Where to get 50k Ohm resistor?

[Ocelot]

Hey guys,

I'm trying to find some 50 kOhm resistors that aren't ginormous or outragously expensive... My usual stomping grounds, digikey, came up short.

Anyone know of where I can get some similar to the ones in this picture?



Thanks,
Brian

[futmacl]

50k is not a standard preferred value in any tolerance. You will have more luck with 49.9k, say here.

You also almost certainly don't actually need to match the resistance that precisely - what's the intended use? You may be fine with just 47k.

PS. If the bands on that resistor in your photo are green-blue-orange-gold, then it's actually 56k +/- 2.8k. 50k would be green-black-orange-(something).

/mz

[Fr0stAngel]

even if do not find 50k resistors , you could connect 2 or more resistors in series or parallel to reach your desired value...
resistances add up in series, and inverse of resistance adds up in parallel....!

[Ocelot]

50k is not a standard preferred value in any tolerance. You will have more luck with 49.9k, say here.

You also almost certainly don't actually need to match the resistance that precisely - what's the intended use? You may be fine with just 47k.

PS. If the bands on that resistor in your photo are green-blue-orange-gold, then it's actually 56k +/- 2.8k. 50k would be green-black-orange-(something).

/mz

Ah good catch! The website I'm pulling that circuit from said they were 50k ohm resistors.

Here's the intended use.

I'm not exactly sure why she chose those exact numbers.

Anyone know whats going on in that circuit? Its an unfamiliar configuration to me.

Thanks for your help!

[Graynomad]

God only knows where 50k came from, sounds like the result of some calculation. In the unlikely event that you find it is required then put two 100k in parallel, otherwise just use 47k or 51k.

The Rs look like they're being used as the top half of a voltage divider, the bottom half being the stuff on the fingers, if that's the case it's even less likely you need exactly 50k.

The caps may be to limit noise.

______
Rob

[Ocelot]

God only knows where 50k came from, sounds like the result of some calculation. In the unlikely event that you find it is required then put two 100k in parallel, otherwise just use 47k or 51k.

The Rs look like they're being used as the top half of a voltage divider, the bottom half being the stuff on the fingers, if that's the case it's even less likely you need exactly 50k.

The caps may be to limit noise.

______
Rob

I believe the cap is being used as a low-pass filter and the resistor as a pull-up resistor. I get the cap value but not the resistor value... I usually use 1k ohm for my pull up resistors... not sure about the 50k...

Anyone have an insight?

[Graynomad]

The circuit though shows the signals going into analogue inputs and you don't pull up an analogue input, that's why I figured the resistors are half of a voltage divider.

Pressure sensors on your fingertips intended for use by children and their piano teachers to visualize the difference between "p" piano (soft) and "f" forte (hard).

This further implies that we're dealing with analogue signals from the glove and therefore a voltage divider.

As for the caps, no idea apart from what I said before.

______
Rob

[Soeren]

Hi,

The caps may be to limit noise.

I believe the cap is being used as a low-pass filter and the resistor as a pull-up resistor. I get the cap value but not the resistor value... I usually use 1k ohm for my pull up resistors... not sure about the 50k...

How do you get one but not the other?  (They're dependant on each other).

Judging from the time constant of the filter, it's for integrating the potentiometer signal, to be able to capture a signal reasonable close to what it's supposed to be. Obviously it was made by a person lacking a bit in skills, but will probably serve the purpose it's intended for in some way.

What do you intend to use it for?

Not that I'm opposed to knitting and crotchet, but I'd prefer a rubber glove with eg. piezo sensors or SFR's for such a task. (And I sure wouldn't want to play a piano in a pair of knitted gloves )

[Ocelot]

The circuit though shows the signals going into analogue inputs and you don't pull up an analogue input, that's why I figured the resistors are half of a voltage divider.

Pressure sensors on your fingertips intended for use by children and their piano teachers to visualize the difference between "p" piano (soft) and "f" forte (hard).

This further implies that we're dealing with analogue signals from the glove and therefore a voltage divider.

As for the caps, no idea apart from what I said before.

______
Rob

Good point on the pull-up on an analog input. Forgot about that.

I intend on using the same setup to be able to tell how hard some one is gripping an object. How would you condition this signal for use with a MCU? The sensors generally have a resistance range <10k ohms.

The advantage of soft circuit sensors is their comfort when integrated into clothing. I'm building a therapy glove (www.advancertechnologies.com) so comfort, robustness and ability to be cleaned is a big influence in choosing soft circuits as opposed to more traditional circuitry. Plus it's reduces cost. 

[Soeren]

Hi,

I intend on using the same setup to be able to tell how hard some one is gripping an object.

As I get it from your page, what you measure is finger position/knuckle angle, rather than gripping strength?

Is there any pressure sensors used and if so, is there one in each section of a finger and how do you assure repeated positioning?

How would you condition this signal for use with a MCU? The sensors generally have a resistance range <10k ohms.

First of all, I'd use a CCG to get a better linearity. Conditioning has to be made with background in the actual circuit and the possible modes of interference and noise and it has to be targeted to how the A/D-C is used.

