AD 815 as power amp? possibilities?

[rhythmdiy]

Ive been searching about opamps and I found one hi current opamp... the specs are very close to a high power amp...

like slew rate 1kv/us
output voltage 20v
output current .5 amps
load at 10ohms...
noise in nano volts...
well we can parallel 5 chips each having two cores of opamps.. delivering about

5 * 2 * .5 amps = 5 amps
voltage = 20v
power = 100watts

or atleast 25watts if im too paranoid...

what do u say guys any suggestions?

can i put gain of 20db per chip?

[AndrewT]

no, it's not a power amp chip.
It runs from supplies <=+-18Vdc.
It dissipates 4.5W when driving a 50r load to 40Vpp.
Read the applications list. It's a video driver with a few similar load applications.

It might make a good high impedance headphone driver.

[coloradosound]

Second Andrew's!
The chip is meant for video line driver use in
driving 75 ohm or 50 ohm loads. The reason it states 10 ohms
in specs is to allow sufficient drive capability for long lines.

[rhythmdiy]

AD 815 is the chip used in Jeff Rowland Coherence Preamp... so when it is capable of driving such high currents then i should get atleast 10watss with 5 chips in parallel ...

http://www.stereophile.com/solidprea...nd/index1.html

now tell me once i remember i played a speaker at lower volumes using an opamp in my desktop small multimedia speaker ( I remember a 8 pin dip opamp ) i didnt knew the make 8 years ago...

but i think its possible... are there any strong reasons why it cant?

what about gain? is it possible to make up gain of 26db?

[AndrewT]

Originally posted by rhythmdiy
AD 815 is the chip ........... i should get ........ 10watts with 5 chips in parallel ............

that's more credible, if you can make sure it is stable and can work with a reactive load and can be made to match each other to ensure balanced currents and and and.

Have you noticed yet the big difference in performance between driving 50r and 200r? That gives a clue to how stretched this chip is in driving 75ohm cables.

[coloradosound]

JRDG is indeed using them for audio rather than video,
but key point is it is being used as preamp output buffer rather
than power amp output. Indeed it states that it can drive
a low imp. load, normal usage of a pre is around 20K ohm
but them Rowlands can indeed drive your pro balanced stuff
@ 600 ohm with no problemmo. I think if you did want to use
them as small power amp you could get a whole bunch as
Andrew states and try matching up sets. I would definately
try to attach some type of chip heat sinks to them as well.
I would really expect performance not much better than a
powerful computer sound card in the end because most of that
performance will be lost due to the nature of trying to
drive all those chips in parallel with the small differences
between each chip (phasing, internal output imp.,etc., etc.).

[KSTR]

Look here for a real world example of an fine amp, rated 20W@8Ohms, using 20pcs (=40 OpAmp channels) of TPA6120a (==THS6012==upgraded AD815)
http://www.diyaudio.com/forums/showt...39#post1305439

- Klaus

[benchtester]

I think a superior amp is quite achievable. I am a little biased as I bought a batch of AD815 chips with the intent of making a power amp. Time passed and the project is on the far back burner.

Years ago I made a nice tweeter amp out of six .25 amp buffer chips inside a loop controled by a AD811. I don't see any reason the same concept can't be applied to the AD815.

The old AD815AY has a big advantage over the TI chips in that you could get it with a big tab to heat sink (extra wide TO-220). For power amp applications the heat sinking must be addressed. However, I just looked on the Analog Devices site and it doesn't look like it is still available.

Reminder of the Basics: The AD815 is a current feedback amp. So if you use it as a voltage follower, you will still need a feedback resistor. If you use one amp to control the loop, you will need to slow it down to maintain stability. With current feedback amps this is a cap to ground, not across the feedback resistor.

If one wants more power, then one could get up to 40 watts by using a balanced (a.k.a. bridge-tied load) configuration.

If anyone wants to try this I might be willing to part with my AD815AY chips. Send me a PM.

BTW: If anyone doesn't want to parallel chips, you can make a great amp out of a single PA19 power op amp from Apex Microtechologies (but they're like $300 a chip).

[coloradosound]

Hey Ken,

I also posted this on your Missing Element thread.

Interesting AP Note from National:
http://www.national.com/an/AN/AN-1485.pdf

Although AD815 will drive lower imp. I would apply
scale of thinking to amps in the note driving their intended loads.