I just finished installing the new 2x15V transformer and measurements are only slightly different. Does that show that the lower current is because of the certain fets?
I also removed the extra cement resistors leaving only the 10R one, so that I re-evaluate from scratch the sound per stage.
The voltage at the filtered caps is now +-20,3V
Vin: 9,99V and 9,95V
CSS: 1,598V and 1,618V
Offset: 0,9mV and 1,6mV
And after 20 minutes
Vin: 9,97V and 9,92V
CCS: 1,593V and 1,610
Offset: 0,9mV and 1,7mV
Which brings it again to about 160mA almost like before
I also removed the extra cement resistors leaving only the 10R one, so that I re-evaluate from scratch the sound per stage.
The voltage at the filtered caps is now +-20,3V
Vin: 9,99V and 9,95V
CSS: 1,598V and 1,618V
Offset: 0,9mV and 1,6mV
And after 20 minutes
Vin: 9,97V and 9,92V
CCS: 1,593V and 1,610
Offset: 0,9mV and 1,7mV
Which brings it again to about 160mA almost like before
Is there a way to snap the rectifier diodes from the pcb and put them on a separate case with the transformer?
Or will I have to just use a separate pcb ?
Or will I have to just use a separate pcb ?
Last edited:
You can just put them in a seperate case with some sepate caps, then run the wiring to caps directly.
You can mount and sink/isolate them diodes in a separate case with the one pair 4700uF filter capacitors, then use AC power chord like thickness umbilical to the DCB1 empty caps pair pads to feed the remaining capacitors. Can use small value power resistors between bridge and capacitors pair at the PSU case for dropping about 1V. It will smooth the EMI from the heavy current charge cycle.
I increased the CCS current again.
CCS resistors are now ~2R85 (1x10R and 3x12R) and voltage accross them ~1,37V giving ~480mA.
Temperature on the plate after 3 hours is ~36C and on the Fets ~39C.
I think that anything more than that would result in more than the psycological barier of 50C in the closed case.
So to further increase the CCS current I would have to either reinforce the plate with some additional heatsinks, or use some kind of battery powered small fan to circulate some cold air in there. Maybe both.
Either way, I have to finish the case before going further. Unfortunately I am still waiting for my new circular saw...
CCS resistors are now ~2R85 (1x10R and 3x12R) and voltage accross them ~1,37V giving ~480mA.
Temperature on the plate after 3 hours is ~36C and on the Fets ~39C.
I think that anything more than that would result in more than the psycological barier of 50C in the closed case.
So to further increase the CCS current I would have to either reinforce the plate with some additional heatsinks, or use some kind of battery powered small fan to circulate some cold air in there. Maybe both.
Either way, I have to finish the case before going further. Unfortunately I am still waiting for my new circular saw...
Ok I got a little impatient, so....
Adding a 4th 12R in parallel (2,3R CSS res) gave ~1,33V and ~1,37V or ~580mA.
And a 5th one (1,93R) gave ~1,29V and 1,326V or ~670mA
I also changed the fuse to a 5A one. Oh and the light bulb checker gets a nice glow now 😀 (I measure stuff without it of course)
Unfortunately it' s late here and I am getting tired for more testing. I hope I can find some nice heatsinks tomorrow, and hopefully some 5R @10W resistors to avoid all the stacking...
Adding a 4th 12R in parallel (2,3R CSS res) gave ~1,33V and ~1,37V or ~580mA.
And a 5th one (1,93R) gave ~1,29V and 1,326V or ~670mA
I also changed the fuse to a 5A one. Oh and the light bulb checker gets a nice glow now 😀 (I measure stuff without it of course)
Unfortunately it' s late here and I am getting tired for more testing. I hope I can find some nice heatsinks tomorrow, and hopefully some 5R @10W resistors to avoid all the stacking...
And a small additional sound impression.
With the Mills 10R alone, the buffer was impressive. My system had an almost SET big-tube feeling.
Adding parallel 12Rs (unfortunately cheap ones) kept increasing the separation of sounds and width and depth of the stage. Unfortunately things got a bit more weak as well and lost that SET sensation, but it could be because of that freshly soldered chunk of cheap resistors.
With the Mills 10R alone, the buffer was impressive. My system had an almost SET big-tube feeling.
Adding parallel 12Rs (unfortunately cheap ones) kept increasing the separation of sounds and width and depth of the stage. Unfortunately things got a bit more weak as well and lost that SET sensation, but it could be because of that freshly soldered chunk of cheap resistors.
I got some more resistors and heatsinks to further push the circuit 🙂
With 2 of the additional heatsinks placed like that on the plate and the CCS current at ~670mA, the temperature was ~36C on the plate and ~39C on the fets.
So it was time for the next step.
