With modern caps and digital ICs, a 2.2uF X7R in 0805 in // with a 10 to 22uF smd polymer per power pin has been working well for me on a few boards. With short wide tracks and groundplanes.
It's worth having a read through the datasheets of X7R caps btw. Their voltage coefficient is heavily dependant on their voltage rating but also on their package size (the bigger the better usually).
It's worth having a read through the datasheets of X7R caps btw. Their voltage coefficient is heavily dependant on their voltage rating but also on their package size (the bigger the better usually).
Measure the leakage. I have.
Leakage typically gets reduced by more than a thousand factor after a slow reform.
As far as comments from a couple people, here is my configuration:
Transformer: 18v X 2
Main power suppler: 10,000uf x 2
Pre regulator caps: 390uf polymer 25V (one for each regulator)
Regulators: 7815/7915 Sparkos discrete
Post regulator caps: 390uf polymer 25V (one for each regulator)
Individual power supply rail filters for I/V and LPF: 120uf polymer (one for each + or – power line per op amp)
capacitor wired directly to +/- power pins of op amp: 120uf polymer
All pre/post regulators are bypassed with 0.1uf metallized polyester (due to space constraints)
All power rail caps are bypassed with 0.1uf metallized Polyphenylene Sulfide.
Caps wired directly to op amp are bypassed with 0.1uf metallized polypropylene.
I found the discrete Sparkos regulators superior to normal 7815/7915. They are much faster/cleaner.
Transformer: 18v X 2
Main power suppler: 10,000uf x 2
Pre regulator caps: 390uf polymer 25V (one for each regulator)
Regulators: 7815/7915 Sparkos discrete
Post regulator caps: 390uf polymer 25V (one for each regulator)
Individual power supply rail filters for I/V and LPF: 120uf polymer (one for each + or – power line per op amp)
capacitor wired directly to +/- power pins of op amp: 120uf polymer
All pre/post regulators are bypassed with 0.1uf metallized polyester (due to space constraints)
All power rail caps are bypassed with 0.1uf metallized Polyphenylene Sulfide.
Caps wired directly to op amp are bypassed with 0.1uf metallized polypropylene.
I found the discrete Sparkos regulators superior to normal 7815/7915. They are much faster/cleaner.
No one reforms brand-new electrolytic caps in commercial electronics manufacturing that I have heard of. Could be wrong...
Marce has stated the same.
I have no reason to doubt either of you.
It would be unreasonable to expect manufacturers to reform electrolytics before assembly.
But un-reformed electrolytics do leak. And they leak badly if the storage conditions make it so.
The rate of leakage will change as the electrolytic reforms while in circuit whenever a positive voltage is applied.
If that positive voltage is from the peaks of an AC signal waveform, then the leakage and reform current will distort the AC signal.
This reforming while passing a signal current is different from reforming before assembly and inserting that very low leakage capacitor into the signal waveform.
I have no reason to doubt either of you.
It would be unreasonable to expect manufacturers to reform electrolytics before assembly.
But un-reformed electrolytics do leak. And they leak badly if the storage conditions make it so.
The rate of leakage will change as the electrolytic reforms while in circuit whenever a positive voltage is applied.
If that positive voltage is from the peaks of an AC signal waveform, then the leakage and reform current will distort the AC signal.
This reforming while passing a signal current is different from reforming before assembly and inserting that very low leakage capacitor into the signal waveform.
Last edited:
adding a bypass capacitor across another can make ringing worse.
They need some damping resistance to snub the ringing.
Bypassing leaves virtually no room for resistance.
There was a time when I had an electrolytic capacitor wired directly to the op amp. I found this was causing a "blare" to the sound and it just did not sound good. Adding a 0.1uf bypass resolved the issue and it actually sounds much clearer and better.
Can you give me an idea or page I can reference on how you would go about "reforming" the capacitor? It would be interesting reading.
It's indeed not that much a question of voltage changes in circuit. It's just a thing to keep in mind when selecting caps. A 10µF/6.3V cap and a 10µF/25V cap will have very different effective values bypassing a 5V supply.
Picking the highest capacitance available can be a bit overkill. It works though.
I'd rather use a bigger bypass cap rather than adding a damping resistance. The small 100n caps are prone to ringing in // with other caps. But caps at 1µ or more often won't be a problem. Manufacturers (murata in particular for ceramics) provide quite advanced simulation tools.
I have just had first hand experience of some of the issues associated with the use of some of the ultra low ESR polymer caps (SPC). I put a 180uF in parallel with a 2200uF Vref electrolytic in my DAC and it blew the input and output transistors in my Superreg supply, presume a resonance? The polymer capacitor also failed short circuit. I don't believe the polymer capacitor was faulty and there was plenty of headroom too.(16V rated cap versus 6.8V)
There is a need to be cautious with these?!
Sent from my iPhone using Tapatalk
There is a need to be cautious with these?!
Sent from my iPhone using Tapatalk
So there was already a total capacitance of 4400uf being fed by the regulator (Panasonic FR) for the Vref and I added 2 no 180uf polymers which is an insignificant capacitance value in comparison. I think you are right that the regulator has oscillated but I think it is probably the ultra low esr of the polymers that has triggered it?
Sent from my iPad using Tapatalk
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.