On Thu, 15 Nov 2018 11:12:00 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 11/15/18 10:59 AM, DemonicTubes wrote: >> On Wednesday, November 14, 2018 at 7:15:54 PM UTC-7, Phil Hobbs wrote: >>> On 11/14/18 6:49 PM, gnuarm.deletethisbit@gmail.com wrote: >>>> I use a 1210 X5R 25V 10% MLCC for input DC blocking. Supply is >>>> drying up and I want to place an order to support production for a >>>> while. In performing a Digikey/Mouser search I see there are now X7R >>>> devices in the same footprint. I know the X7R will be more stable >>>> under DC voltage which is an advantage *if* there is much DC on the >>>> input. The amplifier circuit has a 6 volt DC bias on the input (half >>>> way to the +12 volt rail). The caps prevent the input from seeing >>>> this bias. The design passes all functional tests in production >>>> (including frequency response) with the X5R devices. >>>> >>>> I'm thinking of making the X7R an alternate device going forward. >>>> There are only four devices on the board, so a small increase in >>>> price is not so significant. I'm just wondering if there is any >>>> technical downside to using the X7R formulation in place of the X5R. >>>> Or is the X7R formulation all upside relative to the X5R? >>> >>> Apples to apples, X7R is better, for sure. However, high-density MLCCs >>> are all over the map in performance depending on both manufacturer and >>> part number. You really have to find the characteristic curves to know. >>> >>> We've all been feeling the MLCC shortage this year, so I put up a blog >>> post with links to a bunch of makers' characteristic curves. >>> >>> <https://www.electrooptical.net/News/high-value-ceramic-capacitors-they-stink-and-you-cant-get-them-anyway> >>> >>> Samsung has decent characteristic curves published on Digikey. (Note >>> that it's the characteristics link and not the datasheet link that you >>> want. Cute eh?) >>> >>> Cheers >>> >>> Phil Hobbs >>> >>> -- >>> Dr Philip C D Hobbs >>> Principal Consultant >>> ElectroOptical Innovations LLC / Hobbs ElectroOptics >>> Optics, Electro-optics, Photonics, Analog Electronics >>> Briarcliff Manor NY 10510 >>> >>> http://electrooptical.net >>> http://hobbs-eo.com >> >> I was just informed 2 days ago that my Digi-Key order of Samsung caps is on back order :-( >> >> Going to run a production batch with untested no-namers...wish me luck! >> > >I'd test them first. It's not that hard--put two in series and bias the >middle. > >Cheers > >Phil HobbsI thought I invented that! (Actually, I did.) https://www.dropbox.com/s/ckiw55vw20avtbl/C-V_measurement.JPG?dl=0 https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0 -- John Larkin Highland Technology, Inc lunatic fringe electronics
X5R vs X7R MLCC
Started by ●November 14, 2018
Reply by ●November 15, 20182018-11-15
Reply by ●November 15, 20182018-11-15
John Larkin wrote>https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0Terrible. Better use tantalum? What am I missing here?
