Lately I have become curious about the DS Audio optical pickups and I think sooner or later I will get me one.
They are a bit pricey, but it's possible to purchase a DS-pickup without the equalizer, so DIYing the equalizer safes a fair amount of money.
Here I just want to gather some ideas on the construction of such an equalizer…
There is a pdf "Optical cartridge technical info" which can be downloaded on the DS website https://ds-audio-w.biz/products_info/
I've got all my information from there.
PSU requirements
The pickup consists of an IR-LED (one for both channel for the 2nd generation pickups and one each channel for the 3rd generation pickups) which needs to be powered
and two photo-diodes -one each channel- which need an negative bias voltage.
An LED (the “front light”) is in parallel to the IR-LED and seems to be an on-indicator without further purpose.
The IR-LED is powered from 8V via 68Ohm resistor. Diode Current is said to be 65mA for the DSE-1 and 80mA for the DS003.
There is a resistor in front of the IR-diode shown in the pdf, which probably serves to decouple the IR-LED from the indicator LED.
So the IR-LEDs forward voltage is likely to be approx. 2,9V.
The Photo-Diodes are reverse biased (photoconductive mode) and need a -10V bias voltage each channel.
Output voltage of the DS-E1 pickup is said to be 50mV. The schematic has a 16kOhm resistors to convert the diode output current to a voltage.
Diode-current therefore must be 50mV/16kOhm=3,125µA (5cms/sec at 1kHz???)
In the DS Audio pdf the RIAA equalization is shown as a passive network right behind the 16kOhm resistor.
I’m not shure if this is the optimal position. Also the I-V-conversion could benefit from an active solution.
In the following schematic I use a transimpedance amplifier to transform the current to a voltage.
I “lifted” the output of the TIA to -10V by putting this voltage on the non-inverting input- which in turn sets the bias for the photo-diodes.
This reduces the headroom of the Opamp. One solution could be, to make the Opamp supply asymmetrical: +5 to -25V.
The following capacitor removes DC from the signal and can be part of the succeeding equilazation-network (can we call this still a RIAA-network?).
The channels are phase inverted to each other, so the right channel still needs an inverter stage.
Any comments so far?
Cheers, Boris
They are a bit pricey, but it's possible to purchase a DS-pickup without the equalizer, so DIYing the equalizer safes a fair amount of money.
Here I just want to gather some ideas on the construction of such an equalizer…
There is a pdf "Optical cartridge technical info" which can be downloaded on the DS website https://ds-audio-w.biz/products_info/
I've got all my information from there.
PSU requirements
The pickup consists of an IR-LED (one for both channel for the 2nd generation pickups and one each channel for the 3rd generation pickups) which needs to be powered
and two photo-diodes -one each channel- which need an negative bias voltage.
An LED (the “front light”) is in parallel to the IR-LED and seems to be an on-indicator without further purpose.
The IR-LED is powered from 8V via 68Ohm resistor. Diode Current is said to be 65mA for the DSE-1 and 80mA for the DS003.
There is a resistor in front of the IR-diode shown in the pdf, which probably serves to decouple the IR-LED from the indicator LED.
So the IR-LEDs forward voltage is likely to be approx. 2,9V.
The Photo-Diodes are reverse biased (photoconductive mode) and need a -10V bias voltage each channel.
Output voltage of the DS-E1 pickup is said to be 50mV. The schematic has a 16kOhm resistors to convert the diode output current to a voltage.
Diode-current therefore must be 50mV/16kOhm=3,125µA (5cms/sec at 1kHz???)
In the DS Audio pdf the RIAA equalization is shown as a passive network right behind the 16kOhm resistor.
I’m not shure if this is the optimal position. Also the I-V-conversion could benefit from an active solution.
In the following schematic I use a transimpedance amplifier to transform the current to a voltage.
I “lifted” the output of the TIA to -10V by putting this voltage on the non-inverting input- which in turn sets the bias for the photo-diodes.
This reduces the headroom of the Opamp. One solution could be, to make the Opamp supply asymmetrical: +5 to -25V.
The following capacitor removes DC from the signal and can be part of the succeeding equilazation-network (can we call this still a RIAA-network?).
The channels are phase inverted to each other, so the right channel still needs an inverter stage.
Any comments so far?
Cheers, Boris
Would this work?
I-U-conversion + step-up with an DAC-output or MC-transformer…
Note that the photo-diodes are in reverse direction, so no DC current is flowing through the transformer coils.
In the end I would prefer a solution with tubes.
Phase inversion of the right channel can be done by swapping the primary connection of the transformer.
I-U-conversion + step-up with an DAC-output or MC-transformer…
Note that the photo-diodes are in reverse direction, so no DC current is flowing through the transformer coils.
In the end I would prefer a solution with tubes.
Phase inversion of the right channel can be done by swapping the primary connection of the transformer.
Since this pick-ups have no coils, where the signals are induced (like MM/MC/MI-systems), the output signal depends only on the movement of the tip of the needle and not of the acceleration of the tip. This is similar to strain gauge pick-ups. So the usual RIAA network won’t work to equalize the output.
Also, these pickups can generate output from 0Hz upwards- so a high pass, to remove contend below the audibility threshold, is advisable.
Here is the plot of the passive filter from DS-Audio. They use a shelving high pass from 500 to 2120Hz combined with a 50Hz high pass.
In the green graph the network is fed by a linear ac-source, whereas in the red graph it is fed by a RIAA-curve-source (left upper & lower filter).
I modified the filter to use capacitor values where I have a lot from (filter on the left). This is of cause not tested!
I have the calculations for the shelving high pass from https://linkwitzlab.com/filters.htm#6
Could somebody here be so kind and create an ac-source with the correct functional equantion for velocity pickups?
Because this is beyond my skills
Also, these pickups can generate output from 0Hz upwards- so a high pass, to remove contend below the audibility threshold, is advisable.
Here is the plot of the passive filter from DS-Audio. They use a shelving high pass from 500 to 2120Hz combined with a 50Hz high pass.
In the green graph the network is fed by a linear ac-source, whereas in the red graph it is fed by a RIAA-curve-source (left upper & lower filter).
I modified the filter to use capacitor values where I have a lot from (filter on the left). This is of cause not tested!
I have the calculations for the shelving high pass from https://linkwitzlab.com/filters.htm#6
Could somebody here be so kind and create an ac-source with the correct functional equantion for velocity pickups?
Because this is beyond my skills
The reference design using 16K I/V has -10V output.
Is that reasonable to reduce the I/V (3.3K) to get -2V output, and DC feed a single stage pentode (200-400X gain) with transformer output (10K:600).
No extra needs to be done for inverted output on transformer stage.
I've already built best pentode circuit using Lundahl 1660 to drive LCR 600 with very good result.
Is that reasonable to reduce the I/V (3.3K) to get -2V output, and DC feed a single stage pentode (200-400X gain) with transformer output (10K:600).
No extra needs to be done for inverted output on transformer stage.
I've already built best pentode circuit using Lundahl 1660 to drive LCR 600 with very good result.
Actually I was wrong, -10V after 100k/33k is only -2.5V, perfect to dc couple to tube stage.
