I read the datasheet: http://www.mouser.com/ds/2/405/ne5534-407641.pdf
I see 3 places where it talks about needing a compensation capacitor but it doesn't say anywhere how to calculate it's value based on the gain/frequency bandwidth you are using the chip for. This makes me frustrated, do they expect you to just guess what it should be? 😡
I've read that the value of the compensation capacitor (or lack of having one) is linked to the op amp's stability. Can anyone with experience with this chip point me in the right direction for something to read about this? My searches have come up empty.
EDIT: See this post for measurements of NE5534 at unity gain with and without compensation caps: http://www.diyaudio.com/forums/anal...acitor-value-ne5534-op-amp-4.html#post4547178
I see 3 places where it talks about needing a compensation capacitor but it doesn't say anywhere how to calculate it's value based on the gain/frequency bandwidth you are using the chip for. This makes me frustrated, do they expect you to just guess what it should be? 😡
I've read that the value of the compensation capacitor (or lack of having one) is linked to the op amp's stability. Can anyone with experience with this chip point me in the right direction for something to read about this? My searches have come up empty.
EDIT: See this post for measurements of NE5534 at unity gain with and without compensation caps: http://www.diyaudio.com/forums/anal...acitor-value-ne5534-op-amp-4.html#post4547178
Last edited:
It depends on the gain it's operated at, for gains of 3 or more comp cap is not necessary.
Data sheet: http://www.ti.com/lit/ds/symlink/ne5534.pdf
Mike
Data sheet: http://www.ti.com/lit/ds/symlink/ne5534.pdf
Mike
Last edited:
How are you planning to use it? It might make more sense to use a different chip for a low gain circuit.
Mike
Mike
@Michael Bean -I saw the datasheet, but I didn't see where they give you the answer for the cap value based on bandwidth and gain. Was it in the graphs? (How irritating that they bring it up in sections but don't refer you to the formula!) If it's there, you're going to have to talk to me like I am 4 and tell me where because I don't get it.
I'm planning on using it for a kind of low shelf boost circuit. Some lower frequencies can have a possible gain of >15dB while others are attenuated. Makes picking the overall gain value to plan for confusing.
A curve like this for example but the frequency and boost will be adjustable.
@nigelwright7557
I have an oscilloscope now but have not gotten to see an op amp circuit's waveform under oscillation yet. Will you point me to an example of what to look for in what frequencies for signs of instability?
Is this just a generally unstable chip and I should keep it at a fixed gain when using the NE5534 (22pF compensation if under a gain of 5)?
I'm planning on using it for a kind of low shelf boost circuit. Some lower frequencies can have a possible gain of >15dB while others are attenuated. Makes picking the overall gain value to plan for confusing.
A curve like this for example but the frequency and boost will be adjustable.
@nigelwright7557
I have an oscilloscope now but have not gotten to see an op amp circuit's waveform under oscillation yet. Will you point me to an example of what to look for in what frequencies for signs of instability?
Is this just a generally unstable chip and I should keep it at a fixed gain when using the NE5534 (22pF compensation if under a gain of 5)?
Last edited:
Does this help as an example.
http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
Mooly, good to see you again! As usual, thanks for posting the goods. I'm curious: should I always look for a constant high frequency or can instability show itself in other ways like ringing in transients?
I think I recently encountered oscillation when replacing the op amp chips that sum the dac output on my Asus Essence STX but I am not sure. I replaced a couple JRC2114's with a couple LM4562's. The sound was most unpleasant. I scoped some source material and found that the JRC's output higher voltage with the same exact volume settings. 500mV per div:
The above is with music
Below is with triangle wave synthetic bass
I didn't think to analyze sine waves or look at their behavior at idle before changing the 4562 chips out again. Oops.
I think I recently encountered oscillation when replacing the op amp chips that sum the dac output on my Asus Essence STX but I am not sure. I replaced a couple JRC2114's with a couple LM4562's. The sound was most unpleasant. I scoped some source material and found that the JRC's output higher voltage with the same exact volume settings. 500mV per div:
The above is with music
Below is with triangle wave synthetic bass
I didn't think to analyze sine waves or look at their behavior at idle before changing the 4562 chips out again. Oops.
Last edited:
Instability generally shows as high frequency oscillation. If a circuit were on the verge of instability then fast transients (I'm thinking more generator produced than music) could trigger that instability.
Under compensating typically gives a peaky edge to signals with fast rise and fall times such as shown here,
http://www.diyaudio.com/forums/analog-line-level/196461-different-opamp-compensation-technique.html
Hard to say what is going on in your scope shots tbh. You would need to expand the trace and see what was in there. The last two shots do look like typical oscillation (assuming we are looking at the opamp output) but you would need to be sure on the signal grounds used for the measurement and the vertical scale... whats the amplitude.
