Have not seen this thread in a while!
Just thought I would stop in an say how much I enjoy my amps. They are stock still and sound great. I am currently using them with the Direct Coupled B1 buffer with Shunted PSUs or DCB1 for short. WOW this is a potent combination. I would head on over and check it out if you haven't already. Adding this will be the best "tweak" yet
Just thought I would stop in an say how much I enjoy my amps. They are stock still and sound great. I am currently using them with the Direct Coupled B1 buffer with Shunted PSUs or DCB1 for short. WOW this is a potent combination. I would head on over and check it out if you haven't already. Adding this will be the best "tweak" yet
Last edited:
I use several different digital sources, a PC, an iRiver 120, and a CAL Audio transport through an AMB Labs Gamma 1 (soon to be 2) DAC. My speakers are NHT Super Ones or Polk Monitor Series 5. Before the DC-B1 there I used a conrad johnson PV-7a and a Diamante.
The difference is considerable. This is not the B1 from Nelson Pass although the buffer circuit is the same, this is direct coupled, and features a purpose made shunt regulator. Hers is the thread. Also, I am using a Lightspeed attenuator in place of the pot. That being said... Wow the music just shines with this combo. I have to say of this that I do not want to stop listening, it is one of the few components I have heard that I forget that I am listening to my gear and I listen to the music.
The difference is considerable. This is not the B1 from Nelson Pass although the buffer circuit is the same, this is direct coupled, and features a purpose made shunt regulator. Hers is the thread. Also, I am using a Lightspeed attenuator in place of the pot. That being said... Wow the music just shines with this combo. I have to say of this that I do not want to stop listening, it is one of the few components I have heard that I forget that I am listening to my gear and I listen to the music.
Got around to taking some measurements
For those who care about such things... I took one of the boards of my very first build and measured it in a mono configuration (just one channel) using the excellent ARTA software. Bottomline here is that thing measures really well. The one thing that I continue to have (minor, really; but it bugs me no less) trouble with is the rectifier hash. It is at -100dB on the output but it is there in all it spectral glory for many many MANY harmonics to go... ARRRRGH.
Below is the overall distortion at -5dB input (my measurement chain; going to -4dB on a pure sine wave makes it go into clipping; boy... NASTY waveform) and a picture of the noise spectrum with no input. All the spikes are neatly arranged at 60Hz intervals. Still at -100dB (or there abouts). My measurement chain (if the loopback test is to be believed is at -130dB noise floor).
More pictures at Picasa Web Albums - Peter - RevC Measurem...
Peter
For those who care about such things... I took one of the boards of my very first build and measured it in a mono configuration (just one channel) using the excellent ARTA software. Bottomline here is that thing measures really well. The one thing that I continue to have (minor, really; but it bugs me no less) trouble with is the rectifier hash. It is at -100dB on the output but it is there in all it spectral glory for many many MANY harmonics to go... ARRRRGH.
Below is the overall distortion at -5dB input (my measurement chain; going to -4dB on a pure sine wave makes it go into clipping; boy... NASTY waveform) and a picture of the noise spectrum with no input. All the spikes are neatly arranged at 60Hz intervals. Still at -100dB (or there abouts). My measurement chain (if the loopback test is to be believed is at -130dB noise floor).
More pictures at Picasa Web Albums - Peter - RevC Measurem...
Peter
Attachments
Last edited:
Good old noise issues
As mentioned in the previous post I fiddled again with the first mono block I had put together. I don't understand the noise issues...
The board has a star ground setup so that one only needs to connect the toroid (I have a toroid with a center tapped secondary and two primaries which are run in parallel). First I had the mains earth not connected anywhere (I fully realize the safety implications and used appropriate laboratory etiquette to avoid injury). I had this loud hum on the output. This was with a twisted pair of short input wires (8cm) just coming off the board into mid air. That was just a receiver antenna for this type of hum. Getting my hand near it it increased. Ok. So, I short the input and the hum goes down, but not by much. It still is quite loud. Since the board has a star ground integrated and I add nothing further I am quite certain this is not a loop problem.
