I noticed that for LDOs which usually require a tantalum output cap to ESR-stabilize the loop, SMD chip tantalums seem to be expensive like woah, particularly in the larger values (> ~6.8uF) and more so when you would want to de-rate them 2 or 2.5x on voltage.
I'm wondering if "hybrid polymer" Al-electrolytics would work in the application? They seem to often have an ESR in the appropriate range, which is stable over temp.
Polymers have very low ESRs, milliohms, whereas tantalums are higher,
20-200 mohm maybe. So many LDOs (like LM1117) will load-step ring pretty good when loaded with a polymer.What I have to remind myself is that the only cap isn't the one on the sheet with the regulator. There may be dozens of ceramic bypasses on other sheets of the schematic.
We buy 10uF 16V tantalum for 5 cents. 33u 16V is 16 cents. Most LDOs are happy with a tantalum plus a heap of distributed ceramics. I derate tants 2:1 or 3:1 on voltage, and they seem to be reliable. There seems to be no reason to derate polymers.
You can buy LDOs that are stable ay zero ESR, and go all ceramic.
One alternative at low voltage is to use a giant ceramic, like 47u or more, and deliberately add some ESR, a resistor or some PCB traces. I breadboard to evaluate stability.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Yep, the standard Al-polymers are very low. There also seems to be something called a "hybrid" process which looks to have ESRs somewhat higher, in the 25m to 75m range I think.
Trace L should help isolate the distributed bypass caps from the output a bit, yeah?
Am I looking at the wrong stuff for these tantalums? I don't have much experience with 'em, all the "SMT chip" tantalums in those ranges that I saw on Mouser were like over a buck.
That seems like a lot. We get 4.7uf @35 V for ~$0.30 each. (purchased in
1k lots) and those are through hole. My only tant recommendation is buy from a "good" manufacturer. (Kemet, Murata.. someone like that.) we bought some few cent cheaper tant's from multicomp. (Newarks trade brand) And after a few failures I threw the whole batch out. (~30% failed before 20V.)George H.
Ok, yeah, the thru-holes are cheaper.
Fortunately I'm not a cell phone manufacturer so that should be OK.
I usually like to put zero-ohm jumpers in series with the supplies to make first-article testing easier, and it's easy to put it between the regulator and the output bypass. To avoid instability, it can be swapped out for a small resistor. The required resistance is usually pretty small if not zero.
The other approach is to put it in series with the bypass cap. The idea is to make a lead-lag compensation network with the distributed bypasses--the supply bypass drags the gain way down and then turns into a very small resistor, so that the distributed bypasses make a 1-pole rolloff out at high frequency.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
FYI, last time I did a market overview of polymers and such, I found that:
The populations were something like,
10k total Ta-MnO2, 10% of which are under 50mohm 2k total Ta-Poly, 30% under 50mohm 5k total Al-Poly, 50% under 50mohm, 20% over 200mohmI don't remember the exact numbers anymore, and they will have changed now anyway. Easy enough to select values on Digikey and see for yourself what's out there. Go! Do it!
In any case: search for what you _need_. There is no hard cutoff about any of the above types. You may find a particular part (C, V, ESR) has limited (and expensive) selection among one type. In that case, pick another type and try again; and if that's still not good enough, look at using ceramic with an external resistor.
Nothing hacky about using an external resistor; it guarantees great stability over time!
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com "bitrex" wrote in message news:nRVvA.20618$Bp6.7231@fx36.iad... >I noticed that for LDOs which usually require a tantalum output cap to >ESR-stabilize the loop, SMD chip tantalums seem to be expensive like woah, >particularly in the larger values (> ~6.8uF) and more so when you would >want to de-rate them 2 or 2.5x on voltage. > > I'm wondering if "hybrid polymer" Al-electrolytics would work in the > application? They seem to often have an ESR in the appropriate range, > which is stable over temp.
At these frequencies, not much. It's trace resistance that adds ESR. A trace 5 mils wide, 500 long, 1 oz, adds about 50 mohms. It's unusual to actually get 1oz when you specify it, so a little more ESR is free.
MFR1 AVX TAJB106K016RNJ 10u 16V RoHS 10% 5 cents
is a good one. You need to buy a bunch to get that price. Digikey and Mouser tend to be pricey.
33u 16V shows up asMFR1 AVX TAJD336K016RNJ MFR2 KOA TMC1AETTE336KR MFR3 DIGIKEY 478-3930-1-ND MFR4 PHILIPS 49MC336D016KOAS
and I show an average price of 16 cents.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.