searching for 96.25 ohm resistors

[sneily]

I need a couple of resistors to repair my Krohn-Hite 4200 Oscillator.
96.25 ohm, and 112.7 ohm.  0.5% or better,,,,   1/4 watt or higher.

Digikey requires 5000pc minimum order. I need one or two each.

Mouser no luck.

Seeking suggestions for sourcing these resistors.

Cheers!

[Gyro]

For repairing something, rather than production use, just use series / parallel combinations of standard values. For instance, 100R in parallel with 2k5 will give you 96.1534R, which is [Edit: within around] 0.1%... assuming that you use 0.1% resistors!

I'll let you work the other one out.

[sneily]

Indeed!
I agree, math will show me the way forward.
Cheers

[edpalmer42]

There's got to be an online calculator for this.

Remember that you can also play the tolerances to get closer.  If you only have 5% tolerance resistors, check through them to find the ones that combine to get you as close as possible to the desired value.

[sleemanj]

3.9 kΩ parallel 7.5 kΩ  parallel 100 Ω  gets you to 96.248


Of course 112.7 is trivial (110 in series with 2.7)

[David Aurora]

There's got to be an online calculator for this.

Remember that you can also play the tolerances to get closer.  If you only have 5% tolerance resistors, check through them to find the ones that combine to get you as close as possible to the desired value.

There is- https://sengpielaudio.com/calculator-parallel.htm (and probably zillions more, but this one has been in my bookmarks for eons)

[EC8010]

There's got to be an online calculator for this.

There is: Look up ResCalc.

http://www.pmillett.com/rescalc.htm

[ledtester]

Recent forum thread:

https://www.eevblog.com/forum/programming/software-for-calculating-best-resistors/

[Andree Henkel]

I made a Maple Worksheet for me,
I combine 0.1% 25ppm E24 or E96 and 1% 100ppm E24

best result for 96.25Ohms:
97.6Ohm // 6.8kOhm; 0.1126% Tol, 26.06ppm; nominal value is 96.22Ohm

112.7Ohms:
113Ohm // 43kOhm; 0.1023% Tol, 25.20ppm; nominal value 112.704Ohm

[sneily]

This FORUM makes my repair life a joy. Thank you all.

[ivo]

I've got this one bookmarked, because it's very good at showing you standard resistor series/parallel combinations for a value you want, and how close they'll get

https://www.qsl.net/in3otd/parallr.html

[Overspeed]

Hello

resistor with ''special '' value need to considerate : the tolerance of the resistor as IF you really need a specific value the best way is to sort them with an accurate Ohmmeter or built a assembly serie / paralell to measure what is the real result .

Regards
OS

[coppercone2]

you can buy 5k quantity and put it on ebay. For a KH it might be worth the money to get it right

or just get long lead time one and wait too

I get irked and sometimes buy components in large quantity to fix things as original if their cheap enough instead of building wiccan resistor sculptures. but those resistor look expensive @ 70 cents each too much. if it was 7 cents at quantity for 250$ it might be worth buying 

I often wonder if you can micro tig weld or braze SMT resistors into TH form and make a good resistor that way too.


The most interesting process for metrology would be electroplate welding, that is... you form copper around parts until they are 'welded' together via slow electrochemistry. it might be a way to attach a lead to a resistor without heating it too much and have it reliable.


they are causing so many problems because of component cost misering, it must be possible to add manufacturing quality leads to whatever you want. I think EP-welding might be the bet if someone can figure out how exactly. I think you just need a good fixture

I imagine a flow cell, perhaps out of wax built around a fixtured component and lead with low current and filtering. that might make a precise enough joint. I think you just set them up on a taper like welding and fill it up then sand it down

[ArdWar]

I often wonder if you can micro tig weld or braze SMT resistors into TH form and make a good resistor that way too.

Nah, negative. The binding between the metalized terminals and ceramic substrate usually aren't that strong. A little tug is often all it need to peel out and detach. You need something else to relieve potential stress. Either encapsulating it or putting it in carrier board.

[ch_scr]

I often wonder if you can micro tig weld or braze SMT resistors into TH form and make a good resistor that way too.

Nah, negative. The binding between the metalized terminals and ceramic substrate usually aren't that strong. A little tug is often all it need to peel out and detach. You need something else to relieve potential stress. Either encapsulating it or putting it in carrier board.

For "flat" SMD (0805, 1206, ...) I agree, but I think the MELF family parts have the ends quite firmly attached. Would just solder them though.

[coppercone2]

soldering them sucks because its not durable at all compared to a TH part.

Some of them have end caps and I think it might work. The ones that are "bare" on the sides won't (ultra flimsy). I avoid the SMT that don't have metalization like a full endcap, thats liable to fail... capacitors usually have it. resistors I am not sure

[Doctorandus_P]

KiCad has a built in calculator for this in it's "Calculator Tools":



Only silly thing is that it calculates in kilo Ohm, so it's 1000x off in this case.
Also, Parallel resistors are convenient in SMT. You can just solder resistors on top of each other.

[sneily]

I was able to get "close enough" resistor pairs to complete the repair of this attenuator board on my Krohn-Hite 4200. It's working well.

I managed 96.5 ohms for the 96.25 replacement, and 112.4 ohms for the 112.7 replacement.

As these resistors are on the Attenuator board, they affect the amplitude of the sine wave and not the frequency.
And since this Oscillator has a variable output range for output voltage and can easily adjust this amplitude dial to achieve my desired rms voltage.

So the precision of the replacement resistors are not so critical to me after all.

I enjoyed playing with the various online and downloadable calculators suggested in this thread.