I’m currently using the below EEPROM for my application. In the data sheet 0.1uF is recommended (not a constraint, just roughly mentioned). I have a 1uF decoupling capacitor, couldn’t get a 0.1uF due to availability issues. Can I go ahead with the 1uF, will it cause any problems?
3
-
3\$\begingroup\$ Can't you just tear apart any random electronics junk and salvage a 100nF from it? It's by far the most common cap value. It is really strange that you can't get the most common cap value due to availability either. That means you can't get any caps at all. \$\endgroup\$Lundin– Lundin2025-12-05 07:53:55 +00:00Commented yesterday
-
4\$\begingroup\$ Silly question, maybe, but have you tried looking for 100nF capacitors rather than 0.1uF? They're the same thing. \$\endgroup\$Spehro 'speff' Pefhany– Spehro 'speff' Pefhany2025-12-05 12:13:43 +00:00Commented yesterday
-
2\$\begingroup\$ Or 82nF, or 120nF, or... \$\endgroup\$the busybee– the busybee2025-12-05 12:21:29 +00:00Commented yesterday
Add a comment
|
4 Answers
\$\begingroup\$
\$\endgroup\$
It is very unlikely that using 1uF instead of 100nF causes problem under typical assumed use scenario.
But then again, you also don't state your scenario, i.e. which kinds of capacitors you are going to use and if the memory chip connects to MCU with a few meter cable with only one capacitor on the memory chip side, and how close it will be.
What may be a problem is the rest of the circuit, if you change all 100nF caps in the design into 1uF caps, as in some cases it could be better to have a 10nF capacitor than 1uF. It's also weird that 100nF caps are extremely common so there should be no issues sourcing 100nF caps, or other nearby value.
\$\begingroup\$
\$\endgroup\$
I don't think it will cause a problem; the pin specified on the data sheet is only the supply voltage. Therefore, attaching a 1 uF capacitor (ceramic) to your EEPROM's supply pin will not change anything.
\$\begingroup\$
\$\endgroup\$
When decoupling digital (or any high-frequency, spiky, sensitive) components, use low inductance capacitors, e.g., disc or multi-layer ceramic caps, and place them close the the components in question. Add additional bypass caps, e.g., electrolytics, to handle lower frequency noise. Using only a 1uf cap, if it is an electrolytic, without any additional bypassing is asking for trouble! I do not know whether the poster used an electrolytic (1uf is common), or a ceramic (1uf available, but maybe less common) cap. If a ceramic cap (low inductance) was used, then 1uf (as opposed to .1uf) should not cause any problem. If an electrolytic was used, leave it in place, but add a good ceramic bypass cap to the circuit (.1 uf is a good value). Ensure there is a good ground plane. If this does not solve the problem, check signal line lengths, stray pickup, ground loops, and so on. Just take a look at the waveforms on a scope! That often reveals the problem. If you have a good scope, you can even check the effectiveness of bypassing. I do this all the time.
\$\begingroup\$
\$\endgroup\$
2
The datasheet states:
In order to secure a stable DC supply voltage, it is recommended to decouple the VCC line with a suitable capacitor (usually of the order of 10 nF to 100 nF) close to the VCC/VSS package pins.
You should rather place a 10 nF decoupling capacitor, instead of 1 μF. If in doubt, always follow the instructions provided by the datasheet.
-
\$\begingroup\$ Blind trust in datasheets is unfounded. Many of them are poorly written and contain misleading advice. \$\endgroup\$jpa– jpa2025-12-05 16:40:44 +00:00Commented yesterday
-
\$\begingroup\$ Besides, the advice given here by ST isn't really outlandish. It's not like they recommended 10..100 μF. \$\endgroup\$Velvet– Velvet2025-12-05 17:29:34 +00:00Commented yesterday
