I've been giving the LM4562NA a good thorough listening for almost a week now.
My verdict is in. the LM4562NA is very good as a MM opamp. I enjoy thern more than my hybrid.
The only downside is using the LM4562NA with MC selected, I could hear the distinct sound of an opamp being pushed too hard.
It sounded like this:
tick tick tick tick..............tick tick.... tick tick tick...... tick tick tick.
Out of all the opamp I have tried out (about 18 something pairs now I think) this is the third pair to exhibit this type of behavior at higher gain.
In short LM4562NA are pretty hard to beat for MM. But there are better options for MC.
My verdict is in. the LM4562NA is very good as a MM opamp. I enjoy thern more than my hybrid.
The only downside is using the LM4562NA with MC selected, I could hear the distinct sound of an opamp being pushed too hard.
It sounded like this:
tick tick tick tick..............tick tick.... tick tick tick...... tick tick tick.
Out of all the opamp I have tried out (about 18 something pairs now I think) this is the third pair to exhibit this type of behavior at higher gain.
In short LM4562NA are pretty hard to beat for MM. But there are better options for MC.
Have you tried upgrading the phono output caps and the main output caps? I heard a unit with Mundorf Supremes, just the basic Supremes, and the line stage sounded pretty good. I still preferred my tube phono stage to the phono in the 8500, but I don't use headphones.
The 8500 really sounded sweet when I put in a set of Telefunken smooth plates.
One thing you have to watch, depending on your line voltage, is the voltage on the tube filaments. If it drifts too high you might want to change out the filament dropping resistor to keep it below 12.6 for maximum tube life.
Tube collection isn't that extensive. I did the noise testing 'cause I was selling a bunch off and didn't want to stiff people with noisy or microphonic tubes.
S.
The 8500 really sounded sweet when I put in a set of Telefunken smooth plates.
One thing you have to watch, depending on your line voltage, is the voltage on the tube filaments. If it drifts too high you might want to change out the filament dropping resistor to keep it below 12.6 for maximum tube life.
Tube collection isn't that extensive. I did the noise testing 'cause I was selling a bunch off and didn't want to stiff people with noisy or microphonic tubes.
S.
I got the Mundorf Supreme on the phono and output, and NOS
Mullard Blackburn smooth plates. And I got NOS Telefunken smooth plate and GE 5 start waiting in line once I've depleted the Mullard.
That is very honest! I got stiffed on my first pair of telefunken from Ebay, they were not good. I've only been buying from reputable stores like Viva tubes, etc. since.
I never heard it with the 2227P.
I googled "LM4562 popping" and "LM4562 tick" last night and other folk are reporting the same problem. I couldn't hear it at less than 50% volume, this is with no signal going through the opamp, increasing the volume past 50 made it moe audible.
But in my past searches when I first encountered this problem was that the opamp has become unstable, who knows what it's doing to the signal going through the opamp when it's unstable like that.
Another thing I read last year was from an engineer saying that they have to screen some of the LM series opamps in a test circuit before using them due to some of them having strange issues that can show up for what they were using it for..
who knows... maybe some batches come out a little funky from batch to batch.
I just bought some Muses01 off Mouser
I remember these got some serious hate when they first came out due to the cost,(including me) But when you are using unconventional materials and manufacturing technics, that are too costly for mass production it can get expensive, As what those techniques are, they seem to be a trade secret. very common in the semiconductor industry.
They should be here Saturday
I remember these got some serious hate when they first came out due to the cost,(including me) But when you are using unconventional materials and manufacturing technics, that are too costly for mass production it can get expensive, As what those techniques are, they seem to be a trade secret. very common in the semiconductor industry.
They should be here Saturday
Ah............ then yes, I have heard it with the 4562. I used the TU-8500 as an MC phono-stage to drive one of the inputs to the SLP-98P. To do so I had to turn both up to about 60% to get an adequate volume. Thought the 'tick' was an oscillation due to a mismatch between the two. I've not had to turn the volume up above 11 - 12 o'clock when using the TU-8500 on its own, even at the 1.1x gain, and no tick there. Most interesting. I'll try again with the 2227's back in the system at some point.
Will look forward to your report on the Muses01 performance. At the price, they should be good.
