The 4 batteries are in series. You correctly identified the ball switch. I need to get a magnifying glass and a hell of alot more light to be able to get the chip ids on the two chips you identified. Bifocals suck! I will take apart the stem to see if I can get an id on the hall sensor. I will have time tomorrow night to work on this.
Take Care.
After a couple of false starts, I was finally able to make some current draw measurements.
When you initially activate the handle, the current draw is in the 40-50mA range. After it is activated, the current draw seems to stabilize at 25-29mA (with or without the LEDs flashing or ON) until the sensor handle turns itself off after about 30 sec. When the handle is in the “off” state the current draw is not detectable (<1mA)
This seems to confirm your impressions.
I still haven’t had a chance to disassemble a post to get to the hall sensor but I do have a broken handle in hand.
The magnetic strip use is about 1in wide and .25in thick. THe magnetic field must be quite strong as I wa unable to get the handle to trigger with any of the magnets I had at home. Hard to imagine errant magnetic field causing the false triggers we are seeing in the field.
Ok. I was finally able to read the chip ID you identiified.
The one left middle that you have circled is a TC7660.5. It looks like it is a voltage convertor.
The chip in the top left is a MCP41100.
I was surprised to see the current draw drop to “undetect” after the handle resets and before it is activated.
Our experience is that the battery packs are left plugged in in the battery packs die quicky, within a couple of weeks.
Take Care.
I wa unable to determine the type of hall switch used in the handle it look like it is epoxied partway into the threaded rod. Three wires but that it a common configuration for these sensors. I doubt it is a latching version though as there are no instructions to make sure one the the strips is North side up and the other north side down. As I understand it, unlatching requires the opposite magnetic field to unlatch.
Ian.
Well that's ... ummm ... perplexing. I can't even put forth a good guess as to what function they serve. It makes me think there's an important part of the "big picture" that I'm missing. I'll guess the TC7660 is there to make a negative voltage, but for what use ? Perhaps the output of the Hall sensors isn't a binary one, but rather an analog voltage and there's an op-amp circuit I've not seen (needing a negative supply) ? Perhaps the digital pot sets it's (variable) gain ? Or a tunable threshold ?ijourneaux:
The one left middle that you have circled is a TC7660.5. It looks like it is a voltage convertor.The chip in the top left is a MCP41100.
Up to now I’d have said that designing a newer, simpler circuit … that runs off 5v, or even 3.3v, thereby saving battery life … would be easy. Now I’m unsure.
Given the 1000 mAh rating, you might expect 40 hrs (1000/25) of continuous run-time. And actually quite a bit less as that 1000 mAh rating is only for a 0.2 mA draw, not a 25 mA one. Still at 30secs per use, that seems like a lot of activations due to bouncing in transport. :think:ijourneaux:
I was surprised to see the current draw drop to “undetect” after the handle resets and before it is activated. Our experience is that the battery packs are left plugged in in the battery packs die quicky, within a couple of weeks.
How many activations (average/stone) would you expect in a tournament ?
Once the 30 secs has elapsed, do you need to tip the stone over to get power back on ? Just gripping the handle doesn’t do it ?
Is there any common element to the false alarms ? Are they always a fault indication or ? Do they happen on the ice or off ? Do they occur around the 30 sec mark ? I can see external RF fooling the capacitive grip sensor but not the Hall. And Hall sensors are directional, not any magnetic field will do. :?
ijourneaux:
… as there are no instructions to make sure one the the strips is North side up and the other north side down.
Are there any instructions as to which side is up ? If not that may indicate that the Hall output is a field strength, an analog signal.
Any chance the PIC coding is available ?
I am going to take the circuit board into work and have someone look at it under a better microscope to confirm the Chip ids. I think I got it right but want to confirm.
Each handle/stone is activated about 10x per game and about 10 games per tournament. The stone gets reactivated by fliping the stone over to clean the bottom. Touching the handle is not sufficient. It may have been designed that way to provide a clear trigger that the stone was about to be thrown. Thinking about it briefly, I can’t see why you couldn’t just use touch with appropriate logic.
