Recommend SMD resistor

[mtl777]

I need 4.64K SMD resistors for the circuit that controls input sensitivity of the left and right input channels of an AD converter unit. I guess these resistors must be high precision (1% tolerance or lower) in order for the left and right input channels to match very closely in level. What brands/series would you recommend? Also what type (metal film, thin film or thick film) is best?

I remember Vishay-Dale CMF or RN series were highly recommended for through hole. What do you think of soldering a small enough CMF resistor on the SMD pads instead of using an SMD one? Any drawbacks if I do this? How does the quality of through hole resistors compare to SMD?

Any inputs would be greatly appreciated. Thanks in advance!

[stoc005]

For SMT, here is my opinion.
1. Any metal film MELF resistors, like Vishay .25W SMM0204 series. MELFs are round and can roll off your board before they are soldered down.
2. Panasonic ERA series 1210 size, .25W, .1%, 25 PPM and pricey.
3. Susuma RGH or RR series. RGH series is 0805 and .25W. RR series is 0805 and .1 W

For through hole....Panasonic ERO52C series. Small and metal film.

All are metal film.

[mtl777]

Originally Posted by stoc005

For SMT, here is my opinion.
1. Any metal film MELF resistors, like Vishay .25W SMM0204 series. MELFs are round and can roll off your board before they are soldered down.
2. Panasonic ERA series 1210 size, .25W, .1%, 25 PPM and pricey.
3. Susuma RGH or RR series. RGH series is 0805 and .25W. RR series is 0805 and .1 W

For through hole....Panasonic ERO52C series. Small and metal film.

All are metal film.

Thank you so much for these recommendations!

[jcx]

watch power ratings - use over rated, low TCR types for critical feedback gain setting R

but really large chips can be a reliability problem due to cracking


melf may have the dreaded steel endcaps


thin film - NiCr or TaN are ~$1 each


vishay bulk metal foil in smt >$10 each


I'm sure someone will be along to point out that the most accurate resistors may not "sound best"

[mtl777]

Originally Posted by jcx

watch power ratings - use over rated, low TCR types for critical feedback gain setting R

What is TCR?

Originally Posted by jcx

melf may have the dreaded steel endcaps

What's the problem with steel end caps?

Originally Posted by jcx

thin film - NiCr or TaN are ~$1 each

vishay bulk metal foil in smt >$10 each

Man, those are expensive! I don't think I want to spend that much for resistors. The $1 thin film I can manage. So you think they're better than metal film?

Originally Posted by jcx

I'm sure someone will be along to point out that the most accurate resistors may not "sound best"

You mean accurate but too clinical sounding? For an ADC, my preference would be to capture the source as transparent as possible, without any coloration.

Thanks!

[jcx]

Temperature Coefficient of Resistance - a possible source of low frequency distortion with signal power heating of the resistor

steel/magnetic materials in resistor construction are pointed at by the hard core subjectivist crowd as possibly causing poor sound - as a industrial/scientific instrument designer I've never seen any other place where this is mentioned as a error source at audio signal levels - an effect can be just barely be measured at RF and 100s of watts in Ni plated connectors

I expect it to be really hard to measure distortion of the common "1% metal film" when self heating is low but the specs of thin film and bulk metal foil are even better

[mtl777]

Originally Posted by jcx

Temperature Coefficient of Resistance - a possible source of low frequency distortion with signal power heating of the resistor

So a lower TCR value is better?

Originally Posted by jcx

steel/magnetic materials in resistor construction are pointed at by the hard core subjectivist crowd as possibly causing poor sound - as a industrial/scientific instrument designer I've never seen any other place where this is mentioned as a error source at audio signal levels - an effect can be just barely be measured at RF and 100s of watts in Ni plated connectors

Do you have any specific brands/series recommendations that do not have steel end caps?

Originally Posted by jcx

I expect it to be really hard to measure distortion of the common "1% metal film" when self heating is low but the specs of thin film and bulk metal foil are even better

Thanks for this info!

[Xoc1]

Rhopint Stock a range of precision SMD resistors

http://www.rhopointcomponents.com/pr...=Precision+(sm)

I have some old Rhopoint metal foil resistors from when they were rated at 5ppm (current spec is 0.14ppm!). They ones I have are magnetic. They sound good to me, but they were horrendously expensive!
These days, the thin film resistors rated at 10ppm are good enough for most precision circuits.

[mtl777]

Originally Posted by Xoc1

Rhopint Stock a range of precision SMD resistors

http://www.rhopointcomponents.com/pr...=Precision+(sm)

I have some old Rhopoint metal foil resistors from when they were rated at 5ppm (current spec is 0.14ppm!). They ones I have are magnetic. They sound good to me, but they were horrendously expensive!
These days, the thin film resistors rated at 10ppm are good enough for most precision circuits.

Thanks, Rhopoint sure has good high precision resistors in their product line! One more high quality option to add to my list.

Anyone, I have a question about MELF resistors. What's the reasoning behind their cylindrical shape? I would think that a flat surface would make for more contact area (hence better for soldering) than cylindrical. So why cylindrical?

I would also like to know the color coding of MELF resistors. The resistors I am trying to replace are MELF with 4 bands. They are still in the circuit and I would like to confirm the value I read from my DMM with the value based on their color coding.

Thanks!

[stoc005]

MELFs were the among the first SMT parts. Since they had a round body, they were manufactured a lot like the TH parts except the end caps with leads were replaced by just end caps. Some semiconductor diodes started out as MELFs too as TH parts shifted to SMT. MELF resistors use the same color code as the leaded parts. 4 colors? Probably 10-->5% parts. 1%er use 5 bands, I think. Same color code as the old carbon comp. Allen Bradley resistors of old.

Flat SMT resistors are "new". Made on ceramic slabs and cut to size, value trimmed w/ a laser and coated for protection of the resistive element. Metal film is not the most common resistive element. Thin and thick film is common.