Yes i was wondering if maybe a bass/treble control would be more help here
Baxandall tone control using op-amp.
Audio Designs With Opamps -2
Last edited:
Hi guys thanks for suggestions.
Ne5534 should be amazing as is. It's better than ne5532 from what's people say and specs.
I'm away and had not the opportunity to reply or test further. But as suggested earlier doubling input cap to improve bass. Parts are already very good so design must be looked at.
Does anyone know how to measure DC offset from the chip ne5534? Thanks
Ne5534 should be amazing as is. It's better than ne5532 from what's people say and specs.
I'm away and had not the opportunity to reply or test further. But as suggested earlier doubling input cap to improve bass. Parts are already very good so design must be looked at.
Does anyone know how to measure DC offset from the chip ne5534? Thanks
Really?? I don't find the '5534 sounds any better than the '5532. They sound quite similar.
Just exactly who are these "people say and specs."?
A 5534 compensated for unity gain with a 22pF cap will be almost identical to one channel of a 5532, which is internally compensated for unity gain. Where a 5534 differs is that you can under-compensate it and get more gain-bandwidth, as long as your closed loop gain is sufficiently high. The 5534 is stable for a gain of 3x without a compensation cap, and you can use values smaller than 22pF for gains between 1x and 3x. The data sheet probably defines this. When run that way, a 5534 will behave better. But, with a 22pF compensation cap, it's basically the same as a 5532.
You can measure offset with a voltmeter. Nicer ones are more accurate, but offsets in the mV range aren't too tough to measure. Just read it like you'd read the voltage output of a battery.
You can measure offset with a voltmeter. Nicer ones are more accurate, but offsets in the mV range aren't too tough to measure. Just read it like you'd read the voltage output of a battery.
Sounds interesting. I read that Ne5534 had lower noise or was just better but I guess from whats been discussed here I am wrong!
I'm not sure even what compensation actually does! Also why would I need more gain. The volume is fine the detail is fine everything is fine just needs slightly more bass and slightly less treble.
You say I can measure the DC offset like a battery? I just had a thought about this and I think I figured it out. Since the amp is both a pre and power in one I can probably disconnect the connector at the back and measure the DC offset like a normal amplifier!
I appreciate all the input, I still haven't tried any suggestions due to half term! But I'm really intrigued about this compensations stuff. My aim is to improve the Preamp as much as possible as its the easiest way for me to boost sound quality!
Thanks!
Last edited:
If no progress, then look in a different place? Unless it's built poorly, that amp ought to be nearly dead flat over a generously wide bandwidth. Lacking bass or whatever is most likely a problem somewhere else. Just because a mod is easy doesn't mean that it'll produce results!
Compensation changes the internal speed of an amplifier to make it stable under different closed loop conditions. Most IC op amps are unity gain stable, so you can operate them at any gain you desire and they'll remain stable. Not the case with some older amps - gain was difficult to obtain then, but they were able to trade unity gain stability for extra gain. You don't need it, but other people need gain, and that's why they allowed you to compensate some amplifiers. With a dual op amp and 8 pins, there's no room for compensation pins, so that's out of the question.
For measuring DC offset, don't detach too many things - an open input will behave differently than one with input connections. You measure voltage across two nodes anyway - no need to interrupt the circuit. Attach one probe to the output, the other to ground, read the value.
Compensation changes the internal speed of an amplifier to make it stable under different closed loop conditions. Most IC op amps are unity gain stable, so you can operate them at any gain you desire and they'll remain stable. Not the case with some older amps - gain was difficult to obtain then, but they were able to trade unity gain stability for extra gain. You don't need it, but other people need gain, and that's why they allowed you to compensate some amplifiers. With a dual op amp and 8 pins, there's no room for compensation pins, so that's out of the question.
For measuring DC offset, don't detach too many things - an open input will behave differently than one with input connections. You measure voltage across two nodes anyway - no need to interrupt the circuit. Attach one probe to the output, the other to ground, read the value.
The old, never-updated, data shows '5534 hiss a wee bit better than '5532.
There's no implementation reason they should be different hiss. I've always assumed that they left a little slack in the dual's spec to cover a little dirty Silicon, but if you paid essentially the same price for just one opamp they'd run it on clean Silicon. Dirty Silicon has not been an issue in a long time.