There is no indication of which side is up on the magnetic strips. I agree it is probably field strength.
The false "faults occur then the handle is clearly released and the handle starts outputing a solid red (as opposed to a solid green or blinking red/green (low battery).
Unfortunately the PIC code is not available.
There are no field adjustments to the sensors (i.e. sensitivity). Our handles are going on 10 years old. Perhaps age is coming into play? This is an expensive proposition. Each handle costs in the high 3 digits new. We have 6 set of 16 handles with about 16 additional spares.
The handle is a commercially avaialable lexan handle that has been machined so the batteries fit, wit holes for the LEDs. The hand grip has a durable conductive coating and the underside of the handle is metalized and is connected to the ground on the circuit board.
The battery connector (Harwin style 2pin) sucks and is a ridiculously frequent failure point. There on positive connection clips on the isde of the female connector that are easilt broken if special care isn’t taken. I have atleast 10 handles in the shop where the connector on the circuit board is mangled in some way to make it non-functional. Or course the circuit board is not available as a separate item.
I had someone verify the chip numbers again. Looks like I was close enough to get the chip purpose.
TC76608
E0A0308
(symbol)OVJ
MCP41100
I/SN0318
(symbol)2TJ
So they occur on ice, in the course of play and clearly before the line crossed ? Even way way before ? I wonder if there's a sensitivity problem, that it's picking up the hogline magnet inches or a foot away ? In any case that's got to be bad for play and instilling distrust in the system.The false faults occur then the handle is clearly released and the handle starts outputing a solid red (as opposed to a solid green or blinking red/green (low battery).
Tell you what, if I have a few spare hours this weekend, I’ll draft up a preliminary design for a testbed/prototype based on an Arduino to start with. Coding should be 1 hour for a HS kid, shouldn’t take me more than 3. :mrgreen: But before that perhaps a set of requirements and goals needs to be drafted. Here are my thoughts below.
Detect presence of hand on handle and signal it
Improve battery life
add ‘hard’ on/off switch (keep use of tip switch)
reduce power consumption
possibly new battery type
improve ease of battery replacement
Eliminate false alarms
Reduce cost, initial and maintenance
Continue use of stock handle
may be remachined for new circuitry & batteries
no difference in look and use in play
Add ability to reprogram in circuit
Add some diagnostics
some form of built-in test (BIT) ?
battery level indication
some link to PC or smartphone ?
I thought I had posted a reply last week but obviously didn’t. See comments in your list of requirements.
Detect hogline violations or declare no foul and signal it
Detect presence of hand on handle and signal it
After stone is activated (by flipping over), the leds are off is the hand is touching the handle and rapidly blink green when the hand is not touching.If the stone is released before the hog line, the leds blink green rapidly until the stone crosses the hog line where the LEDs turn a solid green.
Improve battery life - These are really the key.
add ‘hard’ on/off switch (keep use of tip switch)
reduce power consumption
possibly new battery type
improve ease of battery replacement
Eliminate false alarms. I am still waiting on more clarification on the false alarms. It is possible that they are observing the low battery red/green blinking. If the battery is low and the handle is activated (increasing the current draw) is low power indicated?
Reduce cost, initial and maintenance - I have toyed with using flush mounted coin battery holders that would become part of the battery holder “snowflake” so batteries could just be snapped into place. Also looking at a more robust connector between the battery pack and the circuit preferably something that would have a 90deg. wires coming in from the side and connector going straight down. I have permission to use a 3D printer to print the snowflakes so that changes in the design requirements would be easy to implement
Continue use of stock handle
may be remachined for new circuitry & batteries
no difference in look and use in play
Add ability to reprogram in circuit
The handles are bolted onto the stones so there should not be a security issue.
