I'm working on a project and need some .1 and .22uF polypropylene film caps and .1uF polyethylene film caps; the problem is that digikey has no polypropylene caps in this size at all! And the polyester caps they have have a 10-item minimum quantity. What could I substitute?
Do you not see yourself building another project like this in the future? Or at least another project needing similar parts? A 10 piece minimum is really not that bad, as long as it doesn't cost an arm and a leg (which in this case, I doubt it does).
Have you thought about other sources? Have you checked Mouser? Can you source thru surplus (All Electronics, Electronic Goldmine, etc)?
I have no problems in getting them here in Germany.
WIMA FKP are available upto 220nF (2$)
WIMA MKP-10 even upto 3u3 (5$)
What voltage rating do you need? 600volts ? 1200 volts?
For high pulse resistance I think there is no alternative...
FKS (Polycarbonat) seems to be only available upto 10n/600 volts...
I'm working on a project and need some .1 and .22uF polypropylene film caps and .1uF polyethylene film caps; the problem is that digikey has no polypropylene caps in this size at all! And the polyester caps they have have a 10-item minimum quantity. What could I substitute?
Could you share your application requirements with us and possibly explain why you think one dielectric would be better then another in your application? I'm not saying all caps (of the same value) are the same, but there is a lot of misconceptions, hype, myths, etc. One only has to look at the DIY golden ear audiophile crowd to see that many are being encouraged to waste a lot of money on some very questionable properties, which some questionable suppliers are only too happy to supply at outrageous prices.
Again I'm not disputing that some caps do have properties that suit some applications better then other types, ESR, high frequency characteristics, etc. but often this is overblown or just not required in most applications.
Lefty
Have you thought about other sources? Have you checked Mouser?
Yes, Mouser doesn't have them either...they list them, but they are discontinued. The only other film caps they have are Kema or something, and they are 5% tolerance and not 1% like the Panasonics. I don't think Mouser does $2.41 first class shipping like digikey either.
Could you share your application requirements with us and possibly explain why you think one dielectric would be better then another in your application?
I'm building a very high gain (60dB) photo preamp and the design specs the Panasonic caps. I wanted to stick with the spec'd caps to be safe, but there is another equivalent cap at digikey that you think would be equivalent I would try it. When I was searching digikey I didn't see any other 50V film caps at all besides the discontinued Panasonics.
So we are coming closer:
- 50 volts; that is not something where polypropylen caps were made for. I expected you to need 500Volts or so...
- 1% ;D
I doubt that this exists; I think you might be lucky to get 5%, which could be realized by extensive selection of parts.
Perhaps if you could link to a schematic drawing and we see how the caps are actually being used so we might further comment on if it seems any specific cap has a unique dielectric requirement.
Lefty
It's an RIAA phono preamp, for a record player. So it's really high gain, and I think the spec'd 2% tolerance caps are actually important because a 5% difference might be like 1/2 decibel or more of difference, just a guess. Film caps are always used in audio when possible, but I'm not sure what difference there is between polyester vs. polypropylene.
This is the circuit.
what difference there is between polyester vs. polypropylene
I though I told you: high voltage and very good pulse resistance, and lowest ESR. They can be produced upto 5% tolerance. And they cost 8 times as much
It's an RIAA phono preamp,
Ok, you said photo preamp before and that would be another kind of animal. The main RIAA preamp requirement along with gain, is to linearize the signal's frequency response. The process of making a record requires that one amplify the frequencies below 1Khz and attenuate the frequencies of above 1Khz to a very specific response curve. This is call RIAA equalization ( RIAA equalization - Wikipedia ). The overall gain of your preamp circuit is probably less then 30db, not particularly high when spread over three opamps, however the more accurate the capacitor values are the better the equalization will be. So I don't think the type of dielectric used in the caps is as important as their accuracy. 1% caps would be nice, but may be hard to find. If one had a large quantity to choose from then of course by measurement one could hand select from the batch.
Lefty
I think you will also struggle to find the resistors mentioned 1K43 and 316R are not exactly standard values. It looks like what they have done is selected a capacitor, measured it and then generated a resistor value based on the required time constant.
So do a bit of reverse engineering, calculate the time constant R14 * C3 with the given values. Then get a capacitor and measure it and then work out what resistor you need. You will have to make the resistors up with several values in series or parallel and measure them rather than relying on the printed value.
Then when you have done that make another pre amp with standard 5% capacitors and the closest 1% resistors you can get and arrange a double blind listening test. I would be astonished if you could tell the difference even if there is a difference in the frequency response curves.
By the way, what has this got to do with an Arduino?
- 1% Grin
1% capacitors exist, I have used 0.3% tolerance capacitors before.
I cannot understand how they can be produced... What type of caps?
I presume you mean you had ordered - say - 10 pcs and what you had gotten did not differ more than 0.3% from the value printed on them? So if you say so...
they were silver mica types, and no, they were marked 0.3%
I cannot understand how they can be produced... What type of caps?
http://industrial.panasonic.com/www-data/pdf/ABD0000/ABD0000CE16.pdf
Finding manufacturers' part numbers is not a problem. Finding distributors that stock and sell small quantities is (sometimes) the show-stopper.
Sorry that I can't help with that.
Regards,
Dave
Quite interesting (and impressive)! I would have doubted before knowing that lists that there are enough costumers to make production of so many thousend different values profitable.
This is just a guess - derived from my idea of production and the offered(?) range of values - that they produce around 5% and hand-select when customer requests come up. This would also explain why there are literally no distributors for it..
I agree, I don't think manufactures plan on making batches of 1% or less tolerance passive components, but rather make a large batch run and then measure and sort to the various tolerances offered. This extra sorting labor is what increases the price for the better tolerance components.
I once worked for a mini-computer company that would pay extra for 7400 series ICs that would be guaranteed to work over a larger Vcc range then the standard offering. The manufacture would just test to that higher spec and paint a blue dot on them and ship them off to us.
Lefty
well, a lot of semiconductors are sold like that, RAM chips and processors are all made in big batches and then automatically tested for how fast they will go, they then mark them with the speed. This is why SRAM chips are often printed on both sides, they print what part it is on the bottom, then test it and print the final part number on the top.
GPUs are similar, they sell "faulty" chips that have some faulty pipelines as lower power graphics cards. Tri-core processors are the same, they are just quad cores with one of the cores borked.
Transistors used to even be the same, they would test them and sell them at different prices depending on the hfe they have. They probably did similar things with valves before that.
I should be very interested in the answer to what accuracy film capacitors (or any other capacitors..) are produced in the first place... Can it be that they just produce ANYTHING around 10 nF and then select whether its closer to 5nF or 20nF, and leave it at the (typical!) 20% error?
I would think - as said - that a PRESELECTION for 1% seems wasteful.. Who would take all those values to be kept in store in the meantime?? So may be order triggered (!) selection will take place...
I would think - as said - that a PRESELECTION for 1% seems wasteful.. Who would take all those values to be kept in store in the meantime?? So may be order triggered (!) selection will take place...
Every manufacture has natural variation in their processes. To try and design different processes to make the various tolerances offered is near impossible and would be even a greater cost for the customers, it's just a selection process to grade for desired specifications.
Lefty