Will look forward to your report on the Muses01 performance. At the price, they should be good.
Last edited:
🤔 Maybe that's what causes the oscillation? The volume going up changes lowers the output impedance causing a mismatch. which probably means no oscillation at lower volumes.. Just a guess after some more on other forums. I haven't had the chance to hook up thhe scope, I've been busy enjoying the LM4562.
That's what I'm looking forward to finding out. Specs-wise, they don't seem too special except for a couple of things, Like a really low bias current What little material I have found, mostly all in Japanese, NJR put a lot of time and effort in the development the Muses.
That's what I'm looking forward to finding out. Specs-wise, they don't seem too special except for a couple of things, Like a really low bias current What little material I have found, mostly all in Japanese, NJR put a lot of time and effort in the development the Muses.
Well, the MUSES01 have arrived!
Normally UPS doesn't get to my house until later in the day, so I was very surprised that they arrived this early in the morning, exactly when my coffee finished brewing.
Im going to be writing this as I go through the listening process.
After swapping out the LM4562's for the MUSES01's, I'm going to let my TU-8500 warm up for about 30 minutes.
Before beginning my listening session I wanted to check out the noise using the same method as in my previous post. The MUSES01 measure at 56dB, That is 1dB over the stock JRC2068, and 2dB under the OPA2227. So its a very low subtle hiss with my HD800, at my normal listening gain, and my ear has to be against the speaker to hear it. So Noise performance is spot on! Starting my listening session now.
I just finished side A of "Seven + Mary" by RKS. This is going to be difficult to describe. It has turned the TU-8500 into something else. The way that the MUSES01 amplify and transmit sound, could be best compared to what barrels do to whiskey after 30 years of aging, it smooths and widens the flavors, it brings out vanilla with hints tobacco, and salted caramel. The Muses01 sound is smooth, delicate and high in resolution, with a wide sound stage. But that's not even the best part of MUSES01, the dimension is overwhelmingly good.
I think Im done opamp rolling! The MUSES01 are staying for good, out of all the opamp I have tried in the TU-8500, these are the best! If you can afford them I highly recommend these. This entire listening session has been like the moment I powered on my TU-8600 for the first time and music flowed out of it and I discovered a new reference in sound. Just mind blowing! I don't know how these would sound in other circuits, but In the TU-8500, they are in perfect harmony. I can't get over just how good they are and I'm trying my best not to over hype these.
I went down the rabbit hole on Japanese blogs and audio sites during the listening session in search of what sets the MUSES01 apart from your everyday opamp. I'll have to write about what I found out later. These are not ordinary opamp by any measure. There's a lot of information about the design that for some reason didn't make it into translation.
There's one big takeaway from this. You cannot tell how an opamp is going to sound through a spec sheet, The engineers at NJR knew exactly what was needed for a pure audio only opamp. I feel like dumb passing these up years because my thought was "you can get better spec opamps for less". The MUSES01 are proof that those specs won't equate into better sound. You need the right tool for the job.
Normally UPS doesn't get to my house until later in the day, so I was very surprised that they arrived this early in the morning, exactly when my coffee finished brewing.
Im going to be writing this as I go through the listening process.
After swapping out the LM4562's for the MUSES01's, I'm going to let my TU-8500 warm up for about 30 minutes.
Before beginning my listening session I wanted to check out the noise using the same method as in my previous post. The MUSES01 measure at 56dB, That is 1dB over the stock JRC2068, and 2dB under the OPA2227. So its a very low subtle hiss with my HD800, at my normal listening gain, and my ear has to be against the speaker to hear it. So Noise performance is spot on! Starting my listening session now.
I just finished side A of "Seven + Mary" by RKS. This is going to be difficult to describe. It has turned the TU-8500 into something else. The way that the MUSES01 amplify and transmit sound, could be best compared to what barrels do to whiskey after 30 years of aging, it smooths and widens the flavors, it brings out vanilla with hints tobacco, and salted caramel. The Muses01 sound is smooth, delicate and high in resolution, with a wide sound stage. But that's not even the best part of MUSES01, the dimension is overwhelmingly good.
