Thick film or Printed Carbon Resistors for the Hobbyist

#1

Hello,

I am attempting to prototype a device that has very restrictive cost and space constraints. To this end, I’ve identified a suitable angle sensing technique, but am having a hard time sourcing it as a hobbyist.

Taking apart several potentiometers has revealed that some of them contain a separate wafer of some sort of phenolic material that has a carbon resistive track printed onto it. This thin wafer sits on top of the usual pcb material that many potentiometers use as a body, and looks like the following:
image
Please note that the black track is physically separate from the material underneath it.

Some research has turned up two companies that offer to print custom carbon resistive tracks:
https://www.piher.net/sensors/index.php?option=com_content&view=article&id=59&Itemid=13
http://www.panda-pcb.com/html/Products/Thick-Film-Resistors-PCB.html

Unfortunately it seems that as an individual that is not associated with a company or business entity, there is no way to order anything from these companies in small quantities.

Does anyone have any ideas of where to source something like this? I would like to avoid having to take apart many more potentiometers in the future, especially since only some are manufactured this way.

Thank you.

#2

I worked in the precision potentiometer industry in the 70s and back then there were dozens of manufacturers that would be glad to sell you standard, semi-custom and full custom resistive track elements in 100 piece volumes. As far as I know all of the companies fell on hard times due to potentiometers being the fail point for so many systems because of their limited rotational life. Only the large companies that specialized in larger volume lower cost potentiometers survived the 1980’s intact (e.g. Bourns, Spectrol, Phier). They along with Honeywell bought up all of the other companies that offered low volume production. To survive the industry ended up becoming exclusively traditional electronic component suppliers with standard product lines and 1000+ piece minimums for semi-custom and full custom orders. I suspect if you could search long enough globally you could find a manufacturer who would do full custom for less quantity. However you’d likely get sticker shock on the price. Even in the 70’s a 100 piece order of a semi custom low to mid accuracy decent life track element had a per part price of $25.00 or more.

Given the life span issues of potentiometers used as sensors (every tiny movement counts against a relatively short total life) nearly all sensor applications moved to digital encoders (optical, magnetic, or mechanical) or linear hall effect for the sensing technology. For ultra-miniature size, a raw die linear hall sensor with a very tiny neodymium magnet would be smaller than any resistive track system and have theoretically infinite rotational life due to being non-contact. The ordinary small standard component surface mount hall sensor packages should be able to compete with resistive elements for ordinary small space applications.

5 Likes
#3

Thank you for such a prompt and thorough answer, you’ve given me a very positive first impression of these forums. While I will likely end up choosing a hall effect sensor as you have suggested, I was wondering if your expertise might lend you some insight on a few things that have come up during my research.

I never actually saw a price for those carbon tracks, but several factors suggested that they may have become less expensive and more durable over time. Most game system controllers seem to contain a similar mass produced joystick module for the thumb sticks, shown below.


The green parts are small potentiometers, consisting of a carbon track printed onto a pcb encapsulated in a plastic body. These modules as a whole are available for less than a dollar in one-off quantities (from aliexpress), which provides a decent price reference for the individual components they are made of. In terms of durability, I would assume that they must have an appreciable life span in order to keep up with the average gamer.

Unfortunately these particular potentiometers all have small sweep angles of roughly 60° while I require 180°, and they are still simply too large. Efforts to find a supplier to order from directly have been unsuccessful. This no longer matters though, because hall effect sensors provide a satisfactory alternative, but now my academic curiosity is piqued.

The thread below suggests that these particular potentiometers are likely some custom part, specific to this purpose, however that does not explain how they can be made so cheaply.


Is this a simple case of mass production for the gaming industry driving costs down? Or is the industry making a comeback?

#4

The lifespan of potentiometers has been sufficient for use as consumer grade human interface devices since the 1920s. Large production volumes, short consumer life expectancy, and the very low accuracy required for that application has kept potentiomenters a very cost effective solution for many high (nearly infinite) resolution human interface devices. (many volume controls now use encoders instead of pots because the low resolution and logarithmic curve requirements makes them cheaper than pots for that application)

While it is a modern phenomenon to have very low cost multi axis joystick pots, other very low cost potentiometers have been around since at least the 60s when low cost transistor radios came on the market (less than $0.10 in medium quantities).

The overall lifespan of potentiometer elements has not significantly improved since the late 1970s to early 80s. It was during that time frame that potentiometer manufacturers hit up hard against the limits imposed by the laws of physics. This a is why nearly all sensor applications for potentiometers have been replaced with other technologies. A sensing application will often have 1000 or more times the movement during its expected life than a human interface application will have.

2 Likes
#5

What an interesting situation! It makes much more sense now, thank you for your time.

#6

Just realized that a few months ago I discovered some items I’d saved from my time in the 70’s working on precision pots for NEI (New England Instrument Co. now a division of Honeywell) and put them on display under one of my monitors.
Left to right:

  • Cover from an 78FTS-233 precision pot.
  • A molded flat track element (originally a Raytheon product line that NEI acquired). This particular element has a very non-linear first half of rotation and then the rest was linear, I think it went in a missile or military plane.
  • Headless horseman sculpture given to me by the amatuer artist in my department. The black hooves are painted with resistofilm. (back then one of the large component makers was putting ads in EE magazines that featured sculptures made from components) Resitofilm was NEI’s patented spray coating for thin substrates, e.g. mylar & Kapton, that allowed very high precision potentiometers in virtually any size, shape, or output function.

1 Like
#7

Sorry for the delay, classes got in the way towards the end of the semester.

It’s always fun to see the collection of mementos one accumulates throughout a rich and rewarding career; I thank you very much for sharing yours here.

#8

Well, progress has been made, and as usual complications have arisen.

The device that I am designing already contains several fairly powerful neodymium magnets, and despite my best efforts to arrange their magnetic fields perpendicular to the hall effect sensor’s axis of detection I was still finding a significant variation between trials. This alone does not bother me, because the sensor doesn’t have to be very precise, only small. However, I’m interested in exploring another avenue that recently came up: capacitive angle sensors. I’m not talking about the readymade sensors that can be found on this website, but something closer to this:

In this particular instance, it serves me well to have a maximum rotation angle of 180 degrees, because that greatly simplifies design. It is possible that this approach may be even more cost effective than a comparable hall effect and magnet pair.

Of course, doing so just shifts complexity from the hardware to the software, but I’m allowed to do that as a programmer; Engineers attempting to do the same, however, are not tolerated :wink:

At this point, I’m only really bumping this thread in order to give subsequent visitors more options and information.