The fuse makes a difference.
I've read a lot of testimonials to that effect over the years, but was always very skeptical. It seems unreasonable that a tiny piece of wire that carries mains voltage could possibly be a factor in the perception of sonic qualities of an amplifier.
With my last order to Mouser, I figured, what the heck, it costs less than a buck each to try a ceramic fuse to replace the glass cartridges I'd been using since building my first MyRefs several years ago.
I almost forgot about them until last night. Dead of winter here, so there's plenty of leisure to fiddle with audio doodads. I put a new Schurter ceramic fuse in each monoblock and played music while eating dinner, not paying close attention. Later in the evening, I finally sat down in front of the speakers to soak up some sound. I played a CD I have heard many times, one I had listened to only a few days before, and was immediately aware of a certain clarity, depth of soundstage, and palpable imagery that had been present but not pronounced before. The sound wasn't really different, but it was definitely better, a refinement. I was completely surprised and still somewhat skeptical of what I was hearing. Played another CD, same results: leading edges of everything--highs, mids, bass--all more pronounced, without any increase in brightness. There was clearly more presence, depth, clarity, somewhat similar to the difference between a standard metal film resistor and a naked Z-foil.
So now you might think I'm as nuts as I thought other fuse proponents had been. If you can hear the difference in resistors, you probably will hear the improvement a sixty cent fuse can make. I am not prepared to spend twenty or more bucks on an audiophile certified fuse, but I think the modest outlay for a simple ceramic fuse is perhaps the best bargain I've ever heard.
Peace,
Tom E
I've read a lot of testimonials to that effect over the years, but was always very skeptical. It seems unreasonable that a tiny piece of wire that carries mains voltage could possibly be a factor in the perception of sonic qualities of an amplifier.
With my last order to Mouser, I figured, what the heck, it costs less than a buck each to try a ceramic fuse to replace the glass cartridges I'd been using since building my first MyRefs several years ago.
I almost forgot about them until last night. Dead of winter here, so there's plenty of leisure to fiddle with audio doodads. I put a new Schurter ceramic fuse in each monoblock and played music while eating dinner, not paying close attention. Later in the evening, I finally sat down in front of the speakers to soak up some sound. I played a CD I have heard many times, one I had listened to only a few days before, and was immediately aware of a certain clarity, depth of soundstage, and palpable imagery that had been present but not pronounced before. The sound wasn't really different, but it was definitely better, a refinement. I was completely surprised and still somewhat skeptical of what I was hearing. Played another CD, same results: leading edges of everything--highs, mids, bass--all more pronounced, without any increase in brightness. There was clearly more presence, depth, clarity, somewhat similar to the difference between a standard metal film resistor and a naked Z-foil.
So now you might think I'm as nuts as I thought other fuse proponents had been. If you can hear the difference in resistors, you probably will hear the improvement a sixty cent fuse can make. I am not prepared to spend twenty or more bucks on an audiophile certified fuse, but I think the modest outlay for a simple ceramic fuse is perhaps the best bargain I've ever heard.
Peace,
Tom E
I thought I had tracked down the problem, but I was wrong. Perhaps you guys will think of something I haven't.
The boards with the AD4627 had good DC offsets once I cleaned up the socket connections. So I tried to play music but I got a few seconds of clear music, then massive garbled distortion. To my ear, the distortion sounds like clipping, definitely not static or hum. Music is not possible to listen to and I rush to turn it off because it sounds like I am hurting something.
My first guess was that the amp was stable with no input, but begins oscillating with a music input. Not having any 5 pF caps, I ordered some and waited for delivery. Then life got in the way and I had to put the project aside for a while.
Today, I rechecked all the power supply voltages and reinstalled the LM318 boards to check them for hum and DC offset. They measured out perfectly and played nicely when back in the system.
On the AD4627 boards, I soldered a 5 pF silver mica across pins 2 and 6 of the opamp. Measuring the DC offset, one channel was 2 mV and the other was 90 mV. I rechecked the continuity and values of key feedback components finding no error. Since one channel had good measurements, I put it into the system with the other channel "on" but not connected to the speaker. I found the same behavior. For a few seconds, it sounds clear, then significant distortion. Also, when I first turn on the amp, there is no hum, but after playing music and getting distortion, a loud hum is coming through the speaker.
Since I clearly haven't fixed the distortion problem, I hooked up the other speaker at low volume to see if the channel with the 90 mV offset board was also distorted. The distortion is the same as the other board.
It appears that the capacitor across pins 2 and 6 caused the 90 mV offset, but there were several weeks between checking offset without the cap, so something could have changed. Note that I am still using C9, so the low frequency gain remains low.
My thoughts are tending two ways. I am using several resistors on sockets. It might be that the extra loop inductance from those resistors is causing the amp to be unstable. That might explain the distortion, but does not explain the DC offset, at least to me.