Reply by ●November 15, 20182018-11-15
On Thursday, November 15, 2018 at 12:04:58 PM UTC-5, John Larkin wrote:> On Thu, 15 Nov 2018 11:12:00 -0500, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > > >On 11/15/18 10:59 AM, DemonicTubes wrote: > >> On Wednesday, November 14, 2018 at 7:15:54 PM UTC-7, Phil Hobbs wrote: > >>> On 11/14/18 6:49 PM, gnuarm.deletethisbit@gmail.com wrote: > >>>> I use a 1210 X5R 25V 10% MLCC for input DC blocking. Supply is > >>>> drying up and I want to place an order to support production for a > >>>> while. In performing a Digikey/Mouser search I see there are now X7R > >>>> devices in the same footprint. I know the X7R will be more stable > >>>> under DC voltage which is an advantage *if* there is much DC on the > >>>> input. The amplifier circuit has a 6 volt DC bias on the input (half > >>>> way to the +12 volt rail). The caps prevent the input from seeing > >>>> this bias. The design passes all functional tests in production > >>>> (including frequency response) with the X5R devices. > >>>> > >>>> I'm thinking of making the X7R an alternate device going forward. > >>>> There are only four devices on the board, so a small increase in > >>>> price is not so significant. I'm just wondering if there is any > >>>> technical downside to using the X7R formulation in place of the X5R. > >>>> Or is the X7R formulation all upside relative to the X5R? > >>> > >>> Apples to apples, X7R is better, for sure. However, high-density MLCCs > >>> are all over the map in performance depending on both manufacturer and > >>> part number. You really have to find the characteristic curves to know. > >>> > >>> We've all been feeling the MLCC shortage this year, so I put up a blog > >>> post with links to a bunch of makers' characteristic curves. > >>> > >>> <https://www.electrooptical.net/News/high-value-ceramic-capacitors-they-stink-and-you-cant-get-them-anyway> > >>> > >>> Samsung has decent characteristic curves published on Digikey. (Note > >>> that it's the characteristics link and not the datasheet link that you > >>> want. Cute eh?) > >>> > >>> Cheers > >>> > >>> Phil Hobbs > >>> > >>> -- > >>> Dr Philip C D Hobbs > >>> Principal Consultant > >>> ElectroOptical Innovations LLC / Hobbs ElectroOptics > >>> Optics, Electro-optics, Photonics, Analog Electronics > >>> Briarcliff Manor NY 10510 > >>> > >>> http://electrooptical.net > >>> http://hobbs-eo.com > >> > >> I was just informed 2 days ago that my Digi-Key order of Samsung caps is on back order :-( > >> > >> Going to run a production batch with untested no-namers...wish me luck! > >> > > > >I'd test them first. It's not that hard--put two in series and bias the > >middle. > > > >Cheers > > > >Phil Hobbs > > I thought I invented that! > > (Actually, I did.) > > https://www.dropbox.com/s/ckiw55vw20avtbl/C-V_measurement.JPG?dl=0Huh, what did you measure? Did you very the frequency till you found the point where C/2 = 50 ohms? ... splitting the voltage between C's and FG 50 ohm source. George H.> > https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0 > > > -- > > John Larkin Highland Technology, Inc > > lunatic fringe electronics
Reply by ●November 15, 20182018-11-15
On Thursday, November 15, 2018 at 11:12:09 AM UTC-5, Phil Hobbs wrote:> On 11/15/18 10:59 AM, DemonicTubes wrote: > > On Wednesday, November 14, 2018 at 7:15:54 PM UTC-7, Phil Hobbs wrote: > >> On 11/14/18 6:49 PM, gnuarm.deletethisbit@gmail.com wrote: > >>> I use a 1210 X5R 25V 10% MLCC for input DC blocking. Supply is > >>> drying up and I want to place an order to support production for a > >>> while. In performing a Digikey/Mouser search I see there are now X7R > >>> devices in the same footprint. I know the X7R will be more stable > >>> under DC voltage which is an advantage *if* there is much DC on the > >>> input. The amplifier circuit has a 6 volt DC bias on the input (half > >>> way to the +12 volt rail). The caps prevent the input from seeing > >>> this bias. The design passes all functional tests in production > >>> (including frequency response) with the X5R devices. > >>> > >>> I'm thinking of making the X7R an alternate device going forward. > >>> There are only four devices on the board, so a small increase in > >>> price is not so significant. I'm just wondering if there is any > >>> technical downside to using the X7R formulation in place of the X5R. > >>> Or is the X7R formulation all upside relative to the X5R? > >> > >> Apples to apples, X7R is better, for sure. However, high-density MLCCs > >> are all over the map in performance depending on both manufacturer and > >> part number. You really have to find the characteristic curves to know. > >> > >> We've all been feeling the MLCC shortage this year, so I put up a blog > >> post with links to a bunch of makers' characteristic curves. > >> > >> <https://www.electrooptical.net/News/high-value-ceramic-capacitors-they-stink-and-you-cant-get-them-anyway> > >> > >> Samsung has decent characteristic curves published on Digikey. (Note > >> that it's the characteristics link and not the datasheet link that you > >> want. Cute eh?) > >> > >> Cheers > >> > >> Phil Hobbs > >> > >> -- > >> Dr Philip C D Hobbs > >> Principal Consultant > >> ElectroOptical Innovations LLC / Hobbs ElectroOptics > >> Optics, Electro-optics, Photonics, Analog Electronics > >> Briarcliff Manor NY 10510 > >> > >> http://electrooptical.net > >> http://hobbs-eo.com > > > > I was just informed 2 days ago that my Digi-Key order of Samsung caps is on back order :-( > > > > Going to run a production batch with untested no-namers...wish me luck! > > > > I'd test them first. It's not that hard--put two in series and bias the > middle.Checking for what, even voltage? Then check with a capacitance meter??? Rick C. Tesla referral code +-+-+ https://ts.la/richard11209
Reply by ●November 15, 20182018-11-15
On 11/15/18 1:53 PM, gnuarm.deletethisbit@gmail.com wrote:> On Thursday, November 15, 2018 at 11:12:09 AM UTC-5, Phil Hobbs wrote: >> On 11/15/18 10:59 AM, DemonicTubes wrote: >>> On Wednesday, November 14, 2018 at 7:15:54 PM UTC-7, Phil Hobbs wrote: >>>> On 11/14/18 6:49 PM, gnuarm.deletethisbit@gmail.com wrote: >>>>> I use a 1210 X5R 25V 10% MLCC for input DC blocking. Supply is >>>>> drying up and I want to place an order to support production for a >>>>> while. In performing a Digikey/Mouser search I see there are now X7R >>>>> devices in the same footprint. I know the X7R will be more stable >>>>> under DC voltage which is an advantage *if* there is much DC on the >>>>> input. The amplifier circuit has a 6 volt DC bias on the input (half >>>>> way to the +12 volt rail). The caps prevent the input from seeing >>>>> this bias. The design passes all functional tests in production >>>>> (including frequency response) with the X5R devices. >>>>> >>>>> I'm thinking of making the X7R an alternate device going forward. >>>>> There are only four devices on the board, so a small increase in >>>>> price is not so significant. I'm just wondering if there is any >>>>> technical downside to using the X7R formulation in place of the X5R. >>>>> Or is the X7R formulation all upside relative to the X5R? >>>> >>>> Apples to apples, X7R is better, for sure. However, high-density MLCCs >>>> are all over the map in performance depending on both manufacturer and >>>> part number. You really have to find the characteristic curves to know. >>>> >>>> We've all been feeling the MLCC shortage this year, so I put up a blog >>>> post with links to a bunch of makers' characteristic curves. >>>> >>>> <https://www.electrooptical.net/News/high-value-ceramic-capacitors-they-stink-and-you-cant-get-them-anyway> >>>> >>>> Samsung has decent characteristic curves published on Digikey. (Note >>>> that it's the characteristics link and not the datasheet link that you >>>> want. Cute eh?) >>>> >>>> Cheers >>>> >>>> Phil Hobbs >>>> >>>> -- >>>> Dr Philip C D Hobbs >>>> Principal Consultant >>>> ElectroOptical Innovations LLC / Hobbs ElectroOptics >>>> Optics, Electro-optics, Photonics, Analog