Post Reply 
[FRAM71] Pre-Production Batch
06-05-2015, 08:39 AM (This post was last modified: 06-05-2015 08:40 AM by Hans Brueggemann.)
Post: #121
RE: [FRAM71] Pre-Production Batch
an FRAM71_V510_HW104 firmware update is now available in post #1. this should fix the issue of a "zombie IRAM" as reported by olivier.

hans
Find all posts by this user
Quote this message in a reply
06-05-2015, 09:44 AM (This post was last modified: 06-05-2015 02:15 PM by Oulan.)
Post: #122
RE: [FRAM71] Pre-Production Batch
What model of programmer could be used to re-program the module ?

- on radisospare, I found nothing ( http://fr.rs-online.com/ )
- on mouser nothing ( http://www.mouser.fr )
- on conrad nothing( http://www.conrad.fr/ )

ACTEL LC Programmer seems not to be manufactured anymore

Thanks in advance

P.S. after many more configuration tests (and memory lost), my fram71 was certainly hard-reseted. I have now a nice config with FORTH at E0000, MATHROM, JPCROM and full memory (in this case I loose F0000-FFFFF segment as FORTH is installed) 64K iram, rest RAM. I will stay with this config until I open the 71 to solder the OD wire ...
Find all posts by this user
Quote this message in a reply
06-05-2015, 11:21 AM (This post was last modified: 06-05-2015 11:36 AM by Hans Brueggemann.)
Post: #123
RE: [FRAM71] Pre-Production Batch
(06-05-2015 09:44 AM)Oulan Wrote:  What model of programmer could be used to re-program the module ?
ACTEL LC Programmer seems not to be manufactured anymore

you can use the MICROSEMI FlashPro4 programmer with a suitable adapter cable instead. Dave Frederickson has done a description here.

as a work-around for the IRAM-bug, you may want to try a hard-reset of the FRAM71 as follows:
1) back up all contents of your HP-71B, especially those, that are _not_ located in ROM or IRAM
2) switch off your HP-71B
3) remove all power supplies (batteries and/or wall adapter) from the HP-71B
4) take your FRAM71 out of the HP-71B and use a straightened metal paperclip to short-circuit pin VDD with pin GND for approx. 2 sec (RTFM, chapter 14, for pin locations!)
5) re-insert FRAM71 into the HP-71B
6) apply power
7) turn the HP-71B on. calculator will show "MEMORY LOST", but contents of IRAMs and ROMs should still be there.

hans
Find all posts by this user
Quote this message in a reply
06-05-2015, 11:23 AM (This post was last modified: 06-05-2015 11:24 AM by Paul Berger (Canada).)
Post: #124
RE: [FRAM71] Pre-Production Batch
(06-05-2015 09:44 AM)Oulan Wrote:  What model of programmer could be used to re-program the module ?

ACTEL LC Programmer seems not to be manufactured anymore

Thanks in advance

On the first page of this thread Dave Frederickson reports using a FlashPro4, which Microsemi's site says supports the IGLOO FPGA used on the FRAM71, it is likely a FlashPro5 would work as well, as it supports pretty much the same families of Actel programmable devices.
Find all posts by this user
Quote this message in a reply
06-05-2015, 11:23 AM
Post: #125
RE: [FRAM71] Pre-Production Batch
(06-05-2015 09:44 AM)Oulan Wrote:  What model of programmer could be used to re-program the module ?

I have not tried it yet because I still need to make the adapter cable, but I bought the Microsemi SoC FLASHPRO5 programmer at DigiKey (1100-1179-ND) for this.
From what I read it should work, but Hans, Dave, Robert or Jean-Fran├žois are better placed to answer this.
Sylvain
Find all posts by this user
Quote this message in a reply
06-05-2015, 11:50 AM
Post: #126
RE: [FRAM71] Pre-Production Batch
FRAM71 uses an FPGA from the IGLOO nano series. as per MICROSEMI website, FlashPro5 should work with these.

hans
Find all posts by this user
Quote this message in a reply
06-05-2015, 04:10 PM
Post: #127
RE: [FRAM71] Pre-Production Batch
(06-05-2015 09:44 AM)Oulan Wrote:  I will stay with this config until I open the 71 to solder the OD wire ...

