**** BEGIN LOGGING AT Fri Jan 06 02:59:57 2006 Jan 06 09:57:23 das beewoolie-afk in the house Jan 06 09:57:24 !!! Jan 06 09:57:40 beewoolie-afk: almost done with this legacy crap and i can get back to doing dev Jan 06 09:58:03 prpplague: coolio Jan 06 09:58:11 beewoolie-afk: thats the bad thing about working for a relatively small company, sometimes you don't get to exclusively work on dev Jan 06 09:58:33 Yeah, but you're in Barbados! Jan 06 09:59:00 beewoolie-afk: not for long, hehe Jan 06 09:59:15 beewoolie-afk: looks like my wife's visa is finally going through Jan 06 09:59:27 What? And leave paradise? Jan 06 09:59:35 beewoolie-afk: might be back in the states as soon as march Jan 06 10:00:01 beewoolie-afk: barbados is a nice place to visit, not so nice to live(especially if you are a geek) Jan 06 10:01:39 Would you mind looking at a web page? Jan 06 10:01:47 http://www.beyondlogic.org/spp/parallel.htm Jan 06 10:01:59 prpplague: I'm trying to understand pin 15. Jan 06 10:02:16 According to everything I've read, it's supposed to be an input. Jan 06 10:02:50 * prpplague checks Jan 06 10:03:29 right for standard lpt interface Jan 06 10:03:49 note the SPP heading Jan 06 10:04:20 Right. Jan 06 10:04:36 I suppose I should check the mode in the bios... Jan 06 10:04:46 bbiam Jan 06 10:11:53 *cough* Jan 06 10:11:56 beewoolie-afk: well? Jan 06 10:12:17 It's in 'output-only' mode Jan 06 10:12:36 I've got a couple of different problems. Jan 06 10:12:48 I want to detect the VCC from the target, just because I'm like that. Jan 06 10:13:02 Trouble is that the line is being driven by the parallel port. Jan 06 10:13:14 The other problem is that the TMS line isn't toggling. Jan 06 10:14:06 It's fluctuating between 2.5 and 2.4 when I toggle the output line. Jan 06 10:15:15 huh? a status line is driven by your parallel port? Jan 06 10:15:24 beewoolie-afk: remember depending on your circuit some lines might need to be pulled up or down Jan 06 10:15:53 beewoolie-afk: when you day you want to detect vcc from the target Jan 06 10:16:00 Hmm. That could be the issue. it could be floating high. Jan 06 10:16:05 beewoolie-afk: you mean you want to know when it is present, or the value? Jan 06 10:16:09 The thing is that this status line is used to detect Vcc. Jan 06 10:16:23 If it doesn't float low, it wont work at all for detecting Vcc. Jan 06 10:16:44 beewoolie-afk: yea Jan 06 10:17:05 This is the xilinx cable. Jan 06 10:17:09 beewoolie-afk: you trying to do a some sort of jtag dongle? Jan 06 10:17:24 I'm not building the circuit myself. Jan 06 10:17:55 According to the schematic, there is a diode to prevent current flow from the parallel port to the target on this line. Jan 06 10:18:46 what does the nError bit read when you have a target connected? Jan 06 10:18:48 beewoolie-afk: you got a url for the schematic, i'd like to have alook Jan 06 10:18:50 with power on? Jan 06 10:18:56 http://toolbox.xilinx.com/docsan/xilinx4/data/docs/pac/appendixb.html Jan 06 10:19:35 I got the schematic from ... http://toolbox.xilinx.com/docsan/3_1i/pdf/docs/jtg/jtg.pdf Jan 06 10:20:26 beewoolie-afk: probably the easiest thing to do is write a small app to monitor the input lines and just do some tests Jan 06 10:21:00 beewoolie-afk: but from the schematic and i would assume that that the system would interpret a float on 15 as a low value Jan 06 10:21:03 prpplague: That's what I'm doing. The !Error line is always high. Jan 06 10:21:20 vmaster: Power on to the target? Jan 06 10:21:23 beewoolie-afk: even when you add voltage to it? Jan 06 10:21:34 beewoolie-afk: add voltage to pin 15? Jan 06 10:21:36 vmaster: The nError bit is always high. Jan 06 10:21:43 strange... Jan 06 10:21:59 interesting Jan 06 10:22:11 the target is 3.3V, when the cable is detached from the parallel port, I get about 3V at pin 15. Jan 06 10:22:15 * prpplague checks his box Jan 06 10:22:16 do you have bit 5 set on the control port? Enable Bit-Directional? Jan 06 10:22:56 vmaster: That was the next thing I was going to check. Jan 06 10:23:03 my parallel port behaves really weird with that bit set Jan 06 10:23:06 beewoolie-afk: might also want to check you dmesg to see which driver was being loaded Jan 06 10:23:13 I haven't seen anything about the control bits, yet. Jan 06 10:23:24 prpplague: it is loading the spp. Jan 06 10:23:37 gotcha Jan 06 10:23:56 beewoolie-afk: hehe. i forget i'm talking to someone who actually knows linux, hehe Jan 06 