LM4562NA/NOPB Texas Instruments | Mouser
Note that the quantity required for that price ($1.11) is 100. But I think it's the best bargain in the history of DIY audio!!
Good points. My inclination towards sockets is the years of troubleshooting where a socketed IC is a god-send, especially on a PC board that's ~40 years old and probably won't take TOO much soldering on. I'm thinking that at this point, the Mill-Max sockets with gold flash are good enough. I really don't plan on doing any swapping; if I WAS planning that, I would go with the thicker gold. But I think the flash will withstand a few changes, and I need the $$ for the hundreds of capacitors I'm gonna need. Now, for you working stiffs, $25 is nothing; but for us old retired farts on a small fixed income, it IS relevant.
I am retired, not by choice either. I'm basically up a creek, you know the one.
What I would do if I were you is just solder the darned things in except for the early stages where you may want to try another op am type for lower noise. I don't think there is any point in replacing many of the NE5532 chips (does it use those? I'm making an assumption).
I used to work on tape machines professionally where some users wanted to improve performance once the gear was out of warranty. Of course, even under warranty was okay as long as they brought the machines in to us. It was safer to wait for the warranty to expire and that is what we recommended.
-Chris
What I would do if I were you is just solder the darned things in except for the early stages where you may want to try another op am type for lower noise. I don't think there is any point in replacing many of the NE5532 chips (does it use those? I'm making an assumption).
I used to work on tape machines professionally where some users wanted to improve performance once the gear was out of warranty. Of course, even under warranty was okay as long as they brought the machines in to us. It was safer to wait for the warranty to expire and that is what we recommended.
-Chris
No, the Otari I have used a HA12017 for the head preamp and NJM4559s for everything else. If it DID have NE5532s, I'd leave 'em be, as I still think that is a fine opamp. But NJM4559s are really only slightly improved 741s (which suck), and the Hitachi chip, although fairly low noise, is slow, has horrible PSRR, and an order of magnitude more distortion than an LM4562.
Mine have a mix of 2043, 4560 and an oddball from Matsushita in the MX-50 that I have not identified other than that it's a bipolar type rather similar to the 4560. They might be a wee bit better than what dotneck found in his machine but not by much. They are all deratives of the 741. The 5532 is probably a substantial improvement parametrically speaking, whether this results in any audible improvement is hard to guess. I'm surprised at how good the MX-50N actually is.
I think the 4560 is a FET front end, isn't it? Nope, the 4560 is an improved 4558 (dual 741). There was a Japanese op amp in this series that was J-Fet. I just can't remember which one it is.
By all means, pop every single one of those out! It will make a difference with the entire lot changed out.
-Chris
By all means, pop every single one of those out! It will make a difference with the entire lot changed out.
-Chris
The MKIII-2 is a comparatively easy test bed. I'm feeling really leery now about doing the op-amp mod on the MX-50N and that is because it's by far the best sounding tape deck I've ever owned and it easily outperforms my 5050s so I am starting to wonder if I might just actually succeed in making it sound worse. It is modified by only a little - the original machine did not have Muse caps in it, other than that and one simple mod to the LF comp circuit it is completely stock.
Hi Kevin,
When I modify a piece of equipment, I generally find that the highs are more "clear". If you look at a good FFT, the junk in the mid to high frequency ranges is greatly diminished. Personally, I don't feel you are in any danger of making your machine sound worse. You will probably reduce the background noise a bit, the distortion numbers will come down a little bit (perhaps even by 6 dB). There was one bit of kit where the distortion improvement was 20 dB, but that was low hanging fruit. A power amp designed and built by a modifier who used to work at Hafler. Some of you will know who I mean.
In short, I doubt very strongly that you would make it sound worse by applying sensible modifications (maybe) and using parts more suited to the task that we have today. I do think that you know what you are doing and this little bit of self doubt will completely evaporate once you hear your updated machine. In other words, just approach this as you would a good customer's machine and you will be a happy man. Besides, you have an RTX. Do before measurements, then after each major change to ensure you remain on track. You will then be able to attribute each improvement (or not) with the things you changed.
-Chris
When I modify a piece of equipment, I generally find that the highs are more "clear". If you look at a good FFT, the junk in the mid to high frequency ranges is greatly diminished. Personally, I don't feel you are in any danger of making your machine sound worse. You will probably reduce the background noise a bit, the distortion numbers will come down a little bit (perhaps even by 6 dB). There was one bit of kit where the distortion improvement was 20 dB, but that was low hanging fruit. A power amp designed and built by a modifier who used to work at Hafler. Some of you will know who I mean.
