Good evening everyone . Just have a question . I would like to built these pre and since GB is still going what would be the best power supply for this pre ?
Shell i go for Ultrabib 2 x positive , 2 x negative or DCSTB ? Has someone tried both and can share experience which one he prefer and why ?
Would appreciate any thoughts on power supply with the best sound of course on mind.
Thank you.
Shell i go for Ultrabib 2 x positive , 2 x negative or DCSTB ? Has someone tried both and can share experience which one he prefer and why ?
Would appreciate any thoughts on power supply with the best sound of course on mind.
Thank you.
@Properdouble
If the box size is comfortable with the heat management of shunt regs you may use Ubibs for that something extra in expression of sound.
Their very low output impedance and high constant current bias makes for a bit more bouncy overall result.
The DCSTB won't be particularly heard, just a blank stare kind of contribution.
Regarding signal to noise performance for a line level preamp application like this one, it is as good with both.
If the box size is comfortable with the heat management of shunt regs you may use Ubibs for that something extra in expression of sound.
Their very low output impedance and high constant current bias makes for a bit more bouncy overall result.
The DCSTB won't be particularly heard, just a blank stare kind of contribution.
Regarding signal to noise performance for a line level preamp application like this one, it is as good with both.
If anyone wonders why DCG3 is very high on the best preamplifiers list, look at the distortion into high impedance, like at amplifier input. At 1 Vrms, there is nothing but pure H2 at very low level and then nothing even 40 dB down at any audio frequency. If that’s not amazing, then I don’t know what is.
For this measurement, I bypassed Muses volume control at DCG input, as it also contributes some distortion.
For this measurement, I bypassed Muses volume control at DCG input, as it also contributes some distortion.
I very much like the sound of this amp, and have one on my desk now as my headphone amp. I used its circuit to become familiar with using LTspice for audio circuits. At first I thought it was an artifact of Spice when I noticed that reducing the input resistance also reduced the distortion. However, Salas does note that the sound changes as one changes the value of the input potentiometer. I came to understand that what I was seeing in simulation was the Miller effect, and that is also what is being heard with changes in the effective input resistance as one changes values of the potentiometer. It would be interresting to see what distortion you measure with the Muses in circuit, not just at 1K but also 10K or 20K.
I have only 1 kHz signal sources with extremely low distortion.
Measurements at post #6943 are with Muses volume control. Muses has distortion at the same level as DCG3, but with some higher order harmonics (still 30 db below H2, with exception of H3).
At normal attenuation levels, all that high order distortion is buried beneath DCG3 noise floor. Example:
There is no way I would ever return to any passive potentiometer volume control. Muses provides perfectly aligned L+R control, remote and balance control. I really appreciate that with volume change, soundstage doesn’t shift in any direction and 1 dB imbalance is easily heard.
AFAIK, potentiometers have imbalance 1-3 dB over various control range positions and they introduce comparable distortion levels as well.
Stepped attenuators are free from imbalance but their volume control is too coarse for my taste. For me, perfect volume level was always somewhere between available steps.
Measurements at post #6943 are with Muses volume control. Muses has distortion at the same level as DCG3, but with some higher order harmonics (still 30 db below H2, with exception of H3).
At normal attenuation levels, all that high order distortion is buried beneath DCG3 noise floor. Example:
There is no way I would ever return to any passive potentiometer volume control. Muses provides perfectly aligned L+R control, remote and balance control. I really appreciate that with volume change, soundstage doesn’t shift in any direction and 1 dB imbalance is easily heard.
AFAIK, potentiometers have imbalance 1-3 dB over various control range positions and they introduce comparable distortion levels as well.
Stepped attenuators are free from imbalance but their volume control is too coarse for my taste. For me, perfect volume level was always somewhere between available steps.
It has R1 for critical damping against oscillating when a volume pot is not used at all or when fully rotated clockwise, thus contributing no extra series impedance (becomes just a contact at max volume point).
I purposefully avoided an input buffer stage for less circuitry in the signal path. Spec was satisfactory already.
If you put a top notch op-amp wired as a buffer between the pot and the line input pcb terminals you should see best electrical spec and steady max bandwidth. Due to source & pot/attenuator impedance isolation.
I purposefully avoided an input buffer stage for less circuitry in the signal path. Spec was satisfactory already.
If you put a top notch op-amp wired as a buffer between the pot and the line input pcb terminals you should see best electrical spec and steady max bandwidth. Due to source & pot/attenuator impedance isolation.
Hi here...
I saw in post Here!, that Salas posted some THD stepped measurement's...
Just today I changed some code in my DCG3 ESP32 (Arduino) controller, so to remove some nasty spike's created when the ESP32 cpu is running >40MHz or so.
I took a THD vs. Level sweep from around 500mV to 950mV (-9dB ---> -3dB) loaded with around 20k (My diy autoranger) Volume at around 2'Oclock
My DCG3 is build in 2016, still going strong having lot's and lot's on hour's on it's back now
THD's are deepdown as can be seen - The noisefloor on my autoranger/soundcard is lower than -140dB so measurement's should be good.
Zooming in the same picture, shows the gainsteep's in my SuperRanger (Autoranger).A close -3dB match is archived (target 900mV into my hacked Magicbus Behringer)
So a good measurement with the Behringer's best level is archived.
Just a picture of my setup in this situ.
Jesper.
I saw in post Here!, that Salas posted some THD stepped measurement's...
Just today I changed some code in my DCG3 ESP32 (Arduino) controller, so to remove some nasty spike's created when the ESP32 cpu is running >40MHz or so.
