Line preamps using battery power supply _ any experience ?

[gino1961]

Many years ago I built for a friend a Hiraga MC pre amp with self set turret boards. External power supply with line filter and + an - regulators. On the pre amp board additoinally caps for further noise reduction. Result was no hum and no additional noise from the power chain. Therefore my opinion is that battery power is not really necessary. But it may help to avoid ground loop.View attachment 366381
View attachment 366382

Very nice indeed Congratulations
I think I have to Start with measurements on power supplies off the shelf
Just to see what noise means
600 USD are a little too much for an audio analyzer but maybe it's a mandatory step
Then when I have seen the effects of different filters on commercial supplies I could move on
I believe that noise is even worse than a little 2nd order distortion
Noise is evil
Thanks a lot again

 

[SSS]

Very nice indeed Congratulations
I think I have to Start with measurements on power supplies off the shelf
Just to see what noise means
600 USD are a little too much for an audio analyzer but maybe it's a mandatory step
Then when I have seen the effects of different filters on commercial supplies I could move on
I believe that noise is even worse than a little 2nd order distortion
Noise is evil
Thanks a lot again

Thank you. To view at noise the soundcard of a PC may be good enough in the first place. Using one of the FFT PC programs which are some for free and others for little money. An oscilloscope with good sensititvity will also help and will show noise beyond the 20 kHz. Myself I use the E-MU Tracker Pre which has a reasonable low noise and did not cost a fortune when getting a used one. Maybe there are no for sale now. But other boxes are available. Oh, I forget. I use also my trained ears when noise was amplified and listened by a speaker or headphone. Easily you may hear the differences and the sound frequency shape.

 

[levimax]

One of the best sources for DIY schematics and PCB is Elliott Sound Products. I have built this pre-amp and it has worked perfectly for many years. https://sound-au.com/project88.htm

With pre-amps the circuits are the easy part, the mechanical layout and wire routing etc are the hard part and they will determine the final noise and crosstalk performance.

Regarding +/- supplies and voltage if you look at the spec sheets on the standard audio op amps you will see that higher rail voltage gives better noise and distortion performance than lower voltages, it is subtle but there. They all quote specs bases on +/- 15 volts... because that is where they perform the best.

 

[antcollinet]

Hi thank you for the very helpful advice
Then I miss the reason for dual power supply
I can understand if they want to avoid the output cap
But many circuits with dual voltage have still the output cap
Why not using a single voltage?

Dual power supply is normally used because audio is AC, and goes positive and negative. So you need a negative supply so the output of the opamps can go negative.

An alternative if you only have a single supply is to create a voltage at half the supply voltage, which you then use as the 0V reference from the point of view of the circuit.

In this case it is normally necessary to capacitively couple the incoming audio so that it can be level shifted from +/- compared to actual 0V to +/- compared with the pseudo 0V (at half supply voltage)

This though makes things generally more complicated, so it is often easier to get a dual rail supply if at all possible.

 

[gino1961]

Thank you. To view at noise the soundcard of a PC may be good enough in the first place. Using one of the FFT PC programs which are some for free and others for little money. An oscilloscope with good sensititvity will also help and will show noise beyond the 20 kHz. Myself I use the E-MU Tracker Pre which has a reasonable low noise and did not cost a fortune when getting a used one. Maybe there are no for sale now. But other boxes are available. Oh, I forget. I use also my trained ears when noise was amplified and listened by a speaker or headphone. Easily you may hear the differences and the sound frequency shape.

Thank you so much again
I would like to test the preamp standalone to be sure that it is not the noise generator
I think i will end buying one of the cheapest usb unit from Picoscope
I was even thinking to find a mic preamp kit using a battery to use at the preamp output to amplify the preamp noise considering that cheap scopes will have not very high sensitivity A mic preamp can have also a 50dB gain It is a lot

 

[gino1961]

One of the best sources for DIY schematics and PCB is Elliott Sound Products. I have built this pre-amp and it has worked perfectly for many years. https://sound-au.com/project88.htm

With pre-amps the circuits are the easy part, the mechanical layout and wire routing etc are the hard part and they will determine the final noise and crosstalk performance.

