clean 11 Vrms for 600ohm load

Status
Not open for further replies.
Jump to Latest
R

regal

Member
Joined 2004
#1
Odd mission, what topology should I look at for an amplifier that can swing 11Vrms but only needs to deliver ~200mA.

Its an unusual deal, with the 600 ohm load one would think you could find a topology with very low distortion. The problem is that most opamps can't quite get to 11 Vrms, most gainclone/chips increase in distortion directly with voltage not current.

So I guess what I am looking for is a discrete direction where the thd/imd drop significantly with an easy load but still can handle 11vrms.

Any thing like this exist?
 
#5
It would be useful to know a few more parameters, like noise and distortion targets, required voltage gain, desired output impedance, driving impedance, space constraints and budget. Supply noise density, too.

I'd be thinking a unity gain buffer driving an inverting gain stage driving an inverting output buffer of no more than gain 2 with 1 or 2 dual opamps in parallel, all with NE5532s on approx. +/-18 V. 5532s are cheap and have good transfer linearity, and 4 in parallel should do an excellent job on 600 ohm loads. You can easily afford 47R..100R combining resistors.
 
Last edited:
#6
11V RMS is 31.1V P-P

A 5532 is only rated at 24V (min) with ±15V supplies, even with another ±3V on the supplies it's going to be close.

The OPA2604 is capable of driving 600Ω loads with power
supply voltages up to ±24V.
 
#7
The 5532 and 5534 are 44v devices.
They can be run on regulated +-21Vdc supplies.
They will easily give 11Vrms into 10k load. and may still work well into a 2k load.

I would suggest a discrete transistor buffer to get the necessary current capability to drive a 600r load.

You could set the buffer to always remain in ClassA even for peak currents and you can modify the sink/source on the opamp output to convert that output stage to single ended ClassA for all peak currents.
 
#8
The LME49600 would do the job, but it is about ten times the price of a NE5534. The OPA2604 is five times the price of the NE5534. The datasheets gives (short circuit) output current for the LME49600 as 550mA, for the NE5534 as 38mA and for the OPA2604 as 40mA. You pay your money and you make your choice.

Then again - 11V into 600ohm gives 18mA of current AFAIK. You should therefore be OK with the NE5534. For the final choice, other things such as slew rate, noise, distortion etc. will have to be considered as well.
 
#12
It just seems the opamps are a bit on their ragged edge at output of 11 Vrms. I don't want to be near that vertical line of distortion on the graph vs Vrms.

What about some sort of low distortion discrete gain stage and a simple mofset source follower?
 
#13
how aobut some clarification of the requirements, application

the drive V, load impedance don't match the stated current

can the load be driven "balanced" by bridged output?

you don't quantify just how low distortion, or at what cost

dynamic headroom can be a "soft" spec - or not - in pro audio +24 dBu is pretty much "the max" for any input while a 600 Ohm monitor headphone may need even more V drive to manage 120 dB SPL dynamic peaks
 
#14
jcx said:
how aobut some clarification of the requirements, application

the drive V, load impedance don't match the stated current

can the load be driven "balanced" by bridged output?

you don't quantify just how low distortion, or at what cost

dynamic headroom can be a "soft" spec - or not - in pro audio +24 dBu is pretty much "the max" for any input while a 600 Ohm monitor headphone may need even more V drive to manage 120 dB SPL dynamic peaks
Click to expand...

I meant 200mW which is ~11Vrms across a 600 ohm load.

The load could be balanced.

Yes I am thinking of the old 680 headphone outputs on mixing boards, but not doing heavy studio work, I need more like 70db range, at least -90db noisefloor. Low distortion with SS means to me where the thd vs vrmsout is horizontal well before it goes exponetial high. So less tha .01 thd feasable ?

I guess considering balanced brings the opamps back into the picture, thanks
 
#15
most headphones are sold with 3-pin TRS connectors that prevent bridged drive because the R/L gnd return wires are shorted together

true floating balanced outputs (xmfr) work with TRS - but not the two amplifer/channel bridged output

people do reterminate with 4-pin or dual connectors to separate the R/L driver wire pairs entirely and be compatable with "balanced"/bridged output amplifiers


the peak current drive of near 30 mA is awkward - a few "audio" op amps can do that without a added buffer - many show substantial distortion increase when supplying even 1/2 of the datasheet rated output current

but if 100 ppm is adequate then single op amp per side of the bridge should be fine with properly selected parts

really simple buffer in the feedback loop, a Class A bjt follower lets you use a much wider selection of op amps

with advanced multiloop/composite op amp circuits sub-ppm, "unmeasureable" distortion is possible
 
Last edited:
#20
"I don't want to be near that vertical line of distortion on the graph vs Vrms."

Try a ±24 opamp, like the OPA2604. While it's rated for 600Ω drive with ±24V, you could parallel it for even more current headroom.
 
Status
Not open for further replies.
Top Bottom