10:24:11 :-) Jan 06 10:24:24 beewoolie-afk: i just checked mine with a parport dongle, my pin 15 is reading low Jan 06 10:24:25 vmaster: Do you have a register level reference for the parallel port? I haven Jan 06 10:24:30 haven't turned one up Jan 06 10:25:46 only what's on beyondlogic Jan 06 10:25:58 beewoolie-afk: wait, are you taking into account that pin 15 is a NOT input? Jan 06 10:27:02 ah, and make sure you removed the lp module - i had all kinds of problems with that loaded, too Jan 06 10:27:08 prpplague: That isn't the problem. If the pin reads 4.5V, it doesn't matter what the Vcc signal does. The pin will always read H. Now, it is possible that the bit is blown. Jan 06 10:27:15 vmaster: No lp module. Jan 06 10:27:25 ppdev? Jan 06 10:28:19 beewoolie-afk: when i apply +5v to pin 15 i get a 0 back Jan 06 10:28:28 vmaster: I'm using ppdev. Jan 06 10:28:28 beewoolie-afk: when i let it float i get a 1 Jan 06 10:28:35 beewoolie-afk: ground i get a 1 Jan 06 10:29:18 ah, but isn't that what you expect from ttl? Jan 06 10:29:21 prpplague: It always reads 1. Jan 06 10:29:59 beewoolie-afk: and you are push 3.3v on to it? Jan 06 10:30:34 beewoolie-afk: try a wire between one of the data pins (2-9) and 15 Jan 06 10:30:48 vmaster: hehe thats was what i was typing Jan 06 10:30:50 that's what macraigor does to detect a real wiggler Jan 06 10:30:54 hehehe Jan 06 10:31:00 Hold on. Jan 06 10:31:16 I can detect that the cable is attached. That's done with a loopback in the db25 housing. Jan 06 10:31:24 i don't see anything on that schematic that would indicate that it would detect a 3.3v Jan 06 10:31:30 What I'm trying to do is detect the Vcc. Jan 06 10:31:46 The top of the schematic is the part I'm looking at. Jan 06 10:31:50 SENSE Jan 06 10:32:01 yeah, but that would show you if your nError works at all Jan 06 10:32:18 vmaster: I don't understand what you mean. Jan 06 10:32:47 beewoolie-afk: you could set ping 2 high and see if you get a corresponding 0 on pin 15 Jan 06 10:33:35 beewoolie-afk: yea i'm looking at the vcc sense, but it doesn't have any 3.3v buffers which would mean it would need +5v to work Jan 06 10:33:39 I don't undersfand that. Jan 06 10:33:57 beewoolie-afk: 1 out on pin 2, wired to pin 15 , you should get a 0 Jan 06 10:34:05 prpplague: Do you mean it has to drive +5 for the parallel port to detect it? Jan 06 10:34:10 beewoolie-afk: yep Jan 06 10:34:10 yes Jan 06 10:34:17 I don't think that can work. Jan 06 10:34:20 Look at TDO. Jan 06 10:34:32 It is driven by one of the 245's buffers. Jan 06 10:34:43 Those devices are powered by the 3.3V of the target. Jan 06 10:34:56 So, the output of TDO is going to be about 3V. Jan 06 10:35:17 beewoolie-afk: but it passes through the 74hc125 Jan 06 10:35:25 beewoolie-afk: the vcc doesn't Jan 06 10:35:28 Oh, I meant that part. Jan 06 10:35:29 Yeah. Jan 06 10:35:38 So the output voltage is going to be the same. Jan 06 10:35:43 Approximately. Jan 06 10:36:15 hmm, i would disagree on the vcc sense Jan 06 10:36:23 How's that? Jan 06 10:36:51 the vcc sense doesn't pass through the 74hc125 there for it is not buffered to 3.3v Jan 06 10:36:59 its still at 5v Jan 06 10:37:05 Huh? Jan 06 10:37:12 The Vcc from the target is 3.3V. Jan 06 10:37:22 It is 3.09V at pin 15. Jan 06 10:37:30 I'm going to measure TDO... Jan 06 10:37:39 beewoolie-afk: which brings up back to the point, you have to have 5v to drive the lpt port Jan 06 10:37:59 prpplague: I don't think so, or this device won't work at all with a 3.3V target. Jan 06 10:38:17 mhh, i guess that's the magic of these buffers - all wigglers work that way Jan 06 10:39:08 beewoolie-afk: the jtag interface would work, but not the vcc sense Jan 06 10:39:30 prpplague: Because...? Jan 06 10:39:50 beewoolie-afk: the vcc sense needs 5v to go low Jan 06 10:40:58 prpplague: Hmm. TDO is about 2V at pin 13 of the db25. Jan 06 10:41:29 So, what you're saying is that the sense line is floating. If I put 5V on it, it will register a zero in the status register. Jan 06 10:41:38 Easy enought to test. Jan 06 10:44:47 prpplague: Actually, we have it backwards. Jan 06 10:44:54 vmaster: It's what I was suspecting. Jan 06 10:45:10 The problem is that the pin 15 line is floating high. Jan 06 10:45:20 It reads as a 1 when nothing is connected. Jan 06 10:45:26 I need to ground it to read a zero. Jan 06 10:45:45 In other words, it won't work to detect Vcc because it always reads as Vcc is present. Jan 06 