Regarding +/- supplies and voltage if you look at the spec sheets on the standard audio op amps you will see that higher rail voltage gives better noise and distortion performance than lower voltages, it is subtle but there. They all quote specs bases on +/- 15 volts... because that is where they perform the best.

thanks a lot again I have come to this wonderful site because i have noticed a great respect for measurements
I am about to buy some instruments to perform noise measurements on units i have around
I guess that i could get an idea of THD from LTSpice But noise i have to measure it Only then i could start trying to put something together

 

[gino1961]

Dual power supply is normally used because audio is AC, and goes positive and negative. So you need a negative supply so the output of the opamps can go negative.
An alternative if you only have a single supply is to create a voltage at half the supply voltage, which you then use as the 0V reference from the point of view of the circuit.
In this case it is normally necessary to capacitively couple the incoming audio so that it can be level shifted from +/- compared to actual 0V to +/- compared with the pseudo 0V (at half supply voltage)
This though makes things generally more complicated, so it is often easier to get a dual rail supply if at all possible.

thanks a lot for the valuable advice Actually i see single supplies as simpler with their just one rail regulation
but i would like to start with noise measurements expecially on power supplies This is very important to me
Only then i will move on with more demanding activities I think that i could get a good idea of the THD only with circuit simulation
I have already tried and got very promising result like a 2nd harmonic below 90 dB with a 1kHz sine

 

[levimax]

thanks a lot again I have come to this wonderful site because i have noticed a great respect for measurements
I am about to buy some instruments to perform noise measurements on units i have around
I guess that i could get an idea of THD from LTSpice But noise i have to measure it Only then i could start trying to put something together

If noise is really your number one preamp issue (rather than the fun of DIY building and experimenting and learning) it is impossible to beat "no preamp" i.e. just go direct from DAC to power amp. I go back and forth on this and currently I have a MOTU UltraLite MK5 in a DIY tri-amp + subs system. The only thing I am missing with this is "manual tone controls" but this of course could be accomplished digitally. For my TT the ADC works great and I apply RIAA digitally. As far as noise goes it is just plain dead quiet. With my previous DIY tri-amp system with analog active crossovers, subs, and preamp all connected unbalanced the sources of noise seemed endless, now the noise is just gone.

 

[gino1961]

If noise is really your number one preamp issue (rather than the fun of DIY building and experimenting and learning) it is impossible to beat "no preamp" i.e. just go direct from DAC to power amp.

yes but i like the idea of having a volume control in between Almost any rigs have a preamp If i had to put together a system i would start from it
Dacs are quite ok noise wise Never had big issues with dacs Even some cheap one sounds ok

I go back and forth on this and currently I have a MOTU UltraLite MK5 in a DIY tri-amp + subs system. The only thing I am missing with this is "manual tone controls" but this of course could be accomplished digitally. For my TT the ADC works great and I apply RIAA digitally. As far as noise goes it is just plain dead quiet. With my previous DIY tri-amp system with analog active crossovers, subs, and preamp all connected unbalanced the sources of noise seemed endless, now the noise is just gone.

i see I hear a very nice explanation of background noise It is like to have carpets with longer or shorter piles
If you drop little objects long piles carpet you will not see them On a short piles carpet you can spot them
The small sonic details will be missing with higher noise floor These are the details that provide soundstage and ambience
I can hear more thing
Now which would be the number one source of noise ? power supply of course
Some brands provide upgraded power supplies for their units like Naim Audio Or power supply capacitance upgrade
When i read about a uV noise from that battery pack some posts above i was amazed

 

[antcollinet]

I think that i could get a good idea of the THD only with circuit simulation

Unless your simulator can simulate the effect of PCB layout then it may not be telling you much about what the noise level is.

 

[MaxwellsEq]

I think I have to Start with measurements on power supplies off the shelf
Just to see what noise means

Research PSRR.

If you design a gain stage which has good PSRR then the PSU does not need to have perfect noise performance. Incidentally, this is the way everyone goes...

 

[Jaxjax]

Done it with 26dht & 27 idht preamps. Series up a bunch of my bigger Dewalt batteries for B+ & did & used a lab supply for my heaters. Also had battery bias as well. 26 is the best of the 2 but tricky to keep quiet. It was just a quick & dirty experiment as I already had separate supplies I built for everything in completely over kill mode...
It was definitely quiet on battery just never pursued it with the high draw heater issue..I'm sure there something out there nowadays that would work just fine.

 

[gino1961]

Unless your simulator can simulate the effect of PCB layout then it may not be telling you much about what the noise level is.

Good morning and thank you for your kind and valuable response
obviously, as you rightly say, the design and construction of an audio device is a very complex operation
I absolutely don't have the skills needed to do something from scratch
so what do I do? what other incompetent people do I try to copy others
I'm really obsessed with pre-amps which I consider the command center of an entire system
if I had to splurge for an element of an audio chain I would do it for an excellent line preamp
Having said this and speaking of projects, I really like the minimalist approach and that is extracting the maximum in terms of performance from very simple circuits.
I started by downloading service manuals from top preamps from the 70s which they say sound good
Their circuits, being commercial products, have a fundamental quality. They work more or less well but they absolutely work
Usually the line stage is made up of three blocks, initial stage - tone controls - output stage usually called often flat amp (usually a buffer with little of no gain)
In short, it's about taking the flat amp, putting a potentiometer in front of it and you have a line pre-amp Done
Now the question is also important.
Can we increase the already decent performance by maintaining the same basic scheme but changing working point components?
the answer I give myself is probably yes
How can the effects of changing a component be verified in a more practical way?
with the help of simulation software
Now if you ask me how close the distortion calculated with LTSpice is to the measured one, I have no idea
It is clear that the final judgment can be given only by a good spectrum analyzer that remains in my dreams given the prices
I am seriously considering purchasing an entry level model by Picoscope
When a friend of mine told him this, he recommended me a Psychoscope instead
Yes, for me audio is a disease. If I were good at mathematics I would have done it for a living

