Out of what you mentioned I would use the 100uF oscon - mostly because of the value. I think any reasonable cap will be ok here because the LDR's do not transfer high frequencies very much.
You could always build the whole thing and try it before you put in these caps - then you get to hear the difference they make.
You could always build the whole thing and try it before you put in these caps - then you get to hear the difference they make.
Thanks.
Just want to hear more Cap related experience...because this Thread is quite difficult to read...maybe I am bored over Christmas and will setup a Lightspeed Wiki 🙂
Just want to hear more Cap related experience...because this Thread is quite difficult to read...maybe I am bored over Christmas and will setup a Lightspeed Wiki 🙂
47uF mkp1837?
Alan,
Assume these are the Roe/Vishay mkp1837 (a.k.a. ERO mkp1830) but nowhere near this value (perhaps 47nF?).
Suggest that when your LS is up and running, then add the different caps - I use 220uF Silmics (stock cap) to good effect - some people find no noticeable benefit.
If you are using the Toole-Reg to drive the Lightspeed, suggest using the current source version at 12volts that could also act as one supply for a buffer, gain stage, xover, etc.
George uses a simple double reg supply (with the 3pin devices) for the production Lightspeeds.
I use the Teddy-Regs (+/- 15V) mainly to drive the k170/j74 buffer.
A few pages back (posts 1524/5) there is a summary that may be useful to you.
Alan,
Assume these are the Roe/Vishay mkp1837 (a.k.a. ERO mkp1830) but nowhere near this value (perhaps 47nF?).
Suggest that when your LS is up and running, then add the different caps - I use 220uF Silmics (stock cap) to good effect - some people find no noticeable benefit.
If you are using the Toole-Reg to drive the Lightspeed, suggest using the current source version at 12volts that could also act as one supply for a buffer, gain stage, xover, etc.
George uses a simple double reg supply (with the 3pin devices) for the production Lightspeeds.
I use the Teddy-Regs (+/- 15V) mainly to drive the k170/j74 buffer.
A few pages back (posts 1524/5) there is a summary that may be useful to you.
Hi jh,
thanks for the hint to the summary on postings 1524/1525.
One Question, what do all the numbers and double crosses
mean in the pdf´s?
Greetings Ulf
thanks for the hint to the summary on postings 1524/1525.
One Question, what do all the numbers and double crosses
mean in the pdf´s?
Greetings Ulf
Hi,
you are right...it was not nF......and not 47......damn.....my fault 🙂
Thanks
I will use Tooles Reg.......not only for Lightspeed.
Do you mean CCS LDR from Peter Roodzant as Current Source Version.....?
regards
Alan
you are right...it was not nF......and not 47......damn.....my fault 🙂
Thanks
I will use Tooles Reg.......not only for Lightspeed.
Do you mean CCS LDR from Peter Roodzant as Current Source Version.....?
regards
Alan
DVBjunky said:
Alan hi, I have 470uf Oscons across all 4 Nsl32sr2s of my original prototype Lightspeed Attenuator that is still in use in my system, I never bothered to take them out, this was way back when people reported hearing improvements with them.
I also have on the prototype pure dc battery power source at the flick of a switch if I want it, again from reported improvements.
Every time I finish a production Lightspeed Attenuator for a customer, it is tested for 48hrs in my system before dispatch, and yes you guessed it, the production one is in distinguishable from my prototype even with battery power, sometimes I even reach for the wrong control when listening. And my system is high rez, if your interested I can post what it consists of.
Guys, don't forget to vote for the Passive Preamp forum.
http://www.diyaudio.com/request/
Cheers George
Thanks George,
I was just unsure because almost all people claim that OScon are just suitable for digital circuits. I strongly believe in your Lightspeed because an old friend is working for Marconi Germany (Space communication) and we talked about the Lightspeed. He showed me several industrial uses of opto-devices....way back to the early 80s. Unbelievable......i have no more doubts about it!
Its funny that most people are so skeptical about the Lightspeed. I think you posted your equipment details already.