I think Im done opamp rolling! The MUSES01 are staying for good, out of all the opamp I have tried in the TU-8500, these are the best! If you can afford them I highly recommend these. This entire listening session has been like the moment I powered on my TU-8600 for the first time and music flowed out of it and I discovered a new reference in sound. Just mind blowing! I don't know how these would sound in other circuits, but In the TU-8500, they are in perfect harmony. I can't get over just how good they are and I'm trying my best not to over hype these.
I went down the rabbit hole on Japanese blogs and audio sites during the listening session in search of what sets the MUSES01 apart from your everyday opamp. I'll have to write about what I found out later. These are not ordinary opamp by any measure. There's a lot of information about the design that for some reason didn't make it into translation.
There's one big takeaway from this. You cannot tell how an opamp is going to sound through a spec sheet, The engineers at NJR knew exactly what was needed for a pure audio only opamp. I feel like dumb passing these up years because my thought was "you can get better spec opamps for less". The MUSES01 are proof that those specs won't equate into better sound. You need the right tool for the job.
Last edited:
I first found out about MUSES 8 or 9 years ago, and at first glance I thought the MUSES were just another audio gimmick, a cheaper rebranded opamp sold for 30 times the price. They are not! Having them in my Tu-8500 have left me with abig impression. I think it's time for a deeper look into what sets these apart from the rest.
So what makes the MUSES01 different from everyday opamp? Well quite a few things, I was able to find more information through online press releases and information put out by NJR at trade shows going back to 2007. Pretty hard to find since it was 13 or more years ago, there were a lot of dead links, and wayback machine was of little help.
From a press release from 2007, Announcing the start of production of the NJM5720 (which is the oldest reference I could find) states that development of the first Muses OPAMP "NJM5720" (Later renamed MUSES01) started in 2005 , and the first prototype of many, appeared in 2006. The only way to get a hold of a NJM5720 sample was to be a Manufacturer, and from what I can tell, only in Japan. Pricing for a sample at this time cost 3000 yen (about $30 in today's money) per chip. A request had to be submitted with NJR to get one. It was like that for several years. Early details of thur sound quality slowly started to make their way into audio forums, I'm guess from people who worked for manufacturers, who got to listen to the samples, this fueled a lot of hype and skepticism due to the price, and difficulty of getting one, but that made the NJM5720 even the more mysterious since nobody except manufacturers could get their hands on it in the early days.
The NJM5720 was the result of "I'f cost isn't a constraint, what kind of audio opamp can we make?" thinking. The 2007 release states that NJR's goal was to get the sound of a discreet opamp in the form and benefits of an IC.
Let's break down its features to compare to the common opamp.
The same 2007 press release of the NJM5720 says that it uses 2 seperate die, one for each channel. This was and still is a very game changing design even by today's standards, No other company used this technique in opamp before, then, and even today 13 years later, all current opamp are on a single die, NJR is the only one using this method, NJR says that is how they achieved the 150dB plus of channel separation, NJR also attributes this construction to sound you get from the NJM5720.
This here is a single die dual channel opamp LM358, all dual channel opamp look something like this on a single substrate. Remember the red circles for later
I couldn't find a die picture of the NJM5720, but I did sort of find one in a MUSES03 announcement, which uses the same dual die construction of the other MUSES, the only difference is that the MUSES03 is single channel, the Gain stage is on a separate die from the output stage, rather than separate channels per die like the NJM5720, its same concept but different execution.
The next feature is die size, The 2007 press release says the the die size of the NJM5720 is 3 times bigger than any other dual channel opamp die. I think this here is the real magic behind the Muses series opamp. production wise, a bigger die means less yield per wafer, traditionally the goal is to keep the die as small as possible to yield more chips from a wafer. With the NJM5720 they are yielding 3 times less per silicon wafer due to it being 3 times the size. But what this does means is that you have more real estate to make features that would otherwise be impossible on a smaller die, now NJR has been tight lipped about the actual construction of the die, they only stated that they are using "new and advanced techniques", I feel this is their bread and butter! For example capacitors are probably one of the biggest features on any chip design because they take up so much space. (see red circles on the LM358 die) since capacitance has a relation to the surface area, I imagine they are capable of making bigger and better capacitors with the extra space, and this also allows them to use different materials for other devices and features on the die, that would also take up more space. Again, they have been tight lipped, so you can only speculate what they are doing with 3 times the extra substrate space.