The other idea is that I hurt the opamp somehow during assembly. Unfortunately, these are the only AD4627 opamps that I have. I could replace them with OPA627 that were left from a different project. No OPA827 either.
Anything else I should try before scrapping the AD4627?
Thanks in advance.
Jac
The boards with the AD4627 had good DC offsets once I cleaned up the socket connections. So I tried to play music but I got a few seconds of clear music, then massive garbled distortion. To my ear, the distortion sounds like clipping, definitely not static or hum. Music is not possible to listen to and I rush to turn it off because it sounds like I am hurting something.
My first guess was that the amp was stable with no input, but begins oscillating with a music input. Not having any 5 pF caps, I ordered some and waited for delivery. Then life got in the way and I had to put the project aside for a while.
Today, I rechecked all the power supply voltages and reinstalled the LM318 boards to check them for hum and DC offset. They measured out perfectly and played nicely when back in the system.
On the AD4627 boards, I soldered a 5 pF silver mica across pins 2 and 6 of the opamp. Measuring the DC offset, one channel was 2 mV and the other was 90 mV. I rechecked the continuity and values of key feedback components finding no error. Since one channel had good measurements, I put it into the system with the other channel "on" but not connected to the speaker. I found the same behavior. For a few seconds, it sounds clear, then significant distortion. Also, when I first turn on the amp, there is no hum, but after playing music and getting distortion, a loud hum is coming through the speaker.
Since I clearly haven't fixed the distortion problem, I hooked up the other speaker at low volume to see if the channel with the 90 mV offset board was also distorted. The distortion is the same as the other board.
It appears that the capacitor across pins 2 and 6 caused the 90 mV offset, but there were several weeks between checking offset without the cap, so something could have changed. Note that I am still using C9, so the low frequency gain remains low.
My thoughts are tending two ways. I am using several resistors on sockets. It might be that the extra loop inductance from those resistors is causing the amp to be unstable. That might explain the distortion, but does not explain the DC offset, at least to me.
The other idea is that I hurt the opamp somehow during assembly. Unfortunately, these are the only AD4627 opamps that I have. I could replace them with OPA627 that were left from a different project. No OPA827 either.
Anything else I should try before scrapping the AD4627?
Thanks in advance.
Jac
Love ceramic fuses
I switched to ceramic fuses at least fifteen years ago. For amps I use SLO-Blo. For source equipment medium blow. The claim is less microphonics. Not sure, but they do sound better to me. BUT, have not tried a glass fuse in a long time for comparison.
I switched to ceramic fuses at least fifteen years ago. For amps I use SLO-Blo. For source equipment medium blow. The claim is less microphonics. Not sure, but they do sound better to me. BUT, have not tried a glass fuse in a long time for comparison.
Problem Solved - Old boards work with the new jfet opamps
With George's help, I was able to isolate the problem and it's something that I did to myself.
George, Thank you again. We are lucky to have you as part of this thread.
If you want to build an old board (mine was V1.2) with the new jfet opamps, it should be no problem.
In my system, I was powering a balanced line receiver from the regulated voltage on the FE board. This has been fine for years with the LM318, but the ADA4627 caused the op amp on the balanced line receiver to oscillate. Actually, it may have been either or both the balanced line receiver and the ADA that oscillates with music, but we will never know. The main thing is that a separate power supply for the balanced line receiver solved the distortion problem.
There are only two people that I know of that would be affected by this problem. Me and Tom E.'s friend where Tom used one of my balanced line receivers in the same way. Everyone using the standard, single ended input should not have this problem.
Dario can probably say for sure, but I understand that he improved the regulator on the v1.5 and later FE boards. Hopefully, that gives even more stability cushion when using the ADA4627.
I'm just getting started listening to the ADA boards. No conclusions yet, but they do sound nice so far.
Back to building a new power supply.
Jac
With George's help, I was able to isolate the problem and it's something that I did to myself.
George, Thank you again. We are lucky to have you as part of this thread.
If you want to build an old board (mine was V1.2) with the new jfet opamps, it should be no problem.
In my system, I was powering a balanced line receiver from the regulated voltage on the FE board. This has been fine for years with the LM318, but the ADA4627 caused the op amp on the balanced line receiver to oscillate. Actually, it may have been either or both the balanced line receiver and the ADA that oscillates with music, but we will never know. The main thing is that a separate power supply for the balanced line receiver solved the distortion problem.
There are only two people that I know of that would be affected by this problem. Me and Tom E.'s friend where Tom used one of my balanced line receivers in the same way. Everyone using the standard, single ended input should not have this problem.
Dario can probably say for sure, but I understand that he improved the regulator on the v1.5 and later FE boards. Hopefully, that gives even more stability cushion when using the ADA4627.
