Capacitor discharge large difference...ESR?

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

Moderators: Kent, luketeaford, Joe.

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Capacitor discharge large difference...ESR?

Post by devinw1 » Sat May 01, 2021 5:00 pm

So, one of the only things I did was change 1 electrolytic cap 2.2uF (C50 below) to a SMT chip when making the SMT boards for my DR-110 module. This is the schematic portion:
Screen Shot 2021-05-01 at 2.56.24 PM.png
Well, I'm confused because doing this drastically changed the time constant of the envelope there at Test Point 12. It should be about 700ms and with the chip capacitor it is more like 100ms. I de-soldered the chip cap and verified it was 2.2uF and then McGeyver'd on a 2.2uF electrolytic, and voila back to about 700ms. I understand the ESR of this chip cap is probably MUCH lower than the electrolytic, but aren't ESR values like 20 ohms maybe..tops? That shouldn't make a difference here because we are talking about discharging through a 68k resistor.

Sooo...what is going on here? :hmm: :hmm:

User avatar
guest
Super Deluxe Wiggler
Posts: 6029
Joined: Mon Aug 19, 2013 11:49 am

Re: Capacitor discharge large difference...ESR?

Post by guest » Sat May 01, 2021 5:09 pm

what is the voltage rating on your SMT cap? im assuming MLCC. ive had issues with time constant shifts between MLCC and electrolytic, but not that drastic. does the curve look different as well? peak voltage, etc.
openmusiclabs.com

User avatar
MikeDB
Veteran Wiggler
Posts: 678
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Capacitor discharge large difference...ESR?

Post by MikeDB » Sat May 01, 2021 5:15 pm

By chip capacitor I assume you mean a X7R ceramic. What voltage is it ? I assume the schematic is with the old electrolytic since it has a '+' sign.

The X7R would still show 2.2uF on measurement but could do anything if being over-voltaged.
Always interested in innovative audio startups needing help and investment, just PM me.

User avatar
KSS
Super Deluxe Wiggler
Posts: 5565
Joined: Mon Jan 25, 2016 7:28 am

Re: Capacitor discharge large difference...ESR?

Post by KSS » Sat May 01, 2021 5:17 pm

Why not measure the ESR of both type caps out of circuit?

Agree with both guest and MikeDB that voltage rating of new cap needs to be checked. And also needs to be significantly higher than if abn elcap. Well known issue with caps.

User avatar
guest
Super Deluxe Wiggler
Posts: 6029
Joined: Mon Aug 19, 2013 11:49 am

Re: Capacitor discharge large difference...ESR?

Post by guest » Sat May 01, 2021 5:27 pm

check out this example datasheet and the capacitance vs voltage graph. its pretty shocking:

https://ds.yuden.co.jp/TYCOMPAS/ut/down ... =Datasheet
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sat May 01, 2021 5:32 pm

This board is using what is supposed to be a 50V cap Y5V.This is the exact part:

https://jlcpcb.com/parts/componentSearc ... Txt=C49217

My THT board also used a 50V cap, polarized electroltyitc like the OG design.

The profile looks good, amplitude is the same, it's just a matter of time constant being off.

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sat May 01, 2021 5:34 pm

guest wrote:
Sat May 01, 2021 5:27 pm
check out this example datasheet and the capacitance vs voltage graph. its pretty shocking:

https://ds.yuden.co.jp/TYCOMPAS/ut/down ... =Datasheet
Wait, so am I reading that correctly in that if you charge that 16V cap to 16V it acts like it is roughly only 10% of it's rated value?

User avatar
guest
Super Deluxe Wiggler
Posts: 6029
Joined: Mon Aug 19, 2013 11:49 am

Re: Capacitor discharge large difference...ESR?

Post by guest » Sat May 01, 2021 5:43 pm

yup.

you can take your capacitor out of circuit, apply a DC bias, and test it, im betting its just not that stable with voltage, which is suprising from an 0805 rated for 50V with only 6V across it, but who knows what that cap really is. heres some more fun reading:

https://www.maximintegrated.com/en/desi ... /5527.html
openmusiclabs.com

User avatar
guest
Super Deluxe Wiggler
Posts: 6029
Joined: Mon Aug 19, 2013 11:49 am

Re: Capacitor discharge large difference...ESR?

Post by guest » Sat May 01, 2021 5:44 pm

also, Y5V is a poor dielectric, so you might get better results from an X7R, or a larger package. or just bite the bullet and use the electrolytic.
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sat May 01, 2021 5:46 pm

guest wrote:
Sat May 01, 2021 5:43 pm
yup.

you can take your capacitor out of circuit, apply a DC bias, and test it, im betting its just not that stable with voltage, which is suprising from an 0805 rated for 50V with only 6V across it, but who knows what that cap really is. heres some more fun reading:

https://www.maximintegrated.com/en/desi ... /5527.html
Heh, I just found this same article from google searching and reading up. I had NO IDEA this was such a thing for MLCCs.

