My DAC provides for single ended and balanced outputs. My current 807 SE amp (which sounds wonderful) uses a single 12AT7 driver. If I build that amp with a balanced input stage and drive with the balanced output of my DAC would there be any benefit in the amplifier's performance, other than CMNR within the amp?
1. A Balanced signal sent to a Single Ended amplifier usually has no significant benefit.
If single ended signal to that same amplifier has problems: hum, noise, distortion, frequency response, etc., then . . .
fix the problem(s) first; before even considering to convert the input circuit to accept a balanced signal in.
Often, adding an additional circuit fixes one problem, but adds new problem(s) too.
2. Balanced signal to a Balanced Push Pull amplifier does have benefit(s).
The phase splitter is in the signal source (like my CD player with XLR balanced outputs).
I love the sound of my latest amplifier design; it is Balanced.
A 12AU7 input/driver, and 5881 output tubes.
I hope to find the time to take pictures, draw a schematic, and Post a thread starter of my new amplifier design on Tubes / Valves.
If single ended signal to that same amplifier has problems: hum, noise, distortion, frequency response, etc., then . . .
fix the problem(s) first; before even considering to convert the input circuit to accept a balanced signal in.
Often, adding an additional circuit fixes one problem, but adds new problem(s) too.
2. Balanced signal to a Balanced Push Pull amplifier does have benefit(s).
The phase splitter is in the signal source (like my CD player with XLR balanced outputs).
I love the sound of my latest amplifier design; it is Balanced.
A 12AU7 input/driver, and 5881 output tubes.
I hope to find the time to take pictures, draw a schematic, and Post a thread starter of my new amplifier design on Tubes / Valves.
"Balanced" is a word that has a lot of meanings. In your DAC it (probably) means two output paths of equal voltage and impedance with opposite polarity, both referenced to signal ground, and likely chassis ground, for each channel. In semi-pro recording equipment it (nowadays) means two outputs of equal output impedance, but only one with signal (the other is a resistor to signal ground). This is all that's necessary for a differential receiver on the other end of the connecting cable to do its magic of ignoring interference, and is called "impedance balanced".
At the receiving end, "balanced" is almost always misused. 6A3sUMMER uses the term correctly when he says that his amplifier is balanced, meaning push-pull throughout and relies (some or mostly) on the OPT to be differential, that is, rejecting anything common to both polarities of signal, to have Common Mode Rejection Ratio (CMRR). Externally originated noises appear on the connecting cables in common mode, the same on both conductors. Many folk conflate this with "differential", meaning having a high CMRR. Similar but not the same.
A single ended input can have some CMRR by standing signal "ground" a little way off from chassis ground, maybe 10R or so, and bringing signal hot, signal return and chassis ground to three separate pins, like a 3-pin XLR, and keeping them separate all the way back to the source. Shield should really only be a shield, not also the signal return. Keep signal hot and return on their own conductors, and shield on its own connector pin. Our historic reluctance to move away from two conductor (RCA) jacks is foolishness on a governmental scale.
All good fortune,
Chris
At the receiving end, "balanced" is almost always misused. 6A3sUMMER uses the term correctly when he says that his amplifier is balanced, meaning push-pull throughout and relies (some or mostly) on the OPT to be differential, that is, rejecting anything common to both polarities of signal, to have Common Mode Rejection Ratio (CMRR). Externally originated noises appear on the connecting cables in common mode, the same on both conductors. Many folk conflate this with "differential", meaning having a high CMRR. Similar but not the same.
A single ended input can have some CMRR by standing signal "ground" a little way off from chassis ground, maybe 10R or so, and bringing signal hot, signal return and chassis ground to three separate pins, like a 3-pin XLR, and keeping them separate all the way back to the source. Shield should really only be a shield, not also the signal return. Keep signal hot and return on their own conductors, and shield on its own connector pin. Our historic reluctance to move away from two conductor (RCA) jacks is foolishness on a governmental scale.
