Almost all DIY. Noir kit from DIYaudio store, modified Elekit TU-8200DX, the Wire SE with autoformer volume control, an LB Audio LB437I-VL Line Amplifier, Burson Audio Conductor 3XP (sounds better bypassing the headphone amp and using it as a DAC through the DCG3). All my opinions are subjective of course. I'm not saying that any of these amps are bad, in fact they are all very good, it is just that I feel, with my system, the DCG3 has the best balance of traits. It connects me with the music like the Noir and the TU-8200 with possibly the same level of detail as the Wire and the LB437I-VL.
@danbmellow
An interesting collection of diy projects indeed. Congrats for being an avid experimenter. For that reason here's a left-field idea for as long your dcg3 remains in open plan state and still easy to tweak:
If you like to try a detail trait boost trick you can put a buffer op-amp between the pot's output and the dcg3's input. This makes a dcg3 run full bandwidth and max slew like no volume pot impedance is there. It may upset its balance of traits positively or negatively depending on the headphones model. Easy to evaluate.
Needs a Fet input type unity gain stable high performance chip. Any product you like the best. Just a wire link between an op-amp's output and its inverting input makes it a buffer.
Simple to put together. A tiny perfboard, +/- rails tapped from dcstb outputs, local decoupling 0.1uF small film caps right next to the chip's power pins, that's all.
An interesting collection of diy projects indeed. Congrats for being an avid experimenter. For that reason here's a left-field idea for as long your dcg3 remains in open plan state and still easy to tweak:
If you like to try a detail trait boost trick you can put a buffer op-amp between the pot's output and the dcg3's input. This makes a dcg3 run full bandwidth and max slew like no volume pot impedance is there. It may upset its balance of traits positively or negatively depending on the headphones model. Easy to evaluate.
Needs a Fet input type unity gain stable high performance chip. Any product you like the best. Just a wire link between an op-amp's output and its inverting input makes it a buffer.
Simple to put together. A tiny perfboard, +/- rails tapped from dcstb outputs, local decoupling 0.1uF small film caps right next to the chip's power pins, that's all.
Fascinating idea! I had observed that, in LTSpice simulations, reducing the input R value also drove the distortion very low. Of couse it is the effect of the Miller capacitance that is being reduced. I'm sure you have considered using a cascode to achieve this same result?
Reducing total Rinp further from where I settled its min value with R1 when I was originally formulating the dcg3 project on a breadboard several years back, unfortunately led to the verge of instability. I don't exactly remember if I experimentally cascoded the input stage too. But possibly I did. There should have happened something I did not like and skipped it.
Warning for idiosyncratic question!
Theoretically speaking and having in mind the ultrasonic hash coming from a NOS DAC -known to take opamps to their limits, which of the following would make sense: The highest recommended value pot that would possibly relax DCG3 or no pot at all to let it pass without struggling?
Theoretically speaking and having in mind the ultrasonic hash coming from a NOS DAC -known to take opamps to their limits, which of the following would make sense: The highest recommended value pot that would possibly relax DCG3 or no pot at all to let it pass without struggling?
Thanks Salas. I will give it a try.@danbmellow
An interesting collection of diy projects indeed. Congrats for being an avid experimenter. For that reason here's a left-field idea for as long your dcg3 remains in open plan state and still easy to tweak:
If you like to try a detail trait boost trick you can put a buffer op-amp between the pot's output and the dcg3's input. This makes a dcg3 run full bandwidth and max slew like no volume pot impedance is there. It may upset its balance of traits positively or negatively depending on the headphones model. Easy to evaluate.
Needs a Fet input type unity gain stable high performance chip. Any product you like the best. Just a wire link between an op-amp's output and its inverting input makes it a buffer.
View attachment 1163542
Simple to put together. A tiny perfboard, +/- rails tapped from dcstb outputs, local decoupling 0.1uF small film caps right next to the chip's power pins, that's all.
Thinking about this buffer option, I 've found a nice case that has two VU meters. Would it possible to get a parallel output from the headphones, feed it into a buffer as described and then:
Something like that:
would that work without affecting performance?
- from the buffer output connect a series resistor and a multi turn trimmer for VU meter calibration
- In series with the trimmer the VU meter
- and then a 50Ohm resistor to simulate a load
Something like that:
would that work without affecting performance?
A buffer isolates so driving a VU meter through one shouldn't affect the main outputs.
*VU driver circuits also involve diodes rectifying the signal. Must be flexible to different meter models for correct needle deflection and motion.
Do investigate prior art for examples.
*VU driver circuits also involve diodes rectifying the signal. Must be flexible to different meter models for correct needle deflection and motion.
Do investigate prior art for examples.
