I do.
I find it very fiddly after the chip is soldered to the PCB.
It's easier for me to add the decoupling caps to the PINS before soldering to the PCB.
I have recently used PIN2 of an LM3886 as a Power Ground Pin.
SMD 100nF from Pin1 to Pin2, smd 100nF from Pin4 to Pin2 both these are on the heatsink side of the pin leadouts. Through hole 100nF from Pin2 to Pin5 on the front side, tucked right in to touch the vertical leadouts at the plastic package.
Solder Pin2 to the top plane and to the bottom plane for least inductance back to the Power Ground.
At this point of the build I can only add the 100nF underneath directly to the tip of the chip leads. In the other hand I don’t have suitable SMD caps to add at the top of the chip leads as have been suggested by many of you and of course that would be ideal.
The type of small film caps I have at hand is ERO 1837 and those are the ones that will go there.
PS Now that I remember I also have some Pana conformals, will check values at hand.
The type of small film caps I have at hand is ERO 1837 and those are the ones that will go there.
PS Now that I remember I also have some Pana conformals, will check values at hand.
Last edited:
remove c1 from the AC side of the rectifier.
Not needed, and can be quite a dangerous thing.
remove the inductor too, L1. not really need.
if someone wants a cleaner supply, should use a regulated one,
or a battery.
c3 and c4 are a bit small.
I'd increase them to somewhere 6800 uf /25 V 105c rated ones.
c2 should be 100 nf too. foil type.
the connection diagram is okay.
it should start properly.
cheers.
Not needed, and can be quite a dangerous thing.
remove the inductor too, L1. not really need.
if someone wants a cleaner supply, should use a regulated one,
or a battery.
c3 and c4 are a bit small.
I'd increase them to somewhere 6800 uf /25 V 105c rated ones.
c2 should be 100 nf too. foil type.
the connection diagram is okay.
it should start properly.
cheers.
Last edited:
Arty,
IMO nothing wrong with C1 if the voltage rating is high enough in this case is a 250V or 600V (don’t recall which) working on a 13V AC secondary, have seen used many times by others (& me) to get a cleaner AC.
C2 is 10,000uF (10mF) used as the first smoothing filter then we have a small choke (only 0.2mH) which helps further smoothing and cleansing action forming a CLC rail filtering system with the last 2x 2,200u. Yes, I could put bigger caps there but we start with space concerns, if we feel bigger is needed I could try to get a slim set there. The caps I’m using on the supply are rated at 35V (which are the ones I have) so if I get the 25V might not be bigger. I have used many times CLC supply’s with excellent results, you should try that Arty. CLC is even cleaner than regulated sandy supply’s and with a good size choke regulation is amazing, have seen this on the scope.
Antonio
IMO nothing wrong with C1 if the voltage rating is high enough in this case is a 250V or 600V (don’t recall which) working on a 13V AC secondary, have seen used many times by others (& me) to get a cleaner AC.
C2 is 10,000uF (10mF) used as the first smoothing filter then we have a small choke (only 0.2mH) which helps further smoothing and cleansing action forming a CLC rail filtering system with the last 2x 2,200u. Yes, I could put bigger caps there but we start with space concerns, if we feel bigger is needed I could try to get a slim set there. The caps I’m using on the supply are rated at 35V (which are the ones I have) so if I get the 25V might not be bigger. I have used many times CLC supply’s with excellent results, you should try that Arty. CLC is even cleaner than regulated sandy supply’s and with a good size choke regulation is amazing, have seen this on the scope.
Antonio
Frank,
If you remember some pages ago my comment was that I was shooting for a high pass filter, sort of, not a WB amp since this were to be used along with a woofer which has its own amp and that’s why I decided to use a rather small coupling caps and see if they did the intended effect if not I will change them to bigger ones. The high pass effect depends mainly on the impedance of the following stage in this case the chipamp which BTW has a lowish input impedance on the order of 25K so we might find that we need bigger caps. In the other hand, the negative part, relaying only on the small caps (first order) makes a high pass with a shallow slope (6 dB octave?)
So, will see what happens.