10:45:45 it's supposed to be inverted Jan 06 10:45:47 Oh, well. Jan 06 10:46:00 i.e. a ground should read 1, +5v should read 0 Jan 06 10:46:01 If I ground pin 15, I read a zero in the register. Jan 06 10:46:16 Doesn't. Jan 06 10:46:51 I'm not really worried, tho. At least I understand why. Jan 06 10:47:07 Perhaps this parallel port implementation is borked. Jan 06 10:47:17 What is more of a problem is that I cannot drive TMS. Jan 06 10:47:35 BTW, the output of this parallel port is 3.3V. Jan 06 10:48:09 heh, take it to a doctor Jan 06 10:48:34 anyway, guess that's why people try to avoid parallel port alltogether Jan 06 10:49:50 It's a laptop afterall. Jan 06 10:50:18 ok, then you're lucky to have a parallel port at all Jan 06 10:50:29 are you able to drive the other pins? tdi and tck? Jan 06 10:50:47 beewoolie-afk: ohhhh, you left out the fact that it was a laptop Jan 06 10:51:03 beewoolie-afk: most laptops don't have a fully compliant lpt Jan 06 10:51:03 Yeah. Jan 06 10:51:18 It looks like one of the resistors on the board may be an open circuit Jan 06 10:51:31 beewoolie-afk: about 2 years i ago a did a series of tests with about 10 different laptops, none of them were 100% Jan 06 10:51:37 I'm not suprised that it isn't compliant Jan 06 10:51:55 Stil, I can drive all of the lines except for the tms Jan 06 10:52:01 I think it's the board, tho. Jan 06 10:53:16 Hmm. It's supposed to be 300Ohm, but I get nothing. Open circuit... Jan 06 10:53:58 beewoolie-afk: which resistor? Jan 06 10:54:15 beewoolie-afk: still you can do a simple test Jan 06 10:54:16 They aren't numbered in the schematic. Jan 06 10:54:19 beewoolie-afk: remove the dongle Jan 06 10:54:30 beewoolie-afk: wire from pin 2 to pin 15 Jan 06 10:54:30 That's what I did, silly. Jan 06 10:54:38 beewoolie-afk: ahh Jan 06 10:54:48 Remember, we're not working on SENSE anymore. Jan 06 10:54:53 beewoolie-afk: oh Jan 06 10:54:56 This is the series resistor for TMS_IN Jan 06 10:55:25 ahh yea that would cause a problem Jan 06 10:55:33 beewoolie-afk: didn't you just get this dongle? Jan 06 10:55:34 I have some 1%'s. I can solder one on and see if it helps. Jan 06 10:55:45 prpplague: It's ka6sox's. Jan 06 10:56:01 He lent it to me last year. Jan 06 10:56:01 beewoolie-afk: ahh Jan 06 10:56:14 I didn't end up using it because I didn't know it was supported by wince jtag. Jan 06 10:56:24 beewoolie-afk: ahh Jan 06 10:56:24 I have another one, but it doesn't use fly leads. Jan 06 10:56:46 So, I could rig up an adapter, or I could see what happens with a 300Ohm resistor... Jan 06 10:57:01 It's kinda wierd, tho. Jan 06 10:57:14 beewoolie-afk: yea the jtag dongle that ships with the lnode80 has 2x7 and 2x10 headers, and .1" spaced header for the fly leads Jan 06 10:57:48 Why would the resistor go like this? Jan 06 10:57:58 beewoolie-afk: hehe, its weird looking back 5 years ago and not having a jtag adapter on my desk Jan 06 10:58:15 beewoolie-afk: usually over voltage Jan 06 10:58:27 ESD? Jan 06 10:58:32 naw Jan 06 10:58:42 never saw esd kill a resistor Jan 06 10:58:44 more like wall-plug Jan 06 10:58:50 hmm. Jan 06 10:59:03 I cannot see how that could have happened Jan 06 10:59:16 not plug, outlet Jan 06 10:59:17 bbiab, going to solder a resistor Jan 06 10:59:45 ahh lpt is soooo much fun Jan 06 10:59:56 not as much as usb though Jan 06 11:01:08 vmaster: i've been experimenting with one of these for jtag - http://people.omnigroup.com/wiml/soft/pic/keyspan.html Jan 06 11:03:28 hmm, you want to bit-bang the ez-usb? Jan 06 11:06:38 vmaster: yea, bit banging the jtag via the ez-usb Jan 06 11:06:51 vmaster: got some pretty good speed compared to parport Jan 06 11:10:28 So, it looks like the resistor was borked. It had cracked in half. Jan 06 11:13:32 beewoolie-afk: doh Jan 06 11:13:33 prpplague: vmaster: looks like that was it Jan 06 11:13:56 Will a resistor break when it overvoltages? Jan 06 11:15:00 I've only seen one or two resistors after they had to take too much current, but those looked "burnt", they didn't break Jan 06 11:16:46 Why don't these jtag dongles support nTRST? There are plenty of parallel port data pins. Argh! Jan 06 11:17:11 you can live without nTRST, but debugging without nSRST isn't much fun Jan 06 11:18:56 vmaster: Sort of. I was reading in the docs for the IXP. There are some tricks you can pull with nTRST that you cannot perform without it. Jan 06 11:19:15 vmaster: Although, to be fair, those tricks require a reset as well as nTRST. Jan 06 11:19:32 beewoolie-afk: the jtag dongle we have with the lnode80 supports nRST and nTRST Jan 06 11:19:58 this is your own design? Jan 06 11:21:11 beewoolie-afk: yea, basically an extended version of the HG dongle Jan 06 11:22:57 Hey, I'm getting data! Jan 06 11:24:00 heheheh Jan 06 11:24:10 beewoolie-afk: isn't it glorious when you get jtag working Jan 06 11:24:42 yipppeee! Jan 06 11:39:35 prpplague: There is definitely something fishy about the parallel port. Jan 06 11:39:45 prpplague: I am unable to drive one of the data lines, now. Jan 06 11:40:06 prpplague: The line that is used to detect whether or not the dongle is attached is now always 3.3V. Jan 06 11:41:13 beewoolie-afk: hehe, sounds pretty typical for laptop lpt Jan 06 11:41:28 beewoolie-afk: they aren't meant to drive much other than a standard printer Jan 06 11:42:33 You'd think it would be a simple enough thing to get working, tho. Jan 06 11:44:16 beewoolie-afk: hehe, you'd think Jan 06 11:45:26 prpplague: OOPS. I was wrong. it still works. I forgot to toggle that bit. Jan 06 11:54:45 heheh Jan 06 12:19:35 prpplague: I was looking over that schematic again because I realized that all we need is a pull-down on the VCC SENSE pin. Jan 06 12:20:01 prpplague: Seems like the schematic, as is, has the 1K pulldown on the wrong side of the diode. Jan 06 12:20:18 prpplague: But, then perhaps I don't understand what that pulldown is supposed to do. Jan 06 12:21:27 prpplague: With a 1K pulldown, it's gonna sink about 3ma from the target. that seems odd. Jan 06 12:21:35 vmaster: You still around? Jan 06 12:22:03 vmaster: I was wondering if you had an explanation for the 1K pulldown on the xlinix cable schematic. On the VCC sense line. Jan 06 12:27:16 sorry, no Jan 06 13:04:08 vmaster: So, do you think that the design may be flawed? Seems unlikely, but then it may work fine on PC parallel ports. Jan 06 13:04:25 re Jan 06 13:05:15 lennert: hi lennert Jan 06 13:05:21 beewoolie-afk: hullo Jan 06 13:05:23 I got the xilinx cable working. Jan 06 13:05:29 One of the resistors cracked. Jan 06 13:05:32 which one, the parallel cable 3? Jan 06 13:05:36 Yep. Jan 06 13:05:39 ah Jan 06 13:05:45 it works quite nicely here Jan 06 13:05:58 With a cracked resistor? Jan 06 13:06:02 when i got it working under linux i got rid of my windows box Jan 06 13:06:12 i don't think i have any cracked resistors on mine :) Jan 06 13:06:22 There is a wierdness, tho. Jan 06 13:06:27 I cannot use the Vcc sense line. Jan 06 13:06:44 how do you mean Jan 06 13:06:45 It looks like the problem is that the status input line on the serial port is floating high. Jan 06 13:06:45 ? Jan 06 13:06:56 serial port? Jan 06 13:07:05 Parallel :-) Jan 06 13:07:19 hum Jan 06 13:07:26 I was wondering if anyone could explain the schematic. Jan 06 13:07:34 what's the link again? Jan 06 13:07:36 There is a pull down for that line, but it's behind a diode. Jan 06 13:07:38 ... Jan 06 13:08:00 http://toolbox.xilinx.com/docsan/xilinx4/data/docs/pac/appendixb.html Jan 06 13:08:14 pin 15? Jan 06 13:09:57 Yeah. Jan 06 13:10:30 The line floats to about 4.5V from the LAPTOP. I'm using a laptop. Jan 06 13:11:00 you're measuring pin 15 on the laptop? Jan 06 13:11:07 Yeah. Jan 06 13:11:08 (without the cable connected?) Jan 06 13:11:12 Right. Jan 06 13:11:41 what gnd are you using? the connector edge? Jan 06 13:11:41 Well, maybe I didn't test pin 15 that way. Jan 06 13:12:08 What connector edge? Do you mean the shell of the db25? Jan 06 13:12:13 yeah Jan 06 13:12:36 Yeah. The sheild ground. Jan 06 13:12:43 shield. :-( Jan 06 13:13:24 Due to the diode, there isn't anything else that could be powering up pin 15. Jan 06 13:13:42 When the target is powered down. Jan 06 13:14:03 I've measured one side of the diode at 0 and the parallel port side at 4V. Jan 06 13:14:13 Target disconnected and powered down. Jan 06 13:15:10 but you said you had the cable disconnected? Jan 06 13:15:20 or do you have the parallel cable connected but not plugged in to the target? Jan 06 13:15:54 right. I don't think I've measured pin 15 at the laptop... let me do that. Jan 06 13:16:32 3.3v Jan 06 13:20:56 so the cable is pulling it up, hm? Jan 06 13:21:27 odd, there's no power coming from the parallel port (if i look at the schematic) Jan 06 13:21:41 (the cable has no external power supply, right?) Jan 06 13:21:57 Right. Wrong. Jan 06 13:22:01 Wait... Jan 06 13:23:03 OK. Jan 06 13:23:10 