I did not ground the -ve: it is left unconnected as shown and works fine.
I have used BJT outputs: MJL4302.
Hallo jacques antoine.
tres bien.
Nice to see more movement around here.
Jac, can you please later confirm some listening impressions after some time burn in piersmas inverted cfp.
Hallo X.
I know, you love it to go different way, but the Mosfet variant is simpler to realize.
After looking Datasheet IRFP250Mi have seen the 5.generation has lower Cob- 315pF!
Between 3.- 5. Gen. must be advancement cognizable.
Because that i favor the newer one.
Regards Bangla
tres bien.
Nice to see more movement around here.
Jac, can you please later confirm some listening impressions after some time burn in piersmas inverted cfp.
Hallo X.
I know, you love it to go different way, but the Mosfet variant is simpler to realize.
After looking Datasheet IRFP250Mi have seen the 5.generation has lower Cob- 315pF!
Between 3.- 5. Gen. must be advancement cognizable.
Because that i favor the newer one.
Regards Bangla
I kept everything the same. MJL4302 was maybe not the best choice due to its high fT of 35mHz...
Cheers,
Jacques
I got my board!
But, then I got questions. . .
Is this the right schematic for it?
Is there availability of schematic with component labels that match the board screen print?
Gain range of SC4793? (the lower one?)
Gain range of BC556? (b?)
Gain range of BC546? (b?)
Suggested startup value for the trimmer?
Value of C-Opt?
When is the 1n (output area) optional?
Rail voltage?
What value goes where the board says "R = 0 - 12"? (10?)
Also, SA1360 is discontinued and run out of stock; so, I need a substitute. I have authentic Philips BD140 (a near match except for twice the capacitance). And, are there support component (resistor value, capacitor value) changes to support it? OR, should I place an order for KSA1381E, which seems to be a match?
And then, I was going to stuff the board with what parts I have (so as to see what is needed to order afterwards); however, most of the component values differ between schematic and board screen print. So, I'm confused. Help?
But, then I got questions. . .
Is this the right schematic for it?
Is there availability of schematic with component labels that match the board screen print?
Gain range of SC4793? (the lower one?)
Gain range of BC556? (b?)
Gain range of BC546? (b?)
Suggested startup value for the trimmer?
Value of C-Opt?
When is the 1n (output area) optional?
Rail voltage?
What value goes where the board says "R = 0 - 12"? (10?)
Also, SA1360 is discontinued and run out of stock; so, I need a substitute. I have authentic Philips BD140 (a near match except for twice the capacitance). And, are there support component (resistor value, capacitor value) changes to support it? OR, should I place an order for KSA1381E, which seems to be a match?
And then, I was going to stuff the board with what parts I have (so as to see what is needed to order afterwards); however, most of the component values differ between schematic and board screen print. So, I'm confused. Help?
Last edited:
Okay, if I'm not to use the screen print firstly, then please mention it before I do that. Gangway????
Daniel, it is true that the information for the assembly are not easy to find for the Piersma's Inverted J-Fet Circlophone. Do you think it would be preferable that all documentation and advice to build this great amp should be all put in a new thread?
That's a good idea because I think Piersma's design is significantly different and it's tough to find in this thread. Probably best to let Piersma create and manage new thread?
Yes, but Piersma would need to start that thread (build and documentation), so that he keeps edit rights for Post#1 (the only post with "forever edit" functionality), which makes indexing and updates possible.
This will be really great for everyone, including you, because on that thread, you can answer questions once instead of many times.
I would have added the materials to the big circlophone build thread; however, I wasn't able to do it because I don't know the relevant answers. Will do some possibly graceful hotlinks, after I learn more, on your build thread.
Even before building, the board has done what I wanted--a Circlophone with lower dissipation and a different tone. That's what I wanted, that's what it does, so it is probably helpful. However, it supports several optional options; and, since the CFP builds require precision, I haven't been able to build it yet. . . because I didn't know how to do it with precision. Problematically for me, the board can do many different things, however what I need is one specifically.