All good fortune,
Chris
I'm just thinking out loud:
If you use a pmos source follower (like the VP0106) on the cathode of the 12AT7, you can drive the triode with a balanced signal: one on the grid and the opposite on the gate of the mosfet.
If you use a pmos source follower (like the VP0106) on the cathode of the 12AT7, you can drive the triode with a balanced signal: one on the grid and the opposite on the gate of the mosfet.
Here you mean a (series'd) differential input to single-ended output (the cathode). Another example where "balanced" and "differential" are used as if they were the same, but they're not really the same. Nit-picking maybe, but confusing to newcomers.
All good fortune,
Chris
All good fortune,
Chris
Hi Chris,
this is what I mean (sketched, to be improved):
V1 and V6 are the balanced input signal.
On a CCS loaded 12AT7 with 700 mVrms balanced input it gives 78Vrms on the output at 3% THD.
This means also that we can drive a SE triode (or pentode, or whatever) with a phase split stage.
This could also mean that we can focus on cancelling the 3rd harmonic of the output stage instead of the 2nd.
this is what I mean (sketched, to be improved):
V1 and V6 are the balanced input signal.
On a CCS loaded 12AT7 with 700 mVrms balanced input it gives 78Vrms on the output at 3% THD.
This means also that we can drive a SE triode (or pentode, or whatever) with a phase split stage.
This could also mean that we can focus on cancelling the 3rd harmonic of the output stage instead of the 2nd.
V1 and V6 form a differential input that's called a Rush cascode. Here's an example without DC offset:
The input stage is the classic John Curl and the second stage is a Rush cascode. Both are differential.
All good fortune,
Chris
The input stage is the classic John Curl and the second stage is a Rush cascode. Both are differential.
All good fortune,
Chris
I'm sorry Chris but I don't get your point.
The OT asked about benefits of having a balanced signal in a SE amp, as he has a DAC with balanced outputs.
I proposed an idea to get benefits of a balanced signal into a SE driver by driving the grid and the gate of the driver with opposite signals and increase the sensitivity of the amp. What is the clue with the solid state circuit you posted?
The OT asked about benefits of having a balanced signal in a SE amp, as he has a DAC with balanced outputs.
I proposed an idea to get benefits of a balanced signal into a SE driver by driving the grid and the gate of the driver with opposite signals and increase the sensitivity of the amp. What is the clue with the solid state circuit you posted?
To expand some: your circuit is truly differential, in that it doesn't require a source signal that is "balanced" to give noise immunity. It is inherently only responsive to the difference between V1 and V6 - anything appearing equally on both is very largely ignored at the output (in this case, the anode). It doesn't require that signal appear on both V1 and V6 to have this CMRR rejection of noise on connecting cables to accomplish this.
Caveat: the noise voltage impressed on a twin (usually twisted for best averaging) conductor pair is most equal when source and load impedances are equal for both polarities' conductors. This is the origin of the modern semi-pro "impedance balancing" - one of the twin signal conductors is driven from an op-amp with a build-out resistor R and the other is just an identical R to ground. Works the same for a truly differential input.
So, to the OP's point, driven balanced outputs aren't really very useful unless you pair them with an appropriate input. The noise rejection all happens at the receiving end of the wires. But even a single-ended input with three pins can get most of the way there, per post #4.
All good fortune,
Chris
Caveat: the noise voltage impressed on a twin (usually twisted for best averaging) conductor pair is most equal when source and load impedances are equal for both polarities' conductors. This is the origin of the modern semi-pro "impedance balancing" - one of the twin signal conductors is driven from an op-amp with a build-out resistor R and the other is just an identical R to ground. Works the same for a truly differential input.
So, to the OP's point, driven balanced outputs aren't really very useful unless you pair them with an appropriate input. The noise rejection all happens at the receiving end of the wires. But even a single-ended input with three pins can get most of the way there, per post #4.
