Tweet
Originally posted by WM6
View Post
Tahnks again
-
Hi WM6 !, ok I'll de-solder the pin that goes from r9 to the 1st pin of uC
Tahnks again -
Chance-Coil
How many uH have your coil and how many Ohm´s?Comment
-
your problem IS VERY CLEAR. do not use any type of 9V battery in all PI detectors!!!!!!Originally posted by solmer2 View Post
they have VERY small capacity and do not able to give a current in charge the PI needs off.
Comment
-
Originally posted by solmer2 View Post
Chance should work (somehow) even whit 6 AA cells, but use 12V supply first. If problem persist, try this "defectation" using just DMM:
First measure positive voltage at pin 7 of U5 (OP37) and negative at pin 4, should be somewhat lower, volt or so, than supply voltage. If negative voltage is absent, close to 0 volts, check circuit around U1, part orientation etc, if necessary replace 7660.Also check surrounding parts around Q4 (inverter disable). Without -V power stage will not work either. Also, if you have slightly lower voltages present at both pins, remove U5 and try again (maybe U5 is damaged, then check orientation and integrity of protecting diodes D7&D8 and replace chip.
If this is ok, check +5V (obviously good, MCU working) at pin8 U6, and also reference at pin 1 U6, should be about 4.1V. If not, check parts around U3, TL431.If this is OK, problem is probably in the power stage whit MOSFETS. Measure current consumption whit and without coil, should be around 100ma whit coil, and, say, around 20 without. If it differs slightly, too high without coil, or too low when coil is connected, check all parts in power stage, around Q 1,2,3,5,6,9,D1, 4,5,6, orientation, resistor values etc. Will be helpful to have scope at this point. Good luck.Comment
-
-
Hi all
been looking at the coil construction which seems to be the one recommended by the author, surely the coil would need some potting to keep the wires stable in the housing
to stop falseing.
Comment
-
An exception to the rule in post # 18:
Comment
-
Hi all.
My Chance PI have same problem. I can change parameters and LCD is working ok. When coil is connected speaker will make constant sound.
Sound will stop only if coil is disconnected, menu is active or volume level is set to zero. Device will not react any kind of metal.
I did de-solder R9 but it did not help --> same sound again.
Battery: 13.5 V car battery!
Coil: R=0.6ohm L=0.427mH
Tests with digital multimeter: U5_PIN7=+12.47V, U5_PIN4=-12.20V, U6_PIN1=+4.6V and U6_PIN8=+5V
I think TL431 is ok (Uref=+4.6V). Is it?Comment
-
4.6V reference is OK (this is actually correct value), measure current consumption whit and without coil and check power stage. If current is in normal limit, or even somewhat higher then "upper" part around Q3 is most probably OK, this is normal "flyback" part, check "lower" part around Q9 (purpose of it is to "short circuit" the coil between pulses in one part of cycle.) If not working correctly, AD will see flyback when it not supposed to be and MCU will then miscalculate everything, producing constant tone and max. signal indication. Remove D6 and check behavior and consumption, if remains the same, this is the case. Then check parts and orientation of D6, Q9,Q5,Q6,D5 and surrounding resistors.Comment
-
current consumption with coil: 127.6mA
current consumption without coil: 34.5mA
Is this in normal limit?
then I removed D6 HER208
current consumption with coil: 67mA (same constant sound and same behavior)
current consumption without coil: 34.0mAComment
-
Actually, i don't know. My remarks about D6 and detector behavior apply only to old software version, new release is radically and completely different in timing, waveforms and all. Just for curiosity, i get my old and partially cannibalized board from junkbox, put it together and updated to new 1.0.1 software release, without any hardware modifications (i will try mods to see if anything changed, but i don't believe anything significant will), and it works. Note that i don't have at hand 7660, so -V is also supplied from PSU, and this is what i measured: 72mA whit coil, 35 without, 48 w. coil and D6 removed (it still detect metal), also 35 w/o coil. (this may vary due to display type, backlight, audio volume and speaker or headphone type etc). Also consumption will be somewhat higher whit 7660 on board, but not much, appears that it is somewhat high in your detector.
My best recommendation is to get access to oscilloscope and check everything step by step, suspect something in power stage.
And, yes,!$@#%!, in meantime, since my last visit to the author's page, there are another 3 software versions available again! I will check them, add all mods, and post if it interesting. In this one, appears that about any piece of metal is placed somewhere on VDI scale no matter what Iron mask setting is.
BTW what software version is in your detector?Comment
-
Small update on this, version 1.2.1 is up and running. Just some small change in timing, consumption is now 92mA, and it works, well, detects metal. I first have to replace 7660 and try this somewhere else, it is just too noisy here for any relevant testing.
I think Silverdog should start selling programmer PCBs whit every Chance kit, if we want to follow this rapid software changes and be updated. Or, perhaps, this may be just author's strategy to undermine any commercial exploitation of his project?Comment
-
Originally posted by Tepco View Post
It just means the author of the project is refining his work, which is great. I would rather follow a FREE evolving project than a dead one because there's always room for improvement.
I agree with your comment about providing a programmer. Some can be found cheap, $12 at ebay. I wouldn't recommend the cheap usBASP clones though.Comment
-
The circuit below is what I used to program the ATmega8 using ponyprog. Too easy, most people would have these parts in their junk boxes waiting for use.
http://www.geotech1.com/forums/attac...1&d=1315628885
MickComment
-
Hi,
Firmware is 1.2.1.
I did this simple test for mosfet's (D6 D4 R5 R1 removed)
"connect the meter negative to the MOSFET's source.
Hold the MOSFET by the case or the tab if you wish, it doesn't matter if you touch the metal body but be careful not to touch the leads until you need to.
First touch the meter positive on to the gate (a).
Now move the positive meter probe to the drain. You should get a low reading (b). The MOSFET's gate capacitance has been charged up by the meter and the device is turned on.
With the meter positive still connected to the drain, touch a finger between source and gate c) (and drain if you wish, it matters not). The gate will be discharged through your finger and the meter reading should go high, indicating a non-conducting device. "
Q9 TEST a) 2.57V b) 0.603V c) goes to 2.56V
Q3 TEST a) 2.57 b) 0.546V c) 0.5V <--- conducting, so maybe it's broken.
I really need to get oscilloscope, is this good enough for fault finding (just for this Change PI case)? http://xoscope.sourceforge.net/Comment
Comment