I'd like to put in a mention of the schematic component labels matching the board screen print when possible. It could make for less questions.
Also, if KSA1381E happens to be an exact match for the discontinued part, then, in that case I suggest a double-label on the schematic (both parts mentioned). Also, on the schematic, the value of C-Opt and the startup trimmer value, if feasible.
Then there's that matter with the BC560A, which is discontinued, leaving the options of either BC560C or BC557A. Which is better for this? That may be another case of needing a double-label on the schematic. Also, Fairchild's BC560A (not available) and BC557A (the current part) won't be different because it is not affordable to have separate tooling for such similar parts.
Edit:
Not actually complaining (that sounds different), but actually was trying to be helpful.
Edit2:
Also wanted to say thanks, because I mean it. . . even though it might not be the very best timing to do that. I'm still grateful though.
Last edited:
Hi Piersma,
Is there a new inverted Jfet c-phone thread yet?
I just started the build stuffing resistors and diodes for past couple of hours. Realizing I don't have several of the kind of "odd" values. So looks like I will need to place order for the following 1% 0.25W axials: 2.2R, 4.7R, 1.21k, 68R. I also need to get the 0.56R 1W, 1R1W, 0.56R 2W. I also see I don't have some of the diodes, how critical is it to use a BAT57 for D8 and are there alternatives? And the two big MBR10H50's. Can conventional 10A silicon diodes be used there?
There are several places where the designation says: R=0-12 does that mean it could be a 0R jumper and no higher than 12R?
And to double check - resistor values are using the letter E vs R as in R=4E7 is the same as 4R7 or 4.7R?
Btw, this is the most complex amp I am building to date. I believe it has 18 transistors and more resistors than any other amp so far.
it's going fast but running into the values not in my standard kit of resistors.
Inverted J-FET Circlophone
hi XRK,
Unforeseen circumstances left me in Germany last three weeks so was unable to start the builders thread. I will start the new thread before this weekend.
The resistors values do not strictly need be 1%. Because of the advised metalfilm resistors i resorted to 1% precision.
Both degeneration resistors with 0-12 Ohm marking, use 10 Ohm.
Do not use a conventional diode for the TO-220 "by pass" diodes, because of the significant higher voltage drop. Use a low forward voltage Schottky.
To confirm: yes 4E7 is 4R7 or 4.7 Ohms.
hi XRK,
Unforeseen circumstances left me in Germany last three weeks so was unable to start the builders thread. I will start the new thread before this weekend.
The resistors values do not strictly need be 1%. Because of the advised metalfilm resistors i resorted to 1% precision.
Both degeneration resistors with 0-12 Ohm marking, use 10 Ohm.
Do not use a conventional diode for the TO-220 "by pass" diodes, because of the significant higher voltage drop. Use a low forward voltage Schottky.
To confirm: yes 4E7 is 4R7 or 4.7 Ohms.
edit: Piersma is so fast!
. I pulled some SB560's from old switching power supplies.
And if you are using the same board as i have: there was an error around a 560p cap, where you have to cut 2 traces and make a bridge.
I attached my testing simulation if you like to measure things. Some components may be a bit different, i have tried many things.
The low value resistors near the output are not very critical, but the values are optimized for minimal distortion. The 1W resistors can be replaced with 500mW resistors as long as you don't use high voltages, max dissipation is about 100-150mW at 35V rails and a 0.5V sine wave. Most of the current goes through the diodes.
Elvee wrote somewhere that it has to be a small signal schottky, see the build thread.
Schottky is key
. I pulled some SB560's from old switching power supplies.I'm not sure about this. Simulation shows good results with 10 ohm resistors, but that is with my 2sk369 models. Simulation with 1 ohm shows oscillation.
Correct.
And if you are using the same board as i have: there was an error around a 560p cap, where you have to cut 2 traces and make a bridge.
I attached my testing simulation if you like to measure things. Some components may be a bit different, i have tried many things
.
Last edited: