Hi--
This is my first post in DIY forum. I'm a newbie. I've read lots of posts about attenuators and can't figure this out. I know this is a common problem but I would appreciate some input as I feel like I could fry my DAC from not knowing what I'm doing. Basically the details below are to help ask the question: How could you use an Autotransformer as a fixed line attentuator at the output of a digital DAC? Won't just inserting the autotransformer almost short out the output of the DAC if you don't insert resistors or something else into the signal path?
I would like to drop the output of my DAC (2.5v 100ohm output) down to the input of a high gain amp with 0.75Vmax in, 51Kohm input impedance. The amp has a 35db gain with 200W to 8ohm speakers that are 91db efficient. I would like to use the DAC (a Topping D70S MQA-- with 124db SNR) as the volume control and just drop the DAC ouptut signal at least 10db. I don't listen to music very loud and sometimes see the gain on the DAC right now at -38 to -40db output volume. I stream ultra HD (24bit, 192khz) I would like as little messing up of the line signal as possible as the system is very transparent (Job225 amp + reference 3a de capo i speakers). For that reason, and because I'm interested in them, I'm considering using an autotransformer as the fixed attenuator and use no preamp. I have a pair of Bob Crites 3636 autotransformers that have the attenuation capacity of -1 to -12db. I'm thinking of using them in place of some good L pad resistors that might be used to reduce gain.
So multiple questions:
1) If I just drop in the autotransformer between the DAC and the power amp, it seems to me like I'm shorting out the DAC in terms of load. Yes the autotransformer will have a little bit of resistance and a little bit of inducatance and capacitance, but that feels like a seriously low load impedance. (I assume the DAC only outputs AC signals but still...not sure about this). Is there something I should do to keep my DAC happy as far as the output load? (Stick in a resistor? but that seems like it might do something to the signal I'm trying to avoid by using the autotransformer.) Is there anything I need to do on the output of the autotransformer to keep the AMP input happy? What does the circuit look like? Unlike most people who are using stepped attenuators or selectors I'm just wanting a fixed drop in voltage that still retains the nauances of high resolution signals.
2) Would it possibly worth a try to reduce the input signal more than the 10.4dB that it would take to drop 2.5v to 0.75v? Because I'm not after high volume (I think I listen to music often at only 60 to 75db SPL)... should I try another autotransformer or would it make sense to drop in a super quality resistor somewhere in the path to further reduce the overall gain? (If so... I know this maybe be hotly debated question: what resistors might I use and where should I get them). I'm asking about resistors also in part because I might simply compare my autotransformer fixed line attenuator to a simple fixed line attenuator using very good quality resistors.
I thought by simply needing a fixed attenuator I might be able to come up with a line attenutor that doesn't distort the nuances and speed and transparency and imaging of the signal very much. Also I want to avoid spending a fortune on a good preamp. (I have looked for all kinds of fixed line attenuators but many of the most recommended ones seem no longer available).
I loved the alive and transparent sound I hear by simply driving an amp with the DAC and using the DAC as a volume control. I have used the DAC with l lower gain amp and less efficient speakers and it works well. With the super high gain amp, it sounds amazingly alive and imaging is great, but I think it's missing some energy or depth by my operating the DAC at about -40db.
Any pearls of wisdom? Thoughts? Suggestions? I'm just getting into hi-fi as I'm a few years from retiring and want to start really enjoying the music.
Thank you very much.
This is my first post in DIY forum. I'm a newbie. I've read lots of posts about attenuators and can't figure this out. I know this is a common problem but I would appreciate some input as I feel like I could fry my DAC from not knowing what I'm doing. Basically the details below are to help ask the question: How could you use an Autotransformer as a fixed line attentuator at the output of a digital DAC? Won't just inserting the autotransformer almost short out the output of the DAC if you don't insert resistors or something else into the signal path?
I would like to drop the output of my DAC (2.5v 100ohm output) down to the input of a high gain amp with 0.75Vmax in, 51Kohm input impedance. The amp has a 35db gain with 200W to 8ohm speakers that are 91db efficient. I would like to use the DAC (a Topping D70S MQA-- with 124db SNR) as the volume control and just drop the DAC ouptut signal at least 10db. I don't listen to music very loud and sometimes see the gain on the DAC right now at -38 to -40db output volume. I stream ultra HD (24bit, 192khz) I would like as little messing up of the line signal as possible as the system is very transparent (Job225 amp + reference 3a de capo i speakers). For that reason, and because I'm interested in them, I'm considering using an autotransformer as the fixed attenuator and use no preamp. I have a pair of Bob Crites 3636 autotransformers that have the attenuation capacity of -1 to -12db. I'm thinking of using them in place of some good L pad resistors that might be used to reduce gain.
