Rob
User
Reged: 04/29/08
Posts: 9
|
|
I have an TSOP48 --> DIL48 adapter (+ some others) from another company.
It is wired 1:1, i.e. pin 1..48 on the DIL corresponds to pin 1..48 on the TSOP.
How should I proceed to modify the configuration files so that I may be able to use my adapter? I can't find this documented anywhere.
Also, I have the 210865 ISP adapter. How shall I edit the configuration file for OpenOCD to support this properly?
I want to learn more about the newly released API. Is there any written documentation available or do I need to read the source code? I'm thinking of having the g5d to be used as an logic analyzer, if possible.
|
jcl
Conitec
Reged: 07/15/05
Posts: 88
|
|
The cfg file is, at the moment, not supposed to be modified by the user. So you can't change the pin configuration at the moment. However we're working on an OpenAlgo software package that allows users to implement devices themselves, and also edit the pin configuration.
You can find the pin assignment of the ISP adapter here (however for a 40 pin socket):
http://www.conitec.com/xpages/adapter/DS210865.pdf
On the 48-pin G5 Socket, TCK is Pin 25, TDO 26, TMS 27, TDI 29 (I haven't tested it - please check). The rest depends on which other signals and voltages are connected to the target.
The GALEP API has no documentation yet except for the comments in the source. It's certainly possible to use G5D as a logic analyzer or signal generator. If you have specific questions about some particular function in the source, please ask here and I'll get the info from the developers.
|
Rob
User
Reged: 04/29/08
Posts: 9
|
|
Hmmm, while I understand that you may not support such a change I'd urgently still would like to know how to modify this section as I think that I've figured everything else out (exerpt from am29lv800bb 16-bit):
Quote:
DefPin += 1, 23, trDontCare, 13, "A21"; //G5.01
DefPin += 1, 22, trHigh, 14, "Vsp"; //G5.02
DefPin += 1, 21, trDontCare, 15, "RY/BY#"; //G5.03
DefPin += 1, 20, trHigh, 16, "A18"; //G5.04
DefPin += 1, 19, trHigh, 17, "A17"; //G5.05
DefPin += 1, 18, trHigh, 18, "A07"; //G5.06
DefPin += 1, 17, trHigh, 19, "A06"; //G5.07
DefPin += 1, 16, trHigh, 20, "A05"; //G5.08
DefPin += 1, 15, trHigh, 21, "A04"; //G5.09
DefPin += 1, 14, trHigh, 22, "A03"; //G5.10
DefPin += 1, 13, trHigh, 23, "A02"; //G5.11
DefPin += 1, 12, trHigh, 24, "A01"; //G5.12
DefPin += 1, 11, trHigh, 25, "A00"; //G5.13
DefPin += 1, 10, trHigh, 26, "CE#"; //G5.14
DefPin += 1, 9, trHighSideDriverPin, 27, "GND1"; //G5.15
DefPin += 1, 8, trHigh, 28, "OE#"; //G5.16
DefPin += 1, 7, trHigh, 29, "D00"; //G5.17
DefPin += 1, 6, trHigh, 30, "D08"; //G5.18
DefPin += 1, 5, trHigh, 31, "D01"; //G5.19
DefPin += 1, 4, trHigh, 32, "D09"; //G5.20
DefPin += 1, 3, trHigh, 33, "D02"; //G5.21
DefPin += 1, 2, trHigh, 34, "D10"; //G5.22
DefPin += 1, 1, trHigh, 35, "D03"; //G5.23
DefPin += 1, 0, trHigh, 36, "D11"; //G5.24
DefPin += 1, 47, trHigh, 37, "Vcc"; //G5.25
DefPin += 1, 46, trHigh, 38, "D04"; //G5.26
DefPin += 1, 45, trHigh, 39, "D12"; //G5.27
DefPin += 1, 44, trHigh, 40, "D05"; //G5.28
DefPin += 1, 43, trHigh, 41, "D13"; //G5.29
DefPin += 1, 42, trHigh, 42, "D06"; //G5.30
DefPin += 1, 41, trHigh, 43, "D14"; //G5.31
DefPin += 1, 40, trHigh, 44, "D07"; //G5.32
DefPin += 1, 39, trHigh, 45, "D15/A-1"; //G5.33
DefPin += 1, 38, trHigh, 46, "GND2"; //G5.34
DefPin += 1, 37, trHigh, 47, "BYTE#"; //G5.35
DefPin += 1, 36, trHigh, 48, "A16"; //G5.36
DefPin += 1, 35, trHigh, 1, "A15"; //G5.37
DefPin += 1, 34, trHigh, 2, "A14"; //G5.38
DefPin += 1, 33, trHigh, 3, "A13"; //G5.39
DefPin += 1, 32, trHigh, 4, "A12"; //G5.40
DefPin += 1, 31, trHigh, 5, "A11"; //G5.41
DefPin += 1, 30, trHigh, 6, "A10"; //G5.42
DefPin += 1, 24, trHigh, 7, "A09"; //G5.43
DefPin += 1, 25, trHigh, 8, "A08"; //G5.44
DefPin += 1, 26, trDontCare, 9, "A19"; //G5.45
DefPin += 1, 27, trDontCare, 10, "A20"; //G5.46
DefPin += 1, 28, trHigh, 11, "WE#"; //G5.47
DefPin += 1, 29, trHigh, 12, "RESET#"; //G5.48
When I bought the G5D I was told that " If you have simple package converters for XXX, they should work with GALEP ".
