Voltage Inverter
A negative 5V can easily be achieved from USB 5V with some 50mV noise by using MAX871 / MAX 870.
This can be avoided. Will post more.
Looks like PCM2902 takes AC signal and works the signal out internally to be able to process the signal with 5V and ground single polarity supply. The maximal amplitude of the signal is to be Vcom = Vccci / 2. Thus, when a 3.3V reference is used, the maximal amplitude is 1.65V and there is another 1.65V above the 3.3 reference voltage to 4.95V.
I have done some schematics which avoid the use of MAX871. Will post in a while, when drawn.
A negative 5V can easily be achieved from USB 5V with some 50mV noise by using MAX871 / MAX 870.
This can be avoided. Will post more.
Looks like PCM2902 takes AC signal and works the signal out internally to be able to process the signal with 5V and ground single polarity supply. The maximal amplitude of the signal is to be Vcom = Vccci / 2. Thus, when a 3.3V reference is used, the maximal amplitude is 1.65V and there is another 1.65V above the 3.3 reference voltage to 4.95V.
I have done some schematics which avoid the use of MAX871. Will post in a while, when drawn.
No, I don't see all those advantages. In the studio, sure I convert to digital as soon as possible, and stay there as long as possible; on outside gigs, reinforcing or recording, the idea that a power spike can leave my console rebooting for over a minute of silence with the musicians and audience all there is terrifying.
If you say "no cables" to a performer he won't think "No loss of quality due to a high impedance guitar pickup driving 25 metres of coax" he'll think radio mics, no clutter on stage, freedom to dance without wires twisting into a group-wide cocoon. Digital radio mics exist, give lots of channels and show a much diminished tendency to go chasing ambulances, but they're not cheap, and when signal drops below acceptable levels they don't go noisy, they drop out completely.
Actually, I don't see any reasonable quality A-D converters arriving, because the chips themselves cost more than your 'dollar or two' that you think suitable. 24 bit smooth conversion means laser trimming each individual chip (admittedly done by robots, but taking time nonetheless) and no, the market for them is not big. The majority of computer users don't care about absolute quality in the sound chain; as long as it's recognisable, and doesn't fizz or square off, it can be as bad as the sound off cable TV. Musicians are expected to be a little more discerning; this is supposed to be their medium, after all.
And just push the 'distortion' button? I've had guitarists agonise over their signal going through an impedance converter, because it was solid state so the pickups weren't driving direct into the grid of the first AX7 in their Marshal. Oh, yes, it changed the sound; it rendered the cable inaudible, so the high frequencies weren't all being absorbed, then put back (with added hiss) by the amp's tone control.
Yeah, and an awful lot of that interference (lighting buzz, mobile phones, police radio) is collected on the electromagnetic pickups, which can't be adequately screened, as they're supposed to be taking in the signal from the strings. Try and persuade a guitar player these need to be changed.
Digital has programmability on its side, and low noise mix busses. It's mastery of delay has made reverb and echo much more manageable than in my youth, but for live a few specific effects on inserts or auxiliary sends cover most of my needs. But for reliability, reparability, general sound quality, and even (though only just, nowadays) price to performance I'll stick to analogue, and gear which is designed for rough handling. I've seen too many concerts rely on PCs that give up during the performance, leaving the drummer the only one undamaged.
If you say "no cables" to a performer he won't think "No loss of quality due to a high impedance guitar pickup driving 25 metres of coax" he'll think radio mics, no clutter on stage, freedom to dance without wires twisting into a group-wide cocoon. Digital radio mics exist, give lots of channels and show a much diminished tendency to go chasing ambulances, but they're not cheap, and when signal drops below acceptable levels they don't go noisy, they drop out completely.
Actually, I don't see any reasonable quality A-D converters arriving, because the chips themselves cost more than your 'dollar or two' that you think suitable. 24 bit smooth conversion means laser trimming each individual chip (admittedly done by robots, but taking time nonetheless) and no, the market for them is not big. The majority of computer users don't care about absolute quality in the sound chain; as long as it's recognisable, and doesn't fizz or square off, it can be as bad as the sound off cable TV. Musicians are expected to be a little more discerning; this is supposed to be their medium, after all.
And just push the 'distortion' button? I've had guitarists agonise over their signal going through an impedance converter, because it was solid state so the pickups weren't driving direct into the grid of the first AX7 in their Marshal. Oh, yes, it changed the sound; it rendered the cable inaudible, so the high frequencies weren't all being absorbed, then put back (with added hiss) by the amp's tone control.
Yeah, and an awful lot of that interference (lighting buzz, mobile phones, police radio) is collected on the electromagnetic pickups, which can't be adequately screened, as they're supposed to be taking in the signal from the strings. Try and persuade a guitar player these need to be changed.