I hope that my Lightspeed mk2 will play in my active crossover system next month......my friend is building his integrated Aleph J with Lightspeed mk2 and we will make listening test over the depressing holidays 🙂
I will post my impressions.
cheers
Alan
I was just unsure because almost all people claim that OScon are just suitable for digital circuits. I strongly believe in your Lightspeed because an old friend is working for Marconi Germany (Space communication) and we talked about the Lightspeed. He showed me several industrial uses of opto-devices....way back to the early 80s. Unbelievable......i have no more doubts about it!
Its funny that most people are so skeptical about the Lightspeed. I think you posted your equipment details already.
I hope that my Lightspeed mk2 will play in my active crossover system next month......my friend is building his integrated Aleph J with Lightspeed mk2 and we will make listening test over the depressing holidays 🙂
I will post my impressions.
cheers
Alan
Ulf'
The #990,1001 numbers are the post numbers in the thread - red ones refer to the "standard" L.S, the blue ones are for the CCS versions and the green are for all the other things - it's a thread summary of "essential" post (IMo!) from start to finish - no direct links
Alan,
The version I did is a variation of Pietjers on post #1001 and It's hardly changed it at all since, which is very unusual!
I'm looking forward to Thomas's finished design (Tolu) using current Dacs for control instead of a pot - I think this will be very useful indeed.
The #990,1001 numbers are the post numbers in the thread - red ones refer to the "standard" L.S, the blue ones are for the CCS versions and the green are for all the other things - it's a thread summary of "essential" post (IMo!) from start to finish - no direct links
Alan,
The version I did is a variation of Pietjers on post #1001 and It's hardly changed it at all since, which is very unusual!
I'm looking forward to Thomas's finished design (Tolu) using current Dacs for control instead of a pot - I think this will be very useful indeed.
@ JamesHillj
I am also planning to build Pietjers Layout on post #1001....I want to check different versions of the lightspeed. If you have any Tipps for Pietjers Version.....threw it in my direction.
One questions: P1/P2 in his Layout are precision trim pots....for what purpose?
cheers
Alan
I am also planning to build Pietjers Layout on post #1001....I want to check different versions of the lightspeed. If you have any Tipps for Pietjers Version.....threw it in my direction.
One questions: P1/P2 in his Layout are precision trim pots....for what purpose?
cheers
Alan
Well Alan, I had a problem - my F3c has a low inputZ of about 9kR that requires a buffer and this also seemd to indicate that the Mk2 version would be most suitable.
I actually matched 2 pairs very closely and so didn't worry about those trimming pots for either series pair or shunt pair and I haven't noticed any channel imbalance at all - if there is any, it's reasonable and doesn't change with volume. I didn't get all fancy about it as I intended doing a pcb for the it plus the buffers but it works so well, I haven't got around to it yet.
Speaking of the pot - I borrowed Peter's DACT idea and got a cheap stereo stepper from Hong Kong to replace the cheap pot - probably not necessary, but avoids any potential problems.
The buffers aren't quite right yet but, like the LS, it's a bit of a rough "lash up" job - I also tried the rear o/p valve pair of the Aikido preamp as a buffer but not a great succes.
- intend to do a proper one with the Cavalli OCL and possibly when Thomas finishes his dac control design, include this into a pcb, too.
The Toole-Reg pcb group buy is growing - I'll also try these regs on the LS, too - expect improvement in buffer performance, not sure about the LS.
Considering the "prototype" style of the device, it's remarkable how well it works and how "neutral" - it doesn't seem to have any sound characteristic of it's own - quite rare, even in "progear".
For those readers looking for more info about Peter design and further details, print out the thread posts 990, 1001, 1017, 1032 and 1118 all together - makes it simpler.
I actually matched 2 pairs very closely and so didn't worry about those trimming pots for either series pair or shunt pair and I haven't noticed any channel imbalance at all - if there is any, it's reasonable and doesn't change with volume. I didn't get all fancy about it as I intended doing a pcb for the it plus the buffers but it works so well, I haven't got around to it yet.