BTW, haven’t had time to go over the circuit these days but will soon catch up…
Antonio
If you remember some pages ago my comment was that I was shooting for a high pass filter, sort of, not a WB amp since this were to be used along with a woofer which has its own amp and that’s why I decided to use a rather small coupling caps and see if they did the intended effect if not I will change them to bigger ones. The high pass effect depends mainly on the impedance of the following stage in this case the chipamp which BTW has a lowish input impedance on the order of 25K so we might find that we need bigger caps. In the other hand, the negative part, relaying only on the small caps (first order) makes a high pass with a shallow slope (6 dB octave?)
So, will see what happens.
BTW, haven’t had time to go over the circuit these days but will soon catch up…
Antonio
for opamps in pre.amps in active xovers i usually suggest to use 2 12 volt bike batteries.
that does give a clean DC, cleaner than any rectifier would ever.
also, even this amp can run from a battery pack.
3x 6 volt SLA in series, dropped with diodes to give something close to 17 volts.
when the amp is off, a drop type charger is refilling the batteries.
at some places it is a cheaper option than a transformator.
there can not be any ground loops, so it is newcommer freindly.
and involves no high voltage stuff.
i hope to see this amp work, i know you will enjoy it
that does give a clean DC, cleaner than any rectifier would ever.
also, even this amp can run from a battery pack.
3x 6 volt SLA in series, dropped with diodes to give something close to 17 volts.
when the amp is off, a drop type charger is refilling the batteries.
at some places it is a cheaper option than a transformator.
there can not be any ground loops, so it is newcommer freindly.
and involves no high voltage stuff.
i hope to see this amp work, i know you will enjoy it
Another angle view of the setup I omit posting earlier…
http://i628.photobucket.com/albums/uu1/apass01/P6217637.jpg
http://i628.photobucket.com/albums/uu1/apass01/P6217637.jpg
+1 for CLC, I too have observed this on the scope. With enough LC sections (I had 5 in my last incarnation) its possible to have a ripple voltage looking not too dissimilar to a 100Hz sinewave.
The L of a CLC does indeed "round" the corners of the PSU sawtooth output.
An R in a CRC does so, but not as well as an L.
A CLrCLrC gives a very rounded saw tooth by attenuating the HF components of the ripple. (the little "r" of Lr is the parasitic resistance of an air core inductor). There are other "r" in the circuit, the resistances of the wires and traces connecting the transformer secondaries to the amplifier. They all add up to create an RC low pass filter. The on board decoupling capacitor reacts with all the preceding resistances to create a very powerful attenuator.
Don't be afraid of using thin wiring for the PSU. It provides an interference attenuating filter that costs less than thick wiring.
Last edited:
Having tried both air core and powdered iron cored inductors, I don't recommend air core for a couple of reasons. Firstly more copper is needed for the same inductance value and it'll have more resistance, second because without a core to contain the flux it'll spray out and induce ripple noise into surrounding circuitry.
Batteries are very good at producing HF ripple. They call it "noise".
Batteries are a chemical source of electricity. It gets produced in tiny packets, discrete packets.
It's these discrete packets that demand a smoothing capacitor to filter out some of the HF ripple.
Same problem as a mains powered PSU !
air cored inductors don't "spray" out flux. It emerges out of one end (if not a toroid) and flows around to the other end. The strongest is close to the inductor. The field gets weaker as distance from the inductor increases.
Your cored inductor will saturate.
How do you avoid saturation, which kills your effective inductance?
Is your cored inductor free from all external fields?
Have you measured them?
You've got me curious, Andrew. Can you point me to some material which elaborates on this, with some measurements? I just looked up a paper, which stated, that up to 100kHz, voltage noise in chemical batteries was many decades better than traditional power supplies.
Argument from authority. Got any reasoning to support the claim?
By choosing the appropriate size of the core and a core material commensurate with the current requirements. Saturation might occur at startup (capacitors charging from empty) but then I'm not worried about having no effective inductance.
Sorry I can't understand this question. Care to rephrase it? Did you mean "Is a cored inductor perfectly immune to external fields?".
What measurement did you have in mind?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.