Going to be doing a bit of experimentation, and I want to use an opamp with external compensation, such as the 5534.
So, is there another opamp suitable for audio with that pinout (yes, I need the same pinout as the 5534)? I've been searching, and so far come up empty-handed...
If not, the 5534 will probably be fine, but I'd love to expand my options here.
So, is there another opamp suitable for audio with that pinout (yes, I need the same pinout as the 5534)? I've been searching, and so far come up empty-handed...
If not, the 5534 will probably be fine, but I'd love to expand my options here.
I wish I could.
This opamp sits at the input of the amplifier section, and the whole driver+output stage is in the feedback loop. I believe that I will need the ability to adjust compensation in order to achieve stabiity in this amp, which is why I brought up the 5534 and (so far) only the 5534 seems to allows this external 'tweak'.
Just wanting to know if there are any others (that might be acceptable for audio).
This opamp sits at the input of the amplifier section, and the whole driver+output stage is in the feedback loop. I believe that I will need the ability to adjust compensation in order to achieve stabiity in this amp, which is why I brought up the 5534 and (so far) only the 5534 seems to allows this external 'tweak'.
Just wanting to know if there are any others (that might be acceptable for audio).
EchoWars said:
Hi Echo
You can use a unity gain stable op amp and use a capacitor (5pF ...10 pF) between the output of the op amp and the input feedback node.
This way, you have feedback from the output of the amp in the audio frequencys and from the output of the Op amp in the Mhz region...This will increase stability...
Regards
I do have the schematics, and I have tried this on another identical amp some time ago with little luck (oscillated like crazy no matter what, and I gave up without battling too much),peranders said:
The original opamp in this Philips AH578 amp is a RCA CA3100S. Very fast, but the amp I was working on then had a dead one, and I was having extreme trouble finding another one, so I was going to find a way to make a OPA604 or NE5534 work. Both oscillated, and in the midst of my frustration, I found a source for a CA3001 (for $20
). I bought it, and called it a day.Now I have another 578, and although the CA3001's in it are functional, memories of being beaten by the other amplifier linger, and I wanted to resume where I left off.
Tube_Dude, I may do this, but I was hoping to use existing traces and component holes. No biggie, and I can do it as you describe. But my question here was as much to see if the 5534 was the only opamp to offer the external compensation pins as to find a solution.
Thanks guys!!
Hello!
I have attached a simplified schematic of the NE5534 circuit.
I have put in a few values of components
that I have been able to find.
I think it can be interesting to compare this circuit to other amplifiers.
The NPN input differential stage, with offset adjust possibility,
drives a dual PNP stage, which is brought together by a mirror.
So far is like many other amplifiers.
But the last stages are all NPN based.
There are no less than 4 small capacitors, that helps amp to do better.
100pF, 40pF, 7pF and 12pF.
This last 12 pF is the internal compensation cap.
This is recommended to be parallelled with external 22 pF ( Pin5 - Pin8 )
when using NE5534 as a buffer. Gain = 1.
I have attached a simplified schematic of the NE5534 circuit.
I have put in a few values of components
that I have been able to find.
I think it can be interesting to compare this circuit to other amplifiers.
The NPN input differential stage, with offset adjust possibility,
drives a dual PNP stage, which is brought together by a mirror.
So far is like many other amplifiers.
But the last stages are all NPN based.
There are no less than 4 small capacitors, that helps amp to do better.
100pF, 40pF, 7pF and 12pF.
This last 12 pF is the internal compensation cap.
This is recommended to be parallelled with external 22 pF ( Pin5 - Pin8 )
when using NE5534 as a buffer. Gain = 1.
Attachments
EchoWars said:
Neither the NE5534 or OPA604 have the required gain-bandwidth product, I suspect they have used the CA3100 because of it's speed - it was never a cheap part. However... have a look at another one from the olden days: the LM318. This is just off the top of my head, I may be dead wrong here - the LM318 is slower (but not that much slower). A schematic of the circuit would help a lot. In general external compensation on OPamps is not standardized, although it may involve pin 1 and 8, often different connections are required with a different OPamp. Also, you may want to look at the single OPamps in general. If it is only the OPamp itself that is being compensated, you may get away with a higher GBW product amp that is internally compensated.
