No Joy from the #27800 2-Axis Stick
Morris
Posts: 10
I was overjoyed when I saw the #27800 2-Axis joystick for sale at my local Radio Shack, and immediately put it to use in an audio project that I had been working on for some time. Two #27800 joysticks later, I'm now pretty bummed out. After a few hours of use, they seem to just completely wear out and it's quite disheartening. I put in a trouble ticket with Parallax's support desk, but it was deleted without any response. Note to Parallax reps reading this - do not ever, ever do this. It's highly infuriating. If you don't have an answer, just write back and say so. Anyway, as I can't get in touch with Parallax directly, I feel the need to express my issues with the #27800 publically. I've noticed that people on this forum use the #27800 for lots of things. There was even a guy using it as a controller for a custom electric wheelchair - I now wonder how many times a joystick failure stranded his user. It's just unbelievable to me that a joystick would repeatedly quit after a couple hours, but that's what I've seen, and I wonder if anyone else has ever had this issue or what I might be doing to cause it.
The symptoms, basically, are that the pot 'wiper's start becoming intermittent when the stick is moved, and eventually fail to make contact at all when the stick is moved from the center position. This is an audio effect device where the stick might be moved with the 'beat', but this is no more aggressive a use of the stick than a typical video game controller.
I should point out that this is an analog application, not digital - while the #27800 may not be specifically designed for control of analog audio, there's no reason that this should be impossible. The circuit I'm using is not high-powered or outside the 90 mW power-dissipation bounds of the #27800 - it's a TL072 op-amp which cannot output much power at all. And while analog audio circuits are more sensitive to intermittent contacts than some digital applications, in the end, when the pot wiper goes 'open' that would cause a digital circuit to fail just as an analog one would.
Both my #27800's were purchased retail at Radio Shack. Perhaps they were mistreated in some way, or were a factory second batch? I don't know. There are no 'batch codes' on these packages as far as I can tell.
I just want to learn SOMETHING about this. Meanwhile I've sunk my own time and effort into creating and etching a PCB layout that it looks like I'll never get to use. At least, I'd like some sort of answer from the Parallax guys.
The symptoms, basically, are that the pot 'wiper's start becoming intermittent when the stick is moved, and eventually fail to make contact at all when the stick is moved from the center position. This is an audio effect device where the stick might be moved with the 'beat', but this is no more aggressive a use of the stick than a typical video game controller.
I should point out that this is an analog application, not digital - while the #27800 may not be specifically designed for control of analog audio, there's no reason that this should be impossible. The circuit I'm using is not high-powered or outside the 90 mW power-dissipation bounds of the #27800 - it's a TL072 op-amp which cannot output much power at all. And while analog audio circuits are more sensitive to intermittent contacts than some digital applications, in the end, when the pot wiper goes 'open' that would cause a digital circuit to fail just as an analog one would.
Both my #27800's were purchased retail at Radio Shack. Perhaps they were mistreated in some way, or were a factory second batch? I don't know. There are no 'batch codes' on these packages as far as I can tell.
I just want to learn SOMETHING about this. Meanwhile I've sunk my own time and effort into creating and etching a PCB layout that it looks like I'll never get to use. At least, I'd like some sort of answer from the Parallax guys.
Comments
Can you post a schematic ?
Bean
Well, it's a relief that I'm not the only one who has ever had this problem, and that there might be a solution to it. I had imagined the TL072 to be far too wimpy to toast a pot, but I may be wrong. And the L/R pot does seem to have failed both times, so perhaps this is the issue. With this pot set towards the '0' end, the current flowing through it will only be limited by the TL072 itself - roughly 10-20 mA peak.
There is no maximum current specified on the #27800 packages from Radio shack, or on the online listing. There is a maximum dissipation of 90 mW mentioned - at the low-end of the resistance scale, the I^2*R will be negligible, I'll never hit that limit with the TL072 driving it. Obviously there must BE a maximum current, but I have no idea what it might be. Is there any unofficial rough guideline there? I could put, for example, a series resistance along with the LR leg. But how big does that have to be? This will affect my design, as it limits the range of resistance that I can have. If you have a rough figure for maximum wiper current, it would be very helpful.
C.W.
Sufficient to clip the TL072s output to full swing (figure roughly 7V P-P)
http://www.parallax.com/Portals/0/Downloads/docs/prod/sens/27800-2-AxisJoystick-v1.2.pdf
That rating is for "normal" operation where the pots are used as voltage dividers and a minimal current flows through the wiper.
C.W.
Is that end not connected to anything?
While the op amp is +8V single sided supply, is input swinging peak-to-peak in a range over 10 volts, say -7 to +7?