First question: Why, oh, why? With shorted input, why do I hear the hum? Looking at the scope it's basically 120Hz and up, i.e., rectifier hash.
Now I got it down to what you see in the previous post by connecting mains ground to the center tap on the secondary (the rest being star grounded this was the only connection to mains earth). That made the hum go away (well, pretty much).
I have plenty of equipment (commercial...) which is doubly insulated, i.e., has no mains earth connection and I can't hear any rectifier hash from it (digital sources, amps, etc.). What is is that these people do that I don't seem to understand? This is really getting my underwear in a bundle. Is there some kind of text book that explains this once and for all in a scientifically verified manner? I find lots of stuff on the net which is often contradictory, so I am ready for the official answer...
Grrrrrh.
peter
PS: I have one other self built amp that is free of this problem in a doubly insulated environment (there is tiny bit of rectifier hash on the output, but with no mains earth connection). It's Michael Renardson's MJR7-mk3. The one difference I can see is that it uses a single supply rail. Does that change the picture?
As mentioned in the previous post I fiddled again with the first mono block I had put together. I don't understand the noise issues...
The board has a star ground setup so that one only needs to connect the toroid (I have a toroid with a center tapped secondary and two primaries which are run in parallel). First I had the mains earth not connected anywhere (I fully realize the safety implications and used appropriate laboratory etiquette to avoid injury). I had this loud hum on the output. This was with a twisted pair of short input wires (8cm) just coming off the board into mid air. That was just a receiver antenna for this type of hum. Getting my hand near it it increased. Ok. So, I short the input and the hum goes down, but not by much. It still is quite loud. Since the board has a star ground integrated and I add nothing further I am quite certain this is not a loop problem.
First question: Why, oh, why? With shorted input, why do I hear the hum? Looking at the scope it's basically 120Hz and up, i.e., rectifier hash.
Now I got it down to what you see in the previous post by connecting mains ground to the center tap on the secondary (the rest being star grounded this was the only connection to mains earth). That made the hum go away (well, pretty much).
I have plenty of equipment (commercial...) which is doubly insulated, i.e., has no mains earth connection and I can't hear any rectifier hash from it (digital sources, amps, etc.). What is is that these people do that I don't seem to understand? This is really getting my underwear in a bundle.
Is there some kind of text book that explains this once and for all in a scientifically verified manner? I find lots of stuff on the net which is often contradictory, so I am ready for the official answer... Grrrrrh.
peter
PS: I have one other self built amp that is free of this problem in a doubly insulated environment (there is tiny bit of rectifier hash on the output, but with no mains earth connection). It's Michael Renardson's MJR7-mk3. The one difference I can see is that it uses a single supply rail. Does that change the picture?
Hi,
the hum you get with both open circuit and shorted input does not seem right to me. It seems to indicate a grounding problem.
I have tested most of my amplifiers without a Safety Earth connection and even without a chassis. They ALWAYS perform quieter than when connected to mains Earth and quieter than when in a steel chassis.
Yes, your first test without a mains Earth connection should give you the quietest test output.
I find <<0.1mVac at the output in test condition; shorted input, without chassis and without mains earth connection.
I find this output noise nearly always increases to between 0.2mVac to 0.7mVac when fully encased in a Safety Earth connected steel chassis. I try to get this <0.3mVac but in a two channel amp I find this difficult. A monoblock nearly always goes to <0.1mVac but never as quiet as an almost totally silent test arrangement.
the hum you get with both open circuit and shorted input does not seem right to me. It seems to indicate a grounding problem.
I have tested most of my amplifiers without a Safety Earth connection and even without a chassis. They ALWAYS perform quieter than when connected to mains Earth and quieter than when in a steel chassis.
Yes, your first test without a mains Earth connection should give you the quietest test output.
I find <<0.1mVac at the output in test condition; shorted input, without chassis and without mains earth connection.
