PSU whistle

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., lisa, luketeaford, Kent

Post Reply
User avatar
Dimitree
Veteran Wiggler
Posts: 654
Joined: Mon Aug 01, 2011 9:59 am

PSU whistle

Post by Dimitree » Thu Mar 19, 2020 3:36 pm

I built this DC-DC converter, and while it seems to work fine, I noticed a subtle but audible whistle on the PCB (probably close to the inductor). I'm not talking about noise on the electronic signal..I'm talking about acoustic noise.
Is that normal? I also noticed that the whistle increases the more circuits I connect to it, and I can hear LFO clicks when I connect a LFO that has a frequency led indicator.
I didn't connect any audio circuit so I don't know if this noise is on the electronic signal too.
You do not have the required permissions to view the files attached to this post.

User avatar
EATyourGUITAR
has no life
Posts: 4051
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: PSU whistle

Post by EATyourGUITAR » Thu Mar 19, 2020 3:41 pm

whatever inductor or transformer you used for L15, that is the problem. I have heard people say that coil wine is related to the potting material breaking down. I always thought it was an electromagnet transducer converting AC into mechanical motion as one does. something has to push on something else. I see many coils on a video card. but if you only have one transformer then maybe the potting material is not %100.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
latigid on
Ultra Wiggler
Posts: 908
Joined: Fri Oct 25, 2013 11:05 am

Re: PSU whistle

Post by latigid on » Thu Mar 19, 2020 3:59 pm

What material are the capacitors made of? I would suspect high-ish value ceramic types of >1uF can be susceptible to a sort of piezoelectric effect. This can be remedied in some cases by switching to tantalums.

Is it a fabbed PCB or something more prototype? I've heard the design for such regulators/supplies can be very sensitive to layout.

User avatar
Dimitree
Veteran Wiggler
Posts: 654
Joined: Mon Aug 01, 2011 9:59 am

Re: PSU whistle

Post by Dimitree » Thu Mar 19, 2020 6:20 pm

thanks guys

for the inductor I used a Coilcraft, this one: https://www.coilcraft.com/msd1048.cfm
the caps are not tantalum, I used ceramics (MLCC) by Murata (10uF the largest)

this is the layout I made (please note, the linear regulators I used are different than those shown on the schematic, but the DC-DC converter is the same)

Image

User avatar
EATyourGUITAR
has no life
Posts: 4051
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: PSU whistle

Post by EATyourGUITAR » Thu Mar 19, 2020 6:23 pm

latigid on wrote:
Thu Mar 19, 2020 3:59 pm
What material are the capacitors made of? I would suspect high-ish value ceramic types of >1uF can be susceptible to a sort of piezoelectric effect. This can be remedied in some cases by switching to tantalums.

Is it a fabbed PCB or something more prototype? I've heard the design for such regulators/supplies can be very sensitive to layout.
Wow I didn't even know ceramic caps can do that. I heard about them being magnetic or maybe microphonic. I heard about the memory effect, the histerysis etc.. But never knew you could generate vibration. So if this is true then you can confirm buy hitting it with a hammer while hooked up to a high impedance oscilloscope. Make sure it triggers on input and save it. It will be a fast transient low amplitude.
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
EATyourGUITAR
has no life
Posts: 4051
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: PSU whistle

Post by EATyourGUITAR » Thu Mar 19, 2020 6:29 pm

Dimitree wrote:
Thu Mar 19, 2020 6:20 pm
thanks guys

for the inductor I used a Coilcraft, this one: https://www.coilcraft.com/msd1048.cfm
the caps are not tantalum, I used ceramics (MLCC) by Murata (10uF the largest)

this is the layout I made (please note, the linear regulators I used are different than those shown on the schematic, but the DC-DC converter is the same)

Image
1.6mm PCB or 2.0mm PCB?
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

User avatar
Dimitree
Veteran Wiggler
Posts: 654
Joined: Mon Aug 01, 2011 9:59 am

Re: PSU whistle

Post by Dimitree » Thu Mar 19, 2020 6:31 pm

1.6mm PCB, ground plane on both layers

KSS
Super Deluxe Wiggler
Posts: 1234
Joined: Mon Jan 25, 2016 7:28 am

Re: PSU whistle

Post by KSS » Fri Mar 20, 2020 2:03 am

Press a rubber pencil eraser against each component while listening. When the noise changes, you *may* have found the guilty part. Keep going anyways and check all parts. Eraser physically damps oscillation, to check for component vibration under power.

Does gentle 'flexing' of the PCB change or eliminate the noise> Twist and bow in both directions.
above checking for physical integrity of components and soldering

Does the noise change if the orientation of the board changes? If you move it to another location in another room, or another physical location entirely? -like somebody else's house, or work building
Checking here for both electrical power issues outside your board and also EMI RFI in the area.

