Yes I did (and do) extend LED's using light-conductors.
No I never did extend LED-references, like in the Paradise PSU, and did not simulate.
The point is, extending the LED with a wire, add's 10...30cm (just guessing) of dual conductor wire, the loop in this wire will add multiple uH's of inductance, and pF's of capacitance, this will never be the smart option.
P.s. [Not recommended] when extending a LED then add a largish capacitor (10uF + 1uF(cer)) in parallel to the extension (on the board).
No I never did extend LED-references, like in the Paradise PSU, and did not simulate.
The point is, extending the LED with a wire, add's 10...30cm (just guessing) of dual conductor wire, the loop in this wire will add multiple uH's of inductance, and pF's of capacitance, this will never be the smart option.
P.s. [Not recommended] when extending a LED then add a largish capacitor (10uF + 1uF(cer)) in parallel to the extension (on the board).
Last edited:
Just tap from the incoming rectified raw DC line to a panel LED. Why mess with the sensitive Vref branch anyway.
IMHO don't use more than 1-1.5mA for a panel LED or it will be too bright in the night. Matter of choice of course.
The formula for determining its current will be Ia=(dc-Vf)/R. The R is a series protection resistor to the panel LED. Example: When having 25V raw DC and a red panel LED of 1.7V forward drop spec, a 22k resistor will allow it about 1mA current. The dissipation on the resistor will be 23.3mW only, so any 1/4W type will do.
IMHO don't use more than 1-1.5mA for a panel LED or it will be too bright in the night. Matter of choice of course.
The formula for determining its current will be Ia=(dc-Vf)/R. The R is a series protection resistor to the panel LED. Example: When having 25V raw DC and a red panel LED of 1.7V forward drop spec, a 22k resistor will allow it about 1mA current. The dissipation on the resistor will be 23.3mW only, so any 1/4W type will do.
)I've done some comparison against my Whest stage and not sure which I like better , Paradise seems to be more 3D , valve like sounding in a good way but Whest has definitely much better bottom end , Paradise bass seems to be a bit blurry .
I wonder if there is a bit of mismatch between Paradise output and preamplfier input impedance , my pre is quoted at 22k input impedance what's Paradise output impedance ?
I wonder if there is a bit of mismatch between Paradise output and preamplfier input impedance , my pre is quoted at 22k input impedance what's Paradise output impedance ?
I'm not sure if I'm doing it right but I run 1kHz and 20Hz sine wave into the input and checked output without and with a load and used following equation :
Z= Rload (Vunloaded/Vload - 1) ,
output changes when I load it with 10ohm resistor and impedance value is around 7ohm at 1KHz and 9ohm at 20Hz , seems to be very low .
Z= Rload (Vunloaded/Vload - 1) ,
output changes when I load it with 10ohm resistor and impedance value is around 7ohm at 1KHz and 9ohm at 20Hz , seems to be very low .
I had found the standard alignment bit overripe in my speakers/room and I used a bit leaner alignment in the sub bass curve myself back then. By utilizing half the standard emitter decoupling capacitance C6 C7.
Without an external buffer it won't be more than the RDS of Q99+R12 I believe, so roughly 40-50 Ω if with a K170BL (RDS per sample is inversely analogous to Q99's IDSS).
In my build I get +0.7dB @20Hz and +0.15dB @ 20kHz , compared to Whest it don't seem like more low bass just bit uncontrolled so I suspected impedance mismatch but it seems like I was wrong. Above impedance was tested with Calvin buffer in place , was I doing it right? I'm not that great with electronics and all I learn/read comes from internet .
Sounds logical its less Zo with the Calvin than with the integral JFET buffer alone. Its not only about Zo, its about available drive current boost too. Good for winning over HF capacitance in long cables.
To verify your Zo measurement use a 1 Meg resistor then a 100 Ohm trimmer as your load. Where your output signal halves for trimmer value vs the resistor level its your Zo.
To verify your Zo measurement use a 1 Meg resistor then a 100 Ohm trimmer as your load. Where your output signal halves for trimmer value vs the resistor level its your Zo.
Your figures seem to match up with mine, response +0.6/+0.1dB 20Hz-20kHz,
I can't remember exactly Zo value but it was fairly low .
To bring response @20Hz closer to ideal parallel 73k5 resistor with 1M as advised by Salas .
Thank you , I will try to take measurements this way , are there any dangers of output impedance being too low with certain loads?
You could max out the output current easier while trying to meet a truly low Zo with the load matching output voltage halving method. Could put a power amp into protection mode or blow something inside if unprotected. Can also burn the test load itself if driven high. Plus it would be difficult to experimentally define in sub Ohm territory. Not that likely in a circa ten Ohm Zo line buffer though.
Just keep the output signal "unloaded" reference level low. The one you get on the Meg load. Say 100mV. Then additionally load it with the 100 Ohm trimmer wired as a rheostat and wind it down until you see 50mV. Remove it and measure its set value. That's surely equal to your Zo because two same value resistors one in series (Zo) the other in parallel (rheostat) create an exactly halving L-Pad attenuator.
There is no too low Zo, its textbook desirable in an AC voltage source, just more damping factor. Subjectively you may want some less damping in the system or objectively more Zo maybe stabilizes some GNFB circuit better (gives more open loop phase margin) but that's a design phase decision. You see in line level circuits a series resistor after an output op-amp many times. Ranging widely between ten Ohm to one kOhm.
Just keep the output signal "unloaded" reference level low. The one you get on the Meg load. Say 100mV. Then additionally load it with the 100 Ohm trimmer wired as a rheostat and wind it down until you see 50mV. Remove it and measure its set value. That's surely equal to your Zo because two same value resistors one in series (Zo) the other in parallel (rheostat) create an exactly halving L-Pad attenuator.
There is no too low Zo, its textbook desirable in an AC voltage source, just more damping factor. Subjectively you may want some less damping in the system or objectively more Zo maybe stabilizes some GNFB circuit better (gives more open loop phase margin) but that's a design phase decision. You see in line level circuits a series resistor after an output op-amp many times. Ranging widely between ten Ohm to one kOhm.
Pete try adjusting your cart loading , dialing back the hf peak can give the impression of more bass. Dropping the cart loading value will reduce the hf.
With my Benz lp loading g makes negligible difference, with some carts like Ortofons, 10 ohms up or down seems to make quite obvious changes.
Seems like low Internal impedance carts, sub 10 ohms are most affected.
With my Benz lp loading g makes negligible difference, with some carts like Ortofons, 10 ohms up or down seems to make quite obvious changes.
Seems like low Internal impedance carts, sub 10 ohms are most affected.
Looking for....
Hi, I am about to start building my own Paradise and after I have read the thread here I realize that someone out there who have also built the Paradise is most likely sitting on a heap of BC327:s and BC337:s. Not that I can't buy an awful lot of these myself and sit for a day or so to measure them all but most likely someone is in the position that they want to get rid of the surplus themselves so..... Feel free to contact me if you have a heap to sell that are measuring nicely and not too much spread.
Hi, I am about to start building my own Paradise and after I have read the thread here I realize that someone out there who have also built the Paradise is most likely sitting on a heap of BC327:s and BC337:s. Not that I can't buy an awful lot of these myself and sit for a day or so to measure them all but most likely someone is in the position that they want to get rid of the surplus themselves so..... Feel free to contact me if you have a heap to sell that are measuring nicely and not too much spread.