My copy just arrived. The 2004 CD reissue, a stunning £2.10 delivered. I'll let you know how I get on. Will also be interested to know what DR the plugin says it has.
It appears Cc has been added to version 2 but Cf and Rf2 are still not employed. Is this correct?
If so and based on your thread at the link below, why have you chosen to do things this way?
http://www.diyaudio.com/forums/chip...ensation-network-cc-rf2-cf-2.html#post4235244
If so and based on your thread at the link below, why have you chosen to do things this way?
http://www.diyaudio.com/forums/chip...ensation-network-cc-rf2-cf-2.html#post4235244
Correct. Adding Cc reduces the bandwidth of the LM3886 which makes it a bit more challenging to stabilize the global loop in the composite amp. However, Cc also causes a zero in the closed loop response of the LM3886 portion alone, which helps with the stability of the global loop of the composite amp.
However, the zero in the closed loop response of the LM3886 is undesired if the LM3886 is used stand-alone. This is why, for a stand-alone LM3886 amp, I recommend adding Cc, Cf, and Rf2. Cf and Rf2 introduce a pole to cancel the zero caused by Cc. In a stand-alone LM3886 amp, Cc is needed for stability near clipping and Rf2, Cf are needed to ensure a flat frequency response, closed loop.
The feedback loops in the Modulus-86 have been optimized for the performance of the Modulus-86 as a whole.
~Tom
Last edited:
Submitted the Parallel-86 board to the fab today. I should have boards in early April.
The initial batch will only be 12 boards. I will build up at least one for my characterization and measurements. All the simulations look good and my early prototype indicates that the circuit should work well. I'd still like to see the actual circuit before committing to specs. In particular as the layout is such a big part of the high performance level I'm getting with this composite amplifier topology.
I conservatively estimate that the Parallel-86 should be capable of delivering about 60 W into 8 Ω and 120 W into 4 Ω. With two boards bridged, output powers exceeding 300 W should be possible, assuming sufficient heat sinking.
I decided to use the LM4780 rather than two LM3886es in parallel as this provides a much cleaner layout. I am also seeing much tighter matching between channels of an LM4780 than you can hope for with two LM3886es. This helps tremendously in ensuring proper load balancing between the two amp halves. I've attached the block diagram for the Parallel-86.
I have opened up for pre-orders on my website: Parallel-86
~Tom
The initial batch will only be 12 boards. I will build up at least one for my characterization and measurements. All the simulations look good and my early prototype indicates that the circuit should work well. I'd still like to see the actual circuit before committing to specs. In particular as the layout is such a big part of the high performance level I'm getting with this composite amplifier topology.
I conservatively estimate that the Parallel-86 should be capable of delivering about 60 W into 8 Ω and 120 W into 4 Ω. With two boards bridged, output powers exceeding 300 W should be possible, assuming sufficient heat sinking.
I decided to use the LM4780 rather than two LM3886es in parallel as this provides a much cleaner layout. I am also seeing much tighter matching between channels of an LM4780 than you can hope for with two LM3886es. This helps tremendously in ensuring proper load balancing between the two amp halves. I've attached the block diagram for the Parallel-86.
I have opened up for pre-orders on my website: Parallel-86
~Tom
Attachments
"Easily bridged for output powers exceeding 300 W." 35V
300W at 8 ohms?
Power at 4 ohms?
I know, I do not need it but I have curiosity.
Input Sensitivity 2.20 V RMS 60 W, 8 Ω
Then with 2 V RMS (DAC output standard) the power will less.
300W at 8 ohms?
Power at 4 ohms?
I know, I do not need it but I have curiosity.
Input Sensitivity 2.20 V RMS 60 W, 8 Ω
Then with 2 V RMS (DAC output standard) the power will less.
Math says: 359 W into 4 Ω, 285 W into 8 Ω for ±35 V, bridged. Reality will probably be a tad lower.
Correct. There are various ways to solve this:
1) Use a DAC with higher output voltage. My Focusrite Saffire PRO 24 has no trouble delivering over 4 V RMS.