Electronics >>>> Briarcliff Manor NY 10510 >>>> >>>> http://electrooptical.net >>>> http://hobbs-eo.com >>> >>> I was just informed 2 days ago that my Digi-Key order of Samsung caps is on back order :-( >>> >>> Going to run a production batch with untested no-namers...wish me luck! >>> >> >> I'd test them first. It's not that hard--put two in series and bias the >> middle. > > Checking for what, even voltage? Then check with a capacitance meter??? > > Rick C. > > Tesla referral code +-+-+ https://ts.la/richard11209 >No, you put a large resistor across the ends so there's no DC on the meter, ground one end, bias the middle via another large resistor, and put a C meter across the ends. No muss, no fuss. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Reply by ●November 15, 20182018-11-15
On Thu, 15 Nov 2018 10:25:17 -0800 (PST), George Herold <gherold@teachspin.com> wrote:>On Thursday, November 15, 2018 at 12:04:58 PM UTC-5, John Larkin wrote: >> On Thu, 15 Nov 2018 11:12:00 -0500, Phil Hobbs >> <pcdhSpamMeSenseless@electrooptical.net> wrote: >> >> >On 11/15/18 10:59 AM, DemonicTubes wrote: >> >> On Wednesday, November 14, 2018 at 7:15:54 PM UTC-7, Phil Hobbs wrote: >> >>> On 11/14/18 6:49 PM, gnuarm.deletethisbit@gmail.com wrote: >> >>>> I use a 1210 X5R 25V 10% MLCC for input DC blocking. Supply is >> >>>> drying up and I want to place an order to support production for a >> >>>> while. In performing a Digikey/Mouser search I see there are now X7R >> >>>> devices in the same footprint. I know the X7R will be more stable >> >>>> under DC voltage which is an advantage *if* there is much DC on the >> >>>> input. The amplifier circuit has a 6 volt DC bias on the input (half >> >>>> way to the +12 volt rail). The caps prevent the input from seeing >> >>>> this bias. The design passes all functional tests in production >> >>>> (including frequency response) with the X5R devices. >> >>>> >> >>>> I'm thinking of making the X7R an alternate device going forward. >> >>>> There are only four devices on the board, so a small increase in >> >>>> price is not so significant. I'm just wondering if there is any >> >>>> technical downside to using the X7R formulation in place of the X5R. >> >>>> Or is the X7R formulation all upside relative to the X5R? >> >>> >> >>> Apples to apples, X7R is better, for sure. However, high-density MLCCs >> >>> are all over the map in performance depending on both manufacturer and >> >>> part number. You really have to find the characteristic curves to know. >> >>> >> >>> We've all been feeling the MLCC shortage this year, so I put up a blog >> >>> post with links to a bunch of makers' characteristic curves. >> >>> >> >>> <https://www.electrooptical.net/News/high-value-ceramic-capacitors-they-stink-and-you-cant-get-them-anyway> >> >>> >> >>> Samsung has decent characteristic curves published on Digikey. (Note >> >>> that it's the characteristics link and not the datasheet link that you >> >>> want. Cute eh?) >> >>> >> >>> Cheers >> >>> >> >>> Phil Hobbs >> >>> >> >>> -- >> >>> Dr Philip C D Hobbs >> >>> Principal Consultant >> >>> ElectroOptical Innovations LLC / Hobbs ElectroOptics >> >>> Optics, Electro-optics, Photonics, Analog Electronics >> >>> Briarcliff Manor NY 10510 >> >>> >> >>> http://electrooptical.net >> >>> http://hobbs-eo.com >> >> >> >> I was just informed 2 days ago that my Digi-Key order of Samsung caps is on back order :-( >> >> >> >> Going to run a production batch with untested no-namers...wish me luck! >> >> >> > >> >I'd test them first. It's not that hard--put two in series and bias the >> >middle. >> > >> >Cheers >> > >> >Phil Hobbs >> >> I thought I invented that! >> >> (Actually, I did.) >> >> https://www.dropbox.com/s/ckiw55vw20avtbl/C-V_measurement.JPG?dl=0 >Huh, what did you measure? Did you very the frequency till you >found the point where C/2 = 50 ohms? ... splitting the voltage >between C's and FG 50 ohm source. > >George H.Right, I tweaked the frequency to find the -3 dB point, with the 50 ohm generator. Of course, that measures C/2. Both caps see the same DC voltage. I feel better doing this than trusting some digital display. As I turn the frequency knob, I can see the smooth change of voltage vs frequency and be confident everything is OK. Maybe see the ESL and ESR at higher frequencies. It's fun to drive one of these nonlinear caps with a big square wave too. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by ●November 15, 20182018-11-15