User avatar
guest
Super Deluxe Wiggler
Posts: 4000
Joined: Mon Aug 19, 2013 11:49 am

Re: PSU whistle

Post by guest » Fri Mar 20, 2020 3:36 am

transformer vibration is very common. they are basically little speakers without a cone, but once you attach them to a pcb you sort of have a cone. as others have mentioned, potting can help here. as that one is already potted, the potting might not have gotten througout all the inner coils. it can happen in capacitors as well, but that is less common if the capacitor is not seeing a lot of AC voltage.
openmusiclabs.com

User avatar
Altitude909
Super Deluxe Wiggler
Posts: 3299
Joined: Wed Aug 24, 2011 5:13 pm
Location: Meesheegan

Re: PSU whistle

Post by Altitude909 » Fri Mar 20, 2020 7:39 am

Interestingly enough, I built the same thing (this is the PSU out of the Korg MS-20 mini) and had the same thing. I assumed that this was the result of the coil in the actual device being a custom wound one, I couldnt find one that matched the spec

User avatar
Dimitree
Veteran Wiggler
Posts: 654
Joined: Mon Aug 01, 2011 9:59 am

Re: PSU whistle

Post by Dimitree » Sat Mar 21, 2020 1:40 pm

that's exactly that one from the MS20!
did you find it noisy when connected to audio circuits? I hope is just an "acoustic" issue

User avatar
guest
Super Deluxe Wiggler
Posts: 4000
Joined: Mon Aug 19, 2013 11:49 am

Re: PSU whistle

Post by guest » Sat Mar 21, 2020 1:45 pm

you can try covering it all with hotglue. the nice thing about hotglue is that you can chip it off if you dont want it on there someday.
openmusiclabs.com

User avatar
Altitude909
Super Deluxe Wiggler
Posts: 3299
Joined: Wed Aug 24, 2011 5:13 pm
Location: Meesheegan

Re: PSU whistle

Post by Altitude909 » Sat Mar 21, 2020 2:07 pm

it was just coil whine. did loading it down reduce it? I shelved mine and never really got into it due to external factors.. It looks like a decent design though, it would be cool to get it working

KSS
Super Deluxe Wiggler
Posts: 1234
Joined: Mon Jan 25, 2016 7:28 am

Re: PSU whistle

Post by KSS » Sat Mar 21, 2020 5:59 pm

guest wrote:
Sat Mar 21, 2020 1:45 pm
you can try covering it all with hotglue. the nice thing about hotglue is that you can chip it off if you dont want it on there someday.
Be careful doing this! Hot glue formulas vary *greatly*. Not all are compatible with PCBs and components from both a thermal and a chemical basis.

Suggest instead electronics grade RTV silicone. These also vary greatly so if you get one that smells like vinegar, do NOT use it on a PCB.

There are any number of other shoe goo's and the like which *may* work. But doing this without first testing over time -in the absence of specific details of prior good results from soneone you trust- is not a good idea.

FWIW, I trust guest, and would have said the same thing many years ago. Hot glue was different then, and we didn't have the years then to now to see the negative results.

edit: and before covering it with anything, do the pencil eraser damping test. That will let you know if damping is going to help. If you do fins damping helps, consider a designed for electronics conformal coating a valid option. /edit

User avatar
guest
Super Deluxe Wiggler
Posts: 4000
Joined: Mon Aug 19, 2013 11:49 am

Re: PSU whistle

Post by guest » Sun Mar 22, 2020 3:38 am

thanks for the tip on hotglue! ive used it a lot for electronics stuff, but rarely is it on something that needs to last too long (lots of prototypes and such). i did do one run of 500 boards that i hotglued. that was 20 years ago, and they still work, but as you mention, the glues have probably changed. do they etch the board or components? i know they cause issues with RF or capacitive circuits if too close, as they change the dielectric properties.
openmusiclabs.com

User avatar
Altitude909
Super Deluxe Wiggler
Posts: 3299
Joined: Wed Aug 24, 2011 5:13 pm
Location: Meesheegan

Re: PSU whistle

Post by Altitude909 » Sun Mar 22, 2020 7:07 am

found this pic of the device for reference.
digital_board.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
EATyourGUITAR
has no life
Posts: 4051
Joined: Tue Aug 31, 2010 12:24 am
Location: Providence, RI, USA

Re: PSU whistle

Post by EATyourGUITAR » Sun Mar 22, 2020 11:18 am

IMO the biggest problem with hot melt glue is that it shrinks a lot. this can move a component sideways and lift it right off the PCB. I thought it was poly vinyl acetate but wikipedia says poly ethyl vinyl acetate with additives. it also says that it could be made out of any thermoplastic but generally not a very stiff material after application. no fully crosslinked long chain polymers.

https://en.wikipedia.org/wiki/Hot-melt_adhesive
WWW.EATYOURGUITAR.COM <---- MY DIY STUFF

Post Reply

Return to “Music Tech DIY”