The target is disconnected. Jan 06 13:23:18 I am getting 4.7on pin 15. Jan 06 13:23:29 At the right side of the diode, I get 2.5V. Jan 06 13:24:15 They don't say which ground is used for the chips. Jan 06 13:24:29 I believe it is the target ground which is floating. Jan 06 13:25:03 the ground is connected to jtag pin 2 Jan 06 13:25:14 as well as parallel pin 20/25 Jan 06 13:25:19 this schematic is kind of unclear Jan 06 13:25:28 I don't think so. Jan 06 13:25:41 I've checked for continuity and there isn't any. Jan 06 13:25:49 The grounds are separate. Jan 06 13:26:13 the schematic is wrong then Jan 06 13:26:31 it clearly shows both parallel pin 20/25 and jtag pin 2 connected to the same ground Jan 06 13:26:44 you can only guess what ground U1, U2 and the capacitors are using Jan 06 13:26:46 Well, the PDF I'm using has a different symbol for the shield ground from the target ground. Jan 06 13:26:52 Let me look at that web page again... Jan 06 13:26:55 there is the shield ground, yeah Jan 06 13:27:04 but parallel pin 20/25 are connected to the 'regular' ground Jan 06 13:27:43 Ok. I see where it's wrong. Jan 06 13:27:56 On my cable, 20, 25, and the shield are connected. Jan 06 13:28:10 They don't connect to the target ground. Jan 06 13:28:17 the schematic is definitely wrong. Jan 06 13:29:18 lennert: OK. I'm wrong again. Jan 06 13:29:28 odd Jan 06 13:29:33 i wonder what they'd connect the shield to Jan 06 13:29:43 The schematic is correct. They bring both grounds to the board, but I don't know what the shield is for. Jan 06 13:29:54 Let's go back, then. Jan 06 13:30:09 Pin 15 is still floating high. Jan 06 13:31:10 Let me see what happens when I toggle prog Jan 06 13:32:17 All that does is prevent TDO from being read. Jan 06 13:34:10 lennert: Well, I just don't get it. When I disconnect the target and read TDO, I get 1's which means that pin 2 of U2 is high. Jan 06 13:35:29 Pin3 of U1 should be high impedence when the nOE is high. Jan 06 13:40:16 i think my cable is much simpler than this one Jan 06 13:40:29 :-) Jan 06 13:40:46 Turns out that this isn't critical. Jan 06 13:40:55 sorry Jan 06 13:41:03 my other half just keeps blathering Jan 06 13:41:10 we just came back from visiting granny Jan 06 13:41:13 NP. I was just wondering if someone who knows more than I do had an idea. Jan 06 13:41:16 i can't concentrate at all Jan 06 13:41:17 Nice. Jan 06 13:41:22 How's GMa Jan 06 13:41:24 five more minutes and i'll throw her out of the window Jan 06 13:41:29 Gma? Jan 06 13:41:51 no, other half :) Jan 06 13:42:04 she says "okay, i'll leave you alone now" because she knows that that'll get her my attention Jan 06 13:42:15 :-) Jan 06 13:42:15 and then she starts yapping about totally unrelated stuff Jan 06 13:42:22 and then five minutes later she says that again Jan 06 13:42:29 urgh Jan 06 13:42:47 and now she walked away angry Jan 06 13:42:48 oh well Jan 06 13:43:01 Is this grandma, or the other half? Jan 06 13:43:08 other half, still :) Jan 06 13:43:11 grandma lives elsewher Jan 06 13:43:12 Ah. Jan 06 13:43:34 we visited grandma for russian orthodox christmas Jan 06 13:43:46 we'll be partying again soon for russian orthodox new year Jan 06 13:44:04 but other half doesn't understand that after many hours of social nonsense, i want some time to work/study :) Jan 06 13:47:17 21:34 < beewoolie-afk> lennert: Well, I just don't get it. When I disconnect Jan 06 13:47:17 the target and read TDO, I get 1's which means that pin Jan 06 13:47:17 2 of U2 is high. Jan 06 13:47:35 Hmm? Jan 06 13:47:50 that's what you wrote Jan 06 13:48:03 Right. Jan 06 13:48:06 but pin 2 of u2 is an input? Jan 06 13:48:25 so just means that u1 isn't pulling it down Jan 06 13:48:30 (u1 ping 3) Jan 06 13:48:32 pin Jan 06 13:48:58 Hang on... Jan 06 13:49:07 i think i just don't understand what it is that you are trying to say :-( Jan 06 13:50:09 i have the digilent parallel jtag cable Jan 06 13:50:15 from what i can tell it's really really simple Jan 06 13:50:33 I have one of those too, but it doesn't have fly leads. Jan 06 13:50:37 indeed Jan 06 13:50:43 What I was trying to say is this. Jan 06 13:51:08 I measure the voltage at the right side of the diode closest to pin 15. Jan 06 13:51:12 I get 2.5 V. Jan 06 13:51:18 okay Jan 06 13:51:26 so jtag pin 1 (Vcc) is 2.5V .. Jan 06 13:51:41 But the target is disconnected. Jan 06 13:51:53 well Jan 06 13:52:04 U1 and U2 are likely CMOS Jan 06 13:52:14 CMOS components can