DI/Reamp Theory Questions

[MrG]

You need a 30dB or 40dB pad. 30dB is 330 0 times so the output leg of the pad needs to be one about one 30th of the input leg. The output leg also needs to be low enough to feed the transformer. Since its is not clear what this is lets go for 100 ohms. The input leg needs to be 30 times this or 3K. Lets make it balanced so split the 3K into a pair of 1K5 resistors. So wire 1K5 from DAC hot to 100 ohms. Other end of 100 ohms to other 1K5 and other end of this 1K5 to DAC cold. Feed the transformer from either end of the 100 ohms.

Cheers

Ian

Thanks very much, Ian! Very enlightening. You say “30db is 330 0 times” and the mention 1/30th and * 30 - is this a typo or could you explain the 330 0?

The transformer in is 200 ohms. So, I suppose it would be 6k resistor, split on balanced lines so two 3ks. Can’t wait to try this and see how it sounds.

Best,
MG

 

[john12ax7]

Thanks for this, John. So when you say an 1176 is 600, do you mean that the input and output impedance are both 600? Is most old school line-level equipment effectively 1:1 in this way? And nowadays it’s a high input bridging impedance of ~10-20k(?), and a lower output impedance of ~600 (or lower?)?

I hear you about 1:1 xfmrs and am already using this a bunch in application. I’m about to install a switch on my WE 111cs to drop the secondary from 600 to 140 and am also trying to think through all of this theory before using that in series with other line level equipment.

Best,
MG

The 1176 input is 600 ohm.  The output I would need to double check,  but in general gear was designed with the expectation of a 600 ohm load. This is why with some vintage gear you connect a termination resistor when interfacing with modern higher impedance  gear. The 600 ohm thing is a remnant of the vacuum tube days.

Things don't need to be 1:1, the important thing is really that it works with 600 ohms.  For example a 2:1 input and then 1:2 output is common in Neve.  This gives overall unity gain but increased internal headroom.  The downside is the output stage now needs to be able to drive 150 ohms (due to the 1:2 transformer). Neves have beefy output stages as a result.  But you wouldn't want to use that same 1:2 transformer with a NE5534.

Vacuum tubes have inherently high impedance,  so they will use even higher ratios to properly interface with 600.

For modern gear, 50-100 ohm output impedance and 10k-20k input impedance is rather common.

 

[john12ax7]

One more thing.  If you have a known load,  in this case your DAC to ADC,  then you can also safely experiment with other ratios. For example if the ADC input is 10k, then you could try a 1:2, 1:3, or 1:4 turns ratio transformer.  The 1:4 would be 1:16 impedance,  so the DAC would see 625 ohms (10k/16).

Just don't use these transformers and then connect an 1176 on the end.

 

[MrG]

The 1176 input is 600 ohm.  The output I would need to double check,  but in general gear was designed with the expectation of a 600 ohm load. This is why with some vintage gear you connect a termination resistor when interfacing with modern higher impedance  gear. The 600 ohm thing is a remnant of the vacuum tube days.

Things don't need to be 1:1, the important thing is really that it works with 600 ohms.  For example a 2:1 input and then 1:2 output is common in Neve.  This gives overall unity gain but increased internal headroom.  The downside is the output stage now needs to be able to drive 150 ohms (due to the 1:2 transformer). Neves have beefy output stages as a result.  But you wouldn't want to use that same 1:2 transformer with a NE5534.

Vacuum tubes have inherently high impedance,  so they will use even higher ratios to properly interface with 600.

For modern gear, 50-100 ohm output impedance and 10k-20k input impedance is rather common.

Great info, thank you. When you say a Neve output has to be able drive 150 ohms in it’s output stage due to the 1:2 xfmr, how is this number arrived at? In my admittedly new understanding, I was thinking simply 600/2 is 300 ohms of load impedance for the output of the Neve. Thanks in advance for setting me straight!

Edit: just saw your post and it confirmed what I was just realizing - 1:2 turns ratio, squared is voltage impedance ratio, so divided by 4 - hope I’m right.

MG

 

[john12ax7]

I should have specified,  yes 1:2 meant turns ratio,  which equates to 1:4 impedance ratio.

In general something like 1:2 is accepted to mean turns ratio.  Impedance ratios are generally shown as numbers related to the working impedance,  600:600, 150:600, 150:10k, etc.

 

[abbey road d enfer]

I’d like to simplify this particular design and not use a pot, but rather a resistor to drop the signal approximately 30-40dB.

Please note that doing so, the xfmr will distort no more. Actually, with your existing xfmr and other equipment, I see only one possibility to make the xfmr distort while operating the DAC "safely", which consists in inserting a resistor of a few hundred ohms (value to be determined using a potentiometer) between output of the DAC and the primary of the xfmr AND a pad between the secondary and the input of the ADC. This pad in its simplest form would be a resistor in series, again value to be adjusted with a potentiometer.
The xfmr would be wired as a step-up (DAC feeding the 200r winding and 50k winding feeding the ADC).

