PO89ZB , an inline DC filter for SMPS wall warts . Preamps, HPA, Korg NuTube, etc

Jump to Latest
C

ClaudeG

Member
Joined 2015
Thanks Joensd

Well, I don't know what Mark did but my guess is coil resonant frequency, simulations and other bits (try and errors with several SMPS?) helped determining the coil choice and the value of the damping resistor in parallel and possibly the other resistor?

Perhaps?
 
Mark Johnson

Mark Johnson

Member
Joined 2011
Paid Member
I think if you can find another inductor in the same package, with the same inductance, and comfortably greater than 5 amps max current, and a DC resistance that's about the same (plus or minus 25%), you can probably ignore the inductor's self resonance frequency. All manufacturers are up against the same materials limits, so if Company X and Company Y both sell a 2.2 uH, 6A inductor in the same package, the two inductors are probably built about the same way and the two inductors probably have about the same self resonant frequency. Which is nice because some manufacturer's don't put the S.R.F. on their datasheets!
 
pinkfloyd4ever

pinkfloyd4ever

Member
Joined 2008
youtube.com/watch?v=dQw4w9WgXcQ
Awesome thanks konst!:cheers: PM incoming

Man and I finally just placed that Mouser order a couple days ago that I was sitting on for several weeks, just to find this neat little project a day or so later. Ah well, such is the life of the audio DIYer these days. First world problems, really. Maybe I'll just have to also add my MoFo BOM to the cart, and order once I get some Christmas money :xmasman:

For anyone else using Mouser, here's some other inductors that should work and are currently in stock
https://www.mouser.com/Passive-Comp...0s4k1Z1z0z7l5&Rl=wpczZer9dZ1z0wr1gZ1z0wocoSGT
 
Last edited:
C

ClaudeG

Member
Joined 2015
Thanks to all of you for your help, really kind!

I have browsed through the various options and sadly I either found parts that were at least twice the DC resistance (+100% or far far more), or far too low max A rating, or, in most cases, OK values from Murata etc. but with wrong dimensions (problem is mainly diameter too large to fit on the PCB, on a lesser extend legs that are too big but that could have been accomodated).

Long story short: this SMPS filter works so well that I wouldn't want to drift too much apart from the initial values, so equivalent replacement coils are not easy to find... and I am better off waiting until Feb.

I tried though thanks to your kind support, thanks for that!

Claude
 
Mark Johnson

Mark Johnson

Member
Joined 2011
Paid Member
Last week I decided to investigate whether this filter could be improved, if I allowed myself to use electrolytic capacitors whose cylindrical bodies are just a little bit longer. The Kemet caps in the Parts List of post #1, are 20mm long. But I found a couple of interesting looking candidates on Mouser's website, that are 25mm long. These candidates will extend a little bit beyond the right hand edge of the PCB, but the leads will still be solderable, the cans will still fit, and the filter will still be useable.

So I ordered some caps to experiment with; they were delivered this afternoon. I am happy to say that both of the new, 25mm long, capacitors are measurably better than the original Kemets. In fact quite a bit better! I used my fairly high-end East Tester ET4510 LCR meter, to measure the ESR of these caps at 100 kHz. I'm interested in high frequency ESR because the unwanted noise from SMPS's, is high frequency. The lower the capacitor ESR, the better the filter will remove high frequency noise.

And boy take a look! The 25mm long caps are both lower ESR than the Kemet 20mm long cap:

Code: 
80-ESH477M050AH4AA    Kemet             470uF 50V  LS=5  D=10  LEN=20  100K ESR  0.045

661-EKYA500E471MJ25S  United Chemi-Con  470uF 50V  LS=5  D=10  LEN=25  100K ESR  0.030

647-UHW1H471MPD       Nichicon          470uF 50V  LS=5  D=10  LEN=25  100K ESR  0.015

So I think, if your physical packaging can accommodate the extra 5mm of capacitor length, you'll get noticeably better filtering -- greater attenuation of SMPS high frequency noise -- if you use the Nichicon caps above. They are better by a factor of 3X. That's a heck of a nice benefit, compared to the cost (+5mm length), in my opinion.

_
 

Attachments

  • Three_caps_small_filter.jpg
    Three_caps_small_filter.jpg
    229.1 KB · Views: 555
pinkfloyd4ever

pinkfloyd4ever

Member
Joined 2008
youtube.com/watch?v=dQw4w9WgXcQ
Nice! That's great news Mark. Just updated my Mouser project to those Nichicons. You should make a note about this in the OP if you're able.