There's no implementation reason they should be different hiss. I've always assumed that they left a little slack in the dual's spec to cover a little dirty Silicon, but if you paid essentially the same price for just one opamp they'd run it on clean Silicon. Dirty Silicon has not been an issue in a long time.
At this point, there are so many "5534" datasheets, that it can require viewing a number of them just to get some of the detailed specs. Not all 5534s are built the same either - some measure differently and sound better/worse, but it's often subtle. I haven't used a 553x in a while, so I'm not up on them anymore, but I do recall that even a 'lowly AP2322' can sort out different vendors' 5532 versions.
I've noticed that TI specs the dual LME49720 a little worse (En and Vos) than the discontinued single LME49710, but it makes little sense to me - they're otherwise identical. These parts were recently moved to a "modern" 8" wafer line from their original 6" line, so dirty silicon shouldn't be an issue. Similarly, I can't imagine that the yield isn't 100% (or damn near close) for complementary bipolar analog op amps in 2018.
The only thing I can think of is that a dual in the same package as a single will have a higher die temperature. With a JFET amp, this would mess up the input bias, but maybe it also messes up bipolar amps? Higher emission from lattice defects causing extra noise? Who knows... never measured actual devices in enough quantity to know if it's just CYA specs or an actual depiction of differing performance.
I've noticed that TI specs the dual LME49720 a little worse (En and Vos) than the discontinued single LME49710, but it makes little sense to me - they're otherwise identical. These parts were recently moved to a "modern" 8" wafer line from their original 6" line, so dirty silicon shouldn't be an issue. Similarly, I can't imagine that the yield isn't 100% (or damn near close) for complementary bipolar analog op amps in 2018.
The only thing I can think of is that a dual in the same package as a single will have a higher die temperature. With a JFET amp, this would mess up the input bias, but maybe it also messes up bipolar amps? Higher emission from lattice defects causing extra noise? Who knows... never measured actual devices in enough quantity to know if it's just CYA specs or an actual depiction of differing performance.
I completely agree - the original amp stage in question is likely to be really close to flat, and only subtle subjective differences are possible by altering components, with the danger that "I placed my modder's hands upon it therefore it's better" bias might be the biggest factor. Rigging up the LF portion of a simple Baxandall by transforming a simple series resistor into an LF boost (or shelving HF shunt) circuit might be more fruitful.
I can confirm changing the input 2.2uf for a 4.7uf made pretty much no difference!
Also changing the 220pf for 330pf made pretty much no difference
Also adding a 5pf across the 1pf was not so good!
No progress!!
What do you REALLY expect here?
Do you, for some bizarre reason, think the the NE5532 and 5534 are the "Holy Grail" of op-amps when it comes to sound quality?
If so, I can tell you without a shadow of doubt that they are NOT having trying many op-amps in many different applications over the years.
Why don't you throw the '5534 and its dual counterpart '5532 in the trash where they belong and move on to other more modern and better sounding op-amps?
You have a LOT to learn about audio op-amp circuits, so I wish you nothing but the best of luck with your venture!
TI Data sheets specs show NE5534 typ voltage noise lower than NE5532.
Note that there are 'A' grades of both so actual figures depend on that too as well as deviation from typical / median value.
I think this is speculated to be due to more transistor stages paralleled in frontt end of 5534 but the details of these circuits is hard to confirm.
Anyway - the data sheets are Googled up in an instant...
What do you REALLY expect here?
Do you, for some bizarre reason, think the the NE5532 and 5534 are the "Holy Grail" of op-amps when it comes to sound quality?
Why don't you throw the '5534 and its dual counterpart '5532 in the trash where they belong and move on to other more modern and better sounding op-amps?
5532/4 may not be the ultimate opamps for signal level low frequency ac (eg audio !) applications but to claim they belong in the 'trash' is a bizarrely nonsensical statement.
Any minor shortcomings of these devices has nothing to do with the OP's issue of 'thin sound'.
Old and obvious argument - but if that were the case then a huge number commercial recordings would be suffering the same fate or do you advocate TL07x as a better option..?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.