Another feature mentioned is the use of oxygen free copper for the lead frame and bonding, so what is this? let's start with the lead frame. The lead frame is where finished die are attached to a stamped metal pads and leads that provides its ground (die are grounded through the substrate) and their leads that provide connections to the outside world from inside the package, aka pins. I found this neat picture to share
How is a OFC lead frame different? well for starters the majority of leads frames used are made of a Iron and Nickel alloy (the cheapest option) due to its strength, then sometimes a copper alloy (expensive), NJR uses pure OFC with tinned leads for the NJM5720, they state that they wanted to use the same materials used in speaker wire, this does make the pins easier to bend, extra care needs to be taken when handling.
Next is bonding, extremely small wires that connect on pads on the die to the pins. Like in this picture. NJR states that they are using OFC here as well, again they wanted to use the same material used in speaker wire. Copper, at the same diameter when used in bonding can transfer higher currents and is a stronger material when compared to gold. here's an example of what bonding wires look like.
Gold has been used for longest time because it's the easiest to work with. During the development of the NJM5720, copper bonding was still under research, due to the increasing cost of gold vs copper, It was mostly used in incredibly rare situations for bonding. Due to the difficulty and expense of using copper bonding, you need a special environment for it free of oxygen to prevent corrosion caused by heating during building, plumbing gas lines alone can run up to a few million in cost for a manufacturer switching over from gold to copper. you can read more about it here.
Wire bonding - Wikipedia. As of 2015 only about ⅓ of IC were using copper bonding.
In the end of the 2007 press release NJR states that they have started development of a Bipolar input version of the NJM5720, that will later be named the MUSES02.
Fast forward to May of 2009, NJR announces that after receiving heavy requests for several years from the public to make the NJM5720 available for consumer purchase, They state that its under consideration, and if it happens it will become available later in 2009 to the general public. They also announce that the NJM5720, will be renamed the MUSES01, and the Bipolar version, The MUSES02 will be released later that year.
December of 2009 I started finding the first Japanese reviews, blog posts, and forum posts of the MUSES01 and MUSES02. They were in high demand and still hard to get a hold of since going on sale publicly. I even found one guy's blog post where he took a day trip to Tokyo to buy a pair.
Sometime in 2010 NJR announced a production (cheaper) versions of the MUSES01, and MUSES02, the MUSES8820 and MUSES8832. These use more commonly used materials for the lead frame and bonding, and who knows what else to lower cost, But I did find a rumor on one forum saying that these are single die but with a big empty space of silicon between channels, I could not verify this, so it's only rumer, These are suitable for use in high volume manufacturing of audio gear. It is around this time that I start seeing posts about MUSES01 and MUSES02 on American and Canadian forums and blogs. And just like their Japanese audio enthusiast counterparts. The price was shocking. It wasn't until around 2012/2013 that the MUSES became easier to get a hold of for Western customers.
I hope this helps shed some light on how the MUSES series opamps differ from conventional opamp, and why the price is steep, The MUSES are like the Ferrari of the OPAMP world, more has gone into it, especially under the hood, Yeah you can buy a brand new car for $15,000 and laugh at the idea of people who spend over $250,000 on a Ferrari, they essentially have the same things, power windows, a radio, 4 wheels, an engine, and they can both get you from point A to point B. But the experience getting there will be completely different!
So what makes the MUSES01 different from everyday opamp? Well quite a few things, I was able to find more information through online press releases and information put out by NJR at trade shows going back to 2007. Pretty hard to find since it was 13 or more years ago, there were a lot of dead links, and wayback machine was of little help.
From a press release from 2007, Announcing the start of production of the NJM5720 (which is the oldest reference I could find) states that development of the first Muses OPAMP "NJM5720" (Later renamed MUSES01) started in 2005 , and the first prototype of many, appeared in 2006. The only way to get a hold of a NJM5720 sample was to be a Manufacturer, and from what I can tell, only in Japan. Pricing for a sample at this time cost 3000 yen (about $30 in today's money) per chip. A request had to be submitted with NJR to get one. It was like that for several years. Early details of thur sound quality slowly started to make their way into audio forums, I'm guess from people who worked for manufacturers, who got to listen to the samples, this fueled a lot of hype and skepticism due to the price, and difficulty of getting one, but that made the NJM5720 even the more mysterious since nobody except manufacturers could get their hands on it in the early days.