I find this output noise nearly always increases to between 0.2mVac to 0.7mVac when fully encased in a Safety Earth connected steel chassis. I try to get this <0.3mVac but in a two channel amp I find this difficult. A monoblock nearly always goes to <0.1mVac but never as quiet as an almost totally silent test arrangement.
Peter, Interesting stuff I think to probably all of us. Thanks for the information and the experiments.
I know I said it before and got all sorts of trouble because the Twisted Pear boys said NOT to do this but I got rid of my hum by experimenting with ground. I know you didnt say you had hum, but grounding to get rid of hum might solve more than one problem at once.
I drilled a hole in my chassis and ran a bolt through it. On that bolt I tied earth ground. Then tied 0V from toroid as well as running that 0V to the board. I ran my pot ground to signal ground. Tied all speaker grounds to each other and tied the two RCA grounds to each other. Tied speaker and signal grounds together and speaker grounds then went to the original bolt in the chassis.
Quiet.
Maybe its technically wrong but I dont think it is and it worked. Oh, I used a 400VAC rated cap in series with earth to chassis ground. I forget the value of the cap.
So you turned your amp into a distance measuring device!?
I was working on a picoammeter the other day and it turns the current into volts you can measure with more accuracy than the amps. I could start from the farthes point in the room, about 14ft, and walk toward the setup. The voltage would steadily increase the closer I got and stop and decrease if I backed up. Really cool.
Uriah
I know I said it before and got all sorts of trouble because the Twisted Pear boys said NOT to do this but I got rid of my hum by experimenting with ground. I know you didnt say you had hum, but grounding to get rid of hum might solve more than one problem at once.
I drilled a hole in my chassis and ran a bolt through it. On that bolt I tied earth ground. Then tied 0V from toroid as well as running that 0V to the board. I ran my pot ground to signal ground. Tied all speaker grounds to each other and tied the two RCA grounds to each other. Tied speaker and signal grounds together and speaker grounds then went to the original bolt in the chassis.
Quiet.
Maybe its technically wrong but I dont think it is and it worked. Oh, I used a 400VAC rated cap in series with earth to chassis ground. I forget the value of the cap.
So you turned your amp into a distance measuring device!?
I was working on a picoammeter the other day and it turns the current into volts you can measure with more accuracy than the amps. I could start from the farthes point in the room, about 14ft, and walk toward the setup. The voltage would steadily increase the closer I got and stop and decrease if I backed up. Really cool.
Uriah
Sorry
Basically all signal grounds are tied and then go to star. Speaker grounds are included with signal grounds. Earth goes to star. 0V references star with small wire from board where 0V goes to board.
And that 400VAC cap is series from earth to star.
Uriah
All you need to do is tie the 0V point (where the transformer connects to the board) to the mains ground (earth) point and you have a star ground, as all the other grounds are starred on the board. Creating a separate star actually raises the risk of a ground loop (because you have two stars).
If you are going to do this, I would recommend using a ground "breaker" on the connection to mains ground, such as two high-current diodes wired back to back, to prevent current flow unless there are a few volts of potential, such as during a fault (there are a few ways to do this).
Yep. Couldn't agree more. I have performed some more experiments today...
Exactly what I am doing. I have a wood chassis just to hold things together (I am prototyping a custom enclosure I want to build).
What bandwidth? I did another set of tests today and apparently a little more carefully than yesterday... ;( Turns out the connection to mains earth doesn't make a difference. The real difference is in an open input (nothing at all connected to the short twisted wires leading to the input point on the board) and a shorted input. The open one produces about 500mV p-p 120Hz on the output (see attached scope photo). When shorting the input I get Mhz type noise in the 20mV (p-p) range. It's "white" when listening to it through a speaker (there is no input pot or anything here).
Sooooo. The diagnostic at this point (and implied question) is: I get nasty 120Hz (and high order harmonics) on the output with an open input. I suppose the cure to this is to put a high Ohm resistor across the input.
I will rerun my spectrum analysis to make sure nothing has changed (my spectrum analyzer, i.e., my portable computer, is in my office...).