Is it possible for you to update the documentation for the 210865 to be compatible with the G5 and also include the schematics (like your other ISP adapter) or to complete the pin-assignments descriptions for the N6, N7, N8 and N9?
When it comes to the API - Do you have any in-house small test/gadget programs that you are able to share with the public as an starter?
|
jcl
Conitec
Reged: 07/15/05
Posts: 88
|
|
There are different algorithms depending on which adapter you're using. You need to select the algorithm for the 210914 adapter, which is 1:1, not for the 210844 adpater, which is 40 pins -> 48 TSOP. I have no information here how to change the pinning in the cfg file, unfortunately. There's no external documentation for this. I'm not sure if this would work at all, because the algo is for 40 pins and the current cfg files are legacy - they will be very different with the upcoming next GALEP-5 software generation.
If you can't select the 210914 adapter for the chip you want to program, please contact the Device on Demand service. If the adapter is indeed not yet supported for that chip, they'll implement it quickly.
We'll update the pinning info on 210865 for 48-pin sockets. At the moment, please just add 4 to all pins. So pin 1 becomes pin 5 etc.
We don't have a small example program for the API yet - only the "big" example program that is OpenOCD. But I'll forward the suggestion to the developers.
|
Rob
User
Reged: 04/29/08
Posts: 9
|
|
Jcl, the 210914 is already supported for my device.
The adapter that my quote came from was from the cfg-file using the 210914 for my chip! That adapter you and the documentation claim to be 1:1 isn't realy what I'd call 1:1.
The reason I say that is that the 210914 is physically NOT mapped pin 1->1, 2->2, 3->3...48->48 (this is what what I would call 1:1). Instead it's mapped something like 1->13, 36->48, 37->1, 48->12. It would be very wrong to call this an 1:1 mapping since it's obviously not.
I have now several defunct(?) adapters from 3'rd party that are physically 1:1-mapped and can't therefor use them with my brand new g5d (I've only verified the tsop-48 so far).
I'll see if DoD will acknowledge my request for 3'rd party 1:1 mapped adapter. Otherwise I'm having some big trouble using the g5d until the revised software is released 
I tried to map the pins for N6, N7, N8 and N9 on the 210865 but gave up after a while. Can you find this information for me? N6 is the most wanted right now.
I'm looking forward to see some example programs using the API from the programmers :-) I'll have a more thorough look during the next couple of weeks while learning JTAG.
The OOCD-docs refer to a cable part# 095540. I can't find this on your web site. Could you please provide me with a pointer to its documentation.
Tnx again.
|
jcl
Conitec
Reged: 07/15/05
Posts: 88
|
|
Yes, the cable is not yet in our price list. We'll update the list soon.
I've now learned that the 210914 is indeed an 1:1 adapter in that sense that all single pins are connected to each other, but it's not mapped 1:1. I admit that this can be confusing.
We will implement new sockets in the software for a physical 1:1 mapping. You can then continue to use the 3rd party adapters.
|
Rob
User
Reged: 04/29/08
Posts: 9
|
|
Tnx, that was great news! Is there a way for me to get notified when this work is done?
When you say that you will implement new 1:1-sockets is this for the tsop-48 only or all sockets in general that doesn't have an physical 1:1 mapping?
|
jcl
Conitec
Reged: 07/15/05
Posts: 88
|
|
For all GALEP 5 adapters that have a different mapping. As far as I know there are 2 other such adapters besides the 210914.
I'll send you an email through your forum address when the new software will be available.
|
Rob
User
Reged: 04/29/08
Posts: 9
|
|
Any news about this?
|
jcl
Conitec
Reged: 07/15/05
Posts: 88
|
|
As to my information, it should be ready in the next time. I'll ask and post the ETA next week.
Update: Yes, it will be ready this week and make it into the next software version after Aug-11.
|
|
Previous
Index
Next
Threaded
Edit
Reply
Quote