The NJM5720 was the result of "I'f cost isn't a constraint, what kind of audio opamp can we make?" thinking. The 2007 release states that NJR's goal was to get the sound of a discreet opamp in the form and benefits of an IC.
Let's break down its features to compare to the common opamp.
The same 2007 press release of the NJM5720 says that it uses 2 seperate die, one for each channel. This was and still is a very game changing design even by today's standards, No other company used this technique in opamp before, then, and even today 13 years later, all current opamp are on a single die, NJR is the only one using this method, NJR says that is how they achieved the 150dB plus of channel separation, NJR also attributes this construction to sound you get from the NJM5720.
This here is a single die dual channel opamp LM358, all dual channel opamp look something like this on a single substrate. Remember the red circles for later
I couldn't find a die picture of the NJM5720, but I did sort of find one in a MUSES03 announcement, which uses the same dual die construction of the other MUSES, the only difference is that the MUSES03 is single channel, the Gain stage is on a separate die from the output stage, rather than separate channels per die like the NJM5720, its same concept but different execution.
The next feature is die size, The 2007 press release says the the die size of the NJM5720 is 3 times bigger than any other dual channel opamp die. I think this here is the real magic behind the Muses series opamp. production wise, a bigger die means less yield per wafer, traditionally the goal is to keep the die as small as possible to yield more chips from a wafer. With the NJM5720 they are yielding 3 times less per silicon wafer due to it being 3 times the size. But what this does means is that you have more real estate to make features that would otherwise be impossible on a smaller die, now NJR has been tight lipped about the actual construction of the die, they only stated that they are using "new and advanced techniques", I feel this is their bread and butter! For example capacitors are probably one of the biggest features on any chip design because they take up so much space. (see red circles on the LM358 die) since capacitance has a relation to the surface area, I imagine they are capable of making bigger and better capacitors with the extra space, and this also allows them to use different materials for other devices and features on the die, that would also take up more space. Again, they have been tight lipped, so you can only speculate what they are doing with 3 times the extra substrate space.
Another feature mentioned is the use of oxygen free copper for the lead frame and bonding, so what is this? let's start with the lead frame. The lead frame is where finished die are attached to a stamped metal pads and leads that provides its ground (die are grounded through the substrate) and their leads that provide connections to the outside world from inside the package, aka pins. I found this neat picture to share
How is a OFC lead frame different? well for starters the majority of leads frames used are made of a Iron and Nickel alloy (the cheapest option) due to its strength, then sometimes a copper alloy (expensive), NJR uses pure OFC with tinned leads for the NJM5720, they state that they wanted to use the same materials used in speaker wire, this does make the pins easier to bend, extra care needs to be taken when handling.
Next is bonding, extremely small wires that connect on pads on the die to the pins. Like in this picture. NJR states that they are using OFC here as well, again they wanted to use the same material used in speaker wire. Copper, at the same diameter when used in bonding can transfer higher currents and is a stronger material when compared to gold. here's an example of what bonding wires look like.
Gold has been used for longest time because it's the easiest to work with. During the development of the NJM5720, copper bonding was still under research, due to the increasing cost of gold vs copper, It was mostly used in incredibly rare situations for bonding. Due to the difficulty and expense of using copper bonding, you need a special environment for it free of oxygen to prevent corrosion caused by heating during building, plumbing gas lines alone can run up to a few million in cost for a manufacturer switching over from gold to copper. you can read more about it here.
Wire bonding - Wikipedia. As of 2015 only about ⅓ of IC were using copper bonding.
In the end of the 2007 press release NJR states that they have started development of a Bipolar input version of the NJM5720, that will later be named the MUSES02.
