Simplistic NJFET RIAA

diyAudio Chief Moderator
Joined 2002
Paid Member
Follow one schematic only. 5120. Q5 2SK170. Not BF245A. Start with the trimmer set at half travel. At 1/4 travel if Q5 is a BL. Debug all orientations. 5120 shows remote sensing, an older board design has not such provision. Debug all connections.
 
Follow one schematic only. 5120. Q5 2SK170. Not BF245A. Start with the trimmer set at half travel. At 1/4 travel if Q5 is a BL. Debug all orientations. 5120 shows remote sensing, an older board design has not such provision. Debug all connections.

Sorry, I don't understand what you mean by remote sensing. Does it mean 2 separate circuits joined together by interconnecting wires as per your cct diagram? I will build it again p2p following your cct. In fact things got a bit complicated because I erroneously etched the pcb a mirror image of the original layout. Thanks again for your suggestions.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I mean exactly as that schematic. 4 wires. 2 that connect the floating sub circuit to the other 2 main circuit's wires only at the receiving end on the load itself. Its beneficial and it does sound it. Yes, make exactly that circuit point to point please and it will work. Some mix up possibly happened between those boards and parts. This is so many times repeated that will surely work if followed correctly. Good luck.
 
Thanks Salas. Just to check that I get my part orientations correct I'd like to confirm some basics which to a newbie like me are not so obvious. In the case of the IRFPs it's the source that gets connected to the positive side of the circuit/s, unlike the SKs where the drain is connected to the +ve side of things. Isn't that so? Thanks -JA
 
The Source pin is the one depicted as an arrow on schematics. As you face Mosfets they are GDS physically. Q1's source goes to the 10R resistor.
Power MosFETs come in two types, Lateral and Vertical.
All Lateral mosFETs have the Source connected to the backplate.
All Vertical mosFETs have the Drain connected to the backplate.

Even the To3 meet this same standard.
 
I got the Shunt regulator working. However, connecting to it the RIA boards for a first listening test I just got a very small output from one channel and nothing from the other. I thought it was a grounding issue but the problem persisted. I took some voltage readings on the two boards in the hope of getting some clues, but in truth I cannot meaningfully interpret them. Perhaps they can help someone trouble-shoot the problem/s. Thanks S for your help so far.

*Attachment removed after member's request to avoid confusion due to possible errors in the circuit.:cop:
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Are the Jfets 4mA GR for sure? Or are they all connected with their drains oriented north towards the B+? They are DGS as you face them. Q2 is CBE respectively. It just drops like there is more IDSS everywhere. The drop over R8 1K derives 11.5mA for instance! Insane for a GR with a 33R source resistor (R6). Verify the orientations first and report back, we will see if its something else.
 
Thanks for being there, Salas. As always, to help me out. All JFets are GRs except those of the buffer stage (Q4-5) which are BLs. All are connected with their drain pointing north towards B+... Now I understand what you mean that the v-drop across R8 results in Q3 pulling a current of 11mA.
 
Last edited:
diyAudio Chief Moderator
Joined 2002
Paid Member
So everything is oriented as I wrote? If yes and still the same voltages, make your B+ 25V now using the trimmer in the reg. What is the voltage on Q2's collector and Q3's drain after that?

P.S. Is your R5 well connected to Q3's gate and ground? Any possibility of having a 27k instead of a 2k7 R9 or another one wrong like R1 for instance? The voltages are weird around Q2 for Vbe drop even.
 
Made B+ 25.1V and took 2nd reading with a different DMM because the first one was taking a long time to settle down (even with a new battery - time to discard it because it has already happened before). The new voltages are marked in the attached diagram. I haven't studied them but there are a few changes from the first readings. Thanks for your patience.
 

Attachments

  • Salas NJFET RIAA V-readings2.doc
    32 KB · Views: 158
diyAudio Chief Moderator
Joined 2002
Paid Member
That one looks logical. Stronger but logical. Shows your Q1 fets like being 6mA jobs, 5.8mA with slight degeneration of the 3R3. The 2nd stage and output look rather normal now it biased up as a whole. Basically it will play. You should be able to hear music. You need to bring Q3 voltages together at 12.5V. Use 100R trimmers instead of R6s. Experiment with some mismatched 4R3-6R8 range for R2s to bring Q2 at 12V in both channels. So to keep the gain at bay a bit and match up channel outputs. Those first 2 stages fets are upper limit GR it seems. Surely not 4mA IDSS and not well matched between channels, but you will bring everything together by doing the moves I just mentioned. Keep us posted.
 
OK Salas, the Lady sings... the blues! Only one channel is working, but I haven't done the changes you suggested. I'm sure we'll get around it. Well done Salas!! It's the channel whose voltages are shown second in the various voltage readings in my last attachment. Incidentally, as Chief Moderator you'd better remove the attachment in post http://www.diyaudio.com/forums/analogue-source/129126-simplistic-njfet-riaa-519.html#post2355020 #5188 because it can mislead readers and cause trouble. Thanks for now and regards. Will keep you posted. Joe A
 
Luckily I didn't kill anything... that channel was dead due to a bad coupling output capacitor. I inserted the 100R trimmer in place of R6 (33R) and now Q3sees 12.5V as you suggested. Biasing Q2 (BC550) has been trickier. I inserted 4R7, 5R6, etc up to 13R in place of R3 (3R3) and all I managed to get at the junction R1-Q2 was a maximum of 8.33V. (Pcb now looks somewhat like a breadboard!) I stopped there, thinking that maybe R1 (2K4) needs to be tweaked/changed. All other components and positions remained intact.

Anyway, the amp is playing music right now and initial impressions are very good. It does compare very well with the Steve Bench amp. The sound is clear, clean, detailed, strong and wide, with no trace of harshness. It's still burning in and maybe I'm biased because I'm happy for getting it working. Layout is in a shoe box cover so in due course I must put in a proper metal case.

I'm not sure it's worth bothering about the voltage at Q2. I'm sure you know better. So far, very good, so Well Done, Salas and many, many thanks for your help!!
Regards.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
That's nice. Mysteriously you have shown 9.7&10.7V at Q2's collector in your second attachment. You are measuring from C Q2 to ground again, right? Make R1 2k2 then and arrange R2s accordingly so to have something 10-12V at Q2's collector. Channels need be matched at those points so to have same gain for best stereo. Let us know more soon along with pics etc. Also try your input load trimmer to find best tone with your cart. Congrats!;)
 
Spent the past few days tweaking the amp to get the required voltages. Finally I got it sounding to my taste and prepared to put it in a case (a humble one at that!) This is the version I'm listening to and it sounds great. No hum or whistles. Output caps are not identical but I like the sound of the (large) Russian PEPT 3.3uF better. Later on I'll change the other (SCR) cap. Meanwhile, I attach a photo of innards of my Salas nj-fet riaa. I encourage other vinyl lovers to build it. Cheap to build (compared to valves), lot of help on this thread, fun (and a learning experience in my case), and above all GREAT SOUND!!! THANKS A LOT AND WELL DONE, SALAS!!!
 

Attachments

  • sALAS njfet RIAA 006.JPG
    sALAS njfet RIAA 006.JPG
    177.4 KB · Views: 379