-
#125 Reply
Posted by
radioman
on 19 May, 2013 16:15
-
Oh... well, data coming out of the PIC not programmer, i was thinking that data was out from programmer. This is very strange indeed. 15pF cap on the line makes me think about the line impedance too high and a small cap value matter. How the supply voltage is applied to the PIC? maybe a decoupling cap on the PIC voltage rail is needed?
Strange, very strange to tell.
-
#126 Reply
Posted by
PA0PBZ
on 19 May, 2013 16:27
-
The PIC ID is coming from the PIC of course
I tried all kinds of decoupling and pull up and down stuff, but that did not make any difference. However, I think I captured the problem part, it looks like a bad clock that appears just before the PIC spits out the ID, see attachment. A bit weird, because all the other clock edges look much better.
-
#127 Reply
Posted by
SeanB
on 19 May, 2013 16:33
-
Probably the extra capacitance delays the signal so the latch internal to the programmer gets a valid data level, most likely it is being latched just as the signal is being changed, and depending on the device, temperature and supply voltage it will read one or the other. The delay slows it down to a time afterwards when it is stable,
Had plenty of fun with this on cards that would nominally be perfectly functional, but just would not work with each other. Change one or the other and it works. Then take to operating temp of 120C and see if it still works, if not try another card. then take to the ATE and let it run through diagnostics for 5 hours and see if it fails, generally not, but it might do so again in a year.
-
#128 Reply
Posted by
radioman
on 19 May, 2013 16:36
-
Looks like a clock edge jitter for me, but what could cause it?
-
#129 Reply
Posted by
radioman
on 19 May, 2013 17:03
-
The clock line is ringing at the rising edge and after a while it stabilizes, but hmm... clock line is driven by the programmer, maybe a series resistor with clock line help?
-
#130 Reply
Posted by
PA0PBZ
on 19 May, 2013 17:05
-
Probably the extra capacitance delays the signal so the latch internal to the programmer gets a valid data level, most likely it is being latched just as the signal is being changed, and depending on the device, temperature and supply voltage it will read one or the other.
The clock goes to the PIC, it looks like the PIC sees an extra clock cycle and starts to spit out the data bits one clock early (01111100 -> 00111110)
I also don't see why the clock is only 4 Volt, this pic is a 5 Volt device so that makes it only worse.
-
#131 Reply
Posted by
PA0PBZ
on 19 May, 2013 17:11
-
The clock line is ringing at the rising edge and after a while it stabilizes, but hmm... clock line is driven by the programmer, maybe a series resistor with clock line help?
This is the clock as it is on the ZIF socket, I can't measure on the PIC because that makes the problem go away. What worries me is that this programmer should do its work without the extra stuff, if I can't trust it I don't want to use it. I'm using the cable that I got with my (chinese clone) PICkit 3, it is not exeptionally long. Compared do what I see in the pictures of the TL866A it is even shorter, although it looks like the wires are thicker in the original.
-
#132 Reply
Posted by
radioman
on 19 May, 2013 17:20
-
The clock goes to the PIC, it looks like the PIC sees an extra clock cycle and starts to spit out the data bits one clock early (01111100 -> 00111110)
I also don't see why the clock is only 4 Volt, this pic is a 5 Volt device so that makes it only worse.
Oh! that make sense, i think that programmer icsp voltage is only 3.3v an this could cause this issue.
This is the clock as it is on the ZIF socket, I can't measure on the PIC because that makes the problem go away. What worries me is that this programmer should do its work without the extra stuff, if I can't trust it I don't want to use it. I'm using the cable that I got with my (chinese clone) PICkit 3, it is not exeptionally long. Compared do what I see in the pictures of the TL866A it is even shorter, although it looks like the wires are thicker in the original.
Well, cheap and bad design, i'm sure they not tested it well for every case...
-
#133 Reply
Posted by
ju1ce
on 28 May, 2013 04:06
-
Is this device able to reset AVR fuses using high voltage programming?
-
#134 Reply
Posted by
c4757p
on 28 May, 2013 04:10
-
Yes.