In short, I doubt very strongly that you would make it sound worse by applying sensible modifications (maybe) and using parts more suited to the task that we have today. I do think that you know what you are doing and this little bit of self doubt will completely evaporate once you hear your updated machine. In other words, just approach this as you would a good customer's machine and you will be a happy man. Besides, you have an RTX. Do before measurements, then after each major change to ensure you remain on track. You will then be able to attribute each improvement (or not) with the things you changed.
-Chris
Interesting....WHY do you think that it sounds better? Does it have superior heads? Or is there something else in the audio circuit path that makes it better?
It might be that it has less rather than more, the signal path is much simpler, and yes I suspect the heads are significantly better. The transport may play a role as well or I could just be all wet and in love with this deck's particular colorations.. LOL It's not perfect and doesn't make copies indistinguishable from the original.
I still have some additional work to do on this particular deck given its history. I am thinking I should probably replace the reel motor bearings as well as those in the rollers. (The supply reel roller is a definite.) Strangely enough I have a pair of MX-50 reel motors so I could just swap, but who knows if they need bearings? LOL
Interestingly the heads are in good shape, I got lucky. Most of the damage may have occurred in storage. I know a little of the history of the deck, but not how it got damaged.
I suspect Chris is right, but unlike the MKIII-2 this deck requires significant disassembly to modify so while the inclination would be to do a little at a time and test, it's a time sink and increases the probability of damaging something.
I still have some additional work to do on this particular deck given its history. I am thinking I should probably replace the reel motor bearings as well as those in the rollers. (The supply reel roller is a definite.) Strangely enough I have a pair of MX-50 reel motors so I could just swap, but who knows if they need bearings? LOL
Interestingly the heads are in good shape, I got lucky. Most of the damage may have occurred in storage. I know a little of the history of the deck, but not how it got damaged.
I suspect Chris is right, but unlike the MKIII-2 this deck requires significant disassembly to modify so while the inclination would be to do a little at a time and test, it's a time sink and increases the probability of damaging something.
I just received the cal tapes a couple of days ago so I will do a thorough demag and do some measurements after that when I feel motivated.
As you might suspect I don't do a lot of measurements unless I think something is broken or I am designing something new in which case I do a lot as you know. I'm really pretty lazy and work a lot of hours at job that frankly wears me out. So forgive me for playing at not being an engineer..LOL Sometimes it is just fun not to obsess about all the things one is supposed to do.
On a more cynical note there isn't that much to learn, by their nature R2R machines suck, the fact that we are gifted with such imprecise hearing mechanisms and the distortions generated are complementary means we derive pleasure from devices that come nowhere close to reproducing the original signal put into them.
The maximum SNR achieved is under 70dB typically and only when distortion levels approach 3%. Tape distortions and head behavior is complex and not very linear. I sometimes find I prefer the sound of the imperfect analog copy to the digital master it was made from. (that says something huh - what is it adding? ) Any decent 16 bit 48kHz sound card is going to run circles around this or any other decent deck. Most vintage R2R are unlistenable IMO, I got rid of a very low hours ReVox G36 MKIII because even after an overhaul I thought it sounded mediocre. (My M-Audio 2496 at the time made much better sounding transfers of analog material)
So I have not invested a lot of time in investigating these machines performance potential. The MKIII I did some basic measurements and it seemed to meet the minimums with some margin to spare. SNR was not great, thd was not great, and the effects of mechanical modulations on single tones was visible in single tone fft.
I'm not doing much with vinyl either these days, I've entered the cynicism phase of my existence. (Or I am suffering from pretty bad burnout.. )
As you might suspect I don't do a lot of measurements unless I think something is broken or I am designing something new in which case I do a lot as you know. I'm really pretty lazy and work a lot of hours at job that frankly wears me out. So forgive me for playing at not being an engineer..LOL Sometimes it is just fun not to obsess about all the things one is supposed to do.
On a more cynical note there isn't that much to learn, by their nature R2R machines suck, the fact that we are gifted with such imprecise hearing mechanisms and the distortions generated are complementary means we derive pleasure from devices that come nowhere close to reproducing the original signal put into them.