 

[JohnRoberts]

Thanks very much, Ian! Very enlightening. You say “30db is 330 0 times” and the mention 1/30th and * 30 - is this a typo or could you explain the 330 0?

I'm no Ian but 30dB is 30x voltage ratio.

I am sure Ian knows this so a typo is likely suspect.

JR

The transformer in is 200 ohms. So, I suppose it would be 6k resistor, split on balanced lines so two 3ks. Can’t wait to try this and see how it sounds.

Best,
MG

 

[ruffrecords]

Thanks very much, Ian! Very enlightening. You say “30db is 330 0 times” and the mention 1/30th and * 30 - is this a typo or could you explain the 330 0?

The transformer in is 200 ohms. So, I suppose it would be 6k resistor, split on balanced lines so two 3ks. Can’t wait to try this and see how it sounds.

Best,
MG

MEGA TYPO.30dB is 30 times.

Cheers


Ian

 

[PRR]

> 30dB is 30 times.

Before we get too stuck on coincidence:

20dB _IS_ 10:1 of voltage (or 100:1 of power),
but,
30dB is 31.623:1 of voltage (1000:1 power).

Because the error is less than 1/2dB, yes we all use "30:1" for fast figuring.

 

[MrG]

I should have specified,  yes 1:2 meant turns ratio,  which equates to 1:4 impedance ratio.

In general something like 1:2 is accepted to mean turns ratio.  Impedance ratios are generally shown as numbers related to the working impedance,  600:600, 150:600, 150:10k, etc.

Thanks much, John.

Please note that doing so, the xfmr will distort no more. Actually, with your existing xfmr and other equipment, I see only one possibility to make the xfmr distort while operating the DAC "safely", which consists in inserting a resistor of a few hundred ohms (value to be determined using a potentiometer) between output of the DAC and the primary of the xfmr AND a pad between the secondary and the input of the ADC. This pad in its simplest form would be a resistor in series, again value to be adjusted with a potentiometer.
The xfmr would be wired as a step-up (DAC feeding the 200r winding and 50k winding feeding the ADC).

Understood re this pad not affecting/increasing distortion. My interest in distortion is not my primary interest, but rather an option I’d like to creatively exploit.

Thank you for the above detailed suggestion. Could you please shed some light on why a 100 ohm resistor on input (in series?) would increase distortion? And if I put that pad on the output, does it affect the impedance that the load will see from my source, or will 50k still be what’s seen? (I know those numbers are mainly theoretical and will alter based on the impedances it’s connected to.

I'm no Ian but 30dB is 30x voltage ratio.

I am sure Ian knows this so a typo is likely suspect.

JR

MEGA TYPO.30dB is 30 times.

Cheers


Ian

Thanks John and Ian! Didn’t mean to bust anyone about this, I just didn’t know what to think of it given my current knowledge.

> 30dB is 30 times.

Before we get too stuck on coincidence:

20dB _IS_ 10:1 of voltage (or 100:1 of power),
but,
30dB is 31.623:1 of voltage (1000:1 power).

Because the error is less than 1/2dB, yes we all use "30:1" for fast figuring.

Very good to know and makes perfect sense - thanks, PRR.

Best,
MG

 

[abbey road d enfer]

Could you please shed some light on why a 100 ohm resistor on input (in series?) would increase distortion?

As I mentioned earlier, a xfmr is supposed to be fed from the lowest possible impedance; that's because increasing the source impedance degrades the low frequency response and increases low-frequency distortion. You have to take it at face value since demonstrating that implies in-depth knowledge of magnetism.

And if I put that pad on the output, does it affect the impedance that the load will see from my source, or will 50k still be what’s seen?

Yes it will; that's why it's impossible to calculate precisely without knowing other parameters.

 

[ruffrecords]

If you want to understand more about audio transformers you could do worse than read this by Bill Whitlock of Jensen:

http://jensen-transformers.com/wp-content/uploads/2014/09/Audio-Transformers-Chapter.pdf

Very readable and hardly any maths.

Cheers

Ian

 

[MrG]

If you want to understand more about audio transformers you could do worse than read this by Bill Whitlock of Jensen:

http://jensen-transformers.com/wp-content/uploads/2014/09/Audio-Transformers-Chapter.pdf

Very readable and hardly any maths.

Cheers

Ian

Thank you, Ian! I will absolutely read this.

As I mentioned earlier, a xfmr is supposed to be fed from the lowest possible impedance; that's because increasing the source impedance degrades the low frequency response and increases low-frequency distortion. You have to take it at face value since demonstrating that implies in-depth knowledge of magnetism.

