We are discusing the power supply for an amplifier, certain batteries are designed for certain jobs (I,m no expert, but for the last few months I have been working on some interesting stuff that is relevant, and with people who understand batteries and how they are used to supply things). And if a power supply isn't up to the job then its not going to perform. Interestingly in medical, aerospace and mil designs, battery supplies are buffered by filter circuits and SMPS's. Have a look at the Vicor site:
Vicor Power | DC-DC Converter Modules, Configurable Power Supplies, and Systems
It is one of numerous companies that do modules for battery supplied equipement. These systems often involve some quite delicate analogue circuitry as well as digital and have to work at extremes of temperature and altitude.
I am also not a supporter of running laptops of batteries, even though buffered by a SMPS the instantaneous current requirements of the thousands of gates switching can be hundreds of amps, better IMO to have the mechanical momentum of the mains behind you to absorb that noise.
I work on PCB's for both totaly isolated battery supplied systems, and main powered systems, noise and EMC problems are almost the same for both, sometimes isolated systems are more probmatic to solve, where do you shunt the noise.
I believe a power supply should be the first design activity of any system (and in many areas it is, though some such as consumer electronics, cost prevails) if you want ultimate fidelity of your signal, be it digital or analogue. If the power supply is modulating you signal you have problems, and if you can hear it, the problems are of a larger magnitude.
Personaly I prefer the bitter sound of lemon juice, zinc and copper battery, it sharpens the sound, and the used electrolyte makes a mean martini.
Vicor Power | DC-DC Converter Modules, Configurable Power Supplies, and Systems
It is one of numerous companies that do modules for battery supplied equipement. These systems often involve some quite delicate analogue circuitry as well as digital and have to work at extremes of temperature and altitude.
I am also not a supporter of running laptops of batteries, even though buffered by a SMPS the instantaneous current requirements of the thousands of gates switching can be hundreds of amps, better IMO to have the mechanical momentum of the mains behind you to absorb that noise.
I work on PCB's for both totaly isolated battery supplied systems, and main powered systems, noise and EMC problems are almost the same for both, sometimes isolated systems are more probmatic to solve, where do you shunt the noise.
I believe a power supply should be the first design activity of any system (and in many areas it is, though some such as consumer electronics, cost prevails) if you want ultimate fidelity of your signal, be it digital or analogue. If the power supply is modulating you signal you have problems, and if you can hear it, the problems are of a larger magnitude.
Personaly I prefer the bitter sound of lemon juice, zinc and copper battery, it sharpens the sound, and the used electrolyte makes a mean martini.
When I was designing my battery powered dc-preamp, I did some impedance testing of batteries. I noticed their output impedance rising with the frequency of current draw, especially in the gel sealed-lead-acid batteries I used. Alkaline batteries had a flatter impedance over a wide frequency range, but the impedance was a lot higher in general (because if its lower capacity).
To measure the battery's impedance, I used a small audio amp (based on a TDA2040) running through a 10 ohm resistor, then coupled to the battery's positive terminal through a 22mF (milli, not micro) cap. As I swept the frequency into the amp, I watched the signal across the resistor and the battery terminals. There was virtually no signal across the battery up to almost 1KHz, but then the amplitude increased at about 6dB/octave. So I calculated that the battery measured in the low milli-ohm range up to about 1Khz increasing to about 250mOhm at about 50KHz.
Because of this, I added 12mF of capacitance, plus Tantalums and mono-ceramics on the supply rails to ensure a very low impedance across the entire audio range. I used an LME49720 for gain, and given that its PSRR is typically 120dB, I doubt I needed much filtering at all, besides the usual decoupling, for spectacular results.
When deciding on regulated AC power or batteries, I compared the two. What I found was that hands-down, the batteries had less noise and a lower impedance than any regulator, monolithic or even a discrete super-regulator. So I went with the batteries.