work even if Vcc isn't connected Jan 06 13:52:24 it can be that one of the data lines is powering them up Jan 06 13:52:27 Where do they get power? Jan 06 13:52:36 Ah. That makes sense. Jan 06 13:52:38 since they'll have EMP diodes Jan 06 13:52:48 or what's that called in english Jan 06 13:52:57 I don't know EMP. Jan 06 13:52:59 in fact, cmos components without vcc connected can work just fine Jan 06 13:53:10 Hmm. That's new to me. Jan 06 13:53:11 static discharge protection Jan 06 13:53:20 ESD. Jan 06 13:53:21 diode from the data line to Vcc Jan 06 13:53:23 right Jan 06 13:53:28 I see. Jan 06 13:53:32 that's likely to be powering Vcc up Jan 06 13:53:43 OK. So, the other question is this. Jan 06 13:53:46 yes Jan 06 13:53:51 Why is the pulldown on the right side of the diode? Jan 06 13:54:07 Wouldn't it make more sense to put it on the other side? Jan 06 13:54:14 Moreover, why is it so small? Jan 06 13:54:38 the 1K? Jan 06 13:54:39 hmmm Jan 06 13:55:30 Yeah. Jan 06 13:56:57 actually, i don't know Jan 06 13:57:02 it doesn't make much sense to me Jan 06 13:57:14 they're just putting the resistor straight across their supply Jan 06 13:57:34 would make more sense if it was a cap, but they already have one on the right Jan 06 13:58:01 let me ask someone Jan 06 13:58:06 maybe it is to drain the cap Jan 06 13:58:55 could be, but i'd use a bigger one Jan 06 13:59:25 (and the buffers could even have enough internal resistance themselves) Jan 06 13:59:42 That's what I'm saying. Moreover, it would work just as well on the other side of the diode. Jan 06 14:00:51 people i asked also think it's to discharge the 10nF cap Jan 06 14:04:43 beewoolie-afk: http://www.artofelectronics.com/quotes/quotel.htm Jan 06 14:07:51 Cool. Jan 06 14:08:12 still somewhat puzzled about the 1k resistor Jan 06 14:09:24 i don't see why they put the 100pF caps on TDI/TMS tbh Jan 06 14:09:51 Imedence matching? Jan 06 14:10:16 That, or it makes the signaling smoother. Jan 06 14:11:26 i assume the 100ohm series resistors are in case someone accidentally hooks up an input as output Jan 06 14:12:26 I think that's also for impedence matching. IIRC, I read something about that somewhere. I don't know much about that, tho. Jan 06 14:12:44 The IO lines to the target are supposed to be 100Ohm Jan 06 14:15:00 neither here Jan 06 14:15:07 impedance is a vague area to me still :-/ Jan 06 14:35:59 beewoolie-afk: ping Jan 06 14:36:20 beewoolie-afk: having the resistor on the left would prevent a full discharge of the cap Jan 06 14:36:37 beewoolie-afk: due to the 'voltage barrier' in the diode Jan 06 14:36:52 what's that called in english.. 'drempelspanning' Jan 06 14:37:00 barrier voltage? Jan 06 14:39:50 "ON voltage" Jan 06 14:39:53 whatever Jan 06 15:39:32 lennert: I wondered about that. Jan 06 15:40:12 lennert: it's called the band-gap, or the ... I I forget. Jan 06 15:41:36 lennert: Thanks for the chat on this. It helps me remember all of those things I once learned. Jan 06 15:42:45 no idea what it's called Jan 06 15:42:55 np, i like looking at stuff like this Jan 06 15:43:03 like taking stuff apart and such Jan 06 15:44:20 Gotta go to a meeting... Jan 06 15:44:22 ttfn Jan 06 18:22:18 hello everyone.. up late, are we? Jan 06 18:25:01 not that late Jan 06 18:25:26 not yet! Jan 06 18:26:19 too late for him! Jan 06 18:26:23 yeah ;) Jan 06 20:15:17 * [g2] hugs lennert (has it quieted down there yet ? ) Jan 06 20:15:34 it's very quiet here now Jan 06 20:15:38 why are you hugging me? :) Jan 06 20:15:58 4:15am and only the sound of my crunching brain Jan 06 20:15:59 <[g2]> seemed like you needed a hug Jan 06 20:16:06 <[g2]> :) Jan 06 20:16:07 hehe Jan 06 20:16:14 or did _you_ need one? :) Jan 06 20:16:33 <[g2]> maybe I needed one and just _stole_ it Jan 06 20:16:45 yeah Jan 06 20:16:50 you _pirated_ it Jan 06 20:16:58 <[g2]> so did you get a little "caving time" ? Jan 06 20:17:00 <[g2]> heh Jan 06 20:17:05 i.e. you took it away from me and now i no longer have it! Jan 06 20:17:13 <[g2]> heh Jan 06 20:17:19 since 11pm or so i've been studying Jan 06 20:17:25 after celebrating russian orthodox christmas Jan 06 20:17:31 <[g2]> I need to pirate your JTAG cable :) Jan 06 20:17:35 and gulping a few bottles of wine Jan 06 20:17:53 however, i got sober again rather quickly, which is kind of disappointing Jan 06 20:18:09 [g2]: i have the digilent parallel cable? Jan 06 20:18:10 <[g2]> ahh... young livers Jan 06 20:18:20 <[g2]> I've