Simplistic NJFET RIAA

With two coax cables you form a ground loop between the parallel shields, allowing differential-mode current noise on the shields, which gets interpreted as differential voltage noise on the signal lines due to IR losses in the shields. The problem can be minimized by twisting the two coaxes together like with twisted pair, reducing the differential-mode coupling to the surroundings, but that's pretty clunky.



Simple twisted pair (in a common shield) is how to pass differential signals around best. The balance signals cancel inside the shield for minimum coupling to the outside.


Plain twisted pair with a separate ground wire is workable too, especially if the balance is accurate.
 
Hello everybody,

After quite a break from DIY I've finally started this project. It's been in the boxes from Teabag for over a year now. I put together the PSU and one board for the preamp. I have a few questions. They are pretty basic and may have been dealt with earlier in this thread but I can’t seem to find answers. Also, even though, I’ve put together quite a few projects, beginning with the DCB1, I’m still learning. I ordered Teabag’s minikits for this build but it’s still probably my most involved build, hence the questions:

1. For the heatsinks, in the guide it says not to use pads, just a drop of thermal grease on the IRF9530, which is what I did. No grease on the 9610 needed, right? Will these sinks be live then?

2. I’m not quite sure what to do on CY2. Should I just leave it blank and if I find the sound too bright solder in the 100pf polystyrene cap or the 150pf Wima cap I got with the mini kit?

3. C3, I’ll probably be using this with the DCB1 so assuming that's less than 50k impedance, so I'll use 100nF, correct? I’ll probably be using this with my Ortofon Quintet Black MC cart, 0.3mV.

4. First time I’ve made an umbilical but it seems to be working. I’ve got 46Vdc on each channel from the Raw PSU, no load yet. Does that sound about right? I’m using the Antek As-1232 100VA. I put some 3W 0.47R resistors in the link since that’s what I had on hand. Is 3W OK? Now I’m worried I might have misread a post that said “RD/Link 2-5W resistors”. I understood between 2W and 5W.

Sorry for such greenhorn questions but I just want to make sure I’ve got it right before powering everything up.

Thank you
 
diyAudio Chief Moderator
Joined 2002
Paid Member
1. Yes they will be live. But they are board level small sinks so no reason to touch them on a grounded surface. Plus they are black anodized so they have a thin insulation layer except where scratched. You avoid the insulation pads thermal resistance and some hassle with shoulder washers when mounting that way, but its your choice to insulate if afraid of an odd incident with a probe or something.

2. In my view solder the 150p for a start. From measurements it takes enough C2Y value to be ruler flat. Up to 15.4nF C2+C2Y. But compatibility with records, cartridge, turntable, and speakers tone is something to be subjectively judged. Enough belt driven TTs run bit slow for instance. Because this circuit is mainly doing second order THD, a slight upwards tilt seems subjectively compatible with vinyl rigs in various installations when listening over two meters away with dome tweeter speakers in non acoustically treated rooms. Talking just small fractions of a dB tailoring.

I attach the RIAA transfer curve and vs flat net outcome response plots of 15.2nF confirmed total. The anti-RiAA filtering curve to produce a net response is ideal math that I put in the software. Introduces no hardware anti-Riaa errors. Vinyl record cuts are not to be expected with an ideal pre-emphasis curve of course. They vary. Phono output load was the audio card's instruments input, so it was high. Circa 1 Mega Ohm. C4's listed value is chosen to create more subsonic attenuation on usual line preamp up to 50k loading.

3. 100nF C3 seems right in that case.

4. Yes, 2-5W means 3W is in the middle of that acceptable power rating range. Your raw DC voltage level is also fine.
 

Attachments

  • FSP_15n2_TR.png
    FSP_15n2_TR.png
    59.4 KB · Views: 510
  • FSP_15n2_FR.png
    FSP_15n2_FR.png
    46.1 KB · Views: 510
Today I finally finished one channel, got it biased easily, and hooked it up to my system. Just the one channel sounded fantastic. Very detailed and no hum whatsoever. Bravo Salas! So, I couldn’t wait to finish the other channel. Unfortunately, in my enthusiasm, I stupidly soldered Q5x in backwards. I’d done that before so I knew what I had to do but this time, while desoldering, two of the leads broke off about a 1mm away from the body. I guess I could try soldering some extension leads on to the nubs but I would like do this properly and put in a fresh part. I have two 2SK117grs left from my Teabag mini kit. One measures 5.43mA for IDSS and the other 2.82mA. Here are the IDSS numbers for ones I’ve already got soldered in:

Right channel(completed):
Q2x: 3.01mA
Q3x: 3.4mA
Q5x: 3.87mA

Left channel:
Q2x: 3.14mA
Q3x: 3.41mA
Q5x: the one I soldered in backwards was 3.54mA.

So, I’m wondering if I can you use my spare 2.82mA for Q5x or if I have to try and find a replacement one with IDSS between 3-4mA.

At any rate, I am really happy with what am hearing so far.

Thanks,
Dan
 
diyAudio Chief Moderator
Joined 2002
Paid Member
You should use the 2.82mA. For that Q5x CCS collector load position a moderate IDSS mismatch won't create practical differences between the two channels PSUs behavior. Its if it was too high IDSS it would have created borderline dissipation for the Q4x bipolar because the Q5x CCS's bias current is multiplying for many mW with the high VCE across Q4x.
 
which build to build?

Hi I'd like to build this circuit and I'm going to build it myself on a perfboard. This thread is awesome but has way too many choices.

My system is a Klisch Fotres with DIY Al Klapeberg crossovers.

An Aleph J amp, stock build_no mods yet.

Currently I'm tweaking a Korg B1 nutube preamp, but also just recieved a mesmerize board and plan to build Zen Mod's Iron Pumpkin early fall.

Cartrige is a Nagaoka MP150 MM 4.5mV out.

I have 2 nice sets of 2sk170BL matced octets one at 9.9mA IDSS and one at 8.0/8.1mA IDSS plus enough spares for the mesmerize.

I'd like to be able to use all of the preamps with little or minor tweaking on the phono stage and plan to start phono off with battery power then add shunts later.

So what circuit should I build? maybe a 36db version? And if so which one there look to be a few.

Jeffrey
 
Ok, I couldn't find a 40 dB schematic, but I found this 41 dB schematic that looks like it'd be fun to build and the IDSS looks in the right range for my JFETs. Is this one ok?


This schematic does not look right.
at a first glance,

The gain device(s) in Q1 place should not be 2sk170.
The last stage should be a nice 2sk170 buffer


For the most updated version, see the FSP pdf guide in thread post #1 and in the pdf, gain section adjust the source and drain resistor for the desired gain.
 
Jim and Escucalin, Yes I would really love to build the circuit from the FSP pdf, however, the 2sk369BL's seem to be near near impossible to sorce from anyone reputable. Is there a suitable substitute? I agree it has a better buffer in the last stage.

Rush thanks, the schematic I posted is in the wiki you linked.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
All those 2SK170BL telescopic cascode earlier circuits are subjectively good, you will not regret building such, but best is the FSP folded cascode variation. For IDSS selected genuine 2SK369BLs ask Tea-Bag if he is willing to help. For MM configuration you only need one per channel.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Make short signal paths and best build against a copper plane for all ground returns. Works well for noiseless prototypes. Veroboard is usually adequate for MM but why not try more promising ways since you you got sensitive speakers. Google dead bug style (ugly) and Manhattan style (presentable) circuits construction.