simple Phono preamp

Hy

I want to build  a simple phono preamp

Ive got thie schematic from somewhere, i think a friend gave it to me and said its a ggod one ;-)

https://www.dropbox.com/s/gxuuu8xxcz2x1ba/Schaltplan.jpg?dl=0

So my question is, C3 and C11 are 47uF Caps, what should i use for the, ive checked WIma, but they are pricey and realy big, should i use bipolar elcos?
Or something different?

Thanks in advance

rico

I have designed several phono preamps over the decades. I see things I like in that design and things I don't.  All those big electrolytic caps are the things I don't love.

I like that the RIAA eq is split up into two parts, and the gain is shared across two gain stages so one op amp doe not have deliver the full 60 dB of gain.

Substituting a modern JFET input op amp with improved DC specifications could eliminate the need for a few of those DC blocking caps. For example the phono cartridge could be corrected directly to the + input of a JFET op amp.  Typical input termination is 47k ohms so one resistor could replace the 2x 100k. 

With low DC offset and  a DC blocking cap feeding the second gain stage C4 might be replaced with a jumper. A few tens of mV at the output of the first stage will not impact headroom.

With the first stage running at 46 dB of LF gain the second stage does not need to provide up to another 40 dB gain on top of that. Nominal LF gain for phono preamps is 60 dB so that preamp is already 6dB hot at 66dB with the second gain stage turned full down. .

While I would substitute more modern op amps, but if you stay with the old school 5534, there is no need to compensate them with full 22pF. The minimum gain at HF of the first stage is 26dB and the second stage minimum is 20 dB so both better the 10 dB closed loop gain for stability without any compensation caps.

R10 and R11 could be replaced with a single 5.0k resistor. I would be inclined to use 1% Rs for RIAA EQ components, and quality film capacitors for C8, C9, and C10.

I like that the 75uSec RIAA pole is passive so the roll off adheres to the RIAA response above the audio band, while this is pretty esoteric and the cutting lathes do not provide RIAA boost up to light frequency.  If the second stage was inverting with the 75usec pole across the feedback R it would also roll off the noise of the second gain stage but that is probably not audibly significant (noise will be dominated by the first stage). 

It seems like C12 and  C16 are redundant but there are many ways to skin the DC cat.

That preamp will probably sound fine as drawn, but looks like several extra parts to me.

JR

dubwarrior said:

Hy

I want to build  a simple phono preamp

Ive got thie schematic from somewhere, i think a friend gave it to me and said its a ggod one ;-)

https://www.dropbox.com/s/gxuuu8xxcz2x1ba/Schaltplan.jpg?dl=0

So my question is, C3 and C11 are 47uF Caps, what should i use for the, ive checked WIma, but they are pricey and realy big, should i use bipolar elcos?
Or something different?

Thanks in advance

rico

Click to expand...

In the actual state of technology, I think you don't have much choice.
However, C3 and C11 can be substantially decreased as they are loaded by 100kohms, giving a -3dB LF corner of 0.03Hz.
C3 could easily be decreased to 1uF with no noticeable effect.
As for C11 and R12, they have no real use, one could easily do without them; I would just connect the output of IC1 directly to the non-inverting input of IC2.
Even C16 could probably be reduced in view it will probably never see a load smaller than 10k, which results in a -3dB LF corner of 0.3 Hz, or a response only 0.0012dB down at 20Hz.
Finally, do they need to be bipolar?
Regarding C16, yes, since one does not know whatever DC offset would present the load to which the preamp is connected; indeed, many commercial products use standard polarised caps because they are perfectly capable of withstanding small amounts of reverse voltage  - WARNING: not true of tantalum caps.
Then regarding C3, the offset at the input of IC1 is always negative because the input transistors are NPN; as a result, you could use polarised caps with the negative side connected towards the opamp's input.
As to C16, it's a little more complex because the DC offest at the output of IC1 is the combination of the effects of the input bias current in the 100k (which is negative as in the first stage) minus that produced in the feedback network (which is less because the resistance is smaller) AND the opamps input offset voltage. IMO, the best option is to short it out, which solves the issue definitely; other options are:
a) a bipolar cap
b) a polarised cap placed in accordance to effective DC measurements
c) a polarised cap placed randomly, knowing that the very low voltage would not have any really significant consequence

I have recently designed a new phono pre for moving coil cartridges only. It uses a true balanced input stage, that is followed by a passive RIAA network. My preference is to avoid equalization networks in the feedback loop of a circuit. The passive eq sounds much cleaner than a feedback  type equalizer. I have designed my preamp as a commercial product to go head to head with units like the Lehmann Black Cube.