So multiple questions:
1) If I just drop in the autotransformer between the DAC and the power amp, it seems to me like I'm shorting out the DAC in terms of load. Yes the autotransformer will have a little bit of resistance and a little bit of inducatance and capacitance, but that feels like a seriously low load impedance. (I assume the DAC only outputs AC signals but still...not sure about this). Is there something I should do to keep my DAC happy as far as the output load? (Stick in a resistor? but that seems like it might do something to the signal I'm trying to avoid by using the autotransformer.) Is there anything I need to do on the output of the autotransformer to keep the AMP input happy? What does the circuit look like? Unlike most people who are using stepped attenuators or selectors I'm just wanting a fixed drop in voltage that still retains the nauances of high resolution signals.
2) Would it possibly worth a try to reduce the input signal more than the 10.4dB that it would take to drop 2.5v to 0.75v? Because I'm not after high volume (I think I listen to music often at only 60 to 75db SPL)... should I try another autotransformer or would it make sense to drop in a super quality resistor somewhere in the path to further reduce the overall gain? (If so... I know this maybe be hotly debated question: what resistors might I use and where should I get them). I'm asking about resistors also in part because I might simply compare my autotransformer fixed line attenuator to a simple fixed line attenuator using very good quality resistors.
I thought by simply needing a fixed attenuator I might be able to come up with a line attenutor that doesn't distort the nuances and speed and transparency and imaging of the signal very much. Also I want to avoid spending a fortune on a good preamp. (I have looked for all kinds of fixed line attenuators but many of the most recommended ones seem no longer available).
I loved the alive and transparent sound I hear by simply driving an amp with the DAC and using the DAC as a volume control. I have used the DAC with l lower gain amp and less efficient speakers and it works well. With the super high gain amp, it sounds amazingly alive and imaging is great, but I think it's missing some energy or depth by my operating the DAC at about -40db.
Any pearls of wisdom? Thoughts? Suggestions? I'm just getting into hi-fi as I'm a few years from retiring and want to start really enjoying the music.
Thank you very much.
Last edited:
Yes, an L pad with metal film resistors will be better than transformers, but you can try both and see.
I'd scale up the resistors so they sum to around 10k, with a series 7k and a shunt 3k.
Locate the attenuator near the amplifier.
Using the transformers, the DAC will have internal output coupling capacitors, so no problems with DC.
I'd scale up the resistors so they sum to around 10k, with a series 7k and a shunt 3k.
Locate the attenuator near the amplifier.
Using the transformers, the DAC will have internal output coupling capacitors, so no problems with DC.
Last edited:
question about diagram
(I've never replied on any forum and hope it works okay..)
Thanks for the diagram PRR... i assume you are simply trying to say: use an Lpad circuit to attenuate? Is that what you are suggesting-- because I don't understand the actual values you have in the diagram. The values themselves seem like they would not result in enough attenuation? -- thank you.
(I've never replied on any forum and hope it works okay..)
clarification questions
Thank you rayma for responding. I do have a few questions if you may.
Do you personally think metal film resistors are the way to go for high end audio? (I don't know anything about resistors but it seems everyone has their opinions. And should I just look in the forum website to find suppliers or are there a few that most DIY folks go to for resistors?).
It's been about 40 years since I played much with Ohms law... but it seems like what you are suggesting by 10K series and 3K shunt would drop the voltage at the input down to about 0.56V or about -13db (Amp impedance is 51K). Does that feel like an okay thing since the amp can handle up to 0.75V? (but i would never play it at max!) What's the reasoning at setting max voltage below max allowed input of the amp?
I did some calculations and it seems that if I used say 1K series and 430ohm shunt it would attenuate down to 0.75V and be an attenuation of about -10.5db. (But then again I would be not want to be using max input to the amp and would be attenuating the DAC output). Why wouldn't you use 1K and say 300ohms? That would seem to me to provide a 10x load over the 100ohm source and the Lpad would be outputting to a 51K amp input impedance. I've heard of some vague rule of trying to have 10 to 1 ratios of source to loads...but don't really understand why. Why the 10K and 3K resistors verses smaller (like 1K and 300) or even larger values for that matter? (I don't have a feel for what the trade offs are).
Last question: I don't know how digital DACs are designed but it seems like you are saying I don't have to worry about "shorting the output" with a transformer because the output is purely AC with no DC component. So in other words I could just use the autotransformer and directly connect it to the output of the dac, and then connect the output taps of the autotransformer for say -12db and run that into the amp input? No resistors needed?
Thanks so much for your patience and help!!! Thanks to both you and PRR.
Thank you rayma for responding. I do have a few questions if you may.
