup is a Thinkpad T410 laptop, a Asus Xonar U7 MK2 USB sound card, Monoprices best USB Audio cable, and cheap RCA cables.
RCAs out of the sound card output to the DUT input at line level, RCAs out of the DUT to a line level 3.5 mm stereo jack input on the sound card. Pretty simple and straight forward. The tone generator and spectrum analyzer are all part of the REW software. You have a good point about the cables.

 

[George S.]

Oh, and I also built the "pmillett sound card interface" to test amps sometime in the future. Found I don't need to use it if I keep the output of the preamp under 200 mV RMS. Anything higher and the sound card crashes. This is a learning process for me. I have a old SS amp my father built for me in the early 70s and a 80s Akai integrated I want to test and look at. Eventually I'll get there.

 

[Gepetto]

So your using BNC to RCA adapters or using installed BNC jacks for product development that later get swapped for RCA jacks? I can easily make Belden 83284, RG-316 cables, have probably 30' on hand. Just need to get the connectors. Think this would be worthwhile?
My setup is a Thinkpad T410 laptop, a Asus Xonar U7 MK2 USB sound card, Monoprices best USB Audio cable, and cheap RCA cables.
RCAs out of the sound card output to the DUT input at line level, RCAs out of the DUT to a line level 3.5 mm stereo jack input on the sound card. Pretty simple and straight forward. The tone generator and spectrum analyzer are all part of the REW software. You have a good point about the cables.

Yes I have small BMC to RCA adapters to convert as required. My analyzer has all BMC outputs on it and I have equipped my load bank with BNC connections as well.

 

[George S.]

Cool, I have a bunch of BNC to RCA adapters as well, just haven't used them because I was using premade cables with the appropriate ends with my HP scope and cheap Koolertron signal generator. I'm going to make up some RG-316 cables, with RCA or 3.5 mm stereo jack as required. Will update results. Thanks for the insights, I appreciate it.

 

[Gepetto]

Cool, I have a bunch of BNC to RCA adapters as well, just haven't used them because I was using premade cables with the appropriate ends with my HP scope and cheap Koolertron signal generator. I'm going to make up some RG-316 cables, with RCA or 3.5 mm stereo jack as required. Will update results. Thanks for the insights, I appreciate it.

Hoping the high grade shielded coax make a difference.

I have a saying that Lee is familiar with, "check your instrumentation first"

 

[laatsch55]

Rule 1----Check your test equipment, Rule 2----Check your test equipment......

LOL!~!!

 

[J!m]

Have you tried completely floating the steel enclosure and attaching the A/C ground lead to that only? (Faraday cage) Maybe that would help? (bear in mind I only know enough to be dangerous). It does nothing when the noise is generated internally, but it might be easy to try.

Also, if AC and signal cables must ever cross, do so at 90 degrees. Laying them parallel will enhance their ability to pick up noise.

When I have my big system up, all the AC (and DC) power runs down the left side of the back of the rack and all the interconnects run down the right side. Separation of church and state. It's free to try. If your interconnects and AC lines look like an angry lobster in a bowl of spaghetti, that won't help.

 

[George S.]

Have you tried completely floating the steel enclosure and attaching the A/C ground lead to that only? (Faraday cage) Maybe that would help? (bear in mind I only know enough to be dangerous). It does nothing when the noise is generated internally, but it might be easy to try.

Also, if AC and signal cables must ever cross, do so at 90 degrees. Laying them parallel will enhance their ability to pick up noise.

When I have my big system up, all the AC (and DC) power runs down the left side of the back of the rack and all the interconnects run down the right side. Separation of church and state. It's free to try. If your interconnects and AC lines look like an angry lobster in a bowl of spaghetti, that won't help.

Been using a single ground screw from the boards input ground plane to the chassis. The transformers shield is connected to this screw, while the the other screws float above ground. Haven't tried completely floating the board above ground, but will try it if the new cables don't solve the issue. That's a good idea I hadn't thought to try, thanks!
On the bench, I'm keeping the cables well separated, not even crossing. Also keeping the preamps enclosure closed with a few screws holding the bottom and top plates on.
I ordered jacks last night. All Switchcraft. It'll be interesting to see what the new cables do.

 

[mlucitt]

What were the series dropping resistors you used on the output of the transformer?
Also, a suggestion to aid your chasing the last remnant of 60Hz hum from the venerable PL2000:
Shorten the AC wires from the IEC connector to the fuse and the switch to the shortest length possible. It may help reduce those transmitters.

 

[George S.]

Mark L., I can't remember or find that post right now for the resistor values. After reading about R/C circuits I think the series resistors are fine but not sure about those capacitor values until I do the math. Read a PDF article about calculating the value, but I'm on the road, working.
Good thing is, I do have a way to "check my instrumentation". I have that Koolertron signal generator I can use to bench mark the spectrum analyzer. I can even use it instead of REWs signal generator. Will have to learn more about converting peak to peak voltages to RMS and dB(who ever knew there are so many types). Going to get the cables made, the instrumentation bench marked, and see what I have for 60 Hz. Will try to get this done coming weekend.

 

[mlucitt]

My sense is that typical values of 220nF and 10 Ohms for the series snubber (across the transformer secondary coil) would be a big improvement after adding the capacitors across the bridge rectifier.
I don't want to drop my B+ and B- voltage too much, but a little resistance to the 60Hz supply may help to tamp down the noise.
This is the reason I put a 2.2 Ohm resistor from the STAR chassis ground to the circuit board ground in my headphone amp. Just a little float above chassis ground helped, even using a grounded IEC connector, I got no hum (that I could hear) in the headphones.

 

[George S.]

Understood. Yeah, this is going to take some sorting out. The 60 Hz I see when measuring the original unmodded PL2000 is very similar.
From what I read about R/C networks a series resistor is preferred, but undesirable if it drops the secondary voltage too low. Those series resistors are there because because the voltage was over 18 volts, needed to get it down to 17 volts to save the opamps. So replace them with 10 ohms and put the voltage drop on the center tap or or further out on the rails? I don't know at this point.
I have lots of options for testing the test equipment. I can insert a 60 Hz sine wave of known amplitude or vRMS from within or outside REW and see the results. Once I get that nailed, then I can proceed. Thanks. I'll get this solved with all your help and I'll work up a comprehensive PDF for the PL2000. My current modded unit is the pig and the unmodded unit once modded will be proof of concept. It's a fun and enjoyable hobby for me. Just have little time to devote to it so it's going to take awhile.

 

[Gepetto]

The series resistors also dramatically reduce the current pulse into the bulk caps on each 60 Hz charging cycle . That in turn reduces the primary current pulse as well.

The higher the current pulse in all your wires, the more they radiate at 60 and 120 Hz The softening of the current pulse significantly impacts the 2nd harmonic content as well.

 

[George S.]

Well, I found a error. I was using 24 bit resolution. REW supports 16 bit. Getting good results with a 48 kHz sample rate and 16 bit resolution. Sampling the input level 48,000 times per second should be more than adequate. From what I'm reading, a 48 kHz sample rate should allow me to sample signals up to 24 kHz because for each frequency 2 samples are needed to reproduce it. Damn, this gets complicated! Since I'm testing from 20 Hz to 20 kHz, the 48 kHz sample rate appears appropriate.

 

[mlucitt]

I have (2) 10 Ohm 1/2W 1% resistors that I can try as series resistors on both legs of the PL2000 transformer secondary output. I'll measure the voltage before and after and report back.