Digital has programmability on its side, and low noise mix busses. It's mastery of delay has made reverb and echo much more manageable than in my youth, but for live a few specific effects on inserts or auxiliary sends cover most of my needs. But for reliability, reparability, general sound quality, and even (though only just, nowadays) price to performance I'll stick to analogue, and gear which is designed for rough handling. I've seen too many concerts rely on PCs that give up during the performance, leaving the drummer the only one undamaged.
True, the huge input impedance will be more noise prone. When the digitisation is near the source, there shouldn't be so much of a problem. Input resistance can be introduced and the user can decide what the input impedance should be.
I drive a huge impedance with a 10 feet microphone coaxial cable from the microphone in a very noisy environment. No noise FOR THE PURPOSE. THE PURPOSE IS NOT VERY PROFESSIONAL. I also drive the same impedance with a literally $1 microphone literally from the $1 shop. Not bad for the purpose. I drove this impedance from a guitar. Also, not bad for the purpose.
24 bit ADC may be expensive now but may not be in the near or far future. Unlikely for the manufacturers to do what Intel does with processors. I would prefer 16 bit sound even 12 or 10 than analogue, though.
Computer users care what quality sound they have unless they talk on Skype where they don't. Most people these days would use PC only as their main source of sound as opposed to some years ago when the PC sound was mainly useful for porn movies.
Most of the PC users can find the difference between an MP3 and a WAV. I am not very musical and do not have any musical hearing and EVEN I can hear the difference.
PC's shouldn't give up but they still may do. This won't be for long.
Level sensitivity should be possible to turn off or adjust the level.
Laser trimming is a nasty thing, I agree. This could be expensive for a while. You have a very strong point.
I drive a huge impedance with a 10 feet microphone coaxial cable from the microphone in a very noisy environment. No noise FOR THE PURPOSE. THE PURPOSE IS NOT VERY PROFESSIONAL. I also drive the same impedance with a literally $1 microphone literally from the $1 shop. Not bad for the purpose. I drove this impedance from a guitar. Also, not bad for the purpose.
24 bit ADC may be expensive now but may not be in the near or far future. Unlikely for the manufacturers to do what Intel does with processors. I would prefer 16 bit sound even 12 or 10 than analogue, though.
Computer users care what quality sound they have unless they talk on Skype where they don't. Most people these days would use PC only as their main source of sound as opposed to some years ago when the PC sound was mainly useful for porn movies.
Most of the PC users can find the difference between an MP3 and a WAV. I am not very musical and do not have any musical hearing and EVEN I can hear the difference.
PC's shouldn't give up but they still may do. This won't be for long.
Level sensitivity should be possible to turn off or adjust the level.
Laser trimming is a nasty thing, I agree. This could be expensive for a while. You have a very strong point.
Reliability of PC and software for live playing is definitely an issue.. it seems to be getting better but I have not been 100% comfortable with full out digital for live either. I could do it right now, but think that I would implement some backup redundancies to my rig. I have played using dedicated POD by Line 6 which is all digital hardware, just not PC based. Has not broken down once yet. I have had cables break, PA speakers fry and mains breakers go out at shows..
So if I go with a PC for live, it will probably have a bypass switch on my pedalboard. If the PC locks up at least I can bypass to my regular analog effects with POD setup.
So if I go with a PC for live, it will probably have a bypass switch on my pedalboard. If the PC locks up at least I can bypass to my regular analog effects with POD setup.
Single Power
PCM2902 seems to be able to get two type of signals at the VinL / R inputs: an AC without an offset and a maximum amplitude of Vcom or the same but offset by Vcom to swing from 0 to Vccci = 2 Vcom. A good idea may be to have a reference of Vccci = 3.3V, thus Vcom = 1.65V. A good idea is to buffer this with a 5V and Analogue Ground supplied buffer.
There many combinations possible inclusive of these which give a voltage which spans from 0.5Vcom to 2.5Vcom. Most likely, TI would use the same resistors in the two of the schematics thus making Vcom – ui and Vcom + ui when ui is offset by Vcom and 2Vcom - ui and ui when not offset.
In case they would subtract Vcom, they would receive Vcom – ui and ui – Vcom which would be 0 in single polar works. Then Vcom – ui can be digitised in the 0 to Vccci range.
Also possible is to work out the resistors and processing and digitise the negative way separately from the positive thus allowing a Vccci maximum amplitude ( 3.3V ) and increase the precision.
The 4.7uF capacitor can be replaced with a 330nF, > = 5V capacitor which cuts off at 16Hz with the 30K resistor.
The first amplifier at the analogue inputs of PCM2902 has a capacitor in parallel which, hopefully, cuts off at 25KHz.
Before the PCM2902, many possibilities exist for a single power supply using heavily Vcom. These can be combined in many combinations, variations and permutations.
Also, a voltage inverter can be used to make – 5V from USB Vbus +5V.
Although a bit more noisy, a good idea would be to introduce high power negative supplies in standard PC power supplies, so USB can provide + - 12V or, at least, + - 5V.