 

[gino1961]

Research PSRR.
If you design a gain stage which has good PSRR then the PSU does not need to have perfect noise performance. Incidentally, this is the way everyone goes...

Thank you and yes PSRR is very important I guess this is a reason for more complex topologies ? a higher PSRR ?
A simpler circuit could make the power supply design more challenging
But i would prefer to keep the circuit simple and work more on the power supply like in the Bride of Zen design that should have a low PSRR
However the circuit i have in mind uses feedback and this should increase the PSRR if i understand well

 

[gino1961]

Done it with 26dht & 27 idht preamps. Series up a bunch of my bigger Dewalt batteries for B+ & did & used a lab supply for my heaters. Also had battery bias as well. 26 is the best of the 2 but tricky to keep quiet. It was just a quick & dirty experiment as I already had separate supplies I built for everything in completely over kill mode...
It was definitely quiet on battery just never pursued it with the high draw heater issue..I'm sure there something out there nowadays that would work just fine.

Hi thank you for your kind advice I dont know these projects I guess they use tubes ?
maybe tubes can provide a more musical sound but they are too complex for me
i think that decent sound can be obtained with bjts or fets

 

[MCH]

To the OP,

Not exactly what you are looking for but in case it helps. Years ago I had a Trigon Vulcano II battery power supply for my matching Trigon phono stage, the later I still use. I sold the battery because I could not hear any difference and it had no other benefit for my use case.

It had two lead batteries and the functionality was quite good: you could chose battery or mains pressing a button, it would do its charging when needed… There are pictures online of the insides, maybe you can figure out something. I think it is MCU operated.

 

[StigErik]

I used to have a Kaneda preamp. Schematic below. I powered it with six 12V lead-acid batteries decoupled with some large capacitors to bring the self-noise of the batteries down. The batteries were trickle charged, but the charger could be disconnected if desired. It's been a long time since I had it, but I do remember it sounded really good. Distortion was very low, around -100 dB for the line stage, and mainly 2.order.

 

[MaxwellsEq]

Thank you and yes PSRR is very important I guess this is a reason for more complex topologies ? a higher PSRR ?
A simpler circuit could make the power supply design more challenging
But i would prefer to keep the circuit simple and work more on the power supply like in the Bride of Zen design that should have a low PSRR
However the circuit i have in mind uses feedback and this should increase the PSRR if i understand well

Every topology has a different set of strengths and weaknesses. When you study these in college (I had to do this before software simulators were available), you get an intrinsic understanding of the constraints of simple designs. Basically, all simple designs are flawed for real-world use. So all circuits necessarily have to be complex to support real-world use, including: sufficient gain; stability in all possible end-user scenarios; noise performance (when warm, when cold); low frequency performance; high frequency performance; sufficient power handling; stable for all normal thermal conditions; imperturbability from external environment; sufficient working lifetime.

 

[gino1961]

To the OP,

Not exactly what you are looking for but in case it helps. Years ago I had a Trigon Vulcano II battery power supply for my matching Trigon phono stage, the later I still use. I sold the battery because I could not hear any difference and it had no other benefit for my use case.

It had two lead batteries and the functionality was quite good: you could chose battery or mains pressing a button, it would do its charging when needed… There are pictures online of the insides, maybe you can figure out something. I think it is MCU operated.

View attachment 366590

Hi thanks a lot Nice unit indeed
You mean that there weren't benefits switching to battery power?
Good I abandon the idea
I like something more practical for sure
Life is already too complex

 

[gino1961]

I used to have a Kaneda preamp. Schematic below. I powered it with six 12V lead-acid batteries decoupled with some large capacitors to bring the self-noise of the batteries down. The batteries were trickle charged, but the charger could be disconnected if desired. It's been a long time since I had it, but I do remember it sounded really good. Distortion was very low, around -100 dB for the line stage, and mainly 2.order.

Hi thank you for the very interesting reply
I guess the line stage is the circuit on the right? Nice but I have a feeling that similar thd can be obtained also with more basic circuits
That is my dream goal