But are batteries audible? Will different battery chemistries sound different? In my opinion, based on my tests and experience, with a properly designed audio circuit, no. I don't believe you will ever hear any difference in an audio circuit powered with various batteries. Even my preamp, with its rail capacitance and textbook audio circuit, should perform perfectly whether powered by NiCd's, LiPo's, alkaline's, or the SLA's I ended up using.
Of course, some people on this and other forums claim to be able to hear some pretty amazing things that I've never been able to hear. Maybe some can hear how different battery chemistries affect audio in a chain of equipment. I'd love to see tests and measurements backing the claims.
Also, you may want to search this forum for batteries used in audio. I think all this has been discussed previously on this forum.
To measure the battery's impedance, I used a small audio amp (based on a TDA2040) running through a 10 ohm resistor, then coupled to the battery's positive terminal through a 22mF (milli, not micro) cap. As I swept the frequency into the amp, I watched the signal across the resistor and the battery terminals. There was virtually no signal across the battery up to almost 1KHz, but then the amplitude increased at about 6dB/octave. So I calculated that the battery measured in the low milli-ohm range up to about 1Khz increasing to about 250mOhm at about 50KHz.
Because of this, I added 12mF of capacitance, plus Tantalums and mono-ceramics on the supply rails to ensure a very low impedance across the entire audio range. I used an LME49720 for gain, and given that its PSRR is typically 120dB, I doubt I needed much filtering at all, besides the usual decoupling, for spectacular results.
When deciding on regulated AC power or batteries, I compared the two. What I found was that hands-down, the batteries had less noise and a lower impedance than any regulator, monolithic or even a discrete super-regulator. So I went with the batteries.
But are batteries audible? Will different battery chemistries sound different? In my opinion, based on my tests and experience, with a properly designed audio circuit, no. I don't believe you will ever hear any difference in an audio circuit powered with various batteries. Even my preamp, with its rail capacitance and textbook audio circuit, should perform perfectly whether powered by NiCd's, LiPo's, alkaline's, or the SLA's I ended up using.
Of course, some people on this and other forums claim to be able to hear some pretty amazing things that I've never been able to hear. Maybe some can hear how different battery chemistries affect audio in a chain of equipment. I'd love to see tests and measurements backing the claims.
Also, you may want to search this forum for batteries used in audio. I think all this has been discussed previously on this forum.
One thing I want to do when I get time (ha ha) is to measure the terminal voltage of a standard automobile lead-acid battery while it's driving a typical high-power car sub amplifier. These amplifiers have relatively small internal reservoir capacitors because they use high-frequency SMPSes, so the required input current closely matches the power output. It's pretty well acknowledged that such amplifiers cause car headlights to dim in response to the bass peaks, and adding a multi-farad capacitor at the amplifier power input greatly reduces this effect. Some people claim that this is because of the impedance of the power feed from the battery to the amp, but adding the capacitor at the battery end of the feed produces the same effect.
The implication is that lead-acid batteries take a significant amount of time to "ramp up" to an increased current demand. I'd measure this by using a switchable load and scoping the terminal voltage as the load is switched in / out. But before I re-invent the wheel, has anyone else done any work in this area?
With relevance to the subject at hand, dynamic characteristics of batteries should not affect the performance of a circuit designed with those characteristics in mind, or where the circuit topology is insensitive to such effects. For example, dynamic variation of voltage with load is not very relevant when driving class A preamplifier circuits where the current drawn from the supply is essentially constant.
The implication is that lead-acid batteries take a significant amount of time to "ramp up" to an increased current demand. I'd measure this by using a switchable load and scoping the terminal voltage as the load is switched in / out. But before I re-invent the wheel, has anyone else done any work in this area?
With relevance to the subject at hand, dynamic characteristics of batteries should not affect the performance of a circuit designed with those characteristics in mind, or where the circuit topology is insensitive to such effects. For example, dynamic variation of voltage with load is not very relevant when driving class A preamplifier circuits where the current drawn from the supply is essentially constant.
Batteries do have the advantage of isolating the mains and make managing ground loops a little easier.