Speaking of the pot - I borrowed Peter's DACT idea and got a cheap stereo stepper from Hong Kong to replace the cheap pot - probably not necessary, but avoids any potential problems.
The buffers aren't quite right yet but, like the LS, it's a bit of a rough "lash up" job - I also tried the rear o/p valve pair of the Aikido preamp as a buffer but not a great succes.
- intend to do a proper one with the Cavalli OCL and possibly when Thomas finishes his dac control design, include this into a pcb, too.
The Toole-Reg pcb group buy is growing - I'll also try these regs on the LS, too - expect improvement in buffer performance, not sure about the LS.
Considering the "prototype" style of the device, it's remarkable how well it works and how "neutral" - it doesn't seem to have any sound characteristic of it's own - quite rare, even in "progear".
For those readers looking for more info about Peter design and further details, print out the thread posts 990, 1001, 1017, 1032 and 1118 all together - makes it simpler.
Well,
for my Aleph J (242Kohm) the lightspeed Mk2 (CCS or not) is okay. But my Tripath Class D will nee a buffer...but I will solve this with the B1 Buffer.
Another question:
I did read the whole thread.....at nights....had a "head-drops-on-notebook" Incident at 4 am 🙂 Did anyone post pictures about the matching workflow?
thanks
Alan
for my Aleph J (242Kohm) the lightspeed Mk2 (CCS or not) is okay. But my Tripath Class D will nee a buffer...but I will solve this with the B1 Buffer.
Another question:
I did read the whole thread.....at nights....had a "head-drops-on-notebook" Incident at 4 am 🙂 Did anyone post pictures about the matching workflow?
thanks
Alan
As we say in Germany: Don't shoot with canons at sparrows. 😉 or for the english ones: don't break a butterfly on a wheel
If you use a well built PS with not too much ripple, the effect of additional caps isn't audible. Don't expect wonders by using caps! The discussion about the brand hasn't any influence, too. Remember, the caps aren't in the signal path and the LEDs are fed with DC (more or less!🙄 ).
I am not so experienced with LS as George but I did some experiences, too.
1. The main thing is device matching, device matching, device m...!
2. Keep output impedance low!
3. A well built PS with 7805 is sufficient (for my wooden ears).
4. Use thermal coupling of the optocouplers.
5. Omit relais for channel switching.
6. If you want remote control and perfect channel matching with defined steps -> use DACs
7. Avoid buffers until they aren't absolutely necessary
Enjoy airlight listening!
If you use a well built PS with not too much ripple, the effect of additional caps isn't audible. Don't expect wonders by using caps! The discussion about the brand hasn't any influence, too. Remember, the caps aren't in the signal path and the LEDs are fed with DC (more or less!🙄 ).
I am not so experienced with LS as George but I did some experiences, too.
1. The main thing is device matching, device matching, device m...!
2. Keep output impedance low!
3. A well built PS with 7805 is sufficient (for my wooden ears).
4. Use thermal coupling of the optocouplers.
5. Omit relais for channel switching.
6. If you want remote control and perfect channel matching with defined steps -> use DACs
7. Avoid buffers until they aren't absolutely necessary
Enjoy airlight listening!
Hi Tolu,
my Pal "Aykman" did told me your advices.Thanks a lot. I will also build a basic version and later maybe just for fun a CCS Version.
I didnt get the matching thingie and have to read the PDF from this Thread.
Cheers
Alan (Aykman says Hi over ICQ)
my Pal "Aykman" did told me your advices.Thanks a lot. I will also build a basic version and later maybe just for fun a CCS Version.
I didnt get the matching thingie and have to read the PDF from this Thread.
Cheers
Alan (Aykman says Hi over ICQ)
One more for Thomas's List
As George said a long time ago.....
Pot the LDRs in wax. It prevents thermal effects caused by draughts etc If you don't believe me just blow on them when you have a constant current passing through them and measure the resistance.