A fraction of a mV? Wow! At audio frequencies? How do you measure that? My scope picks up that much noise from the air...
So my assertion yesterday that I can only get rid of the high volume hum when making a connection to mains earth was wrong (at least today I can't replicate the problem... very embarrassing
)Peter
Attached are scope traces with open input and with shorted input. This is ch1 on my scope; horizontal and vertical resolution are stated below. This is measured across a 8 Ohm high watt resistor on the output.
Attachments
I wish I had a scope like that.
But is it misleading you.
The noise bandwidth should be restricted to ~20Hz to 20kHz using low pass and high pass filters.
I cannot do that, I just use my DMMs (both have a 200.0mVac scale) which probably has a 10Hz to 500Hz bandwidth one of which I know still reads quite well to 50kHz, but is probably 10dB to 20dB down at that end of the frequency range. I am therefore reading Hum and LF noise.
Try clipping probe and earth leads across a 50r resistor or even just to each other.
But is it misleading you.
The noise bandwidth should be restricted to ~20Hz to 20kHz using low pass and high pass filters.
I cannot do that, I just use my DMMs (both have a 200.0mVac scale) which probably has a 10Hz to 500Hz bandwidth one of which I know still reads quite well to 50kHz, but is probably 10dB to 20dB down at that end of the frequency range. I am therefore reading Hum and LF noise.
surely that is with the probe open circuit.
Try clipping probe and earth leads across a 50r resistor or even just to each other.
Last edited:
Dear Andrew,
in what sense is it misleading? The high amplitude trace with open input shows high frequency crap AM modulated with 120Hz. The result with shorted input shows something outside the audio range (though what I hear on the loudspeaker is "white" [probably more "pink"] noise).
You mean to learn what the scope itself contributes?
peter
Haven't posted here in quite some time, so here is an update for those still interested.
I could not repair my non-working amp. A friend who knows his way around a circuit had a look at it and could not find the fault. I replaced both chips and several other parts, and finally succeeded in wrecking the board, so I gave up. I bought another kit from mindcrime (thanks again) and I'm starting over with duplicating my working monblock with half of the new kit, using the more valuable components over again.
The one that works is extraordinary. Another friend who is also an audiophile with more money than I have has a pair of Herron M-150 monblocks and Thor speakers (with very expensive, high quality SEAS drivers), all Siltech wire, outrageously expensive TT and preamp. I took my one working amp to his house and we replaced one of his amps with My Ref. His jaw almost hit the floor, and he didn't know what to say. The My Ref totally destroyed the $3,000 (each!) Herron in spatiality and musicality. The Herron had more detailed, perhaps even etched highs (to the point of what I consider brightness, but that's a matter of taste) and deeper, better defined bass, but My Ref owned the midrange and imaging. Instruments and voices had a tonality and three dimensional quality that was very realistic. I admit it sounded better in his system than it does in mine, and that was with only one amp. It was not even close, and both of us were thoroughly impressed, he even a little embarassed.
Now, that wasn't my objective, but it did serve to vindicate my efforts and expense to optimize this amp's performance. That WAS my goal. I was beginning to suffer doubts about taking all this trouble to build these things and get them to work. In the end, I believe all this effort of buying and building so many kits (I'm on my fifth one!) and using expensive components will be worthwhile.
Those of you who are using this amp in a compromised system have no idea what you're missing. It is truly worthy of being attached to the best quality components, but I'm sure it sounds good with any combination.
Peace,
Tom E
I could not repair my non-working amp. A friend who knows his way around a circuit had a look at it and could not find the fault. I replaced both chips and several other parts, and finally succeeded in wrecking the board, so I gave up. I bought another kit from mindcrime (thanks again) and I'm starting over with duplicating my working monblock with half of the new kit, using the more valuable components over again.