Secondly, in providing DC, they have the advantage of not requiring rectification of AC to DC, itself a noisy process. Just think of the noise generated as a filter capacitor is charged by the rectifier circuit. In even line level equipment many amps can be drawn over a very short period, the spikes generated can propagate in the wiring and will radiate into the air and surrounding circuits. Maybe some chemically generated noise isn't so bad after all...
One can always filter and regulate after the batteries.
Secondly, in providing DC, they have the advantage of not requiring rectification of AC to DC, itself a noisy process. Just think of the noise generated as a filter capacitor is charged by the rectifier circuit. In even line level equipment many amps can be drawn over a very short period, the spikes generated can propagate in the wiring and will radiate into the air and surrounding circuits. Maybe some chemically generated noise isn't so bad after all...
One can always filter and regulate after the batteries.
Repeat of one battery noise study below. Sounds like Batteries are Extremely noisy.
http://tf.nist.gov/general/pdf/1133.pdf
http://tf.nist.gov/general/pdf/1133.pdf
I've compared and like LiFePo4 best
I've had a chance to do a lot of comparing: different batteries (LiFePo4, SLA, LiIon, vs various linear and smps power supplies, using 2050 based t-amps such as the HifiMeDiy, Sure, Arjen Helder and Virtue amps. T-amps seem to be VERY sensitive to power supply issues.
I checked to see if there was any correlation of perceived quality vs type of battery, or if it was more an issue of low ESR at DC and AC. In every instance the batteries with the highest instantaneous peak current were the ones that sounded best.
Low ESR across the audio band seems to be the key, and in that context the LiFePo4 batteries seem to pack the most punch, along with Li-Ion. But the LiFePo4 is whole lot safer to charge and has some other attributes such as you can discharge 'em almost completely with very little harm, which make them ideal for (safe) home audio applications.
The best implementation I have seen for the LiFePo4 battery packs is the Black Lightning from Vinnie Rossi at RedWine Audio. They have a really low ESR.
Like many audio cynics, I started with the opinion that "batteries are batteries" , but then compared various SLA's after I heard how the better SLA's handily surpassed even very expensive high current (is 20amps enough?) smps and linear power supplies. I was thinking that LiFePo4 would be in the same ball-park, and that's true for most of them, except for the ones that have ultra-low ESR cells.
When I compared I immediately recognized there is a significant difference, and it's in favor of the ultra-low ESR type of LiFePo4. (Yes, there are specialty SLA batteries with very low ESR and I had some really good ones, but the LiFePo4 was clearly better.)
As it turns out, the measurements support the subjective experience: the RedWine supplies had the lowest ESR of any batteries which I tested.
When it comes to t-amps, I would definitely vote for LiFePo4. YMMV with other types of amps, although I suspect good batteries will trump regular supplies there too: have a look at RedWine's new high power battery powered amps.
As of this writing, LiFePo4 RULES!! (imo)
Yesterday I found some impedance & phase measurements on SLA batteries:
http://www.telepower.com.au/INT95b.PDF
The interesting part (to me
So since e.g. LiFePO4 batteries may be more convenient to me (smaller, less polluting) it would be very interesting to me if someone had listened to the batteries.
Greetings, Jesper
I've had a chance to do a lot of comparing: different batteries (LiFePo4, SLA, LiIon, vs various linear and smps power supplies, using 2050 based t-amps such as the HifiMeDiy, Sure, Arjen Helder and Virtue amps. T-amps seem to be VERY sensitive to power supply issues.
I checked to see if there was any correlation of perceived quality vs type of battery, or if it was more an issue of low ESR at DC and AC. In every instance the batteries with the highest instantaneous peak current were the ones that sounded best.
Low ESR across the audio band seems to be the key, and in that context the LiFePo4 batteries seem to pack the most punch, along with Li-Ion. But the LiFePo4 is whole lot safer to charge and has some other attributes such as you can discharge 'em almost completely with very little harm, which make them ideal for (safe) home audio applications.
