Q: What about hardware causes hiss?

[SoupRKnowva]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Oh no. I leave that to the folks who have the knowledge, experience and equipment for that. For me that's CineMag. No. Just the opposite in fact. Interstage and output transformers in tube amps are step-down. Their primary function is impedance matching, but a step-down transformer also reduces the signal by the turns ratio. Thank you for the kind words. I'm glad you've found some of my posts useful. se

maybe im just confused

but do you have a headphone amp that uses a transformer for voltage gain? cuase you said that the only commercial one you know of is the SWGPA and you dont make any? or is cine mag another DIY'er?

 

[Steve Eddy]

Quote:

Originally Posted by SoupRKnowva /img/forum/go_quote.gif maybe im just confused but do you have a headphone amp that uses a transformer for voltage gain? cuase you said that the only commercial one you know of is the SWGPA and you dont make any? or is cine mag another DIY'er?

Sorry for the confusion.

You'd asked me if I made my own transformers, which I took to mean just the transformers themselves, not an amp that used transformers. CineMag is a manufacturer of transformers, not amps that use transformers for voltage gain.

My personal amplifier (for loudspeakers, not headphones) uses transformers for voltage gain, and I'm working on a headphone amp that uses transformers for voltage gain that will be available for commercial sale, but that won't happen until later this year, making the SWGPA the only commercial amplifier that I'm aware of that uses transformers for voltage gain.

se

 

[SoupRKnowva]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Sorry for the confusion. You'd asked me if I made my own transformers, which I took to mean just the transformers themselves, not an amp that used transformers. CineMag is a manufacturer of transformers, not amps that use transformers for voltage gain. My personal amplifier (for loudspeakers, not headphones) uses transformers for voltage gain, and I'm working on a headphone amp that uses transformers for voltage gain that will be available for commercial sale, but that won't happen until later this year, making the SWGPA the only commercial amplifier that I'm aware of that uses transformers for voltage gain. se

ohh got ya, yeah i was asking if you made your own amps using transformers for the voltage gain. Guess we both got a little confused on that one

this seems like it would be kinda of a no brainer though i would think, do you know why more companies dont use transformers for voltage gain?

 

[Steve Eddy]

Quote:

Originally Posted by SoupRKnowva /img/forum/go_quote.gif ohh got ya, yeah i was asking if you made your own amps using transformers for the voltage gain. Guess we both got a little confused on that one this seems like it would be kinda of a no brainer though i would think, do you know why more companies dont use transformers for voltage gain?

Well, for starters, high quality transformers aren't exactly cheap. You can build an active circuit for voltage gain for a fraction of the cost, save perhaps for certain tube-based circuits.

Also, while they give a lot, they also place certain demands on the source component they're used with. For example, the source's output impedance needs to be fairly low (around 200 ohms or less, which rules out some tube-based output stages), and the source needs to be able to drive a rather lower load impedance than is typical (around 2k ohms). Also, the maximum output level from the source can't be too high (around 2 volts or so).

But when used within those constraints, there's nothing better in my opinion.

Yeah, you can get better raw numbers even with a cheap opamp circuit, but I'm not a bean counter and go with what sounds best to me.

se

 

[SoupRKnowva]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Well, for starters, high quality transformers aren't exactly cheap. You can build an active circuit for voltage gain for a fraction of the cost, save perhaps for certain tube-based circuits. Also, while they give a lot, they also place certain demands on the source component they're used with. For example, the source's output impedance needs to be fairly low (around 200 ohms or less, which rules out some tube-based output stages), and the source needs to be able to drive a rather lower load impedance than is typical (around 2k ohms). Also, the maximum output level from the source can't be too high (around 2 volts or so). But when used within those constraints, there's nothing better in my opinion. Yeah, you can get better raw numbers even with a cheap opamp circuit, but I'm not a bean counter and go with what sounds best to me. se

so by source you mean the DAC or cd player that youd be using correct? and in order to be able to drive a input impedance that low, wouldnt you just have to beef up the power supply of the dac? but getting the output voltage to stay below 2 volts would be tough though right? most get up to 2.5 or 3 volts for peaks. are you making your own dac to use with the amp too?

maybe im asking for too much here, but why are those constraints necessary? why does the voltage have to stay so low, but the source have to supply so much current? isnt that what the power supply in the amp is for?

sorry for my ignorance steve, im just a interested mind

 

[Steve Eddy]

Quote:

Originally Posted by SoupRKnowva /img/forum/go_quote.gif so by source you mean the DAC or cd player that youd be using correct?

Correct.

Quote:

and in order to be able to drive a input impedance that low, wouldnt you just have to beef up the power supply of the dac?

Nah. Most any DAC should be able to drive a 2k load without any problems. You'd only have problems with certain asthmatic tube output stages that have output impedances on the order of several thousand ohms or more.

2 volts into a 2k ohm load means that the source only has to deliver 1 milliamp of current.

Quote:

but getting the output voltage to stay below 2 volts would be tough though right? most get up to 2.5 or 3 volts for peaks.

Well, the 2 volts isn't exactly a hard limit. And the limit will ultimately depend on the specific transformer. Some transformers are designed to take higher levels, some lower levels. I just picked 2 volts as an approximation.

Quote:

are you making your own dac to use with the amp too?

Don't have any plans to, no.

Quote:

maybe im asking for too much here, but why are those constraints necessary? why does the voltage have to stay so low, but the source have to supply so much current?

Transformers, namely line level input and output transformers and microphone transformers, are rated for input levels at the point at which THD reaches 1% at 20Hz. The CineMag transformers that I've used and will be using are rated for around 2 volts. Though with the headphone amplifier I'll be using a pair of them per channel wired in series so they'll be able to handle upwards of 4 volts.