Add some diagnostics
some form of built-in test (BIT) ?
battery level indication - YES
some link to PC or smartphone ? I always thought that adding connectivity would increase the power consumption too much to be worth while. I would love to be wrong on this. Connectivity might allow us to automatically time game play. To keep games moving along.each team has 73min to play their half of the game (10 end game). This process uses a time keeper who is dedicated to a specific game. If there was a way to determine that the stone had stopped moving, you may be able to time from the handle being activated to the time the rock comes to rest. This might be of interest for general club use where games tend to take longer than normal because players are ready. Right now to keep play moving along, 25min before the end of regulation time the teams complete the end that are playing and one more end.
I was able to talk with one of the officials that had commented on the false alarms.
When it occurs, the false alarm is always to Red and it occurs well before the hogline. He commented the they replaced the handle on a stone and the problem persisted. They move the stone to another sheet and the problem seemed to go away for a while but did come back. Eventually they had false alarms on all of the sheets. Perhaps the magnetic field that is causing the problem is in the stone itself (granite)
Sorry for the long delay in responding. I’ve yet to come up with a plausible explanation for the FA behavior above. I might have guessed some powerline running under the sheets but that wouldn’t explain why it spread from sheet to sheet as time went on. I might guess some external influence coupled w/dropping battery levels but I don’t know if that even starts to make sense over the course of 1 tournament. Baffling …
FWIW I watched some Olympic curling. It seems they have a similar but different design. I’ll assume their handle operates on a similar magnetic principle. But I note the LEDs, only 2 of them (no doubt combo R/G LEDs), only come on after the hogline is passed. There’s no “grip indication” (a good feature IMO). I’ll guess that they use a different design. Have you looked into what is used as a possible ready-to-go replacement ?
One large design issue is getting a basic grasp on the expected magnetic field strength. Sensors come in a wide variety of sensitivities. There must be some product number or such info for the embedded strip ??
Any chance that the handle that is gripped could be drilled out to house something the diameter of an AA cell ?
I take it back, it appears the Startco “eye on the hog” handle was used at Socci. The video didn’t show the red LEDs and I never saw a violation, so I missed their presence as clearly shown below.
http://blogs.gazette.com/sticksandstone … AAn9IC.jpg
And I didn’t see any green flashing because it’s not there at the time I was (wrongly) expecting. A reread of the manual straightened me out.
http://www.eyeonthehog.com/Eye%20On%20T … Manual.pdf
So an existing alternate source is out.
If you watched Curling closely, you could see the leds go green when the stone crossed the hog line. The green LEDs are the pair for front leds (one on each side of the handle. The red leds are towards the center of the handle. Since the penalty is really severe (stone removed from play), at that level of play if almost never happens. It is more common at natioanl qualifying events (<10 in a event and sometimes much less than 10). I have the magnetic strip info somewhere. WIll post asap
Is the Startco handle based on, or just a modified version of, a Curlex lexan handle ? With a modified washer and some threaded tube and short screw (made of ??) vs the longer bolt normally used w/the Curlex handle ? There’s a curling club not far from here and if the aforementioned is true, I’ll see if I can’t get a used (or new) handle from them. New ones are listed for a meager $36. It would help me see what’s practical or not. Like a CR123 battery (or two) in the “grip”.
Re: the strip …it has a down side and when so mounted I have to believe there’s only 2 practical magnetic field orientations; either north pole up or south pole up. Then it’s a question of field strength.
Yes the startco handle does start with a off the shelf curlex handle. They machine out space for the batteries, mounting surfaces for the circuit board and holes for the leds. The handles in coated with a conductive material as is the under side of the handle.
In a curlex handle, the bolt attaching the handle to the stone is probably 3-4in long. WIth the Startco handlemost of the 3-4in is taken up with an threaded extension that includes the hall sensor. That gets the hall sensor to within 1-1.5in from the ice.
As you said, since the magnetic strip has a rectangular cross section. there are only two practical orientations for the magnetic strip in the ice. I still haven’t located the info on the field strength.