The best implementation I have seen for the LiFePo4 battery packs is the Black Lightning from Vinnie Rossi at RedWine Audio. They have a really low ESR.
Like many audio cynics, I started with the opinion that "batteries are batteries" , but then compared various SLA's after I heard how the better SLA's handily surpassed even very expensive high current (is 20amps enough?) smps and linear power supplies. I was thinking that LiFePo4 would be in the same ball-park, and that's true for most of them, except for the ones that have ultra-low ESR cells.
When I compared I immediately recognized there is a significant difference, and it's in favor of the ultra-low ESR type of LiFePo4. (Yes, there are specialty SLA batteries with very low ESR and I had some really good ones, but the LiFePo4 was clearly better.)
As it turns out, the measurements support the subjective experience: the RedWine supplies had the lowest ESR of any batteries which I tested.
When it comes to t-amps, I would definitely vote for LiFePo4. YMMV with other types of amps, although I suspect good batteries will trump regular supplies there too: have a look at RedWine's new high power battery powered amps.
As of this writing, LiFePo4 RULES!! (imo)
Last edited:
I agree. The conclusion clearly stated that batteries were substantially more quiet than regulators.
Which, is exactly the same conclusion I came to in my testing of batteries and regulators for my preamp design. Of course my test gear is orders of magnitude worse than what was used in that paper, but I was still able to detect the noise produced by regulators, and not able to detect the noise produced by batteries.
I also agree that LiPo batteries would probably be spectacular at having a low-Z over a wide bandwidth compared to SLA's. When I was doing my testing and design, however, LiPo's were still relatively new and rare.
Batteries -- top trace is a 9V, bottom a 12V SLA you'll have to use cross-correlation to measure the noise, but the impedance is pretty high:
Attachments
Originally Posted by Art M
Repeat of one battery noise study below. Sounds like Batteries are Extremely noisy.
I stated No conclusion.. I said "Sounds", clearly a vague idea, like Sounds good
Compared to a -400dbVrms reference level them batteries are garbage, is that better ?
Repeat of one battery noise study below. Sounds like Batteries are Extremely noisy.
I stated No conclusion.. I said "Sounds", clearly a vague idea, like Sounds good
Compared to a -400dbVrms reference level them batteries are garbage, is that better ?
Would you care to back that up with some data? I've measured Zout of the Jung regulator and it can dip under a micro-Ohm in the range of 10 to 32Hz.
Originally Posted by Art M
Repeat of one battery noise study below. Sounds like Batteries are Extremely noisy.
I stated No conclusion.. I said "Sounds", clearly a vague idea, like Sounds good
Compared to a -400dbVrms reference level them batteries are garbage, is that better ?
Look, I have no intention of starting an argument here or scoring points. My manhood isn't enlarged by being "right" on an internet forum
What seemed clear to me was that you stated your impression, however vague, that batteries were noisy, the article linked I thought was the evidence for that impression. My comment was for the benefit of those who hadn't read the article.
Stop repeating this. It is wrong.
ClassA circuits do not draw constant current.
There are a tiny proportion of ClassA circuits that can be designed to be essentially constant draw on the power supply, but these topologies are comparatively rare.
Last edited:
Hi again,
& thanks for replying in so many ways to my question ;-) I'll take the liberty of extracting some of your comments to focus on my initial question - if anyone had listened to the different battery types:
@Coppertop: Hmmm... inspiring metaphorical words and anticipations on the sound of different battery types - gives a different perspective ;-) ... can I ask if you have listened to the batteries or maybe applied your intuition ... or?
@Jack Caldwell: This is very interesting information, Jack! I was somewhat hoping that the LiFePO4 batteries would be a preferred battery because this type - to my knowledge - is also very environmentally friendly and has a long lifespan. So thanks for sharing your impressions and your observation that the sound (or being "best") seems to correlate well with the impedance of the battery across the audio band as well as peak instantaneous current capability.
