Author |
Message |
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Thu Oct 18, 2018 4:32 pm Post subject:
|
|
|
So the question is... Is it right?. Should I move all the caps before the regulators? What´s the teorical difference?
When I had few modules it made a delicious farting random sounds when power on/off. Now, even with more farads per module dont do it... |
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Fri Oct 19, 2018 4:15 am Post subject:
|
|
|
So if i get your number correct you have 4700uF on each rail before the regulator and 12700uF after the regulators? The latter being all the capacitance summed up with all the power decoupling caps in the several modules correct?
The caps before the regulators are usually choosen that you smooth the rectified voltage as best as possible so that your regulator alsways get's a voltage within it's minimum input limits and as stable as possible...
You can check the output voltage going to your modules with a scope and there should be a stable DC voltage with no drops or rest of AC content on top of it... This depends on the size of the caps and the ability of your transformer or wall wart to provide enough current...
IF you are within the limits (i usually leave 30-40% headroom, means when i draw 600mA i use a 1A transformer or wallwart, with 200mA a 500mA would do and so on...) and have some alternating voltage on the top of your DC output then you could increase the capacitance before the regulator as the caps size is not enough to smooth the voltage enough with the current draw.
Therefore i also build so tha i can stick in another big cap or change to bigger ones, meaning having enough room to upgrade them and also important don't put te caps too close to the heatsinks if they get hot, this will reduce the stability and life exceptance of the caps.
regarding the cap size on the secondary there is only the rule that there is an finite end of capacitative load for each PSU. You know when powering up all those caps on the power rails, will suck current to load, there is an point when it becomes too much for the PSU.
Also the rule is to place the caps on each module and where the voltage rails enter the module, first the electrolytes (typically 10-22uF)at the power rails filtering in combination with a small resistor or ferrite bead and then all the decoupling caps for the IC's as close as possible to their power input pins.
For some heavy LED loaded modules or VCO's or other modules that need much stability some use bigger electrolytes like 47-100uF. And also it's a good idea to have separate PSU'S for all the audio stuff and the logic like sequencers, dividers, etc..
Furthermore one last adivce is that it's best to keep the length of shared power busses for the modules as short as possible and the thicker the rails the better.
The ideal case would be all modules power rails would only interconnect directly at the PSU outputs and have the minimum cable length, sharing just a tiny lenght of the same rail. That would be the theoreticallly the ideal...
You will find a lot of useful and interesting stuff regarding PSU's connecting different PSU's together, etc. over at muffs.
Also, i repeat this very often when usig LED's with todays low current LED's you can improve things also by reducing the current consumption of those, the 1K LED resistor can usually be increased up to 10-15K and they shine bright enough and this will save you a few mA per LED, which then sums up when you have dozens... OR to put it in other words, less change in current consumption improves the stability of the PSU output (especially when it's reaching it's limits or is just a wacky wall wart)
Edit: what's the rating of your transformator and what do you draw at all at current per rail?
Sometimes all those things mentioned above don'T help anymore and the only way to go is to upgrade to a bigger PSU being able to provide more current. _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Fri Oct 19, 2018 5:16 am Post subject:
|
|
|
Tons of useful tips. Thank for the uber detailed response!
The Transf is 15+15 30VAC, and Im using it via 2 LM3x7 to provide +-12v
Yeah, in fact it takes some time since I power on and the modules start to work, so probably I have too much caps out there. Im going to remove about 7uF in total to see the difference. The aim is to get the most perfect and noiseless possible power.
The voltages are mostly correct, but at the very end of one of the busses it drops to 11.98V, which I would say it´s normal (the system has 2 rows of 60cms, its pretty large) |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
|
Back to top
|
|
|
wackelpeter
Joined: May 05, 2013 Posts: 461 Location: germany
Audio files: 10
|
Posted: Fri Oct 19, 2018 8:51 am Post subject:
|
|
|
not having exactly+/-12V or 15V isn't a problem. Regulators also have some differencies in their Output and as Long as it'S in range everything is ok. the important Thing is that it should be stable.
Like you can have for example 7815 Regulators that would put out 14,7V to 15,2V or so in the same PSU with the same current drawn.
That's just a matter of manufacturing accuracy and tolerance.
As asked before, would be interesting to know what your current draw on the + and - rails is. _________________ https://soundcloud.com/bastian-j |
|
Back to top
|
|
|
AlanP
Joined: Mar 11, 2014 Posts: 746 Location: New Zealand
Audio files: 41
|
Posted: Fri Oct 19, 2018 9:44 am Post subject:
|
|
|
PHOBoS wrote: |
btw 4700uF would be 4.7mF not 4u7F |
To muddy the waters even more, some oldschool docs write mF, for micro-farad (and not milli-farad), since the µ symbol was a bag of dicks to try and type. |
|
Back to top
|
|
|
MapacheRaper
Joined: Feb 15, 2018 Posts: 166 Location: Spain
|
Posted: Fri Oct 19, 2018 2:19 pm Post subject:
|
|
|
PHOBoS wrote: | MapacheRaper wrote: | The aim is to get the most perfect and noiseless possible power. |
talk to Graham Hinton
12700uF does seem like a lot, which makes me wonder what size caps you are using on your modules.
btw 4700uF would be 4.7mF not 4u7F |
Yeah, in fact every cap is fatter than my thumb... And there are 6 of them... I think Im going to left only 2. |
|
Back to top
|
|
|
|