-
#135 Reply
Posted by
ju1ce
on 28 May, 2013 15:07
-
-
#136 Reply
Posted by
Eight8
on 31 May, 2013 23:43
-
Awesome work Radioman and all the worked on decrypting the firmware.
I shall have to do some reading.
-
#137 Reply
Posted by
abzman
on 12 Jun, 2013 10:56
-
-
#138 Reply
Posted by
radioman
on 13 Jun, 2013 21:36
-
abzman, i read the whole article and i like it, a true hacker! glad to see you happy at the end.
-
#139 Reply
Posted by
digsys
on 13 Jun, 2013 22:58
-
My A version arrived yesterday. If there's anything you want checked, let me know.
I only bought it as a backup of my Galep IV, who take forever to fix bugs and reneged on making driver software available. bastids
-
#140 Reply
Posted by
abzman
on 14 Jun, 2013 01:48
-
So, here's a thought if this project becomes more community oriented. The MC3406 needs a resistor divider set on the comparator inverting input, originally I thought to use an i2c controlled pot, but the i2c lines are taken up, maybe bit banging a spi digital pot like the ad8402 to control the voltages in a much finer way than the 3bit control we have now. This may require other changes than just stripping off the resistors and transistors that are currently on the inputs and replacing them with this digital pot, but I don't know exactly what. This would of course require the designing of a custom firmware and probably a new GUI, but if we eventually get our own chip database going then we might want a wider range of voltages available and this seemed like a simple (hardware) mod that would allow that.
Just a thought, I'm not sure I'm up to anything as radical as re-writing the PIC's firmware, but I wanted to put it out there.
-
#141 Reply
Posted by
mladen82
on 18 Jun, 2013 11:45
-
Did anyone notice that even no memory chip at programmer, software reading is ''succesfuly''
anyway works good for now
-
#142 Reply
Posted by
sparkyuiop
on 20 Jun, 2013 10:07
-
For all experiencing GUI issues try this:
1. Insert your Windows Xp installation CD.
2. Go to Control Panel->Regional...
Thanks allot Radioman, you are a superstar! Minipro software now displays correctly without resizing the window and that only improved it, not cure it.
-
#143 Reply
Posted by
Prema
on 23 Jun, 2013 20:41
-
I have a few SOIC8 which I am trying to re-program. Since this is an ongoing project and I'll have to re-flash the chips a lot of times, I would like to re-flash in-curcuit.
While a college had no trouble with another programmer the TL866 voltage-protection seams to prevent it on any of the chips:
Any ideas how to disable or work around that protection, or getting it done any other way with the TL866 are more than welcome!
THX in advance.
-
#144 Reply
Posted by
radioman
on 25 Jun, 2013 19:49
-
This is overcurrent protection not voltage-protection. How do you connect the chip to the programmer? The overcurrent protection cannot be disabled because is working at the firmware level.
-
#145 Reply
Posted by
c4757p
on 25 Jun, 2013 19:55
-
Pop it open and check for shorts.
-
#146 Reply
Posted by
Prema
on 26 Jun, 2013 06:33
-
It's connected via a testing clip. The clip works perfect with the TL866 for all chips except in-circuit. I have tried 4 different boards and I get that message with all of them. Even shortened the cable to like 10cm...may have to build an external socket for the project as there are no 150mil sockets except the giant ones that we use in programmers. It's just annoying that I need to do all that work while another programmer just works on the same boards in-circuit, maybe that current protection is just too sensitive...just wish we could manually disable it.
-
#147 Reply
Posted by
digsys
on 26 Jun, 2013 06:49
-
Is the board to program powered up? or are you supplying Vcc? If you're supplying Vcc, then maybe it's reading current draw of the board?
-
#148 Reply
Posted by
radioman
on 26 Jun, 2013 06:57
-
I think that board is drawing too much, if is exceeding 100mA then the overcurrent protection will be triggered in about 100uS.
-
#149 Reply
Posted by
Prema
on 26 Jun, 2013 06:59
-
Boards are not powered. Only source is the TL866. Yeah maybe the boards draw too much and I should use another power source?!?