Although I have not tested many batteries for impedances and peak current I once built a single-ended Class A amplifier where I mounted the batteries directly below the output transistor - hardwired to the transistor without any contacts or relays in between. The distance between the batteries and the transistor probably was about 10-15 centimeters (overall). The sound of this amplifier was second to nothing I'd ever heard when it comes to definition, speed, tautness, freedom, detail, openness, lucidity, timbre etc. - sheer presence. I also built similar amplifiers where the batteries were somewhat farther away - about 40 cms - and although still quite special they were less immediate even if the design was otherwise very similar. So my conclusion also has been that the - at the actual circuitry - impedance & bandwidth of a PSU is indeed very important.
Thanks for sharing Jack & all who replied!
Greetings,
Jesper
& thanks for replying in so many ways to my question ;-) I'll take the liberty of extracting some of your comments to focus on my initial question - if anyone had listened to the different battery types:
@Coppertop: Hmmm... inspiring metaphorical words and anticipations on the sound of different battery types - gives a different perspective ;-) ... can I ask if you have listened to the batteries or maybe applied your intuition ... or?
@Jack Caldwell: This is very interesting information, Jack! I was somewhat hoping that the LiFePO4 batteries would be a preferred battery because this type - to my knowledge - is also very environmentally friendly and has a long lifespan. So thanks for sharing your impressions and your observation that the sound (or being "best") seems to correlate well with the impedance of the battery across the audio band as well as peak instantaneous current capability.
Although I have not tested many batteries for impedances and peak current I once built a single-ended Class A amplifier where I mounted the batteries directly below the output transistor - hardwired to the transistor without any contacts or relays in between. The distance between the batteries and the transistor probably was about 10-15 centimeters (overall). The sound of this amplifier was second to nothing I'd ever heard when it comes to definition, speed, tautness, freedom, detail, openness, lucidity, timbre etc. - sheer presence. I also built similar amplifiers where the batteries were somewhat farther away - about 40 cms - and although still quite special they were less immediate even if the design was otherwise very similar. So my conclusion also has been that the - at the actual circuitry - impedance & bandwidth of a PSU is indeed very important.
Thanks for sharing Jack & all who replied!
Greetings,
Jesper
Last edited:
You had it wrong...
Very funny! Similar to the different "sounds" contributed by Coppertop earlier in the thread. Clearly, an advancement to the general knowledge base.
However, the reason you heard nothing was due to operator error! Here are your revised instructions in how to hear a battery directly:
In order to make it work, you have to connect leads to the battery, then get two small sponge balls and soak them in water with lemon juice. You then insert those into your ears and carefully insert the ends of the connected wires from the battery leads into the wet sponges, one in each ear.
Since you did not hear anything on your first try, you might decide to use a large battery like a 24V/10Amp type.... I guarantee you will not only hear the experience but you will definitely feel it as well!
It will sound and feel something like this:
ZAP !!!!!
You will only feel it for a short time. Killer sound, literally!
WARNING: The above "instructions" are a crude attempt at sarcastic humor, offered especially for anyone who would rather make fun of a useful thread. Do not try this AT ALL unless you want to literally wipe out your brain.
I plugged one into my ears and, I swear, I didn't hear anything
Very funny! Similar to the different "sounds" contributed by Coppertop earlier in the thread. Clearly, an advancement to the general knowledge base.
However, the reason you heard nothing was due to operator error! Here are your revised instructions in how to hear a battery directly:
In order to make it work, you have to connect leads to the battery, then get two small sponge balls and soak them in water with lemon juice. You then insert those into your ears and carefully insert the ends of the connected wires from the battery leads into the wet sponges, one in each ear.
Since you did not hear anything on your first try, you might decide to use a large battery like a 24V/10Amp type.... I guarantee you will not only hear the experience but you will definitely feel it as well!
It will sound and feel something like this:
ZAP !!!!!
You will only feel it for a short time. Killer sound, literally!
WARNING: The above "instructions" are a crude attempt at sarcastic humor, offered especially for anyone who would rather make fun of a useful thread. Do not try this AT ALL unless you want to literally wipe out your brain.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.