Fast forward to May of 2009, NJR announces that after receiving heavy requests for several years from the public to make the NJM5720 available for consumer purchase, They state that its under consideration, and if it happens it will become available later in 2009 to the general public. They also announce that the NJM5720, will be renamed the MUSES01, and the Bipolar version, The MUSES02 will be released later that year.
December of 2009 I started finding the first Japanese reviews, blog posts, and forum posts of the MUSES01 and MUSES02. They were in high demand and still hard to get a hold of since going on sale publicly. I even found one guy's blog post where he took a day trip to Tokyo to buy a pair.
Sometime in 2010 NJR announced a production (cheaper) versions of the MUSES01, and MUSES02, the MUSES8820 and MUSES8832. These use more commonly used materials for the lead frame and bonding, and who knows what else to lower cost, But I did find a rumor on one forum saying that these are single die but with a big empty space of silicon between channels, I could not verify this, so it's only rumer, These are suitable for use in high volume manufacturing of audio gear. It is around this time that I start seeing posts about MUSES01 and MUSES02 on American and Canadian forums and blogs. And just like their Japanese audio enthusiast counterparts. The price was shocking. It wasn't until around 2012/2013 that the MUSES became easier to get a hold of for Western customers.
I hope this helps shed some light on how the MUSES series opamps differ from conventional opamp, and why the price is steep, The MUSES are like the Ferrari of the OPAMP world, more has gone into it, especially under the hood, Yeah you can buy a brand new car for $15,000 and laugh at the idea of people who spend over $250,000 on a Ferrari, they essentially have the same things, power windows, a radio, 4 wheels, an engine, and they can both get you from point A to point B. But the experience getting there will be completely different!
Wow, thanks for all the great feedback and info., so what about tubes in the 8500? I started with the Tungsol 12at7 and then the Sylvania JHS 12at7 (TubeDepot.com | Early 1950’s JHS Sylvania 12AT7 - Military Stock) which I'm listening to now. Never even tried the stock tubes. I'm thinking with the OPA2227's and these tubes, I'm getting some really great sound at budget prices?
I don't think you can go wrong with any NOS tubes in the Tu-8500, I haven't had any tubes I didn't like. but I do have favorites.
I've been interested in some Sylvania blackplates, or even some Mazda tubes as well, I've read a lot of good things about them.
I tried the stock tubes once just to get a feel for them, don't remember much about thern since it's been a while, but they weren't bad. Im using NOS Blackburn Mullard now, My favorite so far, I used NOS Telefunken for a a week and they sounded very nice, very open at the top end.
The GE 5963 five star was very present pleasant too, even though it has a little less gain.
The Burson discreet opamp were on my list to test, but after the Sparkos testing I realized it would be a waste of money. Not because they are bad products, far from that. It's just the wrong application. Both Burson and Sparkos full discreet opamp have too high of a bias current for Phono use. You need an opamp under 3uA of bias current, after that it comes down to the cartridge you are using.
Unlike Sparkos, Burson does sell a hybrid which on paper looks like it would be great for a MM cartridge. it's the Burson V5i hybrid, made of off the shelf IC with a couple of discreet passive on V+ and V-, I don't know which IC they are using, I found out by asking someone who owns a pair of they wouldn't mind removing the metal casing, all information on the IC had been sanded off. But Burson says it It has a JFET input with a current mirror main amplification stage.
There's only one problem.. the right channel on the TU-8500 (double check the schematic) is swapped. Meaning that instead of using Pins 1 through 4 is the input. It is instead using Pins 5 through 8. It will cause the right channel to sound different.
You can either use an adaptor that swaps inputs, or if your handy with a soldering iron charge pin out.
Hi all,
Just a quick reply. I replaced the coupling caps as well as some resistors by Kiwame and followed the suggestion from Partsconnexxion by replacing the NJM2068 DD for Sparkos SS3602. This was a disaster with great hum.
IIn view of my experience with the Sparkos I do not know if the MUSES01 offers an upgrade.
Best regards,
Robert
Just a quick reply. I replaced the coupling caps as well as some resistors by Kiwame and followed the suggestion from Partsconnexxion by replacing the NJM2068 DD for Sparkos SS3602. This was a disaster with great hum.
IIn view of my experience with the Sparkos I do not know if the MUSES01 offers an upgrade.
Best regards,
Robert