Damping transformer ringing (Zobel network question)

[JAY X]

Hi All,

Does anyone know which is the gain output of the OEP A262A2E ( for 1:1 balanced line output) compared to sowter 8403x? I use to drive  the sowter from an opamp output.

The capacitors used in zobel network in pf values ¿can be ceramic? if in NF value ¿X type? ¿polyester?

Thank you very much!
JAY X

 

[CJ]

lot of ceramic in old school stuff, go for poly if you want a less crumbly packaging material,
ever see corona of the ends of a worn out ceramic? the plates get exposed when the thin edge of the cap gets worn down or a chunk gets chipped off,

jensen sent me a poly cap with their 1:10 mic in but i thru it out,

do a listening test and see if you can hear a difference, would be interesting to know,

 

[mrdarwin]

Hi!
Anyone experimented zobel value for ll1540 wired 2:1?  Primary serie, secondary parallel.
Lundahl gives some values for secondary serie, is it available for parallel wired secondary?
I'd like substitute a vtb9046
thx

http://www.lundahl.se/wp-content/uploads/datasheets/1540.pdf

 

[abbey road d enfer]

Hi!
Anyone experimented zobel value for ll1540 wired 2:1?  Primary serie, secondary parallel.
Lundahl gives some values for secondary serie, is it available for parallel wired secondary?
I'd like substitute a vtb9046
thx

http://www.lundahl.se/wp-content/uploads/datasheets/1540.pdf

For parallel connection, you should typically double the capacitor and halve the resistor. You should be pretty close.
Remember that ideal Zobel calculation should include the source and load impedance; that means as soon as you don't load with the specific load the Zobel has been calculated for, you're slightly out of whack. It's the designer's choice to find which optimization conditions he imposes and hope it will cover most practical implementations.

 

[mrdarwin]

Thx Mr abbey.
Load source will be 56ohms, out of an orion 32 soundcard.
Input is a ba283 igor board.
Where do you find Zobel calculation?

 

[abbey road d enfer]

Where do you find Zobel calculation?

https://www.trueaudio.com/st_zobel.htm
The principle is to write the impedance equation of the circuit withtwo branches in parallels, Re and Le in series and Rz and Cz in series, then find the solution for minimizing the imaginary term and keeping the HF value equal to the LF value. Can be done with a few pages of math or a simulation software.

 

[mrdarwin]

https://www.trueaudio.com/st_zobel.htm
The principle is to write the impedance equation of the circuit withtwo branches in parallels, Re and Le in series and Rz and Cz in series, then find the solution for minimizing the imaginary term and keeping the HF value equal to the LF value. Can be done with a few pages of math or a simulation software.

Thanks

 

[benb]

The link in the OP is dead, but if anyone's interested, I found it saved in Archive:
https://web.archive.org/web/20130325202408/http://www.siteswithstyle.com/VoltSecond/Damping_ringing_XFMRS/Damping_ringing_in_xfmrs.html

Hi,

Surfing the net trying to find some article on how to correctly damp transformer ringing, I came across this page.
I think it has some error in it: I tried to calculate the Ro, with no correspondence with the calculations in the text.
What's wrong?

The site cites the following textbook:
"Design Techniques for Preventing Input-Filter Oscillations in Switched-Mode Regulators",
Did anyone experienced it? Any link to a pdf?

Thanks.
:

 

[Gold]

I am trying do tame a high frequency rise in a pair of input transformers pulled from a Neumann SAL74. They are probably Haufe but are unlabeled. There was no zobel attached. I haven't ever seen one used on this transformer. I am repurposing it as a line in transformer for a Quad Eight 312.

When I use it with no zobel I have a peak around 45K. It's about +1dB@15K and +2dB@20k. Doing a frequency sweep using the AP P1.

I built a jig according to post #6. I inserted the jig between the transformer secondary and the QE Line In. I used an Analog Discovery2 and was able to get a nice looking square wave with a 0.01uF and a 2kR. It worked like a champ to flatten out the top end. It's within a couple 10ths of a dB at 15K and pretty close at 20K.

Now when I do a frequency sweep with the AP P1 I'm getting a dip in the midrange now of almost 1dB. Centered around about 3K. I don't like that at all. Am I stuck with the choice of either a rising high end response or a midrange dip? Anything else to try?

 

[HenryL]

I am trying do tame a high frequency rise in a pair of input transformers pulled from a Neumann SAL74. They are probably Haufe but are unlabeled. There was no zobel attached. I haven't ever seen one used on this transformer. I am repurposing it as a line in transformer for a Quad Eight 312.

When I use it with no zobel I have a peak around 45K. It's about +1dB@15K and +2dB@20k. Doing a frequency sweep using the AP P1.

I built a jig according to post #6. I inserted the jig between the transformer secondary and the QE Line In. I used an Analog Discovery2 and was able to get a nice looking square wave with a 0.01uF and a 2kR. It worked like a champ to flatten out the top end. It's within a couple 10ths of a dB at 15K and pretty close at 20K.

Now when I do a frequency sweep with the AP P1 I'm getting a dip in the midrange now of almost 1dB. Centered around about 3K. I don't like that at all. Am I stuck with the choice of either a rising high end response or a midrange dip? Anything else to try?

This is probably a rather naieve suggestion from an amateur - but after several experiments with capacitors across input TX (albeit not quite the same scenario as yours) I've decided against them... I"ve been recently seeking here (rather fruitlessly) some more experienced input on this. I also have gone along the Dale Roche lines with a Neve type preamp TX input stage design to try to tune a new transformer into the existing circuit.