I took a THD vs. Level sweep from around 500mV to 950mV (-9dB ---> -3dB) loaded with around 20k (My diy autoranger) Volume at around 2'Oclock
My DCG3 is build in 2016, still going strong having lot's and lot's on hour's on it's back now
THD's are deepdown as can be seen - The noisefloor on my autoranger/soundcard is lower than -140dB so measurement's should be good.
Zooming in the same picture, shows the gainsteep's in my SuperRanger (Autoranger).A close -3dB match is archived (target 900mV into my hacked Magicbus Behringer)
So a good measurement with the Behringer's best level is archived.
Just a picture of my setup in this situ.
Jesper.
Thank's... I'm also still very happy with the Pre.
I asked John the REW owner to add some settling time in REW for stepped measurement's and autoranger's and other slow stuff.
And between Christmas and Newyear 2023, it was added in the build.
Before that, an autoranger had to be very very fast to archive good results...
I just adjusted it to 2,5sec. and took a shoot, and the autoranger behaved nice and easy during the test.
I asked John the REW owner
to add some settling time in REW for stepped measurement's and autoranger's and other slow stuff.And between Christmas and Newyear 2023, it was added in the build.
Before that, an autoranger had to be very very fast to archive good results...
I just adjusted it to 2,5sec. and took a shoot, and the autoranger behaved nice and easy during the test.
Good day Salas. After a 10 month hiatus, I finally assembled the DCG3 and DCSTB in a casing. I also built an outboard PS which includes a single transformer, soft recovery diodes and (2) 20K capacitors. The preamp is used in my main system and the results are excellent; solid image, musical, quiet and dynamic. The DC offset was measured @ -0.5mv/-0.3 for the L/R channel. I used a newly acquired 50K TKD pot and with the low efficient Magnepan speaker, the pot is usually turned to the 60% position.
I have a second DCG3 board which I want to try a 100K shield TKD pot. I also want to try a shunt resistor with a 50K 10 turn linear pot. Salas, can you suggest a value for shunt resistor?
Again thank you very much for sharing your knowledge and expertise and last but not the least, the patience for someone like myself.
Tom
I have a second DCG3 board which I want to try a 100K shield TKD pot. I also want to try a shunt resistor with a 50K 10 turn linear pot. Salas, can you suggest a value for shunt resistor?
Again thank you very much for sharing your knowledge and expertise and last but not the least, the patience for someone like myself.
Tom
Congratulations for the conclusion of your preamp. You are welcome. Photos are also welcome. Smaller value potentiometer will not change the voltage division behavior vs rotation much, it will mainly change the loading on source equipment and the bandwidth / rise time of this preamp. Smaller here gives faster leading edge to notes subjectively also. For more gain like 4X (12dB) if necessary, you can manipulate the values ratio of the feedback resistors by changing R6 (there's a section in the pdf guide explaining gain set).
Shunt resistor to fake log a linear pot should be 15-25% of its original value.
Shunt resistor to fake log a linear pot should be 15-25% of its original value.
No, R1's value is still technically adequate in this case too.
But if you would also like to test mimic the usual 20kΩ pot's equivalent output impedance for subjective reasons, that's 5kΩ at midpoint. I.e. 3kΩ missing in your 2kΩ Zo attn situation. Thus 3k3 R1 standard+3k=6k3 new R1 for mimicking.
But if you would also like to test mimic the usual 20kΩ pot's equivalent output impedance for subjective reasons, that's 5kΩ at midpoint. I.e. 3kΩ missing in your 2kΩ Zo attn situation. Thus 3k3 R1 standard+3k=6k3 new R1 for mimicking.
Hi Salas,
I am skittish about showing pictures of my DCG3 cause my setup is messy and I didn't use coax wire connection as per your instruction. I have a hard time sourcing Mogami wire from Canada. So instead I use Discovery and VDH hook-up wires.
Also, my PS (raw DC) is 3 ft. away from the main casing and is on 24/7. I use a toggle switch to control/switching the +/- rail. I hope you are not too upset about this improvisation. I remember you suggested using the LED relay for on/off function, but that would mean de-soldering the LED from the DCG3 board and I am worried about making mistakes during the process.
BTW The casing is a leftover from previous project and that's why you see all these extra holes and scratch marks on the front and side wall. Anyway, have a good laugh and feel free to comment.
I am skittish about showing pictures of my DCG3 cause my setup is messy and I didn't use coax wire connection as per your instruction. I have a hard time sourcing Mogami wire from Canada. So instead I use Discovery and VDH hook-up wires.
Also, my PS (raw DC) is 3 ft. away from the main casing and is on 24/7. I use a toggle switch to control/switching the +/- rail. I hope you are not too upset about this improvisation. I remember you suggested using the LED relay for on/off function, but that would mean de-soldering the LED from the DCG3 board and I am worried about making mistakes during the process.
BTW The casing is a leftover from previous project and that's why you see all these extra holes and scratch marks on the front and side wall. Anyway, have a good laugh and feel free to comment.
Attachments
Aesthetics is a non issue in my situation because both the power amp & preamp hide behind the TV. The DCG3 is just 3'' below the top of the TV. It is a space issue. This is the only way I can have music and causal TV viewing in this (reading) room which has very good acoustic properties. The way I built the DGC3, it has input jack @ the extreme left/right side of the casing to accommodate 2 turntables sitting on both side. I actually have dedicated bigger rooms in the basement for movie viewing and music listening with big floor standing speakers. But as I am getting older, I found myself using it less and less because it has this isolated feeling and is a little bit cold especially in the winter.
These days, comfort is no.1 priority. I would considered it a bargain if I can maintain my listening ability for the next 15 years.
These days, comfort is no.1 priority. I would considered it a bargain if I can maintain my listening ability for the next 15 years.