All good fortune,
Chris
Now I understand! You were focusing on CMRR, while my proposal is more focused on having the possibility to use the balanced signal as a 2x in amplitude by driving the grid and the cathode with two signals opposed in phase.
This makes the single stage driver capable of having high gain to directly drive low gm output tubes.
I would suggest adding an input transformer in the amp so you have the options of single-ended and balanced inputs. A high impedance, e.g., 10k:10K, bridging transformer will work for its simplicity without altering the the gain and original circuit. You can also add in a bypass switch for SE signal.
Yes I am going to try this. I have ordered an inexpensive Edcor 15K:15K just to try out. One issue I have is my DAC has a relatively high 2.4K output impedance (spec'd 1.2K but measured at 2.4K in one review).
Could be 1.2k from either side to ground and 2.4 k between hot and cold. I’ve often wondered about the Edcors - I’ve used the $65 Hammonds with success. The $6 Triads that are only rated down to 200 Hz are ok down to 20 if the source impedance is low enough (ie, 5532 op amp), but still might not be high enough fi. And would be eating some low end driven from 2400 ohms. Playing around with PA equipment and driving them off the usual mixing consoles and stuff they’re acceptable. Something in between on price would be worth looking at.
The quality of the transformer is paramount. An inexpensive transformer will generally result in a massive decrease of transparency and dynamics. Add to that the long (several hundred hours) breakin period you will probably abandon the whole concept. Knowing the long break in time and the shortcomings of poor iron I would not waste my time skimping on an inexpensive unit. Get a very high quality one and it will retain a lot more resale value as well as netting a better result. I am much more of a fan of step down on a high Z output and allowing your preamp to handle gain and the low level signal to your amplifier. If you use a passive or a VC on your amp you will be happier with the step down and better VC position as well unless you listen to your amp a close to full power.
I used to use interstage transformers in both my single ended and in my push pull amplifiers.
Although they gave some fairly good results, I 'gave up' using Interstage transformers sometime ago, and I found solutions that I consider to generally be better than using an interstage transformer.
And, if I ever had used a Signal Input transformer (I never tried that), I bet if I did, I would have stopped using them too.
YMMV (Your Mileage May Vary)
Now the only transformers I use are Output transformers, and Power transformers.
My latest amplifier, now my favorite, is my 2 stage Balanced/Differential push pull amplifier. I use my CD player that has balanced/differential XLR outputs.
Just my experiences and opinions.
Although they gave some fairly good results, I 'gave up' using Interstage transformers sometime ago, and I found solutions that I consider to generally be better than using an interstage transformer.
And, if I ever had used a Signal Input transformer (I never tried that), I bet if I did, I would have stopped using them too.
YMMV (Your Mileage May Vary)
Now the only transformers I use are Output transformers, and Power transformers.
My latest amplifier, now my favorite, is my 2 stage Balanced/Differential push pull amplifier. I use my CD player that has balanced/differential XLR outputs.
Just my experiences and opinions.
jderimig,
No. Not like that.
But you gave it a good guess and a good try.
Your schematic should work OK, based on my initial look at your circuit.
I really have to draw up my schematic. Then post schematic, pictures, and measurement results.
And It may disappoint some, it is low power.
I only use it either for near field listening, or in a small living room with the speakers not to far away.
I built one monoblock, need to duplicate for stereo.
It is out in the living room, driving my modified ProAc Signature Tablet 2000 speaker.
No. Not like that.
But you gave it a good guess and a good try.
Your schematic should work OK, based on my initial look at your circuit.
I really have to draw up my schematic. Then post schematic, pictures, and measurement results.
And It may disappoint some, it is low power.
I only use it either for near field listening, or in a small living room with the speakers not to far away.
I built one monoblock, need to duplicate for stereo.
It is out in the living room, driving my modified ProAc Signature Tablet 2000 speaker.