help with LT1028

Hi,
 
I bought LT1028csw from here - http://www.ebay.com.au/itm/Two-2x-Mono-Dual-OPAMP-2x-LT1028-NE5532-OPA2604-/290606645188
 
When I use it in my DAC and connect IEM's directly to the RCA out, there is a lot of distortion, especially with bass.
 
When I connect my DAC to my stereo amplifier everything sounds excellent － why?  This happens with no other op-amp so I'd like to isolate what's happening with the LT1028.
 
Thanks in advance for any help, and links to literature on op-amp's please!
 
Here is a link to start - http://www.sakurasystems.com/articles/Op-amps%20Junji.html
 
Thx.
 
Sincerely,
 
kiteki

Here's the LT1028 datasheet,
 
- http://www.linear.com/product/lt1028
- http://cds.linear.com/docs/Datasheet/1028fa.pdf
 
Anything unusual here which could cause very high distortion － when NE5532, OPA827, AD797 etc. don't?
 
 
Some more literature on op-amps courtesy of turokrocks! - http://www.aurorasound.jp/OPAMP.html

Another link to op-amp data... http://www.sg-acoustics.ch/analogue_audio/ic_opamps/pdf/opamp_distortion.pdf

I suspect if you figure out how much current you need to drive your IEMs versus the volume section in your stereo, you'll find the answer to your problems.

LT1028 isn't strictly unity gain stable - requires compensation for low gains and capacitive loads - there is a external compensation pin, requires unusually large >100 pF compensation C to make a difference
 
it has odd dependency on source, feedback R damping for stability at low gains without compensation
 
you may have oscillation or bursts of oscillation, possibly depending on cable, load properties
 
it also doesn't have very high output current spec, Aol droops heavily with sub 500 Ohm loads - I wouldn't use it to directly drive low Z
 
I have designed a strain gage amp with the LT1028, wrung it out on the bench - datasheet is unusually good but needs to be read thoroughly, with a trained eye for the engineering subtleties they are conveying, or "hiding" in the graphs

Thanks, what could happen if you directly drive low Z like a 16 ohm IEM?
 
What do these specs in the datasheet imply more precisely?
 
"ZO Open-Loop Output Impedance VO = 0, IO = 0 80 80 W
IS Supply Current 7.4 9.5 7.6 10.5 mA"
 
 
What do you think of this summary from the seller?
 
"HOW GOOD LT1028 IS?!
Its CMRR up to 126db and PSRR up to 132db, and it can work at +-22v working voltage.

How it sound?
It have crystal clear transparency and incomparable details. Almost no ther opamp sound better in this two charactristics. While some audiophile say it sound too lean and lack warm mid. Some audiophiles do not like too analytical sound. But I think no one can challenge its 'high fidelity' performance."
 
 
I'm using it in this DAC (Asahi AK4396) - http://www.diyaudio.com/forums/digital-source/186245-dac-2496-ak4393-dac-kit-cs8416-ak4393-5532-a.html
 

 
 
If the cable specs are a variable they are -
 

1. [size=x-small]Conductor resistance: 36 mΩ [/size]
2. [size=x-small]Distribution inductance: 2 μH [/size]
3. [size=x-small]Capacitance: 131PF [/size]
4. [size=x-small]Diameter: OD08 * 2F [/size]
5. [size=x-small]Length: 3m [/size]
6. [size=x-small]Spec: OD8.0mmX2C L=3m [/size]
7. [size=x-small]OFC copper[/size]

 
 
The IEM in the picture is JVC FX500 -
 

HP-FX500 主な仕様

 
 
 
I connect to my stereo receiver via RCA, pretty much the same cable.
 
I'm using FX500 as an example but pretty much every single IEM I've used has really high distortion with the LT1028, so I thought it didn't like the circuit in the DAC, and then it sounds excellent connected to my stereo->IEM, stereo->speakers, heh?

his numbers are a bit jibberish, use the datasheet. it will be current limited driving such low impedance and probably voltage too, as it will not swing anywhere near the rails into that load. the current and unity gain w/o compensation (as jcx mentions) will be your main problem, especially as its likely your 16ohm IEMs present a much lower impedance at some frequencies.
 
in short I wouldnt even bother trying it, pick an opamp that has higher current output into low impedance. its an instrumentation amp and an extremely good one, driving high impedance (like a buffer, or your amps input) its a superb chip, one of my favorites. so your choices are to either use a different chip, or add a buffer stage, if you want to drive lowZ directly

I just started experimenting with op-amps in Jan of this year so I can't read every spec. on the sheet yet.
 
I don't need to drive low Z directly, however I'd like to hear the IC's directly.  Especially in blind listening － a long chain of components is too counter intuitive as far as I'm concerned.
 

Originally Posted by rickcr42 /img/forum/go_quote.gif

there is a slew of buffers available and all sound different.the BUF634 is and always has been known as a "dark" sounding part ［／］ etc.

Click to expand...

 

---

 
So here is my next question.
 
I'm satisfied with the LT1028 connected to my stereo, so now I'd like a pure headphone amp design suitable for OPA627, correct voltage supply etc., portable is ideal, any advice?
 
 
Thanks in advance.

Just a quick update.  The LT1028 improved in directly driving an IEM when I increased the resistance from 16 ohm to 91 ohm, however it still didn't sound perfect, anyway it seems like that was the issue... until I connected this to an amplifier, and on the lowest volume the amplifier would turn on and off rapidly, so the LT1028 is still acting strange compared to all other op-amps and I'm not using it.
 
I like the sound of the OPA827 better anyway from this experience, and most likely AD797 too.