I'm just getting started listening to the ADA boards. No conclusions yet, but they do sound nice so far.
Back to building a new power supply.
Jac
Jac and others:
I just finished rebuilding my friend's amps (installing Evo compensation and putting them in proper enclosures) using your wonderful little converters, running off the regulated LM318 supply with no difficulties. These provide superb fidelity and an excellent opportunity to implement balanced cable input to the MyRef. If you ever consider getting boards made, I'd gladly buy a couple more just to have them around. I don't foresee changing out the LM318, and they could be handy in other applications.
Too bad they don't play nice with other opamps' power supply, so you need to build another complete supply to adequately isolate them from the new opamps? What a weird interaction. Hope it's worth it.
In my friend's amps, I have installed a DPDT switch (signal and return) to allow selection between single-ended and balanced inputs. The switch is high quality, but in either position puts a slight veil over the sound. I compared by bypassing the switch. I'm sure in many systems and with many speakers and to some listeners it would be inaudible. I actually prefer the single-ended connection but he hasn't decided. He spent a LOT of money on the balanced IC's and hates to not use them.
Peace,
Tom E
I just finished rebuilding my friend's amps (installing Evo compensation and putting them in proper enclosures) using your wonderful little converters, running off the regulated LM318 supply with no difficulties. These provide superb fidelity and an excellent opportunity to implement balanced cable input to the MyRef. If you ever consider getting boards made, I'd gladly buy a couple more just to have them around. I don't foresee changing out the LM318, and they could be handy in other applications.
Too bad they don't play nice with other opamps' power supply, so you need to build another complete supply to adequately isolate them from the new opamps? What a weird interaction. Hope it's worth it.
In my friend's amps, I have installed a DPDT switch (signal and return) to allow selection between single-ended and balanced inputs. The switch is high quality, but in either position puts a slight veil over the sound. I compared by bypassing the switch. I'm sure in many systems and with many speakers and to some listeners it would be inaudible. I actually prefer the single-ended connection but he hasn't decided. He spent a LOT of money on the balanced IC's and hates to not use them.
Peace,
Tom E
Hi Kim
You have 4 ampboards, are you planning to make a balanced version ?
I am asking as I have 4 boards unassembled with that (balanced and more power) strategy in mind.
That's great news! It's not clear why the change from the LM318 to the ADA4627 caused the oscillation. The LM318 actually takes a little more power, but the ADA4627 is a little faster.
I was getting about 90 mV of DC offset in one channel only when the balanced line receiver was powered and no input. Per George, that indicates that the balanced line receiver was oscillating even with no input. Both channels distorted badly with music using the ADA4627 and the balanced line receiver, so the power supply issue seems to be a bit worse with music. One possibility suggested by George is that an oscillation on the power supply lines may be powered by the inductance of the wires running between boards and the capacitors on the balanced line receiver.
George,
Anything to add?
[/QUOTE=madisonears;5335224]Too bad they don't play nice with other opamps' power supply, so you need to build another complete supply to adequately isolate them from the new opamps? What a weird interaction. [/QUOTE]
Actually, I'm happy to know what the problem is and how to fix it. It also gives me a chance to build a Jos regulator, my name for it, from the Reliaxed 2 pre-amp. It's perfect for a low current supply like this and gives a measured -120 dB noise.
[/QUOTE=madisonears;5335224]In my friend's amps, I have installed a DPDT switch (signal and return) to allow selection between single-ended and balanced inputs. I compared by bypassing the switch. I'm sure in many systems and with many speakers and to some listeners it would be inaudible. [/QUOTE]
I agree that going balanced isn't really needed in most home audio systems. I went that way because I was originally planning on having each amp channel and crossover sitting behind each speaker. That would make a long, line level IC run which would potentially benefit from the lower common mode noise of the balanced line. Looking back, single ended would probably have been fine. Besides, I have never gotten around to moving the amps out to the speakers.
It looks like my attempt at multi-quote didn't work well. Sorry. Still learning.
Jac
Last edited:
Overview of My Ref and Balanced input
It might be a good time to discuss balanced input and the My Ref amp. Although there are amps that are designed to be balanced input, the My Ref is designed for single ended input (you know, that RCA IC input). There a lots of ways to handle the balanced input with My Ref.
Double the Amps
It sound like Kolby and perhaps Kim are planning on turning the My Ref into a true balanced amp. In this case, they will use one My Ref for the '+' input and another My Ref for the '-' input. The result will be twice the power, balanced design all the way to the speaker terminals, and no need to convert the input to single ended. Very Sweet.
If you don't go that way, you need to convert the input signal from balanced to single ended and you have to make sure that converter is of high quality or you will lose any noise/sound quality improvement that you paid for with the balanced lines.
There are a couple of ways to convert balanced to single ended.