On Thu, 15 Nov 2018 17:14:02 GMT, <698839253X6D445TD@nospam.org> wrote:>John Larkin wrote >>https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0 > >Terrible. >Better use tantalum? >What am I missing here?There's nothing missing here. I just measured a cap. That 4.7u is fine to filter the output of a 1.2 or 3.3 volt switcher, or for AC coupling. Or use the 47u 6.3v version, which has an even worse cv curve. Tantalums have a lot of ESR, which is sometimes good, sometimes bad. They are always big and expensive. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●November 16, 20182018-11-16
John Larkin wrote>On Thu, 15 Nov 2018 17:14:02 GMT, <698839253X6D445TD@nospam.org> >wrote: > >>John Larkin wrote >>>https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0 >> >>Terrible. >>Better use tantalum? >>What am I missing here? > >There's nothing missing here. I just measured a cap. > >That 4.7u is fine to filter the output of a 1.2 or 3.3 volt switcher, >or for AC coupling. Or use the 47u 6.3v version, which has an even >worse cv curve. > >Tantalums have a lot of ESR, which is sometimes good, sometimes bad. >They are always big and expensive.Yes, OK tantalums are more expensive, I use those all the time for linear regulator decoupling. Again, do not save on parts... I could not see me use a capacitor for AC coupling between stages if its value could be 10x out of specified range. Makes me wonder too how microphonic those caps are? With microphonic I mean creating voltages when vibrating? I sort of like to tap on boards with a pen... to find dry joints etc.. Was no there a tread about it here years ago? And the reverse, making sound when AC is applied. But then I am biased, I like tantalums, never had a problem with those.. except when put in the wrong way around...
Reply by ●November 16, 20182018-11-16
On Friday, November 16, 2018 at 4:30:28 AM UTC-5, 69883925...@nospam.org wrote:> John Larkin wrote > >On Thu, 15 Nov 2018 17:14:02 GMT, <698839253X6D445TD@nospam.org> > >wrote: > > > >>John Larkin wrote > >>>https://www.dropbox.com/s/ilppasx0ym7b98y/Ccap_CV.JPG?dl=0 > >> > >>Terrible. > >>Better use tantalum? > >>What am I missing here? > > > >There's nothing missing here. I just measured a cap. > > > >That 4.7u is fine to filter the output of a 1.2 or 3.3 volt switcher, > >or for AC coupling. Or use the 47u 6.3v version, which has an even > >worse cv curve. > > > >Tantalums have a lot of ESR, which is sometimes good, sometimes bad. > >They are always big and expensive. > > Yes, OK tantalums are more expensive, > I use those all the time for linear regulator decoupling. > Again, do not save on parts... > I could not see me use a capacitor for AC coupling between stages if its value > could be 10x out of specified range.Hey, I assume COG/NPO caps don't have any voltage coef. (I use those for AC coupling, signal filters.)> > Makes me wonder too how microphonic those caps are? > With microphonic I mean creating voltages when vibrating? > > I sort of like to tap on boards with a pen... to find dry joints etc.. > Was no there a tread about it here years ago? > And the reverse, making sound when AC is applied. > > But then I am biased, I like tantalums, never had a problem with those.. > except when put in the wrong way around...You have to picky about manufacturers, someone saved a few cents on cheaper tants, and the 35V ones failed at 15-20V... semi my mistake as I used them on a 24V supply line. (older design... in the past I don't think tant's were so flaky.) George H.
Reply by ●November 16, 20182018-11-16