Quote:

sorry for my ignorance steve, im just a interested mind

Please never say you're sorry for your ignorance. It's not like any of us come out of the womb knowing what we know. You're curious and asking questions. That's a good thing, not something to be sorry about.

se

 

[SoupRKnowva]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Correct. Nah. Most any DAC should be able to drive a 2k load without any problems. You'd only have problems with certain asthmatic tube output stages that have output impedances on the order of several thousand ohms or more. 2 volts into a 2k ohm load means that the source only has to deliver 1 milliamp of current.

oh wow, i jsut assumed that since most amps have input impedances in the multiple 10's of thousands of ohms that one with only 2k ohms would have require tons of current lol didnt even think to actually do the math and figure out it wouldnt need that much afterall, just alot in comparison.

Quote:

Transformers, namely line level input and output transformers and microphone transformers, are rated for input levels at the point at which THD reaches 1% at 20Hz. The CineMag transformers that I've used and will be using are rated for around 2 volts. Though with the headphone amplifier I'll be using a pair of them per channel wired in series so they'll be able to handle upwards of 4 volts.

this paragraph is where i get confused. So does distortion go up with frequency? or were you just stating that most transformers have a THD of 1% at 20Hz at 2 volts so you would want to stay lower than that? what causes the distortion to go up with voltage? is it the heating of the wire based on the natural resistance that causes it?

and if thats the case could you use higher voltages by winding your own transformers using silver wire to lower total resistance?

Quote:

Please never say you're sorry for your ignorance. It's not like any of us come out of the womb knowing what we know. You're curious and asking questions. That's a good thing, not something to be sorry about. se

ok

i just thought i might have been asking too many questions, dont want to take up too much of your time, but im happy for what you're willing to donate.

 

[Steve Eddy]

Quote:

Originally Posted by SoupRKnowva /img/forum/go_quote.gif oh wow, i jsut assumed that since most amps have input impedances in the multiple 10's of thousands of ohms that one with only 2k ohms would have require tons of current lol didnt even think to actually do the math and figure out it wouldnt need that much afterall, just alot in comparison.

Yup.

Quote:

this paragraph is where i get confused. So does distortion go up with frequency? or were you just stating that most transformers have a THD of 1% at 20Hz at 2 volts so you would want to stay lower than that?

Actually distortion goes down with frequency.

Here's a plot of the typical distortion behavior of a transformer at fixed levels versus frequency:

Quote:

what causes the distortion to go up with voltage? is it the heating of the wire based on the natural resistance that causes it?

No, it's the saturation of the core material that causes it. Transformers rely on the magnetization of the core material, which can only be magnetized to a certain point, after which it saturates and can't be magnetized any further.

This is illustrated by the B-H curve for magnetic materials.

In the middle, the lines are relatively straight, which means the magnetization is rather linear. But as you get closer to the top and bottom, the lines start curving until they eventually flatten out at the saturation point. Where the lines start curving magnetization becomes less and less linear which produces more and more distortion.

It's not unlike clipping in an amplifier. As you drive the input with higher and higher levels, the amplifier's output swings closer and closer to the power supply voltage, until the point at which the input tries to drive the output beyond the point that it can swing and the signal gets clipped.

Quote:

ok i just thought i might have been asking too many questions, dont want to take up too much of your time, but im happy for what you're willing to donate.

Don't worry about it. There were (and are) those in my life who have taken the time to answer the questions I've had (and have) in order to help me better understand things and to those people I am quite grateful. It would be rather ungrateful of me to turn around and be stingy with my time when it came to answering questions asked of me by others.

Edit: By the way, if you want to know more about audio transformers, I recommend giving chapter 13 of Glen Ballou's book, Handbook for Sound Engineers, written by Bill Whitlock of Jensen Transformers.

Here's the pdf version of that chapter:

Audio Transformers

se

 

[SoupRKnowva]

oh got ya. So how does one deal with the distortion present at the lower frequencies in the transformer? and couldnt you just use a larger core in the transformers to be able to use a larger input voltage so it would take longer to reach saturation?

I think im starting to realize that i may want to get a degree in EE, starting to get interested in the actual design of amps. Started looking into DIY but most of it was using others designs, id rather do it myself

My limited knowledge in electronics is from two years of high school physics, and then after joining the air force two months of electronics principles. Being in the air force makes it easy to get a degree like this to use purely as a hobby.

Thanks for that link, ill be sure to check that out

 

[Steve Eddy]

Quote:

Originally Posted by SoupRKnowva /img/forum/go_quote.gif oh got ya. So how does one deal with the distortion present at the lower frequencies in the transformer? and couldnt you just use a larger core in the transformers to be able to use a larger input voltage so it would take longer to reach saturation?

The simple answer is "yes."

But the art of designing high quality transformers involves the art of compromise.

Yes, you can use a larger core in order to allow larger input voltage, but you're going to pay for it somewhere else. The goal is to achieve the best balance among all the various tradeoffs.

I could have CineMag wind me some trannies on larger cores, but I'm already quite satisfied with their current performance. And as you'll read in the Whitlock piece, raw THD figures for transformers aren't as meaningful as they are for active circuits.

Quote:

I think im starting to realize that i may want to get a degree in EE, starting to get interested in the actual design of amps. Started looking into DIY but most of it was using others designs, id rather do it myself My limited knowledge in electronics is from two years of high school physics, and then after joining the air force two months of electronics principles. Being in the air force makes it easy to get a degree like this to use purely as a hobby.

Coolness! I wish you the best!

se