I had an amp kit once that had a 1458 op amp in the mag phono stage. That damn thing was gone so fast…. Replaced with an LF353.
With the original op amp in there it sounded like the phono needed a new stylus it was so bad on HF sibilance. The hisssss wasnt any worse than the little DJ mixer I had with 2SC458’s in it. And a few of those started sounding like lightning and thunder eventually. 2SC1815’s to the rescue there.
With the original op amp in there it sounded like the phono needed a new stylus it was so bad on HF sibilance. The hisssss wasnt any worse than the little DJ mixer I had with 2SC458’s in it. And a few of those started sounding like lightning and thunder eventually. 2SC1815’s to the rescue there.
2SC1775A or 2SC2240 are also excellent choices.
You actually had a kit with 1458's in it for mag phono? Wow! How old was that kit? A-mazing!
You actually had a kit with 1458's in it for mag phono? Wow! How old was that kit? A-mazing!
This was back in like 1979-80. 1458 phono pre, discrete Baxandall bass/mid/treble (ha ha, the only transistors in it), STK modules. All on one PCB powered by a 48V transformer.
Why the C1815? Pack of 100 for a buck.
Why the C1815? Pack of 100 for a buck.
Last edited:
Just looked at the datasheet for the LM1458 and the schematic is pretty much the standard 3-stage audio power amp schematic with higher value resistances and slightly different bias scheme and OS... Inputs aren't Darlington and the input currents are large - so a lot of current noise (not ideal for MM). Wonder if the sibilance was lack of proper decoupling? Oscillating amps respond to high frequency inputs with a lot of IMD products.
I would have bought tons of them for a buck! Mine came direct from Japan, I knew an importer that travelled there.
Not a whole lot wrong with the 2SC1815, I still use what I have left. I'm out of the others.
Not a whole lot wrong with the 2SC1815, I still use what I have left. I'm out of the others.
Most of those garden variety op amps followed the same plan. Die layout and all kinds of things factored in. The common denominator was that they were ... cheap. They were pounded out quickly, not especially carefully.
LM1458 was a dual uA741 with the same dismal slew rate 0.5V/uS. The TI data sheet does not even list the slew rate but if you google it, google finds it. The TI data sheet shows an "equivalent circuit" that has the uA741 IPS where NPN emitters drive PNP emitters. This circuit was used for early chips that had very poor PNP transistors. I have a copy of Walt Jung's 1978 book "Audio IC OP-Amp Applications" which charts 7 audio performance characteristics of the chips of that time. I should hope there is a newer version, but this book is a seminal work and a must for anyone using op-amps for audio. But that was 40 years ago, so as I posted before, just go to a vendor web site and browse their "audio op-amps". If you look them up at ~Digikey etc. you can compare prices, and Digikey (etc) will list / compare different brands.
https://en.wikipedia.org/wiki/Walt_Jung
https://en.wikipedia.org/wiki/Walt_Jung
All of the sibilant sounds made a horrible “scratch” sound like it was being played through a rectifier. It WAS slew limiting.Just looked at the datasheet for the LM1458 and the schematic is pretty much the standard 3-stage audio power amp schematic with higher value resistances and slightly different bias scheme and OS... Inputs aren't Darlington and the input currents are large - so a lot of current noise (not ideal for MM). Wonder if the sibilance was lack of proper decoupling? Oscillating amps respond to high frequency inputs with a lot of IMD products.
Hi Steve,
I wore that book out, and subsequent publications from Walt. Application notes are the other best source of information.
Anyway, I wrote those off for audio use years ago.
I wore that book out, and subsequent publications from Walt. Application notes are the other best source of information.
Anyway, I wrote those off for audio use years ago.
H
HAYK
The RC4558 can swing 25v with 15v rails up to 30khz. Say that people hear slew rate saturation used in preamp is psychological defect.
The combination of non-linear open-loop and low slew rate produces distortion. The op-amp does not slew rapidly enough to traverse the crossover notch without distortion.
Ed
Ed
Sure it can. It begins slew rate limiting just above that. OLG has dropped to approx. 45 dB at around 30 KHz, so distortion is increasing. That means total circuit distortion is increasing and that people can hear just fine.
You specified the RC device number. Now do some work and look up some others. The Raytheon was 2.5 MHz unity gain. Designers use worst case, also older data sheets from TI may show a slower part with less good specs. Those were in effect at the time. Parts did improve over time.
I don't know how much experience you have, but some folks here have over 40 years in the industry and were using these devices when they were considered good (or not too bad). So don't get all cocky.
Now we are considering a headphone driver or amp. The op amp will probably be loaded heavily. It will need all the open loop gain it can muster. Keep the application in mind. The 4558 is a poor selection for one, and the OP wanted to improve it. So I have to ask, what is your point? Do you simply want to argue to enjoy a debate?
-Chris
You specified the RC device number. Now do some work and look up some others. The Raytheon was 2.5 MHz unity gain. Designers use worst case, also older data sheets from TI may show a slower part with less good specs. Those were in effect at the time. Parts did improve over time.
I don't know how much experience you have, but some folks here have over 40 years in the industry and were using these devices when they were considered good (or not too bad). So don't get all cocky.
Now we are considering a headphone driver or amp. The op amp will probably be loaded heavily. It will need all the open loop gain it can muster. Keep the application in mind. The 4558 is a poor selection for one, and the OP wanted to improve it. So I have to ask, what is your point? Do you simply want to argue to enjoy a debate?