Just a wild idea perhaps, and I'm not familiar with the particular transformer or 'Quad 8 312' but FWIW my own experience so far is that if you are tying to tame input TX ringing/overshoot (That's essentially what the Dale Roche process is about?) then perhaps just using a load resistor, no cap might be a better angle - this can knock out ringing/overshoot, perhaps avoiding some frequency-selective effects from the capacitor (acting in unwanted combination with the following amp input behaviour perhaps).

You seem to be trying to achieve a broader Q EQ cut - but the effect of taming the top end might perhaps also be achieved by load resistor without the additional reactive component of the capacitor to complicate things. Just an idea...

 

[Gold]

This is probably a rather naieve suggestion from an amateur - but after several experiments with capacitors across input TX (albeit not quite the same scenario as yours) I've decided against them... I"ve been recently seeking here (rather fruitlessly) some more experienced input on this. I also have gone along the Dale Roche lines with a Neve type preamp TX input stage design to try to tune a new transformer into the existing circuit.

Just a wild idea perhaps, and I'm not familiar with the particular transformer or 'Quad 8 312' but FWIW my own experience so far is that if you are tying to tame input TX ringing/overshoot (That's essentially what the Dale Roche process is about?) then perhaps just using a load resistor, no cap might be a better angle - this can knock out ringing/overshoot, perhaps avoiding some frequency-selective effects from the capacitor (acting in unwanted combination with the following amp input behaviour perhaps).

You seem to be trying to achieve a broader Q EQ cut - but the effect of taming the top end might perhaps also be achieved by load resistor without the additional reactive component of the capacitor to complicate things. Just an idea...

I’m trying ro get flat frequency response from 20hz-20k hz. I have no special attachment to using a capacitor. If a load resistor might do it then I’ll give it a shot. Do I connect a decade resistor box across the secondary while looking at a square wave?

 

[Brian Roth]

All of the scans I have here for the Ampex 350 along with a schemo I marked up re. the power supply.

I’m trying ro get flat frequency response from 20hz-20k hz. I have no special attachment to using a capacitor. If a load resistor might do it then I’ll give it a shot. Do I connect a decade resistor box across the secondary while looking at a square wave?

IMHO......a great method!

Bri

 

[user 37518]

IMHO......a great method!

Bri

I remember sometime ago reading an article by some analog guru (perhaps it was Jim Williams) and the article was about op-amp stability, and how to stabilize them. I expected something really fancy and perhaps math heavy, but it ended up involving a capacitor decade box.

 

[HenryL]

I’m trying ro get flat frequency response from 20hz-20k hz. I have no special attachment to using a capacitor. If a load resistor might do it then I’ll give it a shot. Do I connect a decade resistor box across the secondary while looking at a square wave?

Personally I wouldn't say that's a great way of checking the overall frequency response - even though there are a lot of frequencies in a square wave and it can tell the experienced eye things about the frequency response... I'd say the square wave is good for looking for ringing/damping as you can see the effect of banging the input with a step change, but for freq response I'd do what you were already doing and use Audio Precision or REW or such like whilst varying the load resistance with a box, or just clipping in a sequence of values.

EDIT: ps I mean measuring the frequency response of the whole system from input to output with an analyzer such as an Audio PRecision (assuming that's what AP means here) or Room Eq Wizard - I mean not measuring at the TX as in the Dale Roche damping technique in post #6. Though I expect that's what you were doing anyway!

 

[abbey road d enfer]

I’m trying ro get flat frequency response from 20hz-20k hz. I have no special attachment to using a capacitor. If a load resistor might do it then I’ll give it a shot. Do I connect a decade resistor box across the secondary while looking at a square wave?

For a mic input, there is no doubt a Zobel is better than a single resistor, because it allows better signal transfer. For a line input, it may not matter so much.
remember that the Zobel is tuned for a specific set of source and load impedances. The discrepancy between AD2 and AP may just be a difference in source Z.

 

[HenryL]

For a mic input, there is no doubt a Zobel is better than a single resistor, because it allows better signal transfer. For a line input, it may not matter so much.
remember that the Zobel is tuned for a specific set of source and load impedances. The discrepancy between AD2 and AP may just be a difference in source Z.

Assume by 'signal transfer' here you mean loss of level?

 

[HenryL]

For a mic input, there is no doubt a Zobel is better than a single resistor, because it allows better signal transfer. For a line input, it may not matter so much.
remember that the Zobel is tuned for a specific set of source and load impedances. The discrepancy between AD2 and AP may just be a difference in source Z.

What's AD2 ?

 

[abbey road d enfer]

Assume by 'signal transfer' here you mean loss of level?

Yes.

 

[abbey road d enfer]

What's AD2 ?

Digilent Analog Discovery 2

 

[abbey road d enfer]

@Gold: Apparently, the AD2 has near-zero output Z, when the Portable One has selectable 40, 150 or 600. That may (or may not) explain the difference.
Do you note a difference when switching Z on the P1?

 

[HenryL]

@Gold: Apparently, the AD2 has near-zero output Z, when the Portable One has selectable 40, 150 or 600. That may (or may not) explain the difference.
Do you note a difference when switching Z on the P1?

This seems a good question! The 'Dale Roche technique' Gold refers to from post 6 in this thread involves setting a realistic generator source impedance, but as you say that also has to be considered for the freq response measurements...