Under compensating typically gives a peaky edge to signals with fast rise and fall times such as shown here,
http://www.diyaudio.com/forums/analog-line-level/196461-different-opamp-compensation-technique.html
Hard to say what is going on in your scope shots tbh. You would need to expand the trace and see what was in there. The last two shots do look like typical oscillation (assuming we are looking at the opamp output) but you would need to be sure on the signal grounds used for the measurement and the vertical scale... whats the amplitude.
Above, the top 2 shots are of the same passage of the same song with the same volume levels taken from the output of the soundcard. 500mV per div.
The middle 2 shots are of the same passage of the song but with only the synthetic bass part being soloed.
The bottom 2 shots is the same synthetic bass part which uses 2 octaves of triangle waves, zoomed in.
In all of the compared shots, there is less output for the LM4562. The sound of those chips was shrill and harsh in the high end while lacking low end, I could immediately notice the difference when listening to reference material. You can see in the top shot how the LM4562 starts fading off in amplitude towards the end of the song passage. You can also notice something like that in the middle shot for the 4562 where 3/4 into the passage, the amplitude grows and then goes back down.
Maybe this is something different than an issue with stability, I don't know.
The middle 2 shots are of the same passage of the song but with only the synthetic bass part being soloed.
The bottom 2 shots is the same synthetic bass part which uses 2 octaves of triangle waves, zoomed in.
In all of the compared shots, there is less output for the LM4562. The sound of those chips was shrill and harsh in the high end while lacking low end, I could immediately notice the difference when listening to reference material. You can see in the top shot how the LM4562 starts fading off in amplitude towards the end of the song passage. You can also notice something like that in the middle shot for the 4562 where 3/4 into the passage, the amplitude grows and then goes back down.
Maybe this is something different than an issue with stability, I don't know.
It definitely looks like something is wrong. It would be helpful if you could post a schematic and maybe some pics.
Mike
Mike
You can get a sense of it here: ASUS Xonar Essence STX soundcard review - Modifications - Modifying the soundcard
I don't have a schematic or anything like that. Wish I did.
Going back to my OP, does anyone have suggestions for the cap value to add for compensation based on my EQ'ed response in post #6?
I don't have a schematic or anything like that. Wish I did.
Going back to my OP, does anyone have suggestions for the cap value to add for compensation based on my EQ'ed response in post #6?
the op amp compensation requirement applies for the feedback gain at the loop gain intercept frequency, audio frequency gain is usually unimportant for stability
the 5534 can even be 2-pole compensated for unity gain, keeping the higher open loop gain for lower frequencies - Bruce Hofer gave a circuit in his master class notes
the 5534 can even be 2-pole compensated for unity gain, keeping the higher open loop gain for lower frequencies - Bruce Hofer gave a circuit in his master class notes
OK, there's your problem, the original chips are duals, two op amps per chip, the 5534 is a single op amp with a totally different pin out.
Mike
Mike
@MBean: I know they are duals, that wasn't the problem. My soundcard actually has an LM4562 for the output buffer after the 2 duals which sum the dac out. I've tried replacing the 2 summers with NE5532 (probably the best sounding so far), OPA2134 (had more noise and less clarity), JRC2114's came stock on the board (sounded good) and the worst was the LM4562's. I'm guessing it had problems due to it's extremely high bandwidth and it didn't like something in the design of the board or maybe had bad interactions with the dacs somehow.
Thanks JCX and Nigel. I am not afraid of this chip now and since I already bought a few of them, I can put them to good use. If I run into a situation where they start singing, I'll be sure to post some scope shots now that I know what to look for.
Thanks JCX and Nigel. I am not afraid of this chip now and since I already bought a few of them, I can put them to good use. If I run into a situation where they start singing, I'll be sure to post some scope shots now that I know what to look for.
Because the datasheet doesn't have information about compensation cap values.
When Mooly posted a link to his page showing op amp oscillations I asked if that was the only sign of instability. Then I showed my scope shots of a possible unstable op amp situation to see if any of you have run into a similar problem and could shed some light on it. It's all on the previous thread page if you are still confused.
When Mooly posted a link to his page showing op amp oscillations I asked if that was the only sign of instability. Then I showed my scope shots of a possible unstable op amp situation to see if any of you have run into a similar problem and could shed some light on it. It's all on the previous thread page if you are still confused.
I recall Bob Pease describing an "instability test," perhaps in a column perhaps named "What's with all this instability stuff, anyhow?" where he used a square-wave generator and a resistor in series connected to the opamp's output to see if it rings (the classic exponentially decaying sine wave) or has the nice classic-looking exponential decay.
Especially read the point (c) (well, read the whole article) here:
What?s All This Capacitive Loading, Stuff, Anyhow? | Power content from Electronic Design
Especially read the point (c) (well, read the whole article) here:
What?s All This Capacitive Loading, Stuff, Anyhow? | Power content from Electronic Design
- Status
- Not open for further replies.