The maximum SNR achieved is under 70dB typically and only when distortion levels approach 3%. Tape distortions and head behavior is complex and not very linear. I sometimes find I prefer the sound of the imperfect analog copy to the digital master it was made from. (that says something huh - what is it adding? ) Any decent 16 bit 48kHz sound card is going to run circles around this or any other decent deck. Most vintage R2R are unlistenable IMO, I got rid of a very low hours ReVox G36 MKIII because even after an overhaul I thought it sounded mediocre. (My M-Audio 2496 at the time made much better sounding transfers of analog material)
So I have not invested a lot of time in investigating these machines performance potential. The MKIII I did some basic measurements and it seemed to meet the minimums with some margin to spare. SNR was not great, thd was not great, and the effects of mechanical modulations on single tones was visible in single tone fft.
I'm not doing much with vinyl either these days, I've entered the cynicism phase of my existence. (Or I am suffering from pretty bad burnout.. )
Spent 2 weeks and replaced every single electrolytic capacitor in that deck. There are some films that might be fair game but where do I hit the point of diminishing or no returns? I know one or two spots where a better capacitor in an EQ circuit will probably make a difference.
The readings will be probably not change in any discernible way, the distortion mechanisms in recording and playback are already orders of magnitude worse than the electronics in this deck. It may end up sounding different and I am going to do something having invested in the op-amps, really only way to know is to try it.
Is a change in thd from 0.05% to 0.005% significant when the tape generates 5 - 6x that distortion at normal levels.
My last experience swapping OP285 with LM4562 resulted in a measurable improvement in performance but overall worse sound. I should probably be sticking OP285 in these things! LOL
Hi Kevin,
Look at the spectrum of the noise floor. That's where you will find your answers, not in the numbers.
In Nakamichi cassette decks, you'll find something that sounds better than most other branded machines. Look a the BX-300 as an example. Have gander of not only the head amps, but also at the local power supply regulation. Even when you're playing tape, the Nak decks sound better and have lower noise levels.
Part of the recording chain include the bias traps and how different op amps deal with HF energy on their output pins. Then you have various FET input op amps that resist RF interference better than BJT models. Don't forget that the feedback loop will route that HF energy right to the diff pair of the op amp.
-Chris
Look at the spectrum of the noise floor. That's where you will find your answers, not in the numbers.
In Nakamichi cassette decks, you'll find something that sounds better than most other branded machines. Look a the BX-300 as an example. Have gander of not only the head amps, but also at the local power supply regulation. Even when you're playing tape, the Nak decks sound better and have lower noise levels.
Part of the recording chain include the bias traps and how different op amps deal with HF energy on their output pins. Then you have various FET input op amps that resist RF interference better than BJT models. Don't forget that the feedback loop will route that HF energy right to the diff pair of the op amp.
-Chris
Man, if you really hear a difference between those fine opamps:
A) Your ears are definitely better'n mine!
B) You REALLY should put the 285s in!
Another good point for sockets, BTW.
The two are pretty similar in noise specs @ 3KHz source Z; distortion of both is VERY low; both are fast (≥20v/µV/sec). But the 285 is a LOT more expensive!
Maybe the Butler front end of the 285 makes for a more pleasing sound stage?
OK. I looked at the BX-300 service manual--NJM072 in the PB head path; NJM4558 in the record path. Nothing very special there. Couldn't spy the power supply regulation. No noise floor spectrum shown---could you provide this? I'd like to know what you're rappin' here.
Certainly you're not comparing the sound of a cassette deck to a 15ips open reel machine---?
I wouldn't. 
I heard something because I was definitely unimpressed with the op-amp swap. This was quite a while ago and in my Lambda Ultralink II which was subsequently rather easily vanquished by a $250 Zhaolu Dac somewhat more than a decade newer.
(I don't really get the fascination with "vintage" digital, the modern kit is so much better sounding IMO.)
The brown dog adapters have not shown up thus far which is frustrating since the op-amps showed up 5 days ago from DigiKey. So no experiments just yet..
I heard something because I was definitely unimpressed with the op-amp swap. This was quite a while ago and in my Lambda Ultralink II which was subsequently rather easily vanquished by a $250 Zhaolu Dac somewhat more than a decade newer.
(I don't really get the fascination with "vintage" digital, the modern kit is so much better sounding IMO.)
The brown dog adapters have not shown up thus far which is frustrating since the op-amps showed up 5 days ago from DigiKey. So no experiments just yet..