The advantage of soft circuit sensors is their comfort when integrated into clothing. I'm building a therapy glove (www.advancertechnologies.com) so comfort, robustness and ability to be cleaned is a big influence in choosing soft circuits as opposed to more traditional circuitry. Plus it's reduces cost. 

Well a knitted glove will be more demanding when it comes to cleaning it. Other setups could be cleaned with white spirit, which would be a necessity for something used by different people. Leather gloves (as you have on your page) will quickly be a bacterial mess as well and you cannot clean them without breaking them down, as white spirit, UV light and auto claving will all take the life out of the leather in no time

A conventional glove will only fit a certain range of patients as well.

Cost reducing? How's that?
You'd have to find seriously cheap labor to make a hand sewn arrangement of sensors cheap.

And what about reliability and repeatability?
I think your glove has quite a way to go still, to reach product maturity state.

Connectors - Try small diameter (metal) snap fasteners. They won't slide off that easily.
Even better than tethered use - incorporate a small transmitter a microcontroller and a lithium button cell in the glove.

[Ocelot]

Hi Soeren,

Thanks for taking the time to look at my page. I'm mainly concerned about getting help pertaining to the tactile sensors in my original post rather than feedback on the overall project but I appreciate your feedback nonetheless! =)

As I get it from your page, what you measure is finger position/knuckle angle, rather than gripping strength?

Is there any pressure sensors used and if so, is there one in each section of a finger and how do you assure repeated positioning?

Of course there are tactile/pressure sensors. The glove does not simply measure knuckle angle. It's prime objective is assisted motion.. as in helping a stroke victim move their hand(s). Knowing how hard the user is gripping an object is paramount in this application. The photos on the site are outdated and the prototype is nearing the end of fabrication. In addition to tactile and flex sensors, there are also PMA actuators on the glove. The power and aux systems are contained in a small container worn on the back.

First of all, I'd use a CCG to get a better linearity. Conditioning has to be made with background in the actual circuit and the possible modes of interference and noise and it has to be targeted to how the A/D-C is used.

Whats a CCG? I'm not following you here...

Well a knitted glove will be more demanding when it comes to cleaning it. Other setups could be cleaned with white spirit, which would be a necessity for something used by different people. Leather gloves (as you have on your page) will quickly be a bacterial mess as well and you cannot clean them without breaking them down, as white spirit, UV light and auto claving will all take the life out of the leather in no time

A conventional glove will only fit a certain range of patients as well.

The glove will be machine washable. I don't think thats too demanding nor will it degrade the glove fabric which is actually made of neoprene.

Cost reducing? How's that?
You'd have to find seriously cheap labor to make a hand sewn arrangement of sensors cheap.

I'll ask Nike who they get to do their work

And what about reliability and repeatability?
I think your glove has quite a way to go still, to reach product maturity state.

Of course, its a prototype. I realize its far from robust at this point.

Connectors - Try small diameter (metal) snap fasteners. They won't slide off that easily.

Not sure what you're referring to here.

Even better than tethered use - incorporate a small transmitter a microcontroller and a lithium button cell in the glove.

I've thought of doing this. Will probably be implemented in the next prototype.

[Soeren]

Hi,

Whats a CCG? I'm not following you here...

A Constant Current Generator.
You are creating a voltage divider to read the voltage at the node sensor/resistor.
If you draw the curve for the resistance change ct. output voltage, you'll get a non-linear curve.
Replace the resistor with a CCG and you get a linear voltage output function of sensor resistance.

The glove will be machine washable. I don't think thats too demanding nor will it degrade the glove fabric which is actually made of neoprene.

OK, you be the judge of that. I'm just talking from the results of extremely decreased number of contamination and transferred diseases in Danish hospitals since the inclusion of hand spirit (not sure what it's called in English) dispensers by every sink.

I'll ask Nike who they get to do their work

Oh, I can tell you... They get Chinese kids to do the work.
Even if not looking at the ethics of that, you have to produce enormous amounts to get production costs down.

Not sure what you're referring to here.

Berg pin connectors have a tendency to slide off with movement. Snap fasteners have been used in both home grown electronics and pro equipment for decades (just look at how disposable body electrodes connects). It's a tried and tested method.

[Ocelot]

A Constant Current Generator.
You are creating a voltage divider to read the voltage at the node sensor/resistor.
If you draw the curve for the resistance change ct. output voltage, you'll get a non-linear curve.
Replace the resistor with a CCG and you get a linear voltage output function of sensor resistance.

Do you have a link to where there's an example of something like this? Sounds like its be a good improvement to my flex sensor circuit as well as the tactile sensor circuit.

OK, you be the judge of that. I'm just talking from the results of extremely decreased number of contamination and transferred diseases in Danish hospitals since the inclusion of hand spirit (not sure what it's called in English) dispensers by every sink.

Ah ok. I see what you're getting at. I think you're referring to hand sanitizer. The glove won't need to be that sterile.

Oh, I can tell you... They get Chinese kids to do the work.
Even if not looking at the ethics of that, you have to produce enormous amounts to get production costs down.