For a moving magnet phono preamp, the first stage would likely need a gain of 40 db or a gain of 100, followed by a passive eq network & an output amplifier with gain to make up the losses in the network. The analog devices AD823 would be a good chip for the output stage.

Bill Wilson said:

I have recently designed a new phono pre for moving coil cartridges only. It uses a true balanced input stage, that is followed by a passive RIAA network. My preference is to avoid equalization networks in the feedback loop of a circuit. The passive eq sounds much cleaner than a feedback  type equalizer.

Click to expand...

Can you substantiate that with measurements, or a theory that woud explain it? Technically speaking, it is challenging.

Bill Wilson said:

I have recently designed a new phono pre for moving coil cartridges only. It uses a true balanced input stage, that is followed by a passive RIAA network. My preference is to avoid equalization networks in the feedback loop of a circuit. The passive eq sounds much cleaner than a feedback  type equalizer. I have designed my preamp as a commercial product to go head to head with units like the Lehmann Black Cube.

For a moving magnet phono preamp, the first stage would likely need a gain of 40 db or a gain of 100, followed by a passive eq network & an output amplifier with gain to make up the losses in the network. The analog devices AD823 would be a good chip for the output stage.

Click to expand...

I haven't designed a MC preamp since the early 1980s... Note: the schematic (Published as a kit article in Popular Electronics) shows JFET inputs but the MC version used some nice low noise PNPs (2sb737s) with bias flipped and more fixed gain. .

First stage is flat gain stage, EQ is wrapped around an inverting op amp stage. WRT NF and RIAA EQ, the open loop pole in the TL07x internal compensation is not that far from the 3180 uSec RIAA pole so the open loop gain is close to being in phase with the closed loop gain.. While this is all pretty subtle esoteric phoolery that I left behind decades ago.  I only ran my MM flat gain stage at 30 dB gain. Since RIAA is nominally 20 dB at 20 kHz, anything more than that in a flat front end will come out of HF headroom. So 40dB front end will clip 20 dB before output at 20kHz... Not a huge problem because 20kHz is not that loud coming off vinyl but there is a trade off with flat front ends.   

My later MM only preamp used a completely open loop first gain stage feeding a passive 75uSec pole before hitting any NF circuitry so you could hit that preamp input with lightning edge rates and never slew limit it. But guess what, vinyl and cartridges to not move fast enough to  overload typical preamps, so save your pennies and use almost any old circuit design with accurate RIAA EQ, just substitute newer  low noise (JFET input) op amps. 

JR

PS an old friend and former regular here (Brad)  has recently designed  a new MC preamp based around on uber low noise JFET and some circuit tricks (like cooled termination), but not my story to share..

Hi Abbey: Listening test I conducted confirmed my conclusion that a passive RIAA network sounded cleaner than a feedback network, that could induce some subtle phase shift.

Bill Wilson said:

Hi Abbey: Listening test I conducted confirmed my conclusion that a passive RIAA network sounded cleaner than a feedback network, that could induce some subtle phase shift.

Click to expand...

Bill, have you ever looked into what the actual difference is? You could null out a passive vs. active RIAA  EQ circuit to determine what the errors are.

I suspect if the amplitude response of both is trimmed accurate to RIAA frequency response within a modest fraction of dB 20-20k, the phase response will probably be pretty close too.

JR

Bill Wilson said:

Hi Abbey: Listening test I conducted confirmed my conclusion that a passive RIAA network sounded cleaner than a feedback network, that could induce some subtle phase shift.

Click to expand...

Bill, this is not what I accept as substantiating.
Listening tests are usually fraught with errors due to incorrect level matching, different frequency response, auditive fatigue, inconscious bias...
PAttributing the differences to the phase response is dubious; phono preamps being minimum-phase devices, as long as the frequency response is the same, so is the phase response. In addition, audibility of phase response is as elusive as the perception of ultra-violet light.