Well, dang, that could CERTAINTLY explain what I'm seeing.

Thanks!

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 12:00 am

This makes me think about something: if we are using MLCC caps for decoupling on ICs, would it not behoove us to use the largest package size, and possibly bump up the value? Granted the actual value doesn't matter a ton, but for example if you're using a 0603 similar to the one in that datahseet, biased at 12V and you think it's a 100nF, it's probably acting more like a 10nF.

User avatar
Mungo
Super Deluxe Wiggler
Posts: 1276
Joined: Fri Jan 04, 2013 8:12 pm

Re: Capacitor discharge large difference...ESR?

Post by Mungo » Sun May 02, 2021 3:08 am

devinw1 wrote:
Sun May 02, 2021 12:00 am
This makes me think about something: if we are using MLCC caps for decoupling on ICs, would it not behoove us to use the largest package size, and possibly bump up the value? Granted the actual value doesn't matter a ton, but for example if you're using a 0603 similar to the one in that datahseet, biased at 12V and you think it's a 100nF, it's probably acting more like a 10nF.
Depends what frequencies that cap is trying to suppress/filter, for bulk low frequency stuff you can usually go for larger parts ok. But MHz and up (like where a 100nF might be specified) then the size of the part and its trace/pad layout starts making enough of a difference that going larger even with more capacitance can end up worse off.

Engineering, the job of n thousand simultaneously competing tradeoffs.

User avatar
KSS
Super Deluxe Wiggler
Posts: 5565
Joined: Mon Jan 25, 2016 7:28 am

Re: Capacitor discharge large difference...ESR?

Post by KSS » Sun May 02, 2021 3:43 am

Mungo wrote:
Sun May 02, 2021 3:08 am
Engineering, the job of n thousand simultaneously competing tradeoffs.
Exactly! :tu:

User avatar
MikeDB
Veteran Wiggler
Posts: 678
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Capacitor discharge large difference...ESR?

Post by MikeDB » Sun May 02, 2021 5:56 am

Bear in mind all capacitors have some form of capacitance change vs voltage curve. For film types it's negligable but electrolytics and tantalums also happily change value with voltage, though not to the extent of Y5Vs which I really don't understand why anybody uses when X7Rs are only a fraction of a cent more expensive.
Always interested in innovative audio startups needing help and investment, just PM me.

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 12:51 pm

KSS wrote:
Sun May 02, 2021 3:43 am
Mungo wrote:
Sun May 02, 2021 3:08 am
Engineering, the job of n thousand simultaneously competing tradeoffs.
Exactly! :tu:
truth. :guinness:

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 12:55 pm

MikeDB wrote:
Sun May 02, 2021 5:56 am
Bear in mind all capacitors have some form of capacitance change vs voltage curve. For film types it's negligable but electrolytics and tantalums also happily change value with voltage, though not to the extent of Y5Vs which I really don't understand why anybody uses when X7Rs are only a fraction of a cent more expensive.
Fair point, and as for the Y5V...yeah, I'm not sure either. I was rapidly firing through JLC's part catalog doing my best to find parts that were on the "basic" list. In some cases there are TDK X7Rs as the basic part, but in some, not. In this case, the only 2.2u was a Y5V (from a fairly unknown chinese manufacturer), which I did not notice.

User avatar
MikeDB
Veteran Wiggler
Posts: 678
Joined: Fri Feb 08, 2019 8:28 am
Location: UK

Re: Capacitor discharge large difference...ESR?

Post by MikeDB » Sun May 02, 2021 1:05 pm

devinw1 wrote:
Sun May 02, 2021 12:55 pm
In this case, the only 2.2u was a Y5V (from a fairly unknown chinese manufacturer), which I did not notice.
That probably added 50% to the droop :-)

Just checked their library and they do seem pretty poor at 2.2uF. I only use Samsung parts, and usually 1206, because of better microphonics so I tend to standardise on the 10uF part C13585
Always interested in innovative audio startups needing help and investment, just PM me.

User avatar
neil.johnson
Super Deluxe Wiggler
Posts: 1717
Joined: Thu Apr 08, 2010 3:51 pm
Location: Cambridge, UK
Contact:

Re: Capacitor discharge large difference...ESR?