Do you personally think metal film resistors are the way to go for high end audio? (I don't know anything about resistors but it seems everyone has their opinions. And should I just look in the forum website to find suppliers or are there a few that most DIY folks go to for resistors?).
It's been about 40 years since I played much with Ohms law... but it seems like what you are suggesting by 10K series and 3K shunt would drop the voltage at the input down to about 0.56V or about -13db (Amp impedance is 51K). Does that feel like an okay thing since the amp can handle up to 0.75V? (but i would never play it at max!) What's the reasoning at setting max voltage below max allowed input of the amp?
I did some calculations and it seems that if I used say 1K series and 430ohm shunt it would attenuate down to 0.75V and be an attenuation of about -10.5db. (But then again I would be not want to be using max input to the amp and would be attenuating the DAC output). Why wouldn't you use 1K and say 300ohms? That would seem to me to provide a 10x load over the 100ohm source and the Lpad would be outputting to a 51K amp input impedance. I've heard of some vague rule of trying to have 10 to 1 ratios of source to loads...but don't really understand why. Why the 10K and 3K resistors verses smaller (like 1K and 300) or even larger values for that matter? (I don't have a feel for what the trade offs are).
Last question: I don't know how digital DACs are designed but it seems like you are saying I don't have to worry about "shorting the output" with a transformer because the output is purely AC with no DC component. So in other words I could just use the autotransformer and directly connect it to the output of the dac, and then connect the output taps of the autotransformer for say -12db and run that into the amp input? No resistors needed?
Thanks so much for your patience and help!!! Thanks to both you and PRR.
I would install a dual potentiometer attenuator, initially, so you can twiddle and find the optimum attenuation setting. Then remove and duplicate that with a fixed pad.
When estimating and starting with a fixed pad, I always seem to find a recording in my collection that has lowish recorded levels and I can't get the volume high enough. You don't want to do that.
Dave.
When estimating and starting with a fixed pad, I always seem to find a recording in my collection that has lowish recorded levels and I can't get the volume high enough. You don't want to do that.
Dave.
question for Dave about dual potentiometer
Thanks Dave for offering that suggestion. I'm new to audio hardware. It seems like when I do a google search on dual potentiometers it brings up something with one input and two banks of potentiometers. Are you suggesting a device that can allow trying out the series resistance and independently trying out the shunt resistance or are they somehow linked in a ratio of some kind? (Is there any dual pot that you can suggest that does what you are talking about?) With such a thing can I dial in pretty much what I want for series, and also then dial in pretty much what I want for shunt.. and play with both values? Do you suggest 10K or 100K or something else for max value? Thanks so much!
Thanks Dave for offering that suggestion. I'm new to audio hardware. It seems like when I do a google search on dual potentiometers it brings up something with one input and two banks of potentiometers. Are you suggesting a device that can allow trying out the series resistance and independently trying out the shunt resistance or are they somehow linked in a ratio of some kind? (Is there any dual pot that you can suggest that does what you are talking about?) With such a thing can I dial in pretty much what I want for series, and also then dial in pretty much what I want for shunt.. and play with both values? Do you suggest 10K or 100K or something else for max value? Thanks so much!
Generally I'd use a total load of 10k, so the sum of the resistors would be about that.
Remember technically you have to include the loading effect of the power amp
(mainly for higher total impedance levels), and the source's output impedance
(mainly for lower total impedance levels).
But for your situation, 2.5V -> 0.75V with a total impedance of 10k,
then a series resistance 7k and a shunt resistance of 3k will be fine.
Because the L pad attenuation is: 3k/(3k+7k) = 0.3
and the attenuation factor needed is: 0.75V/2.5V = 0.3
To be more precise, subtract the 100R source from the 7k, to get 6.9k for the series resistor,
and choose the shunt resistor R so that: R // 51k = 3k, so R ~ 3.19k.
You can use any impedance the source can drive, usually 2k or higher.
Generally 10k is a good all round impedance level for modern audio.
Remember technically you have to include the loading effect of the power amp
(mainly for higher total impedance levels), and the source's output impedance
(mainly for lower total impedance levels).
But for your situation, 2.5V -> 0.75V with a total impedance of 10k,
then a series resistance 7k and a shunt resistance of 3k will be fine.
Because the L pad attenuation is: 3k/(3k+7k) = 0.3
and the attenuation factor needed is: 0.75V/2.5V = 0.3
To be more precise, subtract the 100R source from the 7k, to get 6.9k for the series resistor,
and choose the shunt resistor R so that: R // 51k = 3k, so R ~ 3.19k.
You can use any impedance the source can drive, usually 2k or higher.
