it appears you worry too much about supply ripple. 
LTC1044 runs at 10kHz in the circuit I used. I also used low ESR monolithic ceramic caps for filtering. no problems with ripple.
besides, op amps nowadays have rather high PSRR. what is much more important is proper grounding. even if you have very clean DC supplies, wrong grounding will make your preouts much more noisy than a some ripple on the supply but grounded properly.
LTC1044 runs at 10kHz in the circuit I used. I also used low ESR monolithic ceramic caps for filtering. no problems with ripple.
besides, op amps nowadays have rather high PSRR. what is much more important is proper grounding. even if you have very clean DC supplies, wrong grounding will make your preouts much more noisy than a some ripple on the supply but grounded properly.
Ok, lets not worry about supply ripple.
Looking at the datasheet of LTC1044A right on the first page there is a sample circuit showing how to get -10V out of +10V supply. And the problem is that as the load current increases, output voltage drops significantly. At 45mA load instead of -10V you get only -8V. How to overcome this problem? Should opamp be powered with this asymmetrical voltage (+10/-8V)? If I understand correctly, AC coupling must be done in the output in this case. But the worst thing is that opamp will probably not be a constant load so voltage on the output will also change.
I think I'll have to get some opamps and charge pumps to get more familiar with them.
Looking at the datasheet of LTC1044A right on the first page there is a sample circuit showing how to get -10V out of +10V supply. And the problem is that as the load current increases, output voltage drops significantly. At 45mA load instead of -10V you get only -8V. How to overcome this problem? Should opamp be powered with this asymmetrical voltage (+10/-8V)? If I understand correctly, AC coupling must be done in the output in this case. But the worst thing is that opamp will probably not be a constant load so voltage on the output will also change.
I think I'll have to get some opamps and charge pumps to get more familiar with them.
depending on the op amp you use, op amps consume around 2.5mA each. some get up to 5mA. so to draw 45mA, you'll need to be using a lot of op amps. you'll only need four for front/rear outputs. Eclipse used LTC1044 in atleast one of their radio and I have also used one so it isn't really an issue AFAIK.
Ok, so I'm now putting together a shopping list and came across TC962 charge pump from Microchip. It's cheaper than LTC1044 and looks to be more efficient at converting voltages too. Switching frequency of 24kHz when pin 6 is grounded is also good for audio applications (inaudible). Would your recommend it over LTC1044?
Oh, and found link to this circuit somewhere here in forum
Additional LC filters will help my obsession about voltage ripple Not sure if I will need the extra filtering, but we will see...
Oh, and found link to this circuit somewhere here in forum
An externally hosted image should be here but it was not working when we last tested it.
Additional LC filters will help my obsession about voltage ripple
Not sure if I will need the extra filtering, but we will see...I suggest you try it as is. at default low switching frequency and no output filters. if it works, that's great. when you do get noise, then work your way from there.
higher frequency will lower ripple but it may couple to other sections more easily due to the higher switching frequency. you may not realize (not hear) that it's coupling to the audio sections and before you notice, your tweeters are fried.
higher frequency will lower ripple but it may couple to other sections more easily due to the higher switching frequency. you may not realize (not hear) that it's coupling to the audio sections and before you notice, your tweeters are fried.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.