The one that works is extraordinary. Another friend who is also an audiophile with more money than I have has a pair of Herron M-150 monblocks and Thor speakers (with very expensive, high quality SEAS drivers), all Siltech wire, outrageously expensive TT and preamp. I took my one working amp to his house and we replaced one of his amps with My Ref. His jaw almost hit the floor, and he didn't know what to say. The My Ref totally destroyed the $3,000 (each!) Herron in spatiality and musicality. The Herron had more detailed, perhaps even etched highs (to the point of what I consider brightness, but that's a matter of taste) and deeper, better defined bass, but My Ref owned the midrange and imaging. Instruments and voices had a tonality and three dimensional quality that was very realistic. I admit it sounded better in his system than it does in mine, and that was with only one amp. It was not even close, and both of us were thoroughly impressed, he even a little embarassed.
Now, that wasn't my objective, but it did serve to vindicate my efforts and expense to optimize this amp's performance. That WAS my goal. I was beginning to suffer doubts about taking all this trouble to build these things and get them to work. In the end, I believe all this effort of buying and building so many kits (I'm on my fifth one!) and using expensive components will be worthwhile.
Those of you who are using this amp in a compromised system have no idea what you're missing. It is truly worthy of being attached to the best quality components, but I'm sure it sounds good with any combination.
Peace,
Tom E
There is no sound so sweet as the tiny click of a relay for the first time in an amp you just finished building.
I'm back into stereo. The new kit, built with some new parts and many "quality" parts salvaged from the non-working amp, is playing right now, and it sure sounds good.
One thing I have discovered is that the power cord, at least the two non-matching ones I'm currently using, makes a difference. I switched cords between amps to make sure that learn if that was a factor, and indeed it was, even though each amp has a different input cap. Oh no, another variable!
Will now build two identical cords and start listening to input cap differences after a little more break in.
Both amps sound excellent, but they are different.
Peace,
Tom E
I'm back into stereo. The new kit, built with some new parts and many "quality" parts salvaged from the non-working amp, is playing right now, and it sure sounds good.
One thing I have discovered is that the power cord, at least the two non-matching ones I'm currently using, makes a difference. I switched cords between amps to make sure that learn if that was a factor, and indeed it was, even though each amp has a different input cap. Oh no, another variable!
Will now build two identical cords and start listening to input cap differences after a little more break in.
Both amps sound excellent, but they are different.
Peace,
Tom E
Dario,
Still playing with C13. One amp has Mundorf SIO, which is the one I've been listening to the longest, and was in the amp when favorably compared to Herron monoblocks.
The other amp has Fostex copper foil cap. This one is really unknown, but it is the favorite cap for tweeter crossovers at Madisound, and they sell and get to hear a lot of caps. I thought it was worth a try because it's relatively cheap.
So far, different power cords seems to make a bigger difference than the caps! I'm working on that problem by building two identical cords.
The newest amp needs to settle in a little before I can critically evaluate it. The Blackgate at C9 had been sitting idle in the non-working amp for nine months, and I'm sure it needs to "reform" a little. All the other electrolytic caps are brand new, so they need a little playing time, too. I reused all the resistors and Wima film caps from the non-working amp.
I'm just so happy that I finally have two working amps, and they both sound great! I'll post the final configuration once I get everything settled.
Peace,
Tom E
Still playing with C13. One amp has Mundorf SIO, which is the one I've been listening to the longest, and was in the amp when favorably compared to Herron monoblocks.
The other amp has Fostex copper foil cap. This one is really unknown, but it is the favorite cap for tweeter crossovers at Madisound, and they sell and get to hear a lot of caps. I thought it was worth a try because it's relatively cheap.
So far, different power cords seems to make a bigger difference than the caps! I'm working on that problem by building two identical cords.
The newest amp needs to settle in a little before I can critically evaluate it. The Blackgate at C9 had been sitting idle in the non-working amp for nine months, and I'm sure it needs to "reform" a little. All the other electrolytic caps are brand new, so they need a little playing time, too. I reused all the resistors and Wima film caps from the non-working amp.
I'm just so happy that I finally have two working amps, and they both sound great! I'll post the final configuration once I get everything settled.
Peace,
Tom E