Chipomatic balanced input RIAA

Finally finished soldering. (Of course not as good as dimdim) With the exception of the LT regulators - I promised never do that again! - not too difficult. I used just a magnifying glass and a normal size iron.
It works fine electrically, now to make it a practical preamp to test it.
I used 33R for RG in order to get 45 dB gain. I will use some 12dB transformers I already have, for my cart.
 

Attachments

  • IMG_20190118_190253[1].jpg
    IMG_20190118_190253[1].jpg
    414.5 KB · Views: 406
diyAudio Chief Moderator
Joined 2002
Paid Member
Has -20dB RIAA equalized auxiliary output also, to can bypass the second op_amp stage with different external line level gain flavors. Be it classic triode or Nutube or JFET or BJT. An easy task because there's 30mV RMS flat signal readily provided to start with.
 
Hope you wouldn't mind a few remarks.

I built my first phono 10 years ago also on a Gain-Filter-Gain topology, first using AD797 / OPA637, later discrete opamps.
I always have a feeling that it lags headroom, because the filtering at HF comes after the first stage.
No scientific proof, just a feeling, when listening against other topologies.
https://www.diyaudio.com/forums/pas...screte-opamp-dip-8-package-6.html#post1835752

It is soon to be replaced by Scott's phono which is gain and filtering in one stage.
So the above problem will not exist as long as the JFETs have enough bias.
A simplified universal differential or single ended phono preamp

AD8429 is really a great chip.
We used it for a diff probe with great success.
https://www.diyaudio.com/forums/equ...ential-oscilloscope-probes-3.html#post5091048
Glad to see someones trying it out for audio.

We have good experience with Panasonice ECHU PPS film caps.
They have distortions down to -140dB level (measured).
One can argue that COG can do better in measurements.
But in listening we prefer ECHU in the end.
Again no scientific proof, but maybe worth trying.

I personally would not use a DIP switch for sensitive signals.
Especially for DIY, everything at the phono frontend I try to avoid contacts and solder wherever possible.

Cute little PCB. Can be placed very close to the TT. :)


Patrick
 
diyAudio Chief Moderator
Joined 2002
Paid Member
@Patrick

For the time being, i.e. fun proto between mates, its a mix. C0G 1% in the long time constant (no less than 10% 5mm pitch Wima MKP for C1's value available on Mouser) and MKP//FKP C2 combo of various better % in the short constant. Best to be checked and matched for total value with a good LCR meter. MKP1837 1% is not currently stocked in the majors by the way. Although I got some.

I even have 1% NOS film & foil ERO in a drawer that fits dimensions, but that wouldn't go into a real life BOM for others.
The switch exchanges GND ref not switching the signal directly. Its a must for now regarding BAL/SE tests flexibility though.

I had tested C0G for both the low and high turn points as RIAA caps in the past and they reminded me of Silver Mica. Non aggressive and musical if wee bit veiled vs polystyrene in other words. Given the curve is not doctored to suit them better. For the lows maybe more difficult to distinct.

I had an eye on those ECHU for 100% SMD parts use but their 0.6% tanδ kinda held me back vs MKP//FKP Visahy/Wima through hole in the highs.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I would like to acquire your 2 independent channels in a 70mmx60mm card with all the extras and all the modern comforts.
I do not need 4kTV, sofa, espresso machine and a free trip to Mykonos just your prototype.
If you accept of course ;)

Its not a real proven project yet and I only got three board samples. If its successful we will see.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Got to measure some samples then since they look promising. As to why they state equal to or smaller than 0.6% at 1kHz in a short datasheet (the Sep 2017 one without graphs linked in Mouser that I read when I was searching) as well as they state the same in the spec headlines of the expanded one you linked here, I would not guess.
 
Member
Joined 2004
Paid Member
The dissipation factor of ECHU is well under 0.1% in the audio band :
https://cdn-reichelt.de/documents/datenblatt/A900/DS_ECHU_SERIE.pdf
(See figures in P.3)
At 1kHz it is essentially zero.

PPS capacitors are my first choice for RIAA networks and line-level. But I have made some bad experiences when soldering these caps (ECHU, SMD) by hand. They quickly began to melt and changed their capacitance. Now I use leaded Kemet SMR Series, available in 5% at Mouser.


Boris
 
We have measurements in an oscillator circuit.
They were -140dB whereas WIMA FKP2 was -160dB.
You can argue 10x worse. But -140dB is just fine for me.

They are, like all SMD film caps, fragile.
You need to lower iron temperature to 300°C and use fine (0.3mm) leaded solder.
Then temperature is not so high, and also not so much heat energy applied.
Desoldering requires 2 irons. And you need to discard afterwards.
But otherwise we used them a lot without problems.
https://www.diyaudio.com/forums/analogue-source/54658-jfet-loop-riaa-pre-amp-4.html#post5438472


Patrick
 
diyAudio Chief Moderator
Joined 2002
Paid Member
One populated PCB:

728323d1547254897-chipomatic-balanced-input-riaa-chipomatic-prototype-medium-jpg


Power-up was uneventful, no magic smoke in sight.

Power draw was ~18mA / rail / channel.

Proper testing will be done tomorrow..

There's a layout bug to fix before you do some testing with signal. R6 must be moved between R5's left side and RN1's ground side, else you will get unity gain from IC2 instead of 10X. George called to ask why no gain from stage two and we debugged it over the phone.
 
I see it now. My bad, going from your (correct) schematic to my (Eagle's) schematic.

Funny enough, I was looking at my Eagle's schematic the other day when I was ordering the components and that part of it looked a bit odd, but I didn't pay much attention. It appears that I should have.