Post by neil.johnson » Sun May 02, 2021 2:30 pm

devinw1 wrote:
Sat May 01, 2021 5:34 pm
guest wrote:
Sat May 01, 2021 5:27 pm
check out this example datasheet and the capacitance vs voltage graph. its pretty shocking:

https://ds.yuden.co.jp/TYCOMPAS/ut/down ... =Datasheet
Wait, so am I reading that correctly in that if you charge that 16V cap to 16V it acts like it is roughly only 10% of it's rated value?
Yup, it's an effect that has been known about for years. Kemet has a nice write-up here:
https://ec.kemet.com/blog/mlcc-dielectric-differences/

As long as you know about it you can factor it into the design. For example, for decoupling capacitors:
devinw1 wrote:
Sun May 02, 2021 12:00 am
This makes me think about something: if we are using MLCC caps for decoupling on ICs, would it not behoove us to use the largest package size, and possibly bump up the value?
Sort-of, you factor the capacitance drop in your choice. So if you need, say, 10nF at 3.3V, then you can specify a cheaper capacitor but you must check the DC bias curves and factor that into the selection, so you might choose a 100nF 10V part in some cheap mud knowing that at 3.3V it has dropped down to the desired 10nF. Fractions of a cent per capacitor matters to mobile phone manufacturers.
Oh and you also need to factor in the temperature curves and ageing effects. In the case of mobile phones people replace them after 18 months so it's harder to argue for more stable dielectrics (Another 0.01c won't hurt will it? Oh, you buy capacitors in the 100s of millions? Oh ok then....).

Capacitance variation is of utmost concern when they are used in timing circuits, such as the above.

Neil
Random ramblings :: http://www.njohnson.co.uk
SSI parts and more :: https://www.cesyg.co.uk

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 7:50 pm

Thanks for that link, Neil. One of those things, that well...ya learn something new every day!

I am definitely in the habit of selecting high quality film caps like PS for oscialltors, S/H, etc.. and thinking about cap choice for timing purposes, but in this case I figured it's just an envelope (which is actually adjustable over quite a range in my design), so no need to be *too* picky with the cap...well...that was before I knew that "not being too picky" can mean getting something that's operating at 10% of what I thought it was! This has been eye-opening for sure!

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 7:55 pm

This is a pretty cool online capacitor simulator from the Kemet link posted about:

https://ksim3.kemet.com/capacitor-simulation

User avatar
guest
Super Deluxe Wiggler
Posts: 6029
Joined: Mon Aug 19, 2013 11:49 am

Re: Capacitor discharge large difference...ESR?

Post by guest » Sun May 02, 2021 8:15 pm

maybe we should make some bad filters that you sweep cutoff frequency by adjusting bias on the capacitors!
openmusiclabs.com

User avatar
devinw1
Super Deluxe Wiggler
Posts: 2206
Joined: Tue Aug 07, 2018 11:20 am
Location: Portland, OR
Contact:

Re: Capacitor discharge large difference...ESR?

Post by devinw1 » Sun May 02, 2021 9:04 pm

guest wrote:
Sun May 02, 2021 8:15 pm
maybe we should make some bad filters that you sweep cutoff frequency by adjusting bias on the capacitors!
Hot damn, you might be onto something there! :lol:

User avatar
KSS
Super Deluxe Wiggler
Posts: 5565
Joined: Mon Jan 25, 2016 7:28 am

Re: Capacitor discharge large difference...ESR?

Post by KSS » Sun May 02, 2021 9:47 pm

@guest: I like it!

User avatar
neil.johnson
Super Deluxe Wiggler
Posts: 1717
Joined: Thu Apr 08, 2010 3:51 pm
Location: Cambridge, UK
Contact:

Re: Capacitor discharge large difference...ESR?

Post by neil.johnson » Mon May 03, 2021 6:01 am

guest wrote:
Sun May 02, 2021 8:15 pm
maybe we should make some bad filters that you sweep cutoff frequency by adjusting bias on the capacitors!
IIRC this was discussed on synth-diy several years ago.
At best you get about a 10:1 range, or about 3 octaves, using the worst possible capacitor formulation you can find. And it's highly non-linear so control could be, erm, interesting.

To avoid DC bias feedthrough you'd probably need some sort of balanced arrangement so the DC bias cancels out. Although if you're using such craptastic capacitors in the audio path you might find that getting a useful filter control response is the least of your problems.

Neil
Random ramblings :: http://www.njohnson.co.uk
SSI parts and more :: https://www.cesyg.co.uk

jorg
Ultra Wiggler
Posts: 958
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Re: Capacitor discharge large difference...ESR?

Post by jorg » Mon May 03, 2021 8:07 am

If you replaced an electrolytic with a ceramic X7R (or even worse, Y5V) SMT capacitor, you will see huge performance problems. X7R and Y5V capacitors can have large value in a small space, but are completely unsuitable for signal handling; they are only OK for for power supply filtering (their one saving grace is low ESR). They cause massive distortion (not a few percent, but, depending on signal level, tens of percent). They are also extremely sensitive to temperature and to voltage level; capacitance can vary 2:1 or more, depending on voltage.

I've seen some pretty crazy and unexpected waveforms on these caps.

Post Reply

Return to “Music Tech DIY”