Here is a picture of the underside of the handle. You can see the areas routered out to make room for the batteries. You can also see the holes for the leds and there is alittle more machining to the lexan to provide mounting areas for the circuit board. The thick blue wire goes up to make contact with the handle. The gray area is metalized and is connected to the ground on the circuit board. You can also see the 3 wires going into the stem and up to the hall sensor.
The stem on a regular Curlex handle only extends about 1 inch up from the lexan. then there is a union with male threads on both ends. There is a keyway on both ends of the threaded union with small set screews to prevent the stem from unscrewing. Then a threaded rod which houses the hall sensor. It is epoxied into place so that the sensor is about .75 in from the end of the stem.
I was able to confirm that the magnetic strip has a “this side down” label on the one side. That label is known to be in error on some of the strips so I don’t think it really matters,
Well that's interesting and it indicates that the sensor is bipolar vs unipolar. That makes me wonder why they bothered w/the label ??ijourneaux:
That label is known to be in error on some of the strips so I don’t think it really matters,
In any case I was going to propose the following; that you build up 1 testbed (teststone, testrock ?) from readily available parts to first serve as a measurement tool and a place to test code. The tool will measure the strips magnetic field, strength and spatial extent and polarity(ies). It could also be used to measure the background magnetic environment at different venues. That data can be used to choose the proper sensor or come up w/a signature to distinguish the hogline from background (to reduce FA’s). Secondly actual parts can be rotated into the testbed and verified to work as desired. Then I expect you’d want to make a custom PCB to conform to the interior of the handle, just like the existing e-handle, with all the “proper” components and w/high confidence it’ll work due to the prior testing.
Here’s a list of the major items;
https://www.sparkfun.com/products/12587
It’s small and should run on 2.5 - 5.5v. Perhaps overkill for your task but writing and testing code is simple.
https://www.sparkfun.com/products/12041
I like that this one runs on 1.8 - 5.5v and draw << 1 mA. The board is small too.
http://www.digikey.com/product-detail/e … ND/4552949 (check out datasheet for package sizes)
It runs off 2.8 - 5v And outputs an analog voltage that goes up or down depending on the field polarity. Draws < 9mA (meh) but is cheap $s, small and available. It may not be the optimal sensor or the final choice but for a testbed … seems a good enough choice.
Now it’s my expectation to run the testbed off a single 3.6v Li battery, something like a CR123 (readily available, low $s, 1500 mAh and designed for 10-30 mA draw). It may be that not all of the above will fit internal to a Curlex handle. For the moment that’s OK, it’s only a testbed. I expect some changes btw now and the end, custom, product. It’s not for play. Like an A10, it’s ugly but all business.
If the above idea sounds good to you let me know. I’ll do up the electrical interconnects/wiring/design and detailed parts list (including LEDs, etc). I can do up some code. You’ll have to buy and assemble the testbed. And do the testing and post the data. We can come up w/a good test and development plan and analyze the data.
All leading to a final design, hardware and software*, that you can choose to implement. And on that note you should check into the legalities. Startco/Littelefuse may have a broad patent, that covers the concept of magnetic hogline detection. That may or may not prohibit you from selling your version … or even making your design freely accessible to potential customers of Startco/Littelefuse. Frankly I don’t know and leave that all upon you.
FWIW you’ll need this or another USB-TTL serial converter to program the above Arduino. An FTDI adapter cable costs about the same.
This sounds great. I will order the parts tonight. I agree that the patent might limit what I can do. The original project was a university student project. I believe that patent is owned by the Univ of Saskatewan and licensed to StartCo. I have the librarian where I work doing some government work to locate the patent. I am hoping the players involved will do what is best for the game if we have a better mouse trap.
PS the A10 Warthog was an amazing warplane. Hell of an efficient tank killer. Interesting the pentagon just ditched the A10 in favor of more drones.
I was looking at the 1304. That is a surface mount chip, it might be easier to work with in a SIP through hole package. I am trying to identify an equivalent part.
