P2P Pultec 50hz noise help

[Clbraddock]

I just finished building a P2P pultec and overall am pretty happy with it, but would really like to reduce a noise bump I am getting at about 60hz (see pic).

Things I've tried and observed:

1. Checking that the casing of the DPDT and all pots and switches have continuity to ground.
2. With eq in, boosting the bass boost increases the 50-60hz bump a few db.
3. With eq in, increasing the bass attention all the way increases the 50-60hz bump more like 10db
4. With all eq controls zeroed out there is no different between the eq filters in or out.
5. Swapping each tube made no difference.
6. When sending a test signal, turning the unit off results in the 50-60hz bump dying within about a second even though the test signal continues for more like 8-9 seconds (The upper harmonic distortion keeps climbing with power off until the signal dies)
7. Placing a steel sheet adjacent to the power transformer and grounding it against the cap cans makes no difference. Same for mu-metal

What else should I try? I would like to get the noise down a bit, because I had planned to use a pair of these on the mix bus mid-side. To use them in their proper gain staging though, I would probably be applying 10-15 db of gain AFTER the pultecs to get up to reasonable LUFS. The noise floor is okay where it is, but boosting it 10-15 db it is going to be noticeable. Is there anything I can do, or is this just the way real pultecs are?

 

[ruffrecords]

On your graph, the signal looks to be at about -30dB and the hum around -90dB. You don't say what the absolute signal level is that the -30dB represents but if it is 0dBu then you have less than 1mV of hum which is a good start for a design using AC heaters. The fact that the hum disappears when you turn off before the signal dies away does imply it is a magnetic coupling of some sort.

Heater lead dress and overall layout can make a difference. Do you have a hum dinger fitted or do you just connect one side of the AC heater supply to analogue 0V?

Cheers

Ian

 

[Clbraddock]

Thanks Ian. Test signal is -20dbfs at 1khz. I think that should be about 0 dbvu (interface is an Apollo set to +24 db mode). If -20dbfs is 0 dbvu, doesn’t that mean that the -90 dbfs noise is really more like - 70 dbvu? I found an old pultec manual that quotes the noise spec as -90 db +/- 10.


The heaters are wired as per the vintage layout. No humdinger, but the heater tap on the power transformer has a center tap. The center tap connects to ground through the cathode bias resistor of the output tube (12AU7).

I was under the impression that humdingers were for non-center tapped heaters. Would a humdinger still apply to the heater wiring I described above?

 

[Clbraddock]

As far as lead dress of the heaters they are twisted tight and tucked in the corner crease, except between the input 12AX7 and output 12AU7. There there tubes are so close together there isn’t room to do any sort of twist. I wired it like shown here (heaters in light blue):

 

[ruffrecords]

OK, so with the centre tap there is no opportunity for a hum dinger., but you have elevated heaters which will also reduce hum.

The -90dB noise figure is, I believe, relative to +10dBm output so in reality it is about -80dBu (which is about what you would expect from a well constructed modest gain tube device.

Can you post a picture of the inside wiring/construction?

Cheers

Ian

 

[Clbraddock]

Thanks for your help. I really appreciate it. It's hard to get good photos with the front plate still on. I could take it off, but would have to unscrew all of the pots, switches, etc. If that would be more help let me know. Thanks again!

 

[Clbraddock]

In the 2nd to last pic (IMG_0093) I'm wondering if I should pull the wires away from the back panel to get them further from the tube heaters. There is a tube heater approximately where the red line is Ive drawn below:

 

[Clbraddock]

Also, here is the full wiring layout I more or less worked off:

 

[Ike Zimbel]

I just finished building a P2P pultec and overall am pretty happy with it, but would really like to reduce a noise bump I am getting at about 60hz (see pic).

Things I've tried and observed:

1. Checking that the casing of the DPDT and all pots and switches have continuity to ground.
2. With eq in, boosting the bass boost increases the 50-60hz bump a few db.
3. With eq in, increasing the bass attention all the way increases the 50-60hz bump more like 10db
4. With all eq controls zeroed out there is no different between the eq filters in or out.
5. Swapping each tube made no difference.
6. When sending a test signal, turning the unit off results in the 50-60hz bump dying within about a second even though the test signal continues for more like 8-9 seconds (The upper harmonic distortion keeps climbing with power off until the signal dies)
7. Placing a steel sheet adjacent to the power transformer and grounding it against the cap cans makes no difference. Same for mu-metal

What else should I try? I would like to get the noise down a bit, because I had planned to use a pair of these on the mix bus mid-side. To use them in their proper gain staging though, I would probably be applying 10-15 db of gain AFTER the pultecs to get up to reasonable LUFS. The noise floor is okay where it is, but boosting it 10-15 db it is going to be noticeable. Is there anything I can do, or is this just the way real pultecs are?