got a couple Jan 06 20:18:27 a few more days and there'll be russian orthodox new year.. now that's fun Jan 06 20:18:30 <[g2]> and flying leads to go with them Jan 06 20:18:38 <[g2]> happy new year Jan 06 20:18:46 <[g2]> orthodox new year :) Jan 06 20:18:52 i don't have flying leads yet Jan 06 20:19:01 i'd need a jtag connector - flying leads cable Jan 06 20:19:12 i guess i can just cut one open :-/ Jan 06 20:19:20 or get the digilent one Jan 06 20:19:25 and pay triple the amount in shipping.. heh Jan 06 20:19:26 <[g2]> digilent inc sells leads for their 6 pin connectors Jan 06 20:19:49 <[g2]> the sell 2 pin connectors too Jan 06 20:20:52 i just need to work on my s0ldering sk1lls Jan 06 20:21:33 when i get back home i'm getting a DMM Jan 06 20:21:46 then.. an iron, probably Jan 06 20:22:32 (a soldering iron, not a regular iron :) Jan 06 20:23:02 <[g2]> yeah my better half calls that the welding iron :) Jan 06 20:23:10 uhm Jan 06 20:23:20 the welding iron.. okay then :) Jan 06 20:23:29 <[g2]> are you "welding again?" Jan 06 20:23:52 <[g2]> no honey, i need a torch for that :) Jan 06 20:24:20 hehe Jan 06 20:24:23 some people :) Jan 06 20:27:13 beewoolie-afk! Jan 06 20:27:17 <[g2]> wb Jan 06 20:27:22 you're never near your keyboard, are you? Jan 06 20:27:39 actually, i was lying, it was 05:15 here Jan 06 20:27:39 lennert: yo man Jan 06 20:27:49 04:15 is what my irc client says, which is in a different time zone Jan 06 20:28:04 beewoolie-ntk: d00d! Jan 06 20:28:06 <[g2]> beewoolie is near the DMM and soldering iron Jan 06 20:28:16 lucky him Jan 06 20:28:26 * lennert want my DMM :-/ Jan 06 20:28:30 [g2]: Hey. Looks like the jtag dongle that ka6s0x sent me was broken. I had to replace one of the resistors. Jan 06 20:28:46 * beewoolie-ntk really likes his 179 Jan 06 20:28:59 beewoolie-ntk: a 179 it will be for me Jan 06 20:29:11 beewoolie-ntk: thou hath convinced me Jan 06 20:29:21 <[g2]> beewoolie-ntk yeah I read the log Jan 06 20:29:36 <[g2]> did you get the PIN 15 sorted out ? Jan 06 20:29:39 Soon, I'll be writing flash. Jan 06 20:29:54 [g2]: I think that it isn't solvable with the laptop's parallel port. Jan 06 20:30:08 I could solder another pull-down resistor, but that would be going too far. Jan 06 20:30:20 <[g2]> ah... moved to a desktop port ? Jan 06 20:30:37 No. The only desktop machine I can use is too inconvenient Jan 06 20:30:44 fyi, my laptop handles the digilent jtag cable fine Jan 06 20:30:53 I'll be moving to USB when I get the board assembled. Jan 06 20:30:58 (dell inspiron 8500) Jan 06 20:31:07 lennert: I don't have a diligent...or do I? Jan 06 20:31:16 [g2]: Is the cable you sent me a diligent? Jan 06 20:31:16 dunno Jan 06 20:31:18 * [g2] is wondering if the JTAG lines are no pulled high/low on the Loft Jan 06 20:31:28 The spec is that there are pull-ups. Jan 06 20:31:31 <[g2]> beewoolie-ntk yes I sent you the digilent cable Jan 06 20:31:35 Spec according to IEEE on jtag Jan 06 20:31:46 mine says 'Digilent' on the parallel connector end of it Jan 06 20:31:52 OK. This cable works, it's just that I cannot detect Vcc. Jan 06 20:32:05 BTW, the openwince jtag software doesn't look at that kind of stuff. Jan 06 20:32:45 lennert: In other words, I can use the diligent or this xilinx cable just fine. Some of the features aren't available, but no one cares but me. Jan 06 20:32:45 <[g2]> right but openwince isn't detecting the chip Jan 06 20:32:54 [g2]: What chip? Jan 06 20:33:01 <[g2]> IXP425 Jan 06 20:33:04 It is working now that I fixed the resistor. Jan 06 20:33:15 Oh. I haven't plugged into the slug. Jan 06 20:33:26 My plan is to make it possible to really understand what is going on. Jan 06 20:33:28 <[g2]> it'll probably work Jan 06 20:33:37 Openwince doesn't give much info on what it is doing Jan 06 20:33:43 <[g2]> beewoolie-ntk good plan ! Jan 06 20:33:57 I think that I can make something easier to use in the short term. Jan 06 20:34:03 Just flash programming to start with. Jan 06 20:35:21 So, are you saying that you cannot get the diligent to work with the loft? Jan 06 20:36:27 beewoolie-ntk: i don't think my cable tries to detect vcc Jan 06 20:36:53 beewoolie-ntk: the times i used openwince i wanted to rewrite it.. Jan 06 20:37:34 <[g2]> beewoolie-ntk yes, but I haven't tried very hard yet Jan 06 20:37:57 <[g2]> I've got a working Gateworks cable and sw Jan 06 20:38:09 <[g2]> so I can reflash redboot with that Jan 06 20:38:15 <[g2]> and I have Jan 06 20:38:40 <[g2]> I've actually got the cable apart right now and I'm looking at it Jan 06 20:38:49 <[g2]> I could just write sw to the cable Jan 06 20:38:57 <[g2]> meaning the cables pins Jan 06 20:40:43 beewoolie-ntk: http://direct.xilinx.com/bvdocs/userguides/ug130.pdf Jan 06 20:40:50 beewoolie-ntk: your cable is the one on page 62 Jan 06 20:40:55 <[g2]> is the pin numbering on a 25 pin cable 1-13 14-25 ? Jan 06 20:41:26 <[g2]> or 1 3 5 7 9 ... 