Lol yeah that was a joke. The term cheap is kinda relative in this case. Traditio

Berg pin connectors have a tendency to slide off with movement. Snap fasteners have been used in both home grown electronics and pro equipment for decades (just look at how disposable body electrodes connects). It's a tried and tested method.

Ah ok. I see what you mean now. Yeah, I've thought about using some sort of snap fastener and will probably make that improvement in the next version. This is mostly a proof of concept prototype... some thing to build on in future versions.

Thanks for your help!

[Soeren]

Hi,

A Constant Current Generator.
You are creating a voltage divider to read the voltage at the node sensor/resistor.
If you draw the curve for the resistance change ct. output voltage, you'll get a non-linear curve.
Replace the resistor with a CCG and you get a linear voltage output function of sensor resistance.

Do you have a link to where there's an example of something like this? Sounds like its be a good improvement to my flex sensor circuit as well as the tactile sensor circuit.

Not without Googling, but here are a couple of ways to do it.

The left circuit is the most precise, but I think you could easily get by with the slightly simpler/cheaper one.
R2 and R4 are used for setting the current.
All R's can be calculated when supply voltage and resistance range (i.e. min and max) of sensor is known
One capacitor will do fine for your 5 circuits and is only needed if the supply is not decoupled already.

A downside to using this method (at the supply voltage of the controller) is that it cannot supply the full 5V (or whatever), as there will be a drop of around 1V. Using a 10 bit A/D-C this will mean there4 will be around 820 steps left (~205 steps if 8 bit), which should still be ample though.

[Ocelot]

Hi,

A Constant Current Generator.
You are creating a voltage divider to read the voltage at the node sensor/resistor.
If you draw the curve for the resistance change ct. output voltage, you'll get a non-linear curve.
Replace the resistor with a CCG and you get a linear voltage output function of sensor resistance.

Do you have a link to where there's an example of something like this? Sounds like its be a good improvement to my flex sensor circuit as well as the tactile sensor circuit.

Not without Googling, but here are a couple of ways to do it.

The left circuit is the most precise, but I think you could easily get by with the slightly simpler/cheaper one.
R2 and R4 are used for setting the current.
All R's can be calculated when supply voltage and resistance range (i.e. min and max) of sensor is known
One capacitor will do fine for your 5 circuits and is only needed if the supply is not decoupled already.

A downside to using this method (at the supply voltage of the controller) is that it cannot supply the full 5V (or whatever), as there will be a drop of around 1V. Using a 10 bit A/D-C this will mean there4 will be around 820 steps left (~205 steps if 8 bit), which should still be ample though.

Thanks Soeren. What are the equations for calculating the resistor values in the second circuit?

Also I'm assuming I use Ohm's law to calculate the desired current amperage based on the sensors' resistance range and the ADC's voltage range. Right?

[Ocelot]

Hi,

A Constant Current Generator.
You are creating a voltage divider to read the voltage at the node sensor/resistor.
If you draw the curve for the resistance change ct. output voltage, you'll get a non-linear curve.
Replace the resistor with a CCG and you get a linear voltage output function of sensor resistance.

Do you have a link to where there's an example of something like this? Sounds like its be a good improvement to my flex sensor circuit as well as the tactile sensor circuit.

Not without Googling, but here are a couple of ways to do it.

The left circuit is the most precise, but I think you could easily get by with the slightly simpler/cheaper one.
R2 and R4 are used for setting the current.
All R's can be calculated when supply voltage and resistance range (i.e. min and max) of sensor is known
One capacitor will do fine for your 5 circuits and is only needed if the supply is not decoupled already.

A downside to using this method (at the supply voltage of the controller) is that it cannot supply the full 5V (or whatever), as there will be a drop of around 1V. Using a 10 bit A/D-C this will mean there4 will be around 820 steps left (~205 steps if 8 bit), which should still be ample though.

Thanks Soeren. What are the equations for calculating the resistor values in the second circuit?

Also I'm assuming I use Ohm's law to calculate the desired current amperage based on the sensors' resistance range and the ADC's voltage range. Right?

Based on my calculations...

If,
Iout = Vsource/Rmax, where Rmax is the maximum resistance of the sensor

Then
R4 = (Vled - Vbe)/Iout
R5 = (Vsource - Vled)/(Iled+1.5*Iout/hFE), note the 1.5 is sorta arbitrary just to make sure theres adequate current for the transistor specs

Can anyone verify this?

[Soeren]

If,
Iout = Vsource/Rmax, where Rmax is the maximum resistance of the sensor

Have to correct both you and myself (since the voltage drop will depend on the LED).
I_OUT should be (V_SUPPLY-V_LED)/R_MAX

Then
R4 = (Vled - Vbe)/Iout

Yes.

R5 = (Vsource - Vled)/(Iled+1.5*Iout/hFE), note the 1.5 is sorta arbitrary just to make sure theres adequate current for the transistor specs

Forget the h_FE of the transistor and shoot for around 5mA in the LED (regular 20mA type assumed) - too low a current and stability suffers.