WRT #1, have you checked that all of the places where you have tabs going through to the chassis have actual continuity to ground? The paint on the rear chassis looks a) great, and b) like it might have good insulating properties. You might want to grind off some paint at those points (like your main chassis ground screw in particular). Taking off some paint where you have metal-to-metal contact between the chassis panels (back to sides, sides to front panel etc.) might also help. Star toothed washers can also help get through the paint.

 

[k brown]

50z hum in the US?; it must be 60.

 

[Clbraddock]

50z hum in the US?; it must be 60.

Yes, it's prob 60hz. Its just kind of a blob on the frequency analyzer

 

[scott2000]

Is it bad to maybe parallel some more capacitance to that second 40uf cap after the 6x4?

 

[beatnik]

What audio transformers are you using ? If the input and interstage don't have proper magnetic shielding they might pick up hum radiated from the power transformer.
It's also sometimes beneficial to determine the optimal orientation of the transformers in the initial build phase

My approach is building rack gear without the power supply fitted in, make noise tests and fit the power transformer as the last step, compare results. You can see if the proximity of the PSU is problematic.
When building vintage replica gear I also compare results with some low noise solid state external power supplies and see if there is room for improvement (that's often the case) once you liberate yourself from making period correct replicas there are many benefits that can be achieved.

If you have means to make the PSU external I don't see why not, and if making a dual channel why not making a single PSU to power both (except maybe some crosstalk via the PSU line which never been a problem in my experience)
One could go even further and make a chunky Psu to power all tube gear in the studio, lots of money to be saved and the PSU can be spaced away for minimum noise)

 

[Clbraddock]

Input is a Triad HS-56V. Interstage is Vintage Windings VW-29 (HS-29 reissue). Output is Cinemag CM-S217D

 

[ruffrecords]

Input is a Triad HS-56V. Interstage is Vintage Windings VW-29 (HS-29 reissue). Output is Cinemag CM-S217D

The fact that the hum increases both when you boost and cut the bass implies the interstage transformer maybe be picking up something.. You could try just shorting the primary of this transformer and see if the hum disappears.

Cheers

ian

 

[Clbraddock]

The fact that the hum increases both when you boost and cut the bass implies the interstage transformer maybe be picking up something.. You could try just shorting the primary of this transformer and see if the hum disappears.

Cheers

ian

You mean short the primary to ground? Just want to make sure I do the right thing. Thanks for helping me! I really appreciate it.

 

[ruffrecords]

You mean short the primary to ground? Just want to make sure I do the right thing. Thanks for helping me! I really appreciate it.

Yes, that is exactly what I mean.

Cheers

Ian

 

[Clbraddock]

Shorting the interstage transformer primary kills the test signal (as I would expect), but makes no difference on the noise.

Edit: Something I noticed futzing around in there - the plates of the input 12AX7 are connected to the grids of the output 12AU7 via .25uf capacitors. If you bump either of those capacitors the noise hump around 60hz jumps/wriggles, but the test signal is unaffected. I don't know if that means anything though.

Edit2: Also, if I change the test signal from 1khz to 600 hz the humps start "dancing" (attached vid)

 

[ruffrecords]

Shorting the interstage transformer primary kills the test signal (as I would expect), but makes no difference on the noise.

This implies the noise is entering after the EQ section.

Edit: Something I noticed futzing around in there - the plates of the input 12AX7 are connected to the grids of the output 12AU7 via .25uf capacitors. If you bump either of those capacitors the noise hump around 60hz jumps/wriggles, but the test signal is unaffected. I don't know if that means anything though.

Definitely not right. It looks like you have some exotic capacitors in there. Try changing them for a couple of regular film capacitors

Cheers

Ian

 

[Ike Zimbel]

7. Placing a steel sheet adjacent to the power transformer and grounding it against the cap cans makes no difference. Same for mu-metal

A note on this: The idea is not to shield the PT as that would require an entire steel case around it. The idea is to use the steel to shield the parts that are being affected by the PT's magnetic field. In your case, I would start with the capacitors you mentioned as contributing to the hum. I just posted about this on another thread here recently. In that case, the DUT was a commercially made "Pultec inspired" Eq, a good one. I will mention again that you need to be extremely careful moving a piece of conductive metal around inside a chassis where High Voltage is present (Note: the steel does NOT need to be connected to ground, however...the shielding comes from absorbing/disrupting the magnetic field via the ferrous properties of the steel, not draining it off to 0v). In cases where there is not a lot of room inside (although you have plenty) and also to try and narrow down what's being affected, I have sometimes just used a steel tool that's ready-to-hand, like a large pliers or an adjustable wrench to hold near various parts and see/hear if the noise comes down. Once you identify the area, you can come up with a way to do a permanent shield, like mounting an angle bracket to the floor of the device.
One more thing: Have you tested this with the steel lid screwed firmly in place, or have you been waiting to fix the issue before you close it up? Hint...the lid may fix the problem...