AC heaters ultimate hum limit

[ruffrecords]

sorry to bug you again
i think i got the 2i2 part, but i dont get the adding 7db for converting dbu to dbFS...
you are already reading in dbFS!

I guess I did not explain myself properly. I am trying to convert dBFS to dBu. REW reads dBFS and I want to know what that means in terms of actual signal voltage i.e. dBu. Apologies for the confusion.

Cheers

Ian

 

[kambo]

apology is mine 

 

[kambo]

i did what u did with ur 2i2....
on my setting -10dbFS was equal to 0db on the output...
and adjusted input to match that values on REW too...

on loop back measurement i get -10.5 ; pretty close!
i wonder why u have 7dbFS difference!
may be because ur 7dbFS equal to 0dbu, so u dont need to add anything?


edit: i have higher dist than yours, i have edcors connected on DA side, and loopback input is slightly hot...

 

[kambo]

.
another one

 

[ruffrecords]

You need to make sure:

1. You use the XLR input
2. The input gain is turned right down
3. The monitor gain is turned right up
4. The monitor switch is in the off position.

I also have a 600 oh load in my test box and this does drop the level from the 2i2 a bit.

Cheers

Ian

 

[kambo]

YOy need to make sure:

1. You use the XLR input
2. The input gain is turned right down
3. The monitor gain is turned right up
4. The monitor switch is in the off position.

I also have a 600 oh load in my test box and this does drop the level from the 2i2 a bit.

Cheers

Ian

thats my exact setup. 
except the 600ohm resistor.

 

[ruffrecords]

YOy need to make sure:

1. You use the XLR input
2. The input gain is turned right down
3. The monitor gain is turned right up
4. The monitor switch is in the off position.

I also have a 600 oh load in my test box and this does drop the level from the 2i2 a bit.

Cheers

Ian

thats my exact setup. 
except the 600ohm resistor.

And you also said "and adjusted input to match that values on REW too... "

Cheers

Ian

 

[kambo]

YOy need to make sure:

1. You use the XLR input
2. The input gain is turned right down
3. The monitor gain is turned right up
4. The monitor switch is in the off position.

I also have a 600 oh load in my test box and this does drop the level from the 2i2 a bit.

Cheers

Ian

thats my exact setup. 
except the 600ohm resistor.

And you also said "and adjusted input to match that values on REW too... "

Cheers

Ian

when i adjusted the inputs, i ended up at min gain on the inputs!

 

[ruffrecords]

when i adjusted the inputs, i ended up at min gain on the inputs!

OK, I will do a loop test with my 2i2 on its own with the gain set to minimum and report back. That will at least establish a base line.

Cheers

Ian

 

[ruffrecords]

Here is my 2i2 loop back test. The generator is set to -13dBFS and the input measure -6.5dBFS. I checked the 13dBFS output on my Lindos test set and it is exactly 0dBU. I really do not know why yours is not the same. The date code on the bottom of mine is 05/2014.

Cheers

Ian

 

[ruffrecords]

And this is with the same setting but looped back via my test box. As you can see, the return level drops to -10.1dBFS which is mainly the loss in the transformer when loaded with 600 ohms.

Cheers

Ian

 

[kambo]

thanks Ian,
i think i wasnt clear on my 2i2 setup.
although 2i2 is a great sound_card for a lab,
i wasnt happy with 2i2's loopback levels, with experiment trial/error, i ended up connected 10k:10k transformers direct to DA output pins of Cirrus CS4272 AD/DA chip.
now i dont have a output level control. but levels set to min. on inputs,
on loopback i  only get 0.5dbFS difference. this is accurate enough for me!

and feeding 9.8dbFS from REW is ~0dbu ( 779 mvAC) on the output of 2i2.


edit: typo

 

[Chris_V]

I was fighting with some "whistle noise" at around 15kHz on a tube preamp and it was the heater voltage from a smps that was not enough filtered 
Switching frequency is far more higher than that (around 300k) but there was some subharmonic oscillations coupling at the input triode grid !

 

[ruffrecords]

I was fighting with some "whistle noise" at around 15kHz on a tube preamp and it was the heater voltage from a smps that was not enough filtered 
Switching frequency is far more higher than that (around 300k) but there was some subharmonic oscillations coupling at the input triode grid !

This can happen with linear regs too. I had a a big problem a year ago when I started using a 5A LDO regulator. I got about 1MHz on the heaters from it.

Cheers

Ian

 

[ruffrecords]

In preparation for the receipt of the new (low hum) version of the PCB, I measured the spectrum of the hum and noise of the current PCB layout (attached). I then decided to try a simple dc heater supply just using a bridge rectifier and a 4700uF smoothing capacitor, which should reduce the hum amplitude applies to the heaters by about 20dB. To my surprise the overall hum was higher. I took the spectrum again and found that the 50Hz component at the output was higher than with ac heater but the 100Hz component had practically disappeared - the opposite of what I expected. See next post f or what I did next.

Cheers

Ian

 

[ruffrecords]

The spectrum with the dc heaters is attached.  Puzzled by this result I went back to ac heaters and started experimenting with the secondary winding connections. There are three of them, the HT, the heaters and the phantom power. I tried disconnecting each in turn expecting to see the hum drop away completely when the heaters were disconnected. But again the expected did not happen.  Disconnecting a single winding dropped the hum a couple of dB but the only way it disappeared was when the mains was disconnected. So I am beginning to suspect interactions between the secondaries. I knew this happened but had never observed it before. Then I tried swapping the polarity of each secondary winding. It turned out the HT and phantom power polarity made only a small difference (2dB) but the heater winding polarity made 10dB difference in hum level. Again, the only way the hum disappears is when the mains is disconnected (and the HT and heaters stay active long enough to observe the step change in hum level).

Position of the cables from the mains transformer made no difference nor did the orientation of the mains transformer.

Cheers

Ian

 

[Di_Sousa]

Interesting results, I would expect the DC heaters would dramatically reduce the hum.

How far do you have the Power transformer from the circuit?

could you try a measurement with the transformer 1 meter away?

 

[ruffrecords]

Interesting results, I would expect the DC heaters would dramatically reduce the hum.

How far do you have the Power transformer from the circuit?

could you try a measurement with the transformer 1 meter away?

It is within 5cms of the circuit. I have done some experiments with the humdinger (which is just a pair of equal values resistors). The lowest hum is with the resistors equal. With either resistor shorted the hum is at least 10dB higher. I think the equal value hum dinger position is quite finely balanced - any slight change  elsewhere in the circuit moves the balance point and the result is increased hum. That is why the 100Hz disappears when I use the simple dc supply. Next I will try adding a regulator to the dc supply.

I understand you point about moving the transformer far away but I do not think magnetic induction is the issue as the hum does no vary with the orientation of the mains transformer.

Cheers

ian

 

[Di_Sousa]

I see,
so let us posted on your experiments, this is a quite interesting thread.

The documents I emailed you about Heater supplies were of any use?

 

[kambo]

you need 4700uF for per  300mA / heater... looks like u have two tubes pulling 300mA each!
i would at least triple that amount .... dont forget to ground it either !