Also I had one LDR that changed resistance when the ambient light changed. Again the wax stopped that.
As George said a long time ago.....
Pot the LDRs in wax. It prevents thermal effects caused by draughts etc If you don't believe me just blow on them when you have a constant current passing through them and measure the resistance.
Also I had one LDR that changed resistance when the ambient light changed. Again the wax stopped that.
Tolu said:
Hi Tolu,
May I have some details of your own controller ? I owned 20 LDR from Allied and I'll try to build a balanced volume controller. I was first very anxious since my first try was not a total success and because the min attenuation was not low enough for my system...
Then reading you got -72 dB sounded as good news for me...
Thank you for your feedback,
NSL-OPTOCUPLER
.subckt NSL_32SR3S 1 2 3 4
**NSL_32SR3S
* Electronics Workbench
*
* This file was created by:
* multiSIM to SPICE netlist routine
*
* Generated by: Dubmaster
* 5/2/2003 5:30:12 PM
*
*!!!BEGIN-INTERACT
* : Ion ++++f2 ;
*
* ( Internal state variables )
* 0.0 VARIABLE r1Cur
*
* .( Loading the led form ... ) cr
*
* :BEGIN_PLOT
* 0 ==>_*animation_state
* ;
*
* :OUT_DATA
* GET_INSTANCE Vsource ::V V i ==>_*r1Cur
*
* Ion *r1Cur 0.00004999 f.- f.<= if
* 1 ==>_*animation_state
* else
* 0 ==>_*animation_state
* endif
* ;
*
* : BEGIN_ANALYSIS UPDATE_SETTINGS ;
*
* !!!END-INTERACT
*
xLED2 9 2 LED_yellowLED2
*
.SUBCKT LED_yellowLED2 K A
D1 1 A LED_mod
V K 1 DC 0.0005
.MODEL LED_mod D(Is=1e-14 Rs=0 Cjo=0 Vj=1 Tt=0 M=500m n=1e-6 )
.ENDS
*!!!BEGIN-INTERACT
* : Ion ++++f2 ;
*
* ( Internal state variables )
* 0.0 VARIABLE r1Cur
*
* .( Loading the led form ... ) cr
*
* :BEGIN_PLOT
* 0 ==>_*animation_state
* ;
*
* :OUT_DATA
* GET_INSTANCE Vsource ::V V i ==>_*r1Cur
*
* Ion *r1Cur 0.00004999 f.- f.<= if
* 1 ==>_*animation_state
* else
* 0 ==>_*animation_state
* endif
* ;
*
* : BEGIN_ANALYSIS UPDATE_SETTINGS ;
*
* !!!END-INTERACT
*
xLED1 1 7 LED_yellowLED1
*
.SUBCKT LED_yellowLED1 K A
D1 1 A LED_mod
V K 1 DC 0.0005
.MODEL LED_mod D(Is=1e-14 Rs=0 Cjo=0 Vj=1 Tt=0 M=500m n=1e-6 )
.ENDS
xU1 18 0 4 3 V_CONT__RES__2__28
xU2 7 9 18 16 V_CONT__RES__2__23
rR2 0 18 1k resR2
.model resR2
+r(
+
+ )
vV1 16 0 dc 2.5
+
+
.SUBCKT V_CONT__RES__2__28 1 2 4 5
RIN 1 2 1G
EOUT 4 6 POLY(2) (1,2) (3,0) 0 0 0 0 1
FCOPY 0 3 VSENSE 1
RSENSOR 3 0 959K
* CURRENT SENSE
VSENSE 6 5 0
.ENDS
.SUBCKT V_CONT__RES__2__23 1 2 4 5
RIN 1 2 1G
EOUT 4 6 POLY(2) (1,2) (3,0) 0 0 0 0 1
FCOPY 0 3 VSENSE 1
RSENSOR 3 0 959K
* CURRENT SENSE
VSENSE 6 5 0
.ENDS
.ENDShttp://www.dubmusic.com/