On one hand, a voltage inverter may introduce a bit more noise ( in very high frequencies above 500KHz ) but they save from using more circuitry and screwing up with Vcom or the buffer thereof which must be very accurate and with negligible offset.
PCM2902 seems to be able to get two type of signals at the VinL / R inputs: an AC without an offset and a maximum amplitude of Vcom or the same but offset by Vcom to swing from 0 to Vccci = 2 Vcom. A good idea may be to have a reference of Vccci = 3.3V, thus Vcom = 1.65V. A good idea is to buffer this with a 5V and Analogue Ground supplied buffer.
There many combinations possible inclusive of these which give a voltage which spans from 0.5Vcom to 2.5Vcom. Most likely, TI would use the same resistors in the two of the schematics thus making Vcom – ui and Vcom + ui when ui is offset by Vcom and 2Vcom - ui and ui when not offset.
In case they would subtract Vcom, they would receive Vcom – ui and ui – Vcom which would be 0 in single polar works. Then Vcom – ui can be digitised in the 0 to Vccci range.
Also possible is to work out the resistors and processing and digitise the negative way separately from the positive thus allowing a Vccci maximum amplitude ( 3.3V ) and increase the precision.
The 4.7uF capacitor can be replaced with a 330nF, > = 5V capacitor which cuts off at 16Hz with the 30K resistor.
The first amplifier at the analogue inputs of PCM2902 has a capacitor in parallel which, hopefully, cuts off at 25KHz.
Before the PCM2902, many possibilities exist for a single power supply using heavily Vcom. These can be combined in many combinations, variations and permutations.
Also, a voltage inverter can be used to make – 5V from USB Vbus +5V.
Although a bit more noisy, a good idea would be to introduce high power negative supplies in standard PC power supplies, so USB can provide + - 12V or, at least, + - 5V.
On one hand, a voltage inverter may introduce a bit more noise ( in very high frequencies above 500KHz ) but they save from using more circuitry and screwing up with Vcom or the buffer thereof which must be very accurate and with negligible offset.
What's wrong with the drummer being loud?
In a twenty thousand seat auditorium, nothing (assuming there aren't any instruments/voices on the bottom level limit picking him up).
In the bar of the Holiday inn it gives you a lower limit to how far you can turn down and maintain a balance.
And everywhere it increases the amount of gear you have to carry about with you to match his level.
Besides, he isn't digital; that's the conga player
.Interesting site on building USB guitar, step by step
How to Build a USB Guitar
I'm not a fan of the standard commercial USB A and USB B type plugs, I can just imagine them falling out easily during a show. Here is the industrial version, a bit better!
USB2.0-IP67 Housing Set For Front- Rear Panel Mounting - USB 2.0-IP67 Water Resistant Connector System
How to Build a USB Guitar
I'm not a fan of the standard commercial USB A and USB B type plugs, I can just imagine them falling out easily during a show. Here is the industrial version, a bit better!
USB2.0-IP67 Housing Set For Front- Rear Panel Mounting - USB 2.0-IP67 Water Resistant Connector System
Power supplies and transients are not the issue. We have solved those problems a long time ago. In reality a fully digital solution has been possible for years and I am sure there are vendors who offer them. The reality is that an all digital solution simply does not sound the same and it does not accept overdrive well. It is the limitations of analog that actually create a lot of the sound we expect these days. And the way some analog stuff overloads is a large part of the rock and roll sound. It is why a lot of people still use tubes.
Yes there are vendors that are offering these now, for quite a while. Some use PCs for the processing and some are DSP based dedicated equipment. The digital stuff is getting really close. I don't want to say one better over the other because they both have their advantages. For fun this weekend I used my laptop based Gearbox program, guitar run through Toneport (Audio /USB interface) and back out through it to my solid state amp. My solid state amp is OK for clean but the SS distortion/boost channel , well, sucks. With the Gearbox software I was able to get a very close rendition of a Fender tube amp,then a stacked and dimed Marshall sound, etc.with a little tweaking of the controls in the software it is quite good.
I am a great fan of everything as digital as possible. I think, any sound can be reproduced with digital, but, this has already been discussed.
I think I have posted this but I have forgotten : A few years back I saw a Gibson with a digital converter and a network plug, a network guitar. The price was around $7000 as far as I remember. A few month back, I saw a Gibson ( Epiphone ) made in China for $120.
I think I have posted this but I have forgotten : A few years back I saw a Gibson with a digital converter and a network plug, a network guitar. The price was around $7000 as far as I remember. A few month back, I saw a Gibson ( Epiphone ) made in China for $120.
A document, called Errata.doc has been published here : Errata - Google Drive
The document correct errors I may have made in various projects I have worked on.
The document is to be updated.
Please, report all errors and problems.
The document correct errors I may have made in various projects I have worked on.
The document is to be updated.
Please, report all errors and problems.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.