Dear Salas,
Well, I am determined to try film caps to replace the electrolytics.
I am curious how important is the value 470 uF for C6 and C8?
At what point, of lowering the capacitance, do you think problems could occur?
Would the problem be instability or a narrowing of frequency response?
Any help is appreciated.
Thanks and take care,
Well, I am determined to try film caps to replace the electrolytics.
I am curious how important is the value 470 uF for C6 and C8?
At what point, of lowering the capacitance, do you think problems could occur?
Would the problem be instability or a narrowing of frequency response?
Any help is appreciated.
Thanks and take care,
Battery (2x9V in series), and a spare DIP socket, and the actual FET you are going to use in the board. Configure FET as a current source, and measure voltage across the whole string of LED's, swapping a few in and out of the socket to trim.
So I sort of build that section of the phono stage in the dip socket and change leds around until I get the needed voltage?
I must not have listened carefully. I was reading 57.7v with 5w1r rd/link resistors. Just returned from holidays & replaced 1R rd/link resistor with 5w20r in one side and got a drop only to 56.2v. Have a couple more 20rs and some 25s so I could put some in series and see happens. But would appreciate your advice first. Thanks
Do you load the raw dc output with dummy resistor when measuring those voltage values? The difference you found between 1R and 20R sounds weirdly small.
No. I'm simply measuring the DC voltage at the PS outputs. As noted, I could connect to the pcb's, but am reluctant to do so given the voltage.
This is exactly what Salas pointing you to...
You need to simulate your Phono by connecting LOAD resistor to your PSU output (in order not to damage phono board with out of desire high voltage).
If I recall correctly, you need to use 640R 5W (higher W is more safe) resistor.
Then you need to measure your PSU voltage on resistor.
You need to simulate your Phono by connecting LOAD resistor to your PSU output (in order not to damage phono board with out of desire high voltage).
If I recall correctly, you need to use 640R 5W (higher W is more safe) resistor.
Then you need to measure your PSU voltage on resistor.
Attachments
There can be no instability because its an open loop design between its stages. Narrowing of frequency response will not occur either.
I don't know a point because I haven't tested what you have in mind to modify. You can compensate towards original RC6 time constant by increasing the R12 element and Rail+ but not advisable beyond a point where Q4 drain to ground will surpass 13V and Q6 drain to ground 26V. You may need adjust R11 for that too. For C8 you can't adjust some PCB provided R but maybe you can add off board 33R in series to your large replacement film cap.
They are important for the local filtering of noise and interference. No matter if having a clean rail to start with any layout itself is an antenna. The wider it is the better an antenna. So its productive to re-filter what it may gather along the way at nearest possible to a sensitive node. On the other hand C8 filters the self noise of the LEDS. That R15 C7 Q7 D1-D4 C8 leg is to bias the cascode BJT's base and it is a shunt stabilizer PSU on its own. Better filtering of any kind shows more in the MC version of course.
Thanks for the guidance!
The way I am going to do this would result in minimal leads length but still more than the existing caps.
With the board above the 50 mm diameter cans it would not be a nest of wires which I know would be all wrong.
Maybe I should just use these for the supply bypass caps (47 uf) or at best just substitute something for one electroolytic at a time. (Idea number 2)
I know in the world of tube DIY many have found that one could substitute much smaller film caps for electroyltics with no penalty one could ascribe to the lower capacitance BUT I know that is in pure power supply locations.
Maybe I should use these film caps as bypasses for the large value electrolytics and hope for a good compromise. (Idea number 3)
One thing with the power supply - there is no reduction in bass. Bass transients are immediate. I was listening to some Mahler last night and there was a bass drum thwack that was so loud it took me by surprise. I do not remember ever hearing it like that. I think it is the 2nd movement of the Second but I am not sure.
Sustained bass is firm - there is no running out of breath effect and the rest of the spectrum is just less electronic sounding. I accept the risk of proselytizing and say it is worth giving a try.
Thanks again for your time and wisdom.
The way I am going to do this would result in minimal leads length but still more than the existing caps.
With the board above the 50 mm diameter cans it would not be a nest of wires which I know would be all wrong.
Maybe I should just use these for the supply bypass caps (47 uf) or at best just substitute something for one electroolytic at a time. (Idea number 2)
I know in the world of tube DIY many have found that one could substitute much smaller film caps for electroyltics with no penalty one could ascribe to the lower capacitance BUT I know that is in pure power supply locations.
Maybe I should use these film caps as bypasses for the large value electrolytics and hope for a good compromise. (Idea number 3)
One thing with the power supply - there is no reduction in bass. Bass transients are immediate. I was listening to some Mahler last night and there was a bass drum thwack that was so loud it took me by surprise. I do not remember ever hearing it like that. I think it is the 2nd movement of the Second but I am not sure.
Sustained bass is firm - there is no running out of breath effect and the rest of the spectrum is just less electronic sounding. I accept the risk of proselytizing and say it is worth giving a try.
Thanks again for your time and wisdom.
K. With 5w20R on the r/d link and a dummy load of 450R, getting 47vdc at the PS output. Guess all's well after all. Thanks for the help.
Perfect. Let us know more after you test it in the stereo system
Exactly, and by using the actual FET you have selected as current source (Q7, Q2x) you know you will be getting very nearly the same current and voltage drop in the actual circuit (except the supply voltage will be different, I suppose 4x9V batteries would more nearly duplicate conditions in the final circuit).
Edit: But if you look at the 2sk117 data sheet, you can see that Idss changes only ~0.2mA when you change Vds from 10-35V. Then look at how the Vf drop on an LED changes with a change of 0.2mA in current (from a nominal 3-5mA), it's not much.
Last edited:
Salas, now that the amp is all done I am amazed! Man does it sound good!!! I just ripped out the 200pf cap and changed the 47nf caps for the good stuff. Furthermore I fitted coaxial cable on the in and outputs and changed the umbillical for 2 x 2 x 0.8 with screening. Way, way better! Thanks for all the good advice. That goes to all of you!