LM6171 or AD817 would probably be usable subs for CA3100. CA3100 has fairly low Avol (60 - 70 dB) with 3db corner somewhere around 20kHz. 5534 has higher Avol and lower 3db corner, this may explain the stability problems, a small cap between -in and out as suggested earlier might help.
I came across a rail of about 40 CA3100s awhile back and have tried them in X5 noninverting, Zin=30k with input bias current compensation, they sound OK, a little dry in the bass/lower midrange, pretty good for ca. 1978.
I came across a rail of about 40 CA3100s awhile back and have tried them in X5 noninverting, Zin=30k with input bias current compensation, they sound OK, a little dry in the bass/lower midrange, pretty good for ca. 1978.
10 years ago, while consulting to Magnet in Thailand, I designed a nested chip driven version of their award winning 300F power amplifier. For those of you in Thailand, it was shown in the January '95 Audio Show in Bangkok. I attended.
During prior tests with a panel of 5 we auditioned a range of chips in the input drive position on this otherwise discrete power amplifier of 200W. Four chips were tested - NE5534, AD845,LF357, and OPA627. All gave similar distortion figures on the Audio Precision test set - <0.0025% at 20KHz and <0.001% at 1K.
But the differences were clearly audible. The NE5534 was, in fact the lowest THD + N of the lot but the sound was the worst. It was dull and lifeless. Totally uninspiring. The best was the OPA so we ran with that.
Unanimous.
For what it's worth.
Cheers,
Greg
During prior tests with a panel of 5 we auditioned a range of chips in the input drive position on this otherwise discrete power amplifier of 200W. Four chips were tested - NE5534, AD845,LF357, and OPA627. All gave similar distortion figures on the Audio Precision test set - <0.0025% at 20KHz and <0.001% at 1K.
But the differences were clearly audible. The NE5534 was, in fact the lowest THD + N of the lot but the sound was the worst. It was dull and lifeless. Totally uninspiring. The best was the OPA so we ran with that.
Unanimous.
For what it's worth.
Cheers,
Greg
There was a time when NE5534 was among the top of OP-amps.
But is a long long time (electronic IC time goes fast) since it was first released.
Would be strange if wasnt better OP-amps around now.
Surely are better computers around, than when NE5534 was introduced!
Still after so many years NE5534 is useful
and holds a very very high performance/price value.
Compared to some top class modern JFET OP-amps, always mentioned to be better than old NE5534,
NE5534 costs practically NOTHING.
TL071 is cheap, but how good?
Has anybody tried to run NE5534 with JFET input pair?
( connect pins 2+3 to negative rail to block, disable input transistors
add a differential JFET pair with a current source
and let them drive collector resistors 12 kOhm at pins 1+8
these resistors can be parallelled if you want higher current in input differential
they are connected from pins 1+8 to pin7 (= V+)
the default current is ~170uA giving ~2.0 Volt across collector resistors )
I would guess nobody has tried this, but it is possible to do.
What result would be, I have no idea.
But is a long long time (electronic IC time goes fast) since it was first released.
Would be strange if wasnt better OP-amps around now.
Surely are better computers around, than when NE5534 was introduced!
Still after so many years NE5534 is useful
and holds a very very high performance/price value.
Compared to some top class modern JFET OP-amps, always mentioned to be better than old NE5534,
NE5534 costs practically NOTHING.
TL071 is cheap, but how good?
Has anybody tried to run NE5534 with JFET input pair?
( connect pins 2+3 to negative rail to block, disable input transistors
add a differential JFET pair with a current source
and let them drive collector resistors 12 kOhm at pins 1+8
these resistors can be parallelled if you want higher current in input differential
they are connected from pins 1+8 to pin7 (= V+)
the default current is ~170uA giving ~2.0 Volt across collector resistors )
I would guess nobody has tried this, but it is possible to do.
