Upgrade bad ceramic cap on digital board, any advice?

[TasoTaso]

Hi, I should change the ceramic caps to my Deq2496.
The image below shows the values ​​I have to go to put.
Part of it is missing, capacity sensing work is not done yet.
The reason is that the original ceramic capacitors are bad.
They are the capacitors that are on the "everything" power supplies.

For this work I bought wima MKT capacitors (polyester), 2.5mm pitch.

Is it a good choice I had to choose the MKP (polypropylene)?
Unfortunately polypropylene is a bit too big and I was hoping that the polyester could do the job well.

One last question:
It would be appropriate that the old ceramic capacitors take them off right? Or am I going to put the wima directly in parallel with the ceramics?
I think the ceramics should be removed!

Thanks for your suggestions!

 

[rayma]

Paralleled capacitors can cause resonances, so if you can remove the ceramics without damage,
do so. But if not, just parallel them anyway.

Why are the ceramic capacitors bad?

 

[TasoTaso]

Yes, I think that I can remove the ceramics without too many problems.
On other 3ds here on DIY and also on other forums there has been a lot of talk about the very poor quality of all the ceramics used.
So I'm changing them (as others who have deq have done) for wima polyester.

Do you think polyester is suitable for power supplies (+ 1.2v / + 3.3v / + -15v)? Unfortunately I can't put polypropylene.

Many thanks

 

[rayma]

If the values are under 0.1uF, COG/NPO ceramic types are very good.
Polyester (Mylar) dielectric has easily measurable distortion.
But low inductance is the main figure of merit on a digital board.

 

[chris719]

Replace them with the appropriate value X7R if they are truly bad and don't bother adding film caps. These are decoupling power supply and you may not want tiny high Q caps. I rarely see MLCCs that are bad...

 

[abraxalito]

If the values are under 0.1uF, COG/NPO ceramic types are very good.

Indeed they are good, extremely good. But for decoupling you might find they're too good (i.e too high Q) and X7Rs in practice introduce a bit more loss to help damp resonances.

 

[kevinkr]

I find it hard to believe that any generic X7R ceramic is better/worse than any other except possibly in terms of reliability and flex tolerance. These are used to bypass the supply rails of digital ICs and the important thing is low internal losses at the frequencies of interest.

At most I would replace with appropriate X7R from Kemet, Samsung,Yageo, Vishay or similar. COG/NPO types are available, but will be expensive and likely of no benefit in the application. You do not want film caps for these locations, their construction results in fairly high inductance and very poor performance as bypass capacitors.

MLCC ceramic caps are an excellent choice for the applications where you generally find them, they should not be used in high quality analog circuitry except again for supply decoupling.

 

[TasoTaso]

I was simply copying this work.
Work done by some forumer, I don't remember where I got the photo!



I have no specific knowledge like you on the type of "certain more suitable capacitors".
If the best, for power supplies, is X7R ... I'll use X7R.
I want to make sure I put the best in!

Too bad I bought 100pcs wima mkt for nothing, but if they need X7R I buy them.

Do I proceed with X7R?

Thanks!

 

[kn0ppers]

The reason is that the original ceramic capacitors are bad.
They are the capacitors that are on the "everything" power supplies.

You must be joking?! Or do you seriously want to solder 2.54mm pitch film caps onto 0603 SMD footprints in the digital section? Apart from the technical aspect, simply soldering that will be a frustrating experience.

Edit: 100nF 0603 50V X7R cost less than a cent a piece. For example CL10B104KB8NN from Samsung, I have a full reel of those.

 

[TasoTaso]

I've already started doing it. Ultra short pins, soldering iron and immobile hand. No problem, why?

 

[chris719]

Mechanically questionable over time, also may be less effective due to increased mounted inductance from the lead lengths. Overall just not a good idea to do with that guy did. Just because another forum user did something does not mean it is a good idea. Throwing fashionable film caps at a design will almost never improve it.

 

[TasoTaso]

Ok, i think i will follow your advice using smd cap.
I don't know if it is more frustrating to solder smd or tht on smd pad. I'll find out '.

In an online shop in my country they sell these Kemet, 100nF.

Being that I also have voltages of 15v..I would choose to buy them all equal to 50v work. It is these, they should be fine, right?

Is the size / series I need 0603? I know very little about smd.
Is Automotive Grade..a don't know the difference..I suppose they are better than the "normal" ones.

C0603C104K5RACAUTO KEMET - Capacitor: ceramic | MLCC; 100nF; 50VDC; X7R; +-10%; SMD; 0603 | TME - Electronic components

Datasheet: https://www.tme.eu/Document/538d5a015e3573476a0af80cd70a1929/C0603C104K5RACAUTO.pdf

Datasheet: https://www.tme.eu/Document/81603c358f97099ddac4951c4da5ac68/KEM_C1023_X7R_AUTO_SMD.pdf

Thanks again

 

[chris719]

You can use a higher voltage, no problem. You can tell the size by measuring the existing parts. Automotive just means the part has undergone testing to AEC-Q200 usually, which covers things like operating temperature range, mechanical shock and vibration resistance, etc. You don't need it, but it won't hurt. I doubt the components are any different. Kemet is good quality.

Are you sure the original caps are bad?

 

[JMFahey]

here on DIY and also on other forums there has been a lot of talk about the very poor quality of all the ceramics used.

Where is the [rolleyes] emoticon when you need it?

 

[jean-paul]

Copying behavior is not the best choice in many cases.

 

[TasoTaso]

My own are based on this 3d:

Another (ultimate?) Behringer mod - what can $50 do to DEQ2496 (PART 1)

Just write on google anyway: " deq2496 ceramic capacitors "and endless discussions come out about bad capacitors.

 

[kn0ppers]

People who know little to nothing about electronics trying to impress people who know even less, that is what most of those discussions are probably.

The user (avp1) modding the Behringer knew which caps were in the signal path and didn't just go in and wildly replace caps. He measured, redrew schematics, simulated, modified and tested. Whole different story than just going in and soldering film caps everywhere.

 

[MarcelvdG]

My advice would be not to change the power supply decoupling capacitors at all unless you have reasons to believe that the capacitors actually got damaged. It's a different matter altogether for capacitors in the signal path.

 

[canvas]

I would not replace those ceramics. Film caps will introduce lead inductance and ceramics are better at high frequency.

 

[TasoTaso]

People who know little to nothing about electronics trying to impress people who know even less, that is what most of those discussions are probably.

The user (avp1) modding the Behringer knew which caps were in the signal path and didn't just go in and wildly replace caps. He measured, redrew schematics, simulated, modified and tested. Whole different story than just going in and soldering film caps everywhere.

Yes, it's true. However what Avp1 did I have already replicated on my deq.