Simplistic NJFET RIAA

Member
Joined 2006
Paid Member
Thank you SGregory..... I will do that.

Even without calibration, if I use a small coil, I will be able to determine the stronger fields and shield accordingly :)

Regards

Ricardo

PS: I believe I underestimated the effect of isolating CCS GND from the input. There is a whole new level of detail and presence possible with this simple mod :)
 
Thanks.

I knew I was asking a lot to consider "layout" as an exercise in minimising loop area for each circuit.

Your changes show how extensive this can become and I suspect there is more to be done in improvements yet. Are any of the possible/alledged improvements audible?

Layout is important, both for parasitics and for interference, from external influence and from internal current circulations.
 
Member
Joined 2006
Paid Member
Both mods are clearly audible. The first one reduced overall loop area and I was rewarded with a lower noise floor as well as a lower sensivity to fields.

The second mod (Isolating CCS GND) surprised me with an overall better sound balance, greater detail and larger soundstage.

I will try to isolate GND lines on second stage also :)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Some medium size two box towers of my design with PHL mid, Audax TW, and 2xTangband car audio oval woofers. Closed box in the lows, restricted flow vent in the mids. Here is their true room response with RTA (not windowed MLS).
 

Attachments

  • SalasTsynRTA.gif
    SalasTsynRTA.gif
    14 KB · Views: 332
My build is coming along nicely, however, still waiting on some components to arrive.
FYI the riaa stage will be fully encase in a mild steel box as was discussed before. Arrangement turned out pretty well. I will post pictures later.

In the meantime, I am pondering the best way to measure this circuit. The input impedance of the riaa is 150R which means I should have a signal source impedance less than 15R or so. My intention was to use a simple voltage divider, so that a 0.3V signal from my soundcard would be dropped to 0.0003V (or 0.3mV) for use on the riaa. This would require the top resistor to be approx 15k. I will have to bypass the pmillet soundcard interface for this as it will roll off at the low frequencies needed when I eventually hook up the TT. I want to use the same test arrangement for both to eliminate issues. Luckly for this I think direct to soundcard would be best, once the circuit is proven, with little issue of overloading the soundcard. (P.S. I will use the interface when I initially fire up the circuit and test that it works and is stable).

Is this reasonable thinking or is there a better way to test?
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I use a 1:100 or 1:200 divider routinely for that with adequate Zo given the cart I shoot for. Its good in dividing the audio card's output noise also. Only check your card's real mV when driving the input impedance of the divider with a DVM so you calibrate and actually know you get reliable sub mV numbers after the division. The dividing resistors should be enclosed in a special purpose made terminated interconnect or a metallic adapter. If not shielded, you will see hum that isn't intrinsic in the DUT.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
A soundcard will drive an impedance of 1k to 3k?

My EMU tracker does it, but loses max out V than with say 100k. Can go from 2.17V down to 1.35V. So I check its 0dB non clipped on the actual divider as load and I feed data to the software cal. Then if I tell it to give say 250mV I check it actually listens with that load or having to trim settings. Also under 100mVout most cards don't listen that well so I see if I have to instruct it 67mV to really feed 60mV for instance.
 
A little bit of update for peer review. Attached are photo's of the steel box mounted in the chassis. I soldered a ground connection to the top and bottom and intend to run it to the chassis ground stab.

Another photo shows how the pc boards are mounted in the box with the output caps leading to the RCA connectors. The blue hose is the Force and Sense that will lead to the shunt regulators. They will be mounted on the heat sinks. Initially I will use the toroid transformer shown earlier, but after reading up on the Rcore xfmr I have one on order and will switch it out at some point in the future.

In the photo, I do not have the rear panel attached as I need to mill out the opening for the RCA connectors to pass through and mount the IEC power conector. I still intend to run the AC mains in a shielded cable or in some steel pipe to minimize the pick-ups. I made the box big enough to be able to change out the pc board with new designs if need be.

Let me know what I could do to improve it. Thanks

P.S. I have some temporary riaa caps installed while for circuit testing, I hope the Russian ones arrive soon.
 

Attachments

  • DSC02549.JPG
    DSC02549.JPG
    135.7 KB · Views: 342
  • DSC02551.JPG
    DSC02551.JPG
    133.9 KB · Views: 333
  • DSC02553.JPG
    DSC02553.JPG
    137 KB · Views: 333
  • DSC02552.JPG
    DSC02552.JPG
    145.8 KB · Views: 312
diyAudio Chief Moderator
Joined 2002
Paid Member
I say its a very nice job for housing high gain low level circuits no matter changes in the basic pcbs that might be or not. Including the minimum field R-Core and already having wide voltage adjustment range in the regs you are well equipped.

*Give it a go with the temporary caps to see if everything works without mistakes meantime.