Transformer
The first is an input transformer. These are very nice and can maintain a high level of CMRR across a wide range of frequencies. They do have a few drawbacks. First, they are expensive. Typically $75 or more per channel. Also, they do limit bandwidth, some would say a little more than desirable. And they are a touch finicky. You have to be careful to balance input impedance with ringing of the transformer. And input impedance affects bandwidth. No free lunch here.
Active Converter
The second approach is an active approach, essentially an op-amp that measures the difference between the input '+' and '-' and outputs a single voltage to the single ended input. As I discovered, this needs a high quality power supply that should be isolated from other supplies. You also need to be careful to have matched impedance all the way through. The homemade boards that Tom and I used are the simplest type of balanced line receiver with just a single high quality op-amp and relatively low impedance resistors (1k).
There are other, more complicated designs and IC designed for this purpose. I felt like I could get better specs with discrete precision resistors and a simple design, but some of the more complicated versions (THAT 1200 series, for example) provide bootstrapped CMRR and RF filtering.
After playing with this stuff for a few years, my opinion is that balanced lines are great for pro's using very long IC runs, but not really worth the trip for home audio. But if you find yourself with a nice piece of balanced equipment and want to integrate it with your single ended amps, take the time to learn about converting to single ended and make sure you put together something that is high quality.
Jac
It might be a good time to discuss balanced input and the My Ref amp. Although there are amps that are designed to be balanced input, the My Ref is designed for single ended input (you know, that RCA IC input). There a lots of ways to handle the balanced input with My Ref.
Double the Amps
It sound like Kolby and perhaps Kim are planning on turning the My Ref into a true balanced amp. In this case, they will use one My Ref for the '+' input and another My Ref for the '-' input. The result will be twice the power, balanced design all the way to the speaker terminals, and no need to convert the input to single ended. Very Sweet.
If you don't go that way, you need to convert the input signal from balanced to single ended and you have to make sure that converter is of high quality or you will lose any noise/sound quality improvement that you paid for with the balanced lines.
There are a couple of ways to convert balanced to single ended.
Transformer
The first is an input transformer. These are very nice and can maintain a high level of CMRR across a wide range of frequencies. They do have a few drawbacks. First, they are expensive. Typically $75 or more per channel. Also, they do limit bandwidth, some would say a little more than desirable. And they are a touch finicky. You have to be careful to balance input impedance with ringing of the transformer. And input impedance affects bandwidth. No free lunch here.
Active Converter
The second approach is an active approach, essentially an op-amp that measures the difference between the input '+' and '-' and outputs a single voltage to the single ended input. As I discovered, this needs a high quality power supply that should be isolated from other supplies. You also need to be careful to have matched impedance all the way through. The homemade boards that Tom and I used are the simplest type of balanced line receiver with just a single high quality op-amp and relatively low impedance resistors (1k).
There are other, more complicated designs and IC designed for this purpose. I felt like I could get better specs with discrete precision resistors and a simple design, but some of the more complicated versions (THAT 1200 series, for example) provide bootstrapped CMRR and RF filtering.
After playing with this stuff for a few years, my opinion is that balanced lines are great for pro's using very long IC runs, but not really worth the trip for home audio. But if you find yourself with a nice piece of balanced equipment and want to integrate it with your single ended amps, take the time to learn about converting to single ended and make sure you put together something that is high quality.
Jac
Actually more like 4 times the power, if the load is not too low. Double the voltage out results in four times the power.
Balanced My_Ref
I have considered making a balanced version. Do not have any two amplifiers the are identical.
My main reasoning is my dac/preamps are balanced output only. Been shorting ground to - to go single ended. Not exactly SOTA.
I have a box of Cinimag input transformers left from a bunch of preamp projects. Considered shoe horning a pair into a Full Evo. But have not found the time. Also concerned about EMI and RFI levels at the rear of the chassis. But to downgrade the Sonics.
Would also consider bypassing the input cap to see if a transformer would keep the unit stable. Never liked using an input transformer directly into an AC coupled input.
I have considered making a balanced version. Do not have any two amplifiers the are identical.
My main reasoning is my dac/preamps are balanced output only. Been shorting ground to - to go single ended. Not exactly SOTA.
I have a box of Cinimag input transformers left from a bunch of preamp projects. Considered shoe horning a pair into a Full Evo. But have not found the time. Also concerned about EMI and RFI levels at the rear of the chassis. But to downgrade the Sonics.
Would also consider bypassing the input cap to see if a transformer would keep the unit stable. Never liked using an input transformer directly into an AC coupled input.
If you want to convert balanced to single ended, download the AD797 datasheet. Near the end, they have an app note for a "differential line receiver". That's what I have used with success. Of course, powering it separately, even if it's only it's own regulator, should be considered.
I was going to comment on another project I did with DC coupled input transformers, but on second thought, it was tubes and that's a very different animal.
Jac