Thanks and I understand what you’re saying. I’m not looking to question basic tenets of electronics which I don’t yet understand. What I was curious about is that in this scenario, I was thinking you were saying I could safely distort the 200ohm input of the xfmr by adding a several hundred ohm resistor (value to be determined by a potentiomenter) in series.

So I’m thinking your example means (with a 100 ohm resistor, for example) ultimately increasing the input impedance to 300 from 200, lowering the voltage impedance ratio of the transformer from 250:1 to ~167:1, and thus increasing the impedance that the DAC sees from ~40 to ~60 ohms. Is this correct? Is this effectively lowering the “feeding from the lowest source impedance” that you mention above?

The only thing I’m still confused by is that previously I was getting distortion, in theory, only from having a higher impedance from DAC (~100) than the input to the xfo (~40), so I don’t understand how this is much different than that, or how it will increase distortion.

I very much appreciate your input, and my hope is to be able to add insights to this forum in the future as I learn more, just so you know I’m putting this to good use.

Edit: perhaps I’m misunderstanding that the best way to distort a transformer is via increased level into it, and not by impedance mismatching? Is it essentially _never_ a good idea to distort via “impedance methods”?

Best,
MG

 

[abbey road d enfer]

What I was curious about is that in this scenario, you were saying I could safely distort the 200ohm input of the xfmr by adding a 100 ohm resistor in series.

Did I say (or write) that? Pleasr read again.

So I’m thinking your example means (with a 100 ohm resistor) ultimately increasing the input impedance to 300 from 200, lowering the voltage impedance ratio of the transformer from 250:1 to ~167:1,

The xfmr ratio does not change; what changes is the impedance it is presented with. With 100 ohms in series (which I did not recommend), the xfmr's primary would see about 170-250 ohms. The impedance it sees is the sum of the DAC's output Z (50-100 ohms), the additional resistor (which you insist being 100r) and the DC resistance of the primary (20-50 ohms). The reflected impedance at the secondary would be 42k-62k, to which you must add the DC resistance of the secondary winding, probably about 500-1000 ohms.
Please note that the DAC would not be very happy to be loaded with about 120 ohms.

and thus increasing the impedance that the DAC sees from ~40 to ~60 ohms. Is this correct?

No, the reflected impedance at the input is very dependant on what the secondary is connected to.  If you connect it directly to a 10kohm  input, it's gonna be approximately 200.(10k/50k), or 40 ohms, to which you must add the DCr of the primary, for a total of about 60 ohms.

  Is this effectively lowering the “feeding from the lowest source impedance” that you mention above?

My suggestion of inserting a resistor in series is an answer to your wish of generating some distortion, which, as I mentioned earlier, increasing the source impedance favours.

The only thing I’m still confused by is that previously I was getting distortion, in theory, only from having a higher impedance from DAC (~100) than the input to the xfo (~40),

No. As I mentioned earlier, power is what generates distortion  The ratio of the source impedance to the nominal impedance is irrelevant.

Please note that 200:50k is a nominal impedance; it indicates that the xfmr will provide satisfactory performance (i.e. good frequency response and low distortion) with these impedances within a well-defined frequency range. This nominal value has no material existence. It's like speed limits; they do not mean that you can not go faster or slower, but you have to know the consequences.

 

[MrG]

Did I say (or write) that? Pleasr read again.
The xfmr ratio does not change; what changes is the impedance it is presented with. With 100 ohms in series (which I did not recommend), the xfmr's primary would see about 170-250 ohms. The impedance it sees is the sum of the DAC's output Z (50-100 ohms), the additional resistor (which you insist being 100r) and the DC resistance of the primary (20-50 ohms). The reflected impedance at the secondary would be 42k-62k, to which you must add the DC resistance of the secondary winding, probably about 500-1000 ohms.
Please note that the DAC would not be very happy to be loaded with about 120 ohms.
No, the reflected impedance at the input is very dependant on what the secondary is connected to.  If you connect it directly to a 10kohm  input, it's gonna be approximately 200.(10k/50k), or 40 ohms, to which you must add the DCr of the primary, for a total of about 60 ohms.
My suggestion of inserting a resistor in series is an answer to your wish of generating some distortion, which, as I mentioned earlier, increasing the source impedance favours.
No. As I mentioned earlier, power is what generates distortion  The ratio of the source impedance to the nominal impedance is irrelevant.

Please note that 200:50k is a nominal impedance; it indicates that the xfmr will provide satisfactory performance (i.e. good frequency response and low distortion) with these impedances within a well-defined frequency range. This nominal value has no material existence. It's like speed limits; they do not mean that you can not go faster or slower, but you have to know the consequences.

