choke in HT PSU

[kambo]

i thought using a choke would be better than using a 1k resistor...
with this drawing i get more 60hz bump,
if i replace choke with 1k resistor, 60hz bump is way down!

any trick on using choke  :

 

[johnheath]

What happens if you place the 1k resistor after the choke?


Best regards

/John

 

[abbey road d enfer]

i thought using a choke would be better than using a 1k resistor...
with this drawing i get more 60hz bump,
if i replace choke with 1k resistor, 60hz bump is way down!

any trick on using choke  :

I believe your choke is saturated by the DC current, so instead of 10H, it may actually measure 10 or 20 times less. It is probably not designed for the current you're asking.
Or there is magnetic coupling between the choke and the PT.

 

[kambo]

I believe your choke is saturated by the DC current, so instead of 10H, it may actually measure 10 or 20 times less. It is probably not designed for the current you're asking.
Or there is magnetic coupling between the choke and the PT.

its LL1673, wired series for 10H - 200mA
i am pulling way less then that... running only one half 6SN7

What happens if you place the 1k resistor after the choke?
Best regards
/John

i can try...

do you mean, series  connected to choke, before the last cap ?
(choke + 1k)

 

[johnheath]

No.. but rather C-L-C-R-C

C - Cap
L - Inductor
R - Resistor

Which you already know

Best regards

/John

 

[johnheath]

I also found this on the Lundahl website…
"I have bought some LL1673/10H from you, but the measured inductance (with my inductance meter) is only 6H. What is wrong?

Our power supply chokes are designed to work at 100/120Hz and with voltages around 100V. Most induction meters work at a few volts or less, at around 1kHz. This is not the correct operating point for a power supply choke, and the reading on the induction meter is usually around half the specified value of the choke.

Best way to measure the inductance in a power supply choke is to connect the choke to something like 100V, 50/60Hz, measure voltage and current and calculate inductance. I use a variable autotransformer connected to my mains outlet for this measurement. Be careful however how you connect the autotransformer to the mains, so you do not touch 220V when you think your terminal is at 0V."


Best regards

/John

 

[johnheath]

I guess you already have read it but just in case here's the data sheet:

http://www.lundahl.se/wp-content/uploads/datasheets/1673.pdf


/John

 

[kambo]

No.. but rather C-L-C-R-C

C - Cap
L - Inductor
R - Resistor

Which you already know

Best regards

/John

ohh got it... thanks!
i only have 2 caps to play with... i need to order more caps anyway...

i did try additional resistor after the choke, that helped a lot at mid to high frequency spikes!

 

[johnheath]

Yes I see.

The inductor is mainly used here to clean up the VDC. I have had very good results in using a C-L-C-R-C-R-C-R-C filter and I use Hammond inductors…. cheap but reliable their model 154EA is rated 20H and can handle 500VDC but not more than 20mA… which should be enough for a double triode. And use a 2W bleeder resistor in parallel with the last cap to ground/ 0V.

I use 47µF electrolytes and the final one is a 22µF in some cases using two double triodes.

Best regards

/John

 

[abbey road d enfer]

i did try additional resistor after the choke, that helped a lot at mid to high frequency spikes!

What was the spectrum with the original arrangement?

 

[kambo]

What was the spectrum with the original arrangement?

white lines (ASIO lynx)


edit: mic pre for the mixer has filtered PSU, with C-R-C-R-C-R-c arrangement,
just measured, its way better than the green line.... no spikes, and no 60 hz bump!
i am trying to use those chokes, been sitting in my stash for so long...
i dont think i will be able to sell them, they r pretty prices!

 

[abbey road d enfer]

white lines (ASIO lynx)

OK, I see. Do you measure the actual B+ rail? Hopefully with a high-voltage capacitor in series? Where is the meauremnt probe ground connected?

 

[kambo]

OK, I see. Do you measure the actual B+ rail? Hopefully with a high-voltage capacitor in series? Where is the meauremnt probe ground connected?

i dont know if this gonna me the exact technical setup explanation, but basically 
i am measuring at audio output transformer of actual  mixer test setup.

my test set up is:

balanced line in(10k:10k)---> 6SN7 line input (adjust gain)---> unbalanced out --->
SRPP mic pre input trx (adjust gain out)----> 2:1 output trx

ground is from XLR out.

 

[abbey road d enfer]

i dont know if this gonna me the exact technical setup explanation, but basically 
i am measuring at audio output transformer of actual  mixer test setup.

my test set up is:

balanced line in(10k:10k)---> 6SN7 line input (adjust gain)---> unbalanced out --->
SRPP mic pre input trx (adjust gain out)----> 2:1 output trx

ground is from XLR out.

Aaah! OK, so you're measuring the output of the complete circuit... So, it could be a problem of magnetic coupling between the choke and one of the xfmrs. Because it doesn't make sense attributing the difference to a more or less well-filtered B+. 60Hz should be lower than 120Hz, because full-wave rectification cancels fundamental, and harmonics are filtered at about 20dB/oct. The spectrum does show 60 Hz and does not exhibit the 20dB/oct decrease; that indicates the origin of hum components is not in the filtering of B+.
Maybe the simple presence of the choke deflects the stray magnetic field of the PT? You may want to try using a piece of steel as a shield/field deflector to investigate further.
BTW, swapping resistor and inductor doesn't make significant changes; both cases exhibit a VLF bump (between 2-4Hz) but in both cases it's way below audible frequencies.

 

[CJ]

wire it for max rejection as per data sheet

resistance to ac s/b 6.28 * 10 = 62.6 * 120 = 7.5 K (series)

kind of a waste as this choke will handle 10 times the current you are using,

maybe save it for a power amp and use resistors for the 6SN7 as do many box builders,

 

[kambo]

i will be running 6 x 6SN7  and 2 x 5751  and may be additional 2xEF86 for additional FX.
and i have 2 of these chokes... one for L one for R channel
had them long time ago...

i did try all  available connection options according to data sheet , and i have double folded thin
mu_metal sheets... no measurable  effect that would justify the cost of mu_metal sheets!

in side/outside of the enclosure is not making noticeable difference either....

nothing is really making sense with these chokes

common mode rejection seems to be working better tho.
but, using a 1K resistor after the choke is making more measurable effect!
funny enough, if i use 1k resistor before the choke, not much of the same effect ???

when i first got them zillion years ago for my tube mic pre, i had the same problem.... they ended up unused in my stash...

ordered more caps today... i will report back this weekend.... this is not making sense at all!

 

[CJ]

try measuring ripple with a voltmeter, put the ac probe before the choke and after and measure ac volts,
do the same with the resistor swap,

at this point we question the measurement system and thus try verify this with alternate methods,

are you loading the output?

 

[trobbins]

Kambo,
It appears that you are down near noise floor of measurement system.
How well have you laid out the PT and rectifier and first filter cap for minimal loop area and egress of noise, and is the PT shielded?
Can you reduce the noise floor of measurement system by itself - battery powering - optical link or usb isolator - laptop?
Can you shield the choke ?
Can you replace the OT/load circuitry with a load resistor drawing the same current?

 

[kambo]

WIP - moving the whole PSU to an external enclosure.

 

[DerEber]

Yea, maybe the magnetic field of your choke is coppling into your circuit.
Try to move the choke outside of the chassis.