Generally 10k is a good all round impedance level for modern audio.
Last edited:
calculations from rayma
Thank you again rayma. You have cleared up something in my brain. I wasn't sure how to include the output impedance of the DAC in any calculations. So all my calculations have been off (including my calculations of PRR's circuit). I wasn't realizing I needed to include the 100ohm resistance at the output of the DAC as part of the signal path... meaning that as the DAC gets loaded it will always drop some (small) voltage over that 100ohm resistor. Very helpful. I was able to use your updated values and include the 100ohm resistor to verify in my own mind that you indeed dropped the voltage to 0.75V. Is that what you typically aim for (allowing max input) or do you ever target something lower than max? Thank you!
Thank you again rayma. You have cleared up something in my brain. I wasn't sure how to include the output impedance of the DAC in any calculations. So all my calculations have been off (including my calculations of PRR's circuit). I wasn't realizing I needed to include the 100ohm resistance at the output of the DAC as part of the signal path... meaning that as the DAC gets loaded it will always drop some (small) voltage over that 100ohm resistor. Very helpful. I was able to use your updated values and include the 100ohm resistor to verify in my own mind that you indeed dropped the voltage to 0.75V. Is that what you typically aim for (allowing max input) or do you ever target something lower than max? Thank you!
I have two sick dogs here. They distract me and I can't keep a thought together. As Rayma figured, I swapped the values. Big dog little dog, big resistor little resistor, big/little dump on the carpet.
_I_ would not be afraid to put 2k load on a DAC. But the values Rayma suggests are not much higher and we are nowhere near the hiss level.
Hi, what about reducing the gain of the nameless high gain power amplifier? If that nameless device of unknown topology has an input stage with for instance an opamp then reducing the gain is easy peasy (often just 2 resistors). The source has more or less standardized output level and attenuating that with resistors seems walking backwards.
Of course this issue will not be answered efficiently as long as readers don't know the brand/type/schematic of the amplifier but reducing its gain seems most efficient as you said a low gain amplifier did the job better. So, the issue is the too high gain not the standard signal levels.
The 2 sick dogs then can stay outside and won't dump on the carpet
Last edited:
Thank you Roy for suggesting the Harrison labs attenuators. I've avoided them because some folks in another forum indicated they detected they affected the sound in really transparent systems.
PRR-- I hope your dogs are doing better. Thanks for the input. I once read that the lower the resistors the better in terms of noise because there's more current in the signal path... so your suggestions have about 2K in load..which seemed like a bit better than 10k load but I don't really know. I don't really know what hiss level is.
jean-paul-- thanks for your suggestion. On my very first post I did mention the amplifier but it's quite buried in my long post and my formatting was bad.... a job 225 amp (made by Job somehow related to Goldmund). I don't think I want to tear into that compact device and mess with it but thank you for the suggestion.
Thank you all!
Nobody liking autotransformer attenuation?
PRR-- I hope your dogs are doing better. Thanks for the input. I once read that the lower the resistors the better in terms of noise because there's more current in the signal path... so your suggestions have about 2K in load..which seemed like a bit better than 10k load but I don't really know. I don't really know what hiss level is.
jean-paul-- thanks for your suggestion. On my very first post I did mention the amplifier but it's quite buried in my long post and my formatting was bad.... a job 225 amp (made by Job somehow related to Goldmund). I don't think I want to tear into that compact device and mess with it but thank you for the suggestion.
Thank you all!
Nobody liking autotransformer attenuation?
Yes, most would want the option of fully driving the amplifier to clipping, or at least near it.
For parties or unsupervised users, you could substitute an attenuator that further reduces the level.
Anyone use Autotransformer as fixed attenuator?
Thank you all for your very helpful input in my consideration of my using a DAC as digital volume control and adding possible attenuation after the DAC and before a very high gain amp (or even modifying the amp input gain system). I learned a lot! Thank you.
All of the responses have included resistive devices. Anyone use an autotransformer as a rather expensive fixed attenuator?
Some swear by autransformers/autoformers as attenuators and claim they have less distortion affects on the line-level signal throughout the entire frequency spectrum than using resistive components. Is that a super controversial or debated assertion? -- thanks for any thoughts!
Thank you all for your very helpful input in my consideration of my using a DAC as digital volume control and adding possible attenuation after the DAC and before a very high gain amp (or even modifying the amp input gain system). I learned a lot! Thank you.
All of the responses have included resistive devices. Anyone use an autotransformer as a rather expensive fixed attenuator?
Some swear by autransformers/autoformers as attenuators and claim they have less distortion affects on the line-level signal throughout the entire frequency spectrum than using resistive components. Is that a super controversial or debated assertion? -- thanks for any thoughts!