Doing my homework I 've found this very interesting article here: Practical VU Meter Circuits for anyone interested. Also checked what is readily available and I found out that these cheap VU meters come basically in two flavors:
Another question. When I asked for a suitable heatsink, @Salas suggested for testing the following:
- With integrated control board on the back side of the meter
- stand alone, driven by external board (usually provided with them)
Another question. When I asked for a suitable heatsink, @Salas suggested for testing the following:
Is this one suitable for permanent use (heat-sink-aluminium-black-160x32x62mm)? Operating conditions will be 3x gain and 140-150 mA bias. Headphones AT ATH-M50X for now, and possibly in the future Fostex Dekoni Blue or Dan Clark Aeon Closed X. I am trying to opt for low depth fins to accomodate layout for a 41cm wide, 30cm depth, 74cm height internal dimensions enclosure. Also I would like to use an once off solution for both breadboard testing and final assembly to avoid readjusting the mosfets and to satisfy my eagerness to test it as soon as possibleFor heatsink it doesn't need crazy C/W capability, roughly a 5 cm by 15 cm or somewhat bigger well finned piece should get you by in testing.
Was thinking orientation like in post #4047 if that is the parallel thing with bottom and upper cover. Also the bottom side of the sink screwed to the bottom plate using the threaded holes for better heat dissipation. I can go with one a bit larger, 176x60x33. For larger sink depth it seems marginal with the enclosure. The case has some ventilation holes in the upper cover. There are no pictures of the inside or the bottom though. It's this one: https://www.ebay.com/itm/3742636477...rksid=p2047675.c101113.m2109&autorefresh=true
Should I drop the idea of the heatsink and mount directly on chassis for better results? It would allow for easier layout too.
Should I drop the idea of the heatsink and mount directly on chassis for better results? It would allow for easier layout too.
You will buy a sink for the open test phase anyway. You can try it sitting located along those top openings when the box comes. It adds metal mass and fins to channel the heat towards the top which can't be a bad thing. The box alone can absorb enough heat but you don't want it to saturate and act like a pizza oven. There will be the dual mono PSU with its sinks radiating too plus you plan 0.15A bias. You will see in practice how the total heat exchange situation will behave in that box. Then you decide which configuration to fix.
Closed back planar excellence seems like commanding $$$ still. The $249 Dekoni Fostex semi-open should be fun, have had the basic T-50RP MK3 model. Nice. Not much comfortable to wear for too long nonetheless.Headphones AT ATH-M50X for now, and possibly in the future Fostex Dekoni Blue or Dan Clark Aeon Closed X.
To really get in that game I feel it might be the DC Noire at $900. Difficult to get very good treble resolution with medium priced enclosed planars. Width can be an issue in closed backs at any price though.
I have dynamic closed backs. Beyer DT770 80Ω very sturdy and affordable. Goes low and its wide. Lacks in tuning but with software Equalizer APO that's solved. Peculiar tuning still happens in every price category. There are also affordables with excellent tuning that may lack in resolution and sturdiness. Example AKG K371. I also have the Beyer DT150 250Ω. All other ones I got are open back.
Thanks Nikos for all that information. The most annoying thing with cans is that you seldom can find many in a row and see what fits your head and taste more. Reviews are nice to read, but you can't really know unless you try them. Had the chance to do that once in the pas 10+ years ago and ended with the ATH M50 (also liked the Beyer DT770 and a Sony studio model, but found the M50x more balanced).
Fast forward to today, I now have kids and the M50H are now are my most used drivers, since I either use them during piano practicing with a Roland HP704 or listening to music during work with a focusrite i24. To be honest, I prefer listening to speakers, but it takes to have the whole universe aligned to be able to really enjoy them.
So with the DCG3 I want to have a strong performing amp to be able to drive most cans and have the opportunity to listen to music late at night at normal volumes and high quality. And to be able to play with various power amps in the long run (really thinking trying one of those new class D amps with purifi modules). Regarding cans, although the closed X are tempting, I will probably wait to build up the budget to go to the Noires DC, but I don't see this happening till the end of this year and maybe more.
Btw, regarding my heat sinking options, I finally ordered the larger version of the enclosure with 35.6cm depth for more layout options and better air flow inside the enclosure.
Fast forward to today, I now have kids and the M50H are now are my most used drivers, since I either use them during piano practicing with a Roland HP704 or listening to music during work with a focusrite i24. To be honest, I prefer listening to speakers, but it takes to have the whole universe aligned to be able to really enjoy them.
So with the DCG3 I want to have a strong performing amp to be able to drive most cans and have the opportunity to listen to music late at night at normal volumes and high quality. And to be able to play with various power amps in the long run (really thinking trying one of those new class D amps with purifi modules). Regarding cans, although the closed X are tempting, I will probably wait to build up the budget to go to the Noires DC, but I don't see this happening till the end of this year and maybe more.
Btw, regarding my heat sinking options, I finally ordered the larger version of the enclosure with 35.6cm depth for more layout options and better air flow inside the enclosure.
- Home
- Source & Line
- Analog Line Level
- Salas DCG3 preamp (line & headphone)