You may want to rethink how your circuit operates and what you are trying to accomplish.
At a minimum you do need some series resistance in the L/R path, otherwise at one extreme of travel you are bacially shorting the output of the TL072 to ground via the 0.1 uF cap.
This is why the L/R pot is getting toasted.
The RCTime examples from parallax use a 220 ohm resistor to the wiper with a warning not to exceed 5V.
Since your signal should average to 4V you should be within a safe range if you put a 220 ohm resistor in series with the wiper of the L/R pot, I would probably go to something more like 330 or 470 if that will let you achieve your desired output.
C.W.
Nope. Maybe 3V p-p at absolute maximum. You can't overdrive the inputs of a TL072 anyway, they can get really weird.
I agree with the good advice here, that I screwed up my design by making no provision to limit the current through the L/R pot other than the reactance of the 0.1 uF cap and the built-in ill-defined current limiting of the TL072. It was an oversight and has probably been the main cause of my issues with the #27800.
Certainly that's what I'm going to have to do. I noted the diagram in the attached PDF that Loopy posted and did notice this - 220 ohms on a 5V circuit, going up to 300 or more would be hopefully safe for this one.
I do appreciate everybody's advice and patience with me. I'll keep the thread updated once I make this modification and test the device.
Morris,
Welcome to the forums. Let me first start by assuring you that no person is deleting your support tickets. I was recently made aware of some missing tickets and contacted our IT department. It seems there are some server changes happening and that it is possible some tickets were lost due to that. They are contiuing to look into the issue.
As for your issues, other have already chimed in and explained that the problems are related to excessive current in the device. A potentiometer is not intended to attentuate high power signals, for that you would need a rheostat. The potentiometer should only be used in line-level audio signals or for low voltage/current attenuation. On a high-impedence input (such as an op-amp) current would normally be very low.
Thank you very much for looking into it. I appreciate the help from this forum, I expect to have a modified device working soon and will update with my experiences.
And it seems to be working so far - I put a 380 ohm resistor between the LR wiper and ground, and it sounds fine and is not getting the deadly 'crackle' after a bit of use. I'll still have to see if it survives to the rigor of a typical Musical Instrument application - MI stuff has to take a beating - but so far I'm feeling good about it.
I made the tyro error of thinking that an op-amp just COULDN'T toast a pot. Had it been a big full-sized 1/4" pot, it probably could have taken it all day and all night, but this is a cute little micro thing and all that the op-amp's "short circuit protection" means is that the OP-AMP won't get toasted, it makes no promise to the LOAD. I've had a lot of experience with bigger potentiometers, but this sure isn't one of them.
Thanks very much to everyone for their help.
Glad to hear it is working out so far.
C.W.
However, I AM getting those 'cracklies' again - the little noises which, in the past, happened before the stick completely burned up. This time, they occur only near the CENTER position, and they go away after it's been used for a minute or so. Pots are just weird things, and audio band is a much more demanding application than the digital control the stick is really meant for. I'll find out how well it holds up in the end.
Perhaps I am STILL putting in too much current? But I'm starting to think that the #27800 is just not suitable for this application at all. I dream of a joystick just like it, fitting in a 1" square of PCB, but with more sturdy construction.
I'd take a two pronged approach.
A. Start over with a fresh #27800 and a good plan for current limiting to see if you were working with damage you created while learning.
B. Accept the fact that you require a joystick that allows you to add any audio quality pots you desire. It won't be as small, but you will certainly have more opportunity to resolve wear issues .
In general, pots were never really intended to be in constant motion. They do have wear and noise issues. And most good joysticks provide an extra pair for quick switch over if a sudden failure occurs.
Here you are primarily dealing with that 1/10th watt leaving a very tight margin for quality and wear control. Everything gets bigger the more watts it can handle, and more watts means more capable of handling more abuses. Wire wound pots with the right sweeper would be ideal, but they are not tiny.
From what I have read on the web, these little 'Thumb joysticks' were replacements for XBox 'thumb joysticks'. Parallax primarily provides items that are intended to be low cost for experimental and educational use. Most of the customers are trying out new things and are expense adverse.
Since the proof of concept is achieved, even with a noisy pot; it is time to seek out a better quality alternative. The resistor portion of these are some sort of plastic. It may never wear as well as you wish.
German and Swiss manufacturers make a huge variety of extreme precision joy sticks. And I am sure there are some Japanese and American makers as well.
You likely want something with a Military specification rather than a device made for a Microsoft product.
http://www.blbrasil.com.br/representadas/ultra/joysticks.html