I replaced the 5 12R resistors with 5 5R9 ones pushing the resistance to 1,05R and the current to 1,180mA.
I am currently listening to it like that. The increase in separation, clarity, textures and sentiment was again quite big. What is worth noticing is that instruments especially wind and string ones sound more emotional without sounding sweeter or bigger. Same with voices. Unfortunately it s late again, so I will have to watch its temperature silently 🙁
But first impressions are again that It`s definetely worh the effort to push it to the upper limit...
Another issue that occured is that I started to get some minor vibration from the Hammond transformer (damn hammonds always do vibrate). So I will have to either move it outside the case, or put it on top of a big chunk of wenge to dampen it.
With 2 of the additional heatsinks placed like that on the plate and the CCS current at ~670mA, the temperature was ~36C on the plate and ~39C on the fets.
So it was time for the next step.
I replaced the 5 12R resistors with 5 5R9 ones pushing the resistance to 1,05R and the current to 1,180mA.
I am currently listening to it like that. The increase in separation, clarity, textures and sentiment was again quite big. What is worth noticing is that instruments especially wind and string ones sound more emotional without sounding sweeter or bigger. Same with voices. Unfortunately it s late again, so I will have to watch its temperature silently 🙁
But first impressions are again that It`s definetely worh the effort to push it to the upper limit...
Another issue that occured is that I started to get some minor vibration from the Hammond transformer (damn hammonds always do vibrate). So I will have to either move it outside the case, or put it on top of a big chunk of wenge to dampen it.
And about 2 hours later the temperature is 45C on the plate/heatsinks and 50C on the fets. Sorry for the spam, I just want to keep these somewhere where I will not lose them...
So the next step is 2 additional heatsinks and heat-conductive paste, to drop it to around 40C on the fets.
So the next step is 2 additional heatsinks and heat-conductive paste, to drop it to around 40C on the fets.
Members like detailed step experiments, personal evaluations, and photos. No spam its your story.
P.S. 50C on the cases at 1A is no drama for 0.82C/W RthJC devices with 150C internal breakpoints, especially on Kerafol pads.
P.S. 50C on the cases at 1A is no drama for 0.82C/W RthJC devices with 150C internal breakpoints, especially on Kerafol pads.
Thanks Salas.
According to Mike the pads are no kerafol ones. Just some cheapish from mouser.
syklab nothing significant has changed on my build so far. Just check my previous photos. I have just soldered the extra parallel resistors, and placed 2 heatsinks 10x10x4cm rated at 0,6C/W right on top of the plate. No thermal paste or screws yet.
I will try to post a photo later today.
PS for me, a reminder to get bigger rating resistors. These are getting hot.
According to Mike the pads are no kerafol ones. Just some cheapish from mouser.
syklab nothing significant has changed on my build so far. Just check my previous photos. I have just soldered the extra parallel resistors, and placed 2 heatsinks 10x10x4cm rated at 0,6C/W right on top of the plate. No thermal paste or screws yet.
I will try to post a photo later today.
PS for me, a reminder to get bigger rating resistors. These are getting hot.
With normal Silpads its 1C/W additionally. With 1A you burn 10W in each Mosfet now. So at 50C case temp you should have +10C +8.2C in the core. I.e. 68.2C. IRF240s yawn at that.
Nice 🙂
By the way, I have some of the DIY store PSU boards and I am thinking of using one to move the transformer out of the chassis.
http://www.diyaudio.com/forums/imag.../P-PSU-1V20-psu-cap-diode-board-schematic.pdf
I have used 0,47R for R1-R8, FEP30DP for the rectifiers and 4 22kuF caps per rail (basically prepared it for F5 which is now postponed for a bit).
Would this be suitable? Or will it have any negative affect on the performance if we take into consideration that we are trying to get as much as we can by pushing the circuit's CSS current?
By the way, I have some of the DIY store PSU boards and I am thinking of using one to move the transformer out of the chassis.
http://www.diyaudio.com/forums/imag.../P-PSU-1V20-psu-cap-diode-board-schematic.pdf
I have used 0,47R for R1-R8, FEP30DP for the rectifiers and 4 22kuF caps per rail (basically prepared it for F5 which is now postponed for a bit).
Would this be suitable? Or will it have any negative affect on the performance if we take into consideration that we are trying to get as much as we can by pushing the circuit's CSS current?
It looks very nice actually. The very low value RC helps with the heavy charge cycle's EMI too. I don't know how you will perceive so much capacitance subjectively but its a board I would surely try given it being handy.
I have some non-populated pcbs too. Unfortunately I have no diodes left and no 4700uF caps of decent quality. You think the huge capacitance could be an issue?
To be unnecessarily hard to a given transfo maybe, in general its about going the royal way and testing the subjective impact. Only you can tell bcs its what caps type also.