What result would be, I have no idea.
Attachments
Interesting that this got dug up. Reminds me of working on those Philips disasters. While actually not a terrible design, every component in it (with the apparent exception of the expensive RCA opamp, and the large impressive heatsinks to draw the buyers in) was about the cheapest Philips could find. Other than being capable of a lot of power, as far as I'm concerned there was nothing to recommend it.
It is still likely tough to find an opamp with the distortion specs of the 5532/5534. When I bought my Tek SG505 signal generator, I wasn't suprised to see that it was 5534-based.
If I hadn't found the replacment RCA for the Philips, this thread very likely would have gone a lot longer...
ilimzn, I hadn't thought of the LM318. It might have worked, and was even available in a TO-5 metal can to pop right in place.
It is still likely tough to find an opamp with the distortion specs of the 5532/5534. When I bought my Tek SG505 signal generator, I wasn't suprised to see that it was 5534-based.
If I hadn't found the replacment RCA for the Philips, this thread very likely would have gone a lot longer...
ilimzn, I hadn't thought of the LM318. It might have worked, and was even available in a TO-5 metal can to pop right in place.
lineup said:
I first tried this in '87, later added more JFETS as bootstrapped cascode. Works OK but you're better off with OPA627.
NE5534 with noninverting feedback and fairly high source impedance tends to sound like that. I recently did some subjective comparison of opamps in X5 noninverting along with the CA3100, outputs buffered, with about 30k source impedance. I preferred 5534 to OPA637, the OPA637 offered good low-level detail but with some added glare that the 5534 didn't have, OTOH the 5534's low level detail was not as good but the overall balance seemed more natural. I preferred LM6171 to both of these, but that wasn't available in 1/95. 5534 like most OPAs is much better with inverting feedback and low source impedance, but this isn't always practical.
Gosh lucky you .....'oh yes I've done this.....got a copy of the Siliconix small signal Fet design cat July 1983 cat.. front App note AN81-3 (composite op amp for high performance ). I don't have copying facilities where I am, so You'l have to scate their website for details.
For those who criticise the 5534, this ic was nearly used by xxx millions in every console produced before digital scooped 'em up. It's still one of the only op amps that doesn't fuss with 600 ohms. The 5532 is compensated version perhaps too much OMO
richj
For those who criticise the 5534, this ic was nearly used by xxx millions in every console produced before digital scooped 'em up. It's still one of the only op amps that doesn't fuss with 600 ohms. The 5532 is compensated version perhaps too much OMO
richj
richwalters said:
richwalters
I couldnt find this App Note in the web.
Anybody got a link?
===============================================
Originally posted by lineup:
Has anybody tried to run NE5534 with JFET input pair?
( connect pins 2+3 to negative rail to block, disable input transistors
add a differential JFET pair with a current source
and let them drive collector resistors 12 kOhm at pins 1+8
these resistors can be parallelled if you want higher current in input differential
they are connected from pins 1+8 to pin7 (= V+)
the default current is ~170uA giving ~2.0 Volt across collector resistors )
I would guess nobody has tried this, but it is possible to do.
What result would be, I have no idea.
john curl said:
In principle, the AD797 should run rings around the 5534.
john curl
I say, it would be most strange if it didnt!
In my catalog,
for the price of 4 (four) AD797
I get 120 (hundred and twenty) Philips NE5534N
For the price of 4 (four) OPA627
I get 200 (two hundered) Philips NE5534N
Nice to hear you have used NE5534 with JFET inputs
and found it better sounding.
I have also use these (NE5534's) extensively in mixing desks way back then - and LM301's with feedforward.
Other than that the majority of audio opamp applications are non-inverting so these are far from the most transparent.
They're still excellent and cost effective in low noise reg power supplies.
Cheers,
greg
Other than that the majority of audio opamp applications are non-inverting so these are far from the most transparent.
They're still excellent and cost effective in low noise reg power supplies.
Cheers,
greg
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.