2) Use a preamp with gain. A THAT1646 would be just about perfect for this as it has a gain of 6 dB. One could also configure an LME49724 to provide a bit of gain.
3) Increase the gain of the PAR86 as described in the design doc.
Those who's followed my posts for some time will recognize that I'm a big proponent of good gain structure. 99.999 % of the time, the power amp will operate below a few watt, even when playing relatively loud. No additional gain is needed for these use cases. For the 0.001 % of the use cases that need extra power, I strongly suggest putting the gain where it belongs; in the preamp...
~Tom
Last edited:
I guess input sensitivity 2.2V RMS is via RCA and therefore 4.4 V RMS via XLR.
Suppose we had, exaggerating, 3 V RMS at the RCA, what problems would result if the DAC had no potentiometer to reduce it?
Suppose we had, exaggerating, 3 V RMS at the RCA, what problems would result if the DAC had no potentiometer to reduce it?
Last edited:
No. The gain of the THAT1200 is 0 dB (= 1x). If you apply 2.2 V RMS differential, you'll drive the PAR86 to clipping. If you drive it with 2.2 V RMS single ended, you'll drive it to clipping as well.
That would cause the amp to clip pretty hard. Just like any other amp does when you overdrive the input.
~Tom
TEAC UD-501
TEAC UD-501
If I connect -without an attenuator- this DAC to the amplifier with XLR I will have clipping problems?
Note 1: potentiometer only works with headphones output 🙁
Note 2: Frequency range 5Hz - 80kHz (-3dB); Bandwith *86 is 85 Khz (-3dB)
Note 3: S/N ratio 115dB; *86 S/N???
TEAC UD-501
I suppose then if I have 1,5 V RMS via RCA I will have 3 V RMS via XLR???
If I connect -without an attenuator- this DAC to the amplifier with XLR I will have clipping problems?
Note 1: potentiometer only works with headphones output 🙁
Note 2: Frequency range 5Hz - 80kHz (-3dB); Bandwith *86 is 85 Khz (-3dB)
Note 3: S/N ratio 115dB; *86 S/N???
Last edited:
trying to read between the lines drives me to assume that the output voltages of the XLR and the RCA are the same.
The XLR is balanced and has has two outputs:
Pin2 is 2Vac above the median.
Pin3 is 2Vac below the median.
Pin2 to pin3 gives an effective 4Vac
The RCA is unbalanced it has one output:
Pin is 2Vac above the median and that median is connected to Signal Return/Ground.
The XLR is balanced and has has two outputs:
Pin2 is 2Vac above the median.
Pin3 is 2Vac below the median.
Pin2 to pin3 gives an effective 4Vac
The RCA is unbalanced it has one output:
Pin is 2Vac above the median and that median is connected to Signal Return/Ground.
For that DAC, that is true. The XLR output has 6 dB of gain over the RCA output. That's the case for a differential driver like the THAT1646 as well. That does not mean that differential = 2*single ended, always.
The differential voltage is the difference between the (+) and the (-) of the signal pair. If you tie the (-) input to ground and apply 2 V RMS to the (+) input, as you would for a single-ended drive, the differential voltage is still V(+) - V(-) = 2 V RMS.
With this DAC you would need to handle the volume control either in the digital domain or by inserting a volume control between it and the amp. This is exactly the same as if you'd use this DAC with any other amplifier, so I don't really understand your question....
~Tom
"This is exactly the same as if you'd use this DAC with any other amplifier, so I don't really understand your question...."
Yes, I do not know why I did it 🙁
Yes, I do not know why I did it 🙁
Tom,
Could you test the parallel-86 also in bridged mode
(so it could be used with XLR inputs )
I'm interested what max voltage supply level could be used incase 2 parallel -86 configured in bridge mode
btw
congratulations
Could you test the parallel-86 also in bridged mode
(so it could be used with XLR inputs )
I'm interested what max voltage supply level could be used incase 2 parallel -86 configured in bridge mode
btw
congratulations
Hey hey hey!! Congrats on the new arrival! Very exciting.
I have to finish building my second helping before I come back for thirds, but I am planning on it.