# Accommodation for a capacitor on the + and - outputs (out<->ground)?
# The components on the low side will have its legs sticking up on the "high" side - I see some legs close to the rim of the bigger capacitors - is there a risk for a short here? Even if you don't physically ground the capacitor to keep clearance, it could tilt down to touch the "can" maybe?
//
# The components on the low side will have its legs sticking up on the "high" side - I see some legs close to the rim of the bigger capacitors - is there a risk for a short here? Even if you don't physically ground the capacitor to keep clearance, it could tilt down to touch the "can" maybe?
//
The pins under the caps do worry me. The input cap grounds should have their own trace going straight to the input ground to reduce ground noise. There should also be caps on the output.
Here's what I've done so far on my layout. This is for under 50V. I removed most of the protection diodes because they only help if you do something wrong, and even then they don't seem to do much good. I would say this is a very muntzed version. It can be used for both positive and negative.
So, tell me what you don't like and I'll try and get it ready soon.
Here's what I've done so far on my layout. This is for under 50V. I removed most of the protection diodes because they only help if you do something wrong, and even then they don't seem to do much good. I would say this is a very muntzed version. It can be used for both positive and negative.
So, tell me what you don't like and I'll try and get it ready soon.
Attachments
Sorry for not responding sooner - DIYAudio never mailed me to notify me of this post.
The one with the extra diodes works above 50V and withstands high inrush currents, although that is an unlikely problem.
1: For 20mA-10mA try 150R for R1.
2: Use OnSemi D45H11/D44H11 with a heatsink instead of C5171/A1930, but a 2.2V green LED instead of 3 diodes in the upper schematic. Have a lytic at the output and be careful about using film bypass (if it oscillates the output voltage will probably rise).
3: Not if your ground current is negative or below 15mA.
4: This would be like having a +15 V and +30V output from your supply, except you use the +15V as ground and ground as -15V. To keep ground from having too little or negative current, the -15V rail would always need to draw at leat 10mA more than the +15V rail.
Hi all,
Thanks to MR Keantoken for the circuit design
need to ask here, what is the Difference Result/Performance between those A and B circuit regarding KM regulator ?
The one with the extra diodes works above 50V and withstands high inrush currents, although that is an unlikely problem.
Any options to make it for let say :
1. Low voltage 12 volt, low current 20 -50mA only
2. High voltage 20-25 Volt, high current 3 - 10 Amp.
So we can choose the right Transistor and components value and avoiding any oscillations.
3. is it good to use Two positive circuit for Dual Rail supply + & - ?
4. Any disadvantage using positive circuit into Dual rail supply ?
Sorry too many Q
Thanks for any clue.
regards,
jc
1: For 20mA-10mA try 150R for R1.
2: Use OnSemi D45H11/D44H11 with a heatsink instead of C5171/A1930, but a 2.2V green LED instead of 3 diodes in the upper schematic. Have a lytic at the output and be careful about using film bypass (if it oscillates the output voltage will probably rise).
3: Not if your ground current is negative or below 15mA.
4: This would be like having a +15 V and +30V output from your supply, except you use the +15V as ground and ground as -15V. To keep ground from having too little or negative current, the -15V rail would always need to draw at leat 10mA more than the +15V rail.
Would it be possible to make a version without heatsinks so that no shorting of caps is possible and to house small output caps? Please
//
What about surface mount? You can have a layer of copper act as a heatsink. It'll limit the current, but probably not anywhere near many applications.
<< V2 >>
* only one mounting hole, I hope enough for 5x5 cm, saving space here
* added output caps (C5, C6)
* big caps on the edge of the board, larger footprint possible
* dedicated ground traces for C1-C4 (noise minimization)
* all traces 66 mils (1,67 mm) width
* no pins beneath caps except output caps, but not close
only problem for me is welding C5&C6, pins very close to diodes, but I hope is doable
Please check&comment
PS
You have choice to mount low height component to top or bottom, I put it on bottom but nothing stops you to mount on top.
That way, C5 & C6 will sit on diodes but welding is not a problem.
Eventualy I can make version with low height component to top (no change in cooper traces only silkscreen change).
* only one mounting hole, I hope enough for 5x5 cm, saving space here
* added output caps (C5, C6)
* big caps on the edge of the board, larger footprint possible
* dedicated ground traces for C1-C4 (noise minimization)
* all traces 66 mils (1,67 mm) width
* no pins beneath caps except output caps, but not close
only problem for me is welding C5&C6, pins very close to diodes, but I hope is doable
Please check&comment
PS
You have choice to mount low height component to top or bottom, I put it on bottom but nothing stops you to mount on top.
That way, C5 & C6 will sit on diodes but welding is not a problem.
Eventualy I can make version with low height component to top (no change in cooper traces only silkscreen change).
Attachments
Last edited:
Looks good. How about standing components. This way you save a maybe 50% for a resistor or diod. Except for D1b then..
Maybe a mount hole between C1/C2 and heatsinks so that if one may choose of more rigid mounting and heatsink. Hetasink could still be applied on transistor even if it is not mounted on board.
I would skip these heatsinks and let the user use component mounted heatsinks. They teake to much board space me thinks.
One single standoff does not offer a rigid mounting solution.
//
Maybe a mount hole between C1/C2 and heatsinks so that if one may choose of more rigid mounting and heatsink. Hetasink could still be applied on transistor even if it is not mounted on board.
I would skip these heatsinks and let the user use component mounted heatsinks. They teake to much board space me thinks.
One single standoff does not offer a rigid mounting solution.
//
You can use all 4 heatsink holes as mounting holes.
If you are worried because of traces near holes, use plastic or paper washerand be worry free
.
True!
Good that you ask Kean for permission - kudos! I donated.
//
