Electronics > Projects, Designs, and Technical Stuff
Need an op amp recommendation
hummusdude:
I'm building a simple stereo pre-amp to drive my power amp. I have a 250 VA power supply for the power amp that put's out +- 26 VDC and I'd like to tap that to power a pre-amp if there is an op amp that will handle that high a supply voltage. The power amp only puts out 30w per channel into 8 ohms so seems like there is plenty of headroom on the transformer. My goal is to reduce the complexity and not have to deal with stepping down the voltage if it's not necessary.
I know there are efficient buck converters available for cheap but I'm starting with the op-amp selection in case I can avoid more parts and potentially more noise sources.
Ian.M:
An OPAMP preamp should only need a few tens of mA. Why not use a LM7815 positive regulator and LM7915 negative regulator for nice clean +/-15V rails for the preamp and have a free choice of OPAMPs? If you don't need more than 50mA, you can get away without heatsinking the regulators, or under 150mA small clip-on heatsinks should suffice.
Don't even consider a buck converter or other switching regulator unless you can guarantee a sufficient minimum load to prevent it entering cycle skipping mode, as cycle skipping can easily cause significant noise in the audio band.
hummusdude:
Thanks Ian...that's sounds like it could work. Would the regulator dissipate much power if the input is at 26 VDC? I'm aware that a linear power supply can have issues with this when the input voltage is significantly greater than the output voltage. I'm trying to be mindful of parasitic current draw from the main supply.
Ian.M:
A LM7815 has a quiescent current of about 5mA, so that's 130mW dissipation just sitting there with a 26V input, + load dependent dissipation of its input to output voltage drop times its load current. At 50mA load current the total dissipation would be 680mW.
Similar figures apply to the 7915 negative regulator.
N.B. You *MUST* use the recommended decoupling caps (see datasheet) on the input and output of LM78xx and 79xx regulators and connect them as close as possible to the regulator. Omit the decoupling and if you are unlucky they can act as a RF power oscillator, 'singing' like a canary at a few tens of MHz. Unless you scope their output with a 50MHz or better scope, the only clue you'll get that they are misbehaving that way (apart from the missing or badly located decoupling) is that the output voltage is significantly out of spec, and its load will most likely be malfunctioning extremely weirdly.
Caution: Unlike the LM78xx series positive regulators, the heatsink tab of a LM79xx negative regulator is connected to its inout pin, not to ground. If you bolt one down to chassis or a grounded heatsink, you must use a silpad or other electrically insulating thermally conductive heatsink washer under it and a top-hat insulating washer on the screw.
Mechatrommer:
--- Quote from: hummusdude on July 10, 2020, 01:35:06 am ---My goal is to reduce the complexity and not have to deal with stepping down the voltage if it's not necessary.
--- End quote ---
https://www.digikey.com/product-detail/en/texas-instruments/OPA2544TG3/296-50635-ND/1572287
https://www.ti.com/lit/ds/symlink/opa2544.pdf?HQS=TI-null-null-digikeymode-df-pf-null-wwe&ts=1594361160756
you can get combined power 10W right out of it, 40W+ if you can get it cool. the catch is if you burn it, thats another $23 fix (it has overheat protection)
Navigation
[0] Message Index
[#] Next page
Go to full version