.subckt NSL_32SR3S 1 2 3 4
**NSL_32SR3S
* Electronics Workbench
*
* This file was created by:
* multiSIM to SPICE netlist routine
*
* Generated by: Dubmaster
* 5/2/2003 5:30:12 PM
*
*!!!BEGIN-INTERACT
* : Ion ++++f2 ;
*
* ( Internal state variables )
* 0.0 VARIABLE r1Cur
*
* .( Loading the led form ... ) cr
*
* :BEGIN_PLOT
* 0 ==>_*animation_state
* ;
*
* :OUT_DATA
* GET_INSTANCE Vsource ::V V i ==>_*r1Cur
*
* Ion *r1Cur 0.00004999 f.- f.<= if
* 1 ==>_*animation_state
* else
* 0 ==>_*animation_state
* endif
* ;
*
* : BEGIN_ANALYSIS UPDATE_SETTINGS ;
*
* !!!END-INTERACT
*
xLED2 9 2 LED_yellowLED2
*
.SUBCKT LED_yellowLED2 K A
D1 1 A LED_mod
V K 1 DC 0.0005
.MODEL LED_mod D(Is=1e-14 Rs=0 Cjo=0 Vj=1 Tt=0 M=500m n=1e-6 )
.ENDS
*!!!BEGIN-INTERACT
* : Ion ++++f2 ;
*
* ( Internal state variables )
* 0.0 VARIABLE r1Cur
*
* .( Loading the led form ... ) cr
*
* :BEGIN_PLOT
* 0 ==>_*animation_state
* ;
*
* :OUT_DATA
* GET_INSTANCE Vsource ::V V i ==>_*r1Cur
*
* Ion *r1Cur 0.00004999 f.- f.<= if
* 1 ==>_*animation_state
* else
* 0 ==>_*animation_state
* endif
* ;
*
* : BEGIN_ANALYSIS UPDATE_SETTINGS ;
*
* !!!END-INTERACT
*
xLED1 1 7 LED_yellowLED1
*
.SUBCKT LED_yellowLED1 K A
D1 1 A LED_mod
V K 1 DC 0.0005
.MODEL LED_mod D(Is=1e-14 Rs=0 Cjo=0 Vj=1 Tt=0 M=500m n=1e-6 )
.ENDS
xU1 18 0 4 3 V_CONT__RES__2__28
xU2 7 9 18 16 V_CONT__RES__2__23
rR2 0 18 1k resR2
.model resR2
+r(
+
+ )
vV1 16 0 dc 2.5
+
+
.SUBCKT V_CONT__RES__2__28 1 2 4 5
RIN 1 2 1G
EOUT 4 6 POLY(2) (1,2) (3,0) 0 0 0 0 1
FCOPY 0 3 VSENSE 1
RSENSOR 3 0 959K
* CURRENT SENSE
VSENSE 6 5 0
.ENDS
.SUBCKT V_CONT__RES__2__23 1 2 4 5
RIN 1 2 1G
EOUT 4 6 POLY(2) (1,2) (3,0) 0 0 0 0 1
FCOPY 0 3 VSENSE 1
RSENSOR 3 0 959K
* CURRENT SENSE
VSENSE 6 5 0
.ENDS
.ENDShttp://www.dubmusic.com/
Attachments
jameshillj said:
Any update on this?
I also have a F3 and looking for the most transparent way to control the volume.
ondesx said:
The trick is various input impedance. In order to get -72dB out of a 32SR2 you have to rise resistance of the series part at around 50kOhm.
Telstar said:
Hi Telstar, I now have a few customers with Nelsons amps, and the low input impedance is a problem with the Lightspeed Attenuator.
Nelson has come up with an answer by designing a buffer for it, a few pages back which cures all the problems. My customers have been putting it at the input, inside their Pass amps as there is plenty of room and it then makes the Pass amps more compatible with other high output impedance pre's not just the Lightspeed, like Passive Pots, Dact passives, TVC passives, and many Tube Pre's.
Cheers George