I want you to think about this HAYK, you can attack the idea, but not the person. We do not allow that here.Say that people hear slew rate saturation used in preamp is psychological defect.
-Chris
FWIW I was using 741 in 1971-72, when they were the new kid on the block, so it adds up to 50 years 😱 .I don't know how much experience you have, but some folks here have over 40 years in the industry and were using these devices when they were considered good
There was nothing else widely available (remember I am in Argentina), so we had to work with it, like it or not.
Usable in general (my field is Guitar amps and related stuff, not Home Hi Fi or Recording Studios) but hiss was a deal breaker in early transformerless balanced mic inputs in stage mixers so I had to include a dual PNP balanced input stage to take care of the critical first 10-12dB of channel strip gain.
Hiss went from distracting/annoying to "you can hear it in the background IF you pay attention" which was acceptable.
As of slew rate limiting, it only bit me once: I had designed a new powered Mixer with a simplified gain structure, it worked fine with tests at home, but on the first night it was used in a Rock show at a Theater, meaning it was pushed, voice sibilants and cymbal crashes were accompanied by a background sound similar to tearing 2 Velcro strips apart, very ugly.
It also seemed "noise gated".
Next day at the shop, driving it full blast, found that sinewaves at 10kHz and above became triangular 🙁
To avoid throwing design and already built stuff into a dumpster, I made it usable by lowering signal levels at hot points along signal chain by some 10-12dB and added a 10dB gain discrete miniboard at the output.
A band-aid, does not seem much of an improvement, but now signal could plain squarewave clip against the rails at any frequency within the Audio band, Mixer was now "usable".
Junking the whole lot would have bankrupted me at that early stage 🙁
H
HAYK
Administrators can't be contradicted.Hi HAYK,
That's a simulation. Pure fantasy.
Unsubscibed
741 is used in temperature controllers, I have some with AD707 as well, the relatively polling interval (and simple signal) makes it a very good choice.
And when I made a pre amp for a friend, I used a 8 pin socket, checked the quality with a 358, and a 4558 before settling on 072.
358 was very dull, I used the Eagles song 'Hotel California', as it has some great guitar and drum solos, for comparing the op amp performance.
The first few seconds of Steve Winwood's track 'Higher Love' also has a drum roll, useful for testing.
Mark above seems to have found the issue, maybe some person has extra stock!...
And when I made a pre amp for a friend, I used a 8 pin socket, checked the quality with a 358, and a 4558 before settling on 072.
358 was very dull, I used the Eagles song 'Hotel California', as it has some great guitar and drum solos, for comparing the op amp performance.
The first few seconds of Steve Winwood's track 'Higher Love' also has a drum roll, useful for testing.
Mark above seems to have found the issue, maybe some person has extra stock!...
To clarify HAYK's comment ...
I would hope that any experienced member is afforded the same amount of respect a moderator or administrator might expect. We're no different.
If we have to post in an official capacity, you will normally see this --->
We make our mode very clear.
Sure we can. We post as normal members using our personal experience. We just have extra work to do, that's all.Administrators can't be contradicted.
I would hope that any experienced member is afforded the same amount of respect a moderator or administrator might expect. We're no different.
If we have to post in an official capacity, you will normally see this --->

We make our mode very clear.
Hi JMFahey,
lol!
I was a kid experimenting back then. I used them also, and we used what we could afford. Most of our parts came from stripping dead items for parts. I've got almost 50 years professionally. Some folks here have longer.
Very good solution to your slew problems in the mixer. When using parts not up to the task, you are forced to be creative and think creatively. Back then, almost no one had parts up to the task and even transistors were all over the map.
Today we have it so good with part quality and great test equipment! Don't kid yourself, you can buy junk easily, but good semi's and good test gear is easily obtained. It just costs! lol!
Hi NareshBrd,
Yes, the 741 is still used in industry for applications like that. It is well behaved in most situations.
lol!
I was a kid experimenting back then. I used them also, and we used what we could afford. Most of our parts came from stripping dead items for parts. I've got almost 50 years professionally. Some folks here have longer.
Very good solution to your slew problems in the mixer. When using parts not up to the task, you are forced to be creative and think creatively. Back then, almost no one had parts up to the task and even transistors were all over the map.
Today we have it so good with part quality and great test equipment! Don't kid yourself, you can buy junk easily, but good semi's and good test gear is easily obtained. It just costs! lol!
Hi NareshBrd,
Yes, the 741 is still used in industry for applications like that. It is well behaved in most situations.
If you lightly load an opamp (10k?) its output stage may stay in class A for the normal signal range (the first mW ?). Its takes a good output stage to keep distortion under control when driving low loads like 500 ohms, but many opamps have pretty good OS's these days and can handle it. Most opamps have specs/graphs for 10k and 2k loads, some have them for loader load impedances too. Generally a high quiescent current allows smoother transition in the OS, for bipolar opamps.
Rail-to-rail outputs can be shockingly bad for distortion if you take the output close to the rails - there can be crossover artifacts there as well as for zero output current. Check the datasheet carefully - you often get either good distortion figures or full rail-to-rail swing, but not both.
Rail-to-rail outputs can be shockingly bad for distortion if you take the output close to the rails - there can be crossover artifacts there as well as for zero output current. Check the datasheet carefully - you often get either good distortion figures or full rail-to-rail swing, but not both.
- Home
- Amplifiers
- Solid State
- Improving 4558 sound?