25 on the top and 2 4 6 8 ... 24 on the bottom ? Jan 06 20:42:03 note how the digilent cable doesn't have parallel resistors on the caps Jan 06 20:42:27 it does have the 100pFs on the outputs Jan 06 20:43:17 [g2]: http://www.geocities.com/nozomsite/parallel.htm Jan 06 20:44:41 [g2]: The pins are numbered on the DB25 if you look closely beside the pins. Jan 06 20:45:09 lennert: It's loading Jan 06 20:46:06 lennert: which page has the schematic you're looking at? Jan 06 20:46:11 p62 Jan 06 20:46:49 <[g2]> lennert thx Jan 06 20:47:02 lennert: That's a lot like the wiggler Jan 06 20:47:12 <[g2]> FWIW there are resisters coming from the TARGET JTAG pins Jan 06 20:47:43 buffers Jan 06 20:47:48 oh Jan 06 20:47:55 i thought you wrote reGisters Jan 06 20:48:11 that's probably to deal with accidental reversal of drive direction Jan 06 20:48:17 i.e. treating an input as output Jan 06 20:48:21 not to blow stuff up Jan 06 20:48:46 lennert: Not clear why the dlc5 has resistors on the inputs, on the parallel side. Jan 06 20:48:49 <[g2]> there are little chips further down the line Jan 06 20:49:30 dlc5? Jan 06 20:49:38 lennert: that's the cable that I have. Jan 06 20:49:54 I'll be up and back. Got to make dinner for dollface. Jan 06 20:49:56 <[g2]> z17B ? Jan 06 20:50:00 what dlc5 do you mean? Jan 06 20:50:02 okay Jan 06 20:51:50 06:00 AM here.. time to go flat Jan 06 20:52:12 <[g2]> sweet power sleep :) Jan 06 20:52:35 lennert: nitey nite. Jan 06 20:52:40 beewoolie-afk: g'nite Jan 06 20:52:46 lennert: The cable I have is the xilinx cabl III Jan 06 20:52:51 AKA DLC5. Jan 06 20:53:38 <[g2]> hey did you see google is going to be providing Video for sale ? Jan 06 20:53:57 [g2]: I read that, but I haven't followed the story. Jan 06 20:54:00 What sort of video? Jan 06 20:54:15 <[g2]> CBS NBA etc... Jan 06 20:55:01 <[g2]> Hmmmm maybe I'll start buying all my video can kill the cable Jan 06 20:55:20 <[g2]> except I _NEED_ that for the cable modem :) Jan 06 20:59:00 And your cable speed is awesome. Jan 06 20:59:18 What does it cost for cable modem broadband without also buying cable TV service? Jan 06 21:00:13 <[g2]> like an extra 20 bucks : Jan 06 21:00:29 <[g2]> the cable actually is pretty awesome Jan 06 21:00:42 So, then how much would it be for your speed? Jan 06 21:01:06 <[g2]> I think 65 versus 45 Jan 06 21:01:12 Hmm. Jan 06 21:01:28 But that has no IP addresses, right? You cannot run a server. Jan 06 21:01:38 <[g2]> but I pull 5Mbs Jan 06 21:01:47 <[g2]> and burst near 10 Jan 06 21:01:48 Yeah, but servers! The servers! Jan 06 21:01:56 <[g2]> dude I've got a T1 Jan 06 21:02:15 <[g2]> that's the server ticket but it's expensive Jan 06 21:02:23 Download, but not upload. Or am I wrong? Jan 06 21:02:54 <[g2]> The T1 is 1.5 down/up Jan 06 21:03:08 <[g2]> The cable modem is 5-6Mbs down and 378K up Jan 06 21:37:03 Ah. How much does the T1 cost, eh? Jan 06 21:45:41 <[g2]> _way_ too much :( Jan 06 21:53:49 last time I priced a T-1, it was somewhere around $600 / month running 100% data Jan 06 21:56:20 ByronT: Wow. That's a lot, I got two T1's in Seattle for about $300 each, per month. Jan 06 21:56:39 who's the data service through? Jan 06 21:56:55 BellSouth charges an arm and a leg - and naming rights on first born Jan 06 22:38:42 ByronT: There's a local provider that is giving me point-to-point server. They're called Blarg. I like them because they're professional. They know how to run an ISP. I've never had downtime. I mean never. I've had them for three or four years. Even speakeasy has had downtime. Jan 06 22:41:16 your choices here are limited to BellSouth and whomever is re-selling BellSouth services Jan 06 22:42:10 fortunately, BellSouth is not too bad - 6MB DSL is being rolled out - I've been promised end of April worse case Jan 06 22:52:13 ByronT: We've got that,too, but I haven't signed up since it doesn't increase the upload speed and that's what I care most about. Jan 06 22:52:28 Nice chatting... **** ENDING LOGGING AT Sat Jan 07 02:59:58 2006