Ok, thank you, and sorry for confusion.

Yes, I was misquoting you I see - I thought I initally read you saying 100 ohms (to be precisely determined by a potentiomenter) but now I see you said several hundred ohms to be determined by a potentiometer. I will edit my post above to reflect this.

Thanks for explaining. I understand that the xfmr impedances are nominal, and I understand that distortion is a matter of power, but obviously have more to learn as to how one “typically”, purposely, and safely generates it. I’ll research.

MG

 

[MrG]

The xfmr ratio does not change; what changes is the impedance it is presented with. With 100 ohms in series (which I did not recommend), the xfmr's primary would see about 170-250 ohms. The impedance it sees is the sum of the DAC's output Z (50-100 ohms), the additional resistor (which you insist being 100r) and the DC resistance of the primary (20-50 ohms). The reflected impedance at the secondary would be 42k-62k, to which you must add the DC resistance of the secondary winding, probably about 500-1000 ohms.
Please note that the DAC would not be very happy to be loaded with about 120 ohms.

Just to understand this theory: so, in the above example, the primary of the xfmr (200:50k connected to 10k) sees the impedance of DAC out (100) + resistor in series (100-300, say) + DC resistance of xfmr (50-100)? Then would the DAC see the load as (10k/250) 40 + 100-300r + DC resistance of xfmr? I’m trying to figure out how you reached the 120 ohm load on the DAC.

When it comes to more to gear like say an 1176, are impedances still calculated across what is connected on both ends, i.e. the connection on one end divided or multiplied by the overall ratio of the gear? Or does one just look at the impedance at the source vs. the load more straightforwardly?

Thanks,
MG

 

[abbey road d enfer]

Just to understand this theory: so, in the above example, the primary of the xfmr (200:50k connected to 10k) sees the impedance of DAC out (100) + resistor in series (100-300, say) + DC resistance of xfmr (50-100)? Then would the DAC see the load as (10k/250) 40 + 100-300r + DC resistance of xfmr? I’m trying to figure out how you reached the 120 ohm load on the DAC.

Well, I figure out I just forgot the reflected impedance. So more correctly 40 ohms (reflected Z) + 20 ohms (guestimated DCr) + 100 ohms (additional series resistor) =160 ohms. My comment about the DAC not being happy is still valid.

When it comes to more to gear like say an 1176, are impedances still calculated across what is connected on both ends, i.e. the connection on one end divided or multiplied by the overall ratio of the gear? Or does one just look at the impedance at the source vs. the load more straightforwardly?

Actually, the input and output impedances are real values, that can be measured; that is because these units were designed to operate in matching mode. They are supposed to see and drive 600r impedance for proper operation. When operated in bridging mode, one is supposed to install load resistors in order to compensate the higher impedance they are presented to.

 

[MrG]

Well, I figure out I just forgot the reflected impedance. So more correctly 40 ohms (reflected Z) + 20 ohms (guestimated DCr) + 100 ohms (additional series resistor) =160 ohms. My comment about the DAC not being happy is still valid.
Actually, the input and output impedances are real values, that can be measured; that is because these units were designed to operate in matching mode. They are supposed to see and drive 600r impedance for proper operation. When operated in bridging mode, one is supposed to install load resistors in order to compensate the higher impedance they are presented to.

Very interesting. Thank you!

So, if this xfmr going 200:50k were seeing ADC in at ~25k, then is there potential danger to the xfmr secondary and is it likely to distort the ADC along with attenuating?

And is there a formula for how much singal attenuation happens when the load impedance is lower than the source?

Thanks,
MG

 

[abbey road d enfer]

So, if this xfmr going 200:50k were seeing ADC in at ~25k, then is there potential danger to the xfmr secondary

Not really. The primary would be the most likely to fail; that would happen only if too much power is transmitted.

and is it likely to distort the ADC along with attenuating?

It would distort only if the power exceeds a certain level.

And is there a formula for how much singal attenuation happens when the load impedance is lower than the source?

Ohm's law. Voltage divider. Valid whatever the ratio between source impedance and load.

 

[MrG]

Hey all,

Just wanted to report back that I went with @ruffrecords suggestion of putting resistors on the lo-z side of this xfmr for reamping, so I could keep the hi-z reamp side facing seeing the amp and it worked beautifully. I knocked the signal down by approximately 30dB before reamping and the resulting sound interfaces wonderfully with my gear and sounds very full-featured. Thank you all again!

Now another question that’s a bit more creative:

If I wanted to use this di/reamp box in either direction (200>50k or 50k>200) as a line level insertable tone box in any random AD/DA, presuming DAC is 100 ohms and ADC is 10k, how would you all recommend I create acceptable impedances on either side without losing gain, while keeping it passive?

Thanks in advance!
MG