what follows the 50k pot is critical to the final performance.
10k and 20k are more tolerant of some capacitance with the load.
Yes, if there is a must use 50k at hand.
Quality reasons. The jfets used are capacitive enough to lose some bandwidth with a 50k pot (beyond audibility territory). It will sound sluggish VS up to 20k pots.
Last questions, while I wait for the last parts so I can complete phase 1 (wiring up and testing before preparing the box and all extras) 
Say I've got 6.0-6.4V for the 3 LED string and 9.8-10V for the 5 LED string. I don't see a big problem myself, yet. Am I wrong?
I actually would like to be able to regulate at about 11 or even 12 volts. IIRC there was a post in the original B1 thread that measured performance over different voltages and higher ones did measure better. I still need to be able to drop about 6-7 volts trough the regulator to achieve good performance, right? My traffo will deliver enough, so running up to 12V after the regulator shouldn't be a problem for me.
Say I've got 6.0-6.4V for the 3 LED string and 9.8-10V for the 5 LED string. I don't see a big problem myself, yet. Am I wrong?
I actually would like to be able to regulate at about 11 or even 12 volts. IIRC there was a post in the original B1 thread that measured performance over different voltages and higher ones did measure better. I still need to be able to drop about 6-7 volts trough the regulator to achieve good performance, right? My traffo will deliver enough, so running up to 12V after the regulator shouldn't be a problem for me.
if you have efficient speakers and enough amp gain, you would be better off reducing the shunt resistor (connects between pot pin 2 and gnd) instead to maybe 33k. max volume will be reduced, but will keep the impedance seen by the sk170 fets low enough to help prevent HF roll off.
Hi
I have just started populating my hypnotize pcb, and just wanted to check a couple things before I go any further...
I purchased several LSK170B jfets a while ago and wondered if someone can advise on the best combination from the following?
4 x 6.4 @1%
2 x 8.9 @1%
2 x 6.3 @2%
2 x 6.9 @2%
2 x 6.9 @5%
2 x 10 @5%
Random selection 6.3, 7.6, 7.9, 8.8, 9.4
Thanks
I have just started populating my hypnotize pcb, and just wanted to check a couple things before I go any further...
I purchased several LSK170B jfets a while ago and wondered if someone can advise on the best combination from the following?
4 x 6.4 @1%
2 x 8.9 @1%
2 x 6.3 @2%
2 x 6.9 @2%
2 x 6.9 @5%
2 x 10 @5%
Random selection 6.3, 7.6, 7.9, 8.8, 9.4
Thanks
Get the closest matched ones for the 4 signal transistors for best effect, populate the rest with the random ones, I guess
Hi Atilla
I ended up using the four jfets with an idss of 6.4
I hope that's not too low an idss
Hi Atilla
I ended up using the four jfets with an idss of 6.4
I hope that's not too low an idss
Your Idss is ok. Cool.
OK, got my prototype working for about an hour now.
Supply voltages are +10V/-10,8V. CCS is running around 200 mA and heatsinks on the IRFP's are around 50 degrees Celsius.
Output offset is around 3 mV which is nice because my lm3875 are DC coupled.
Here's my impressions so far compared to my Densen DM10.
Detail level i way up. I hear stuff I've not heard before. I didnt think the DacMagic was up to it.
Sound stage is much wider than before. It now actually stretches beyond speaker positions.
Bass is much tighter. It sound like there's a lot less of it, but separating the bass guitar from drums is so easy.
There seems to be much more control in the mid range and voices sound more natural.
There's also a lot less background noise. The hiss from the Densen is gone. I cannot hear anything without signal no matter how close my ear is to the speakers.
Maybe the Densen is broken ? It's definately coming up for sale
In my setup the LightSpeed acutally sound better WITH the buffer. Mids were more messy without it.
I have to go re-listen to the rest of my albums.
Supply voltages are +10V/-10,8V. CCS is running around 200 mA and heatsinks on the IRFP's are around 50 degrees Celsius.
Output offset is around 3 mV which is nice because my lm3875 are DC coupled.
Here's my impressions so far compared to my Densen DM10.
Detail level i way up. I hear stuff I've not heard before. I didnt think the DacMagic was up to it.
Sound stage is much wider than before. It now actually stretches beyond speaker positions.
Bass is much tighter. It sound like there's a lot less of it, but separating the bass guitar from drums is so easy.
There seems to be much more control in the mid range and voices sound more natural.
There's also a lot less background noise. The hiss from the Densen is gone. I cannot hear anything without signal no matter how close my ear is to the speakers.
Maybe the Densen is broken ? It's definately coming up for sale
In my setup the LightSpeed acutally sound better WITH the buffer. Mids were more messy without it.
I have to go re-listen to the rest of my albums.
Sounds familiar except the bass level thing. I know it by having characteristic slam, the separation comment is still familiar though. Maybe you rewired the speakers in opposite polarity, or did something else in the chain along building this? The Densen is an integrated, you substituted the whole thing with DCB1 and a chip amp, so you refer to buffer+3875 or you used the Densen as a pro comparing line level only? Also 3mV offset is no problem but with better audio fets matching I have seen many 1mV or sub mV references by builders in this thread. Maybe your Vout(s) could come a little closer with one or two less happier Leds on the higher side.
Janpo,
You have found that inserting an active stage between source and receiver improves the sound quality.
That tells me that the source was incapable of driving the cables and receiver properly.
Using that information can you further improve the "system" by relocating the position of the B1?
3mV of output offset from a DCB1 indicates a poorly selected pair of jFETs in the buffer.
You have found that inserting an active stage between source and receiver improves the sound quality.
That tells me that the source was incapable of driving the cables and receiver properly.
Using that information can you further improve the "system" by relocating the position of the B1?
3mV of output offset from a DCB1 indicates a poorly selected pair of jFETs in the buffer.
Oops, my Densen is actually a DM20 pre.
Thanks for all the suggestions. I will incorporate them in the final version. This one is just a prototype in a selfmade MDF enclosure. But if i can convince the wife that it its hip, i wont need to do anything to the speakers either Maybe raw MDF will be the latest fashion some day.
I bought the audio fets matched from AndrewT. They were all labeled 7.4 mA. I took it face value so unless the soldering did something to them, i dont think they can be a better match.
I dont think bass level is low. It just so much more tight.
Thanks for all the suggestions. I will incorporate them in the final version. This one is just a prototype in a selfmade MDF enclosure. But if i can convince the wife that it its hip, i wont need to do anything to the speakers either
Maybe raw MDF will be the latest fashion some day.I bought the audio fets matched from AndrewT. They were all labeled 7.4 mA. I took it face value so unless the soldering did something to them, i dont think they can be a better match.
I dont think bass level is low. It just so much more tight.