ClaudeG said:
Argh! I just measured and they won't fit, everything is tightly packed on the baord.
Click to expand...

I think this one should work? 744772022 Wurth Elektronik | Mouser
It's 2.2uH, 6A, .01Ohm, 7.8mm diameter, 5mm lead spacing.

Self Resonant Frequency is 1.1MHz though...hmm, Mark, is that of big enough concern to not use this inductor?

Wait a second...when you look at the actual datasheet, is says the SRF 75.5MHz. :eek: Quite a difference

Goes to show that Mouser's parametric values aren't always correct, and it's probably good practice to check the actual manufacturer's datasheet to verify any critical parameters.
 
Last edited:
C

ClaudeG

Member
Joined 2015
Pinkfloyd4ever,

That looks like a candidate to me! Let's see what Mark says...

Mouser is quite confusing, looking at the webpage it says indeed that this Wurth coil has a self res freq of 1,1MHz, vs. 63MHz for what Mark recommended. That's why I discarted it first place. But I just looked into the spec sheet and my understanding is also that the Wurth has a self res freq of 75.5MHz. Would amso match Mark's wise remark on construction and specs.

Wether that value is acceptable or not, I will let Mark respond...

Apart from that, dimensions look OK and legs could possibly fit. OK, they are 3x more expensive than what Mark recommended, possibly the biggest difference.

Claude
 
C

ClaudeG

Member
Joined 2015
Hi Mark,

I am now in the process of building 8 (!) more of these filters in one go. So I better get it right ;-)

No hurry at all nor any real need on my side. This is for testing and pleasing others mainly. That means I could wait but I would like to go for the right parts if possible, rather than building 8 lower grade filters. I am about to place an order at Mouser's for another project though, something that can't wait.

Mark, your findings on the caps are VERY interesting and, as I haven't placed my order yet, I will go for your latest recommendation. Size shouldn't be a problem as Gilles is still considering packaging the filters properly. Sounds nearly like series production LOL.

Remains the question of the coil. Again, I don't mind waiting for Mouser to send me the parts in BO in Feb... and I am unlikely to have real needs or time to visit many friends before that anyway. On the other hand, if you say the Wurth coil is a good choice, then job done between Xmas and NYD...

Many many thanks again for your feedback Mark

Claude
 
TNT

TNT

Member
Joined 2003
Paid Member
And boy take a look!...

Hmm, couldn't cause your attached picture managed to miss the ESR values after the dot. But maybe it was on purpose to make us look for it ourselfs? :)

//

Mark Johnson said:
Last week I decided to investigate whether this filter could be improved, if I allowed myself to use electrolytic capacitors whose cylindrical bodies are just a little bit longer. The Kemet caps in the Parts List of post #1, are 20mm long. But I found a couple of interesting looking candidates on Mouser's website, that are 25mm long. These candidates will extend a little bit beyond the right hand edge of the PCB, but the leads will still be solderable, the cans will still fit, and the filter will still be useable.

So I ordered some caps to experiment with; they were delivered this afternoon. I am happy to say that both of the new, 25mm long, capacitors are measurably better than the original Kemets. In fact quite a bit better! I used my fairly high-end East Tester ET4510 LCR meter, to measure the ESR of these caps at 100 kHz. I'm interested in high frequency ESR because the unwanted noise from SMPS's, is high frequency. The lower the capacitor ESR, the better the filter will remove high frequency noise.

And boy take a look! The 25mm long caps are both lower ESR than the Kemet 20mm long cap:

Code: 
80-ESH477M050AH4AA    Kemet             470uF 50V  LS=5  D=10  LEN=20  100K ESR  0.045

661-EKYA500E471MJ25S  United Chemi-Con  470uF 50V  LS=5  D=10  LEN=25  100K ESR  0.030

647-UHW1H471MPD       Nichicon          470uF 50V  LS=5  D=10  LEN=25  100K ESR  0.015

So I think, if your physical packaging can accommodate the extra 5mm of capacitor length, you'll get noticeably better filtering -- greater attenuation of SMPS high frequency noise -- if you use the Nichicon caps above. They are better by a factor of 3X. That's a heck of a nice benefit, compared to the cost (+5mm length), in my opinion.

_
Click to expand...
 
You must log in or register to reply here.
Top Bottom