http://www.diyaudio.com/forums/power-supplies/177516-keantokens-cfp-cap-multiplier.html
With many very constructive comments from Andrew.
With many very constructive comments from Andrew.
Andrew, before I read the whole thread and then realize not to have understand a lot, let me just ask you if you mean the K-multiplier on the site The K Multiplier
Is it that schemata I should build? I'm now at site 14 and there are a lot pages to read, but I'd like to build anything within the next days
Is it that schemata I should build? I'm now at site 14 and there are a lot pages to read, but I'd like to build anything within the next days
To be really sure I have following question... The LME and the outputs have their own PS-connection on the board. So when I wanna use one PS for both I connect one PS to both connections, while if I wanna use a separate PS for the LME I connect it to LME's connection WITHOUT changing anything on the board, right?
As it's not described anywhere I'd like somebody to confirm it
Hi Stammheim,
Both power supplies need to be connected no matter what. There is no on-board connection between the two.
If you are running one supply for both the front end and the output stage, then wire both supply points separately back to the supply. This will prevent the front end voltage from dropping due to the drop across any supply leads.
If you are running separate supplies, simply wire each up to its respective terminal on the board, and you should be ready to go.
If you are running a single supply for both front end and output stage, you might want to try and isolate the front end supply with an additional CRC filter with large caps and a resistance value as high as you can get with only about a volt or two of drop. Because the front end draws a constant current, it will help reduce noise on the front end supply, and isolate it somewhat from ripple caused by the output stage when under heavy load. Somewhere in the ballpark of 10,000uF, 75R, 10,000uF would probably work reasonably well.
To address your earlier post about the supply of the ALF lateral fets, I will need to look into it a little further. You are correct that there does not seem to be trustworthy supplier for these parts at the moment, and everyone appears to be out of stock.
I have a healthy supply of 160V N channel parts (ALF16N16W) if anyone needs them, but that's all I have on hand. I will do a little digging and see if I can figure our why none of these lateral parts seem to be available anymore.
Regards,
Owen
Both power supplies need to be connected no matter what. There is no on-board connection between the two.
If you are running one supply for both the front end and the output stage, then wire both supply points separately back to the supply. This will prevent the front end voltage from dropping due to the drop across any supply leads.
If you are running separate supplies, simply wire each up to its respective terminal on the board, and you should be ready to go.
If you are running a single supply for both front end and output stage, you might want to try and isolate the front end supply with an additional CRC filter with large caps and a resistance value as high as you can get with only about a volt or two of drop. Because the front end draws a constant current, it will help reduce noise on the front end supply, and isolate it somewhat from ripple caused by the output stage when under heavy load. Somewhere in the ballpark of 10,000uF, 75R, 10,000uF would probably work reasonably well.
To address your earlier post about the supply of the ALF lateral fets, I will need to look into it a little further. You are correct that there does not seem to be trustworthy supplier for these parts at the moment, and everyone appears to be out of stock.
I have a healthy supply of 160V N channel parts (ALF16N16W) if anyone needs them, but that's all I have on hand. I will do a little digging and see if I can figure our why none of these lateral parts seem to be available anymore.
Regards,
Owen
Owen, many thanks for your fast response!
OK, I got it right, fine.
The ALF16N16W are easy to get, at least in Germany. So I'll try them out. As I only have the Hypex SMPS600 with +-65V I drive them at the upper end... and still think about using another PS.. also because of the higher voltage for the LME.
Did anybody have success with the DSP500? I read the thread from the beginning and am now on page 90, so still a lot of stuff...
Many thanks,
Stammheim
OK, I got it right, fine.
The ALF16N16W are easy to get, at least in Germany. So I'll try them out. As I only have the Hypex SMPS600 with +-65V I drive them at the upper end... and still think about using another PS.. also because of the higher voltage for the LME.
Did anybody have success with the DSP500? I read the thread from the beginning and am now on page 90, so still a lot of stuff...
Many thanks,
Stammheim
In addition to my older questions...
To achieve the 360mA I'd liketo know where theyhave tobe measured... between both +-rails? Between +/- and GND?
If that's current through the output mosfets, then it's rail to rail. You take off one supply wire from the PSU to one side of the mosfets and either insert a meter set to current measurement, or insert a low value resistor of sufficient wattage, eg 1 Ohm or something, and measure the voltage drop across it and use Ohms law to calculate the current.
Last edited:
Hi guys,
I hope, anybody can help me...
When powering on my amp it just made "klick" and the SMPS showed a short circuit.
As I understood from the schemata, the P-Mosfet has to be soldered to the -rail, so the left side of the PCB.
Right?
I hope I have not killed neither the mosfets nor the LME
Thanks,
Stammheim
I hope, anybody can help me...
When powering on my amp it just made "klick" and the SMPS showed a short circuit.
As I understood from the schemata, the P-Mosfet has to be soldered to the -rail, so the left side of the PCB.
Right?
I hope I have not killed neither the mosfets nor the LME
Thanks,
Stammheim
How is the bias pot set? It should be turned fully clockwise at startup for minimum current. This is opposite of what you would expect, but it's covered in the assembly manual.
If you had the pot fully CCW, then you would be pulling maximum bias current, which is way too high for the SMPS at startup.
Regards,
Owen
If you had the pot fully CCW, then you would be pulling maximum bias current, which is way too high for the SMPS at startup.
Regards,
Owen
Hi Stammheim,
Where did you source your parts from? That's very concerning if you're having issues with multiple parts.
The only other thing I can think of is that the parts were damaged by ESD. Is your workspace set up in a reasonable manner? Do you have an anti-static mat or wear a wrist strap?
Of all the parts I've soldered (and there have been thousands), I have only ever seen one IC failure, and it was an OPA1632 that had excessively high 3rd order harmonic distortion. I chalked it up to me being careless with ESD, and just replaced the part with one from the same rail, which was fine.
The fact that you've had 4 faulty parts now means something is definitely out of the ordinary, either with your parts supplier, or with the way you're assembling/soldering.
Regards,
Owen
Where did you source your parts from? That's very concerning if you're having issues with multiple parts.
The only other thing I can think of is that the parts were damaged by ESD. Is your workspace set up in a reasonable manner? Do you have an anti-static mat or wear a wrist strap?
Of all the parts I've soldered (and there have been thousands), I have only ever seen one IC failure, and it was an OPA1632 that had excessively high 3rd order harmonic distortion. I chalked it up to me being careless with ESD, and just replaced the part with one from the same rail, which was fine.
The fact that you've had 4 faulty parts now means something is definitely out of the ordinary, either with your parts supplier, or with the way you're assembling/soldering.
Regards,
Owen
Hi Owen,
I got all my parts from Digi Mouser or Reichelt (a German distri). I guess I have killed the first reg by myself and by parallel connecting of both amps to one transformer the other one. The defective regs might have demaged the LME49990's. Soldering and the working place are alright. I soldered my most complex amps (SystemD_2k by Chocoholic, see d-amps) without errors.
The LME49830 was in fact a surprise for me, too. However, at the moment everything is fine, and maybe I'll also find an optimum solution for both LPUHP's, like I wrote in the LPUHP-thread
I got all my parts from Digi Mouser or Reichelt (a German distri). I guess I have killed the first reg by myself and by parallel connecting of both amps to one transformer the other one. The defective regs might have demaged the LME49990's. Soldering and the working place are alright. I soldered my most complex amps (SystemD_2k by Chocoholic, see d-amps) without errors.
The LME49830 was in fact a surprise for me, too. However, at the moment everything is fine, and maybe I'll also find an optimum solution for both LPUHP's, like I wrote in the LPUHP-thread