If your 71 has a metal back then the two boards are connected by a delicate flex cable, as opposed to the "golden zebra" pictured in the FRAM manual and used in the later plastic-backed case. It was reported in the PPC Journal that the traces in the cable can easily be fractured.

Consider modifying a module. I've modified 5 modules and can provide pics of my technique if anyone's interested.

Dave
Find all posts by this user
Quote this message in a reply
06-08-2015, 07:13 AM
Post: #128
RE: [FRAM71] Pre-Production Batch
Yes I would be interested with pics for modding a module, but first I need to get a spare module ...
Find all posts by this user
Quote this message in a reply
06-09-2015, 01:32 PM (This post was last modified: 06-09-2015 01:34 PM by Hans Brueggemann.)
Post: #129
RE: [FRAM71] Pre-Production Batch
TL; DR: FlashPro4 needs special programming cable with VJTAG support, otherwise will give error.

I have just received a note from michael that the FlashPro4 programmer is not working on his FFRAM71 when using a programming cable that was made according to dave's suggestions.
the reason is that dave's module, as well some others, but not all modules from the pre-production batch, still has R10, R11 installed (RTFM). the production batch has those resistors removed in order to conserve a couple of microamps in current consumption.
With R10, R11 removed, there is no VJTAG supply to the programmer, and that's why the programmer throws a vjtag error.
Until i have a solution for a programming cable, users who run into this issue can connect CN5:5 to CN3:2 during programming (yep, soldering required and again, RTFM). don't forget to remove the connection after programming is done, will save a bit on battery life.

Hans
Find all posts by this user
Quote this message in a reply
06-09-2015, 05:03 PM
Post: #130
RE: [FRAM71] Pre-Production Batch
(06-09-2015 01:32 PM)Hans Brueggemann Wrote:  TL; DR: FlashPro4 needs special programming cable with VJTAG support, otherwise will give error.

I have just received a note from michael that the FlashPro4 programmer is not working on his FFRAM71 when using a programming cable that was made according to dave's suggestions.
the reason is that dave's module, as well some others, but not all modules from the pre-production batch, still has R10, R11 installed (RTFM). the production batch has those resistors removed in order to conserve a couple of microamps in current consumption.
With R10, R11 removed, there is no VJTAG supply to the programmer, and that's why the programmer throws a vjtag error.
Until i have a solution for a programming cable, users who run into this issue can connect CN5:5 to CN3:2 during programming (yep, soldering required and again, RTFM). don't forget to remove the connection after programming is done, will save a bit on battery life.

Hans

Interesting. Are there any other differences between the HW104 modules, besides the programming header being pre-installed on the later batch?

I didn't mention it before, but my cable is wired as follows:

Code:

CN5     SIGNAL    LCPS      FlashPro4
1       GND       11,12     2,10
2       TMS       3         5
3       VPUMP     7,10      7
4       TCK       2         1
5       VJTAG     7,10      6
6       TDI       6         9
7       TRST      8         8
8       TDO       9         3

... and I get the following warning:

Warning: programmer '73735' : Vpump has been selected on programmer AND an externally provided Vpump has also been detected. Using externally provided Vpump voltage source.

The Microsemi datasheet states that the VPUMP provided by the programmer can supply one device and that VPUMP and VJTAG should be provided by separate supplies. Despite this, has connecting pins 6 and 7 on the FlashPro4 end of the cable been tried?

Dave
Find all posts by this user
Quote this message in a reply
06-09-2015, 06:19 PM (This post was last modified: 06-09-2015 06:21 PM by Hans Brueggemann.)
Post: #131
RE: [FRAM71] Pre-Production Batch
(06-09-2015 05:03 PM)Dave Frederickson Wrote:  Are there any other differences between the HW104 modules, besides the programming header being pre-installed on the later batch?

the pre-production batch was originally fully equipped with an auto power-off feature, that brought the current consumption down to just a few microamps, approx. 10 minutes after the HP-71B had been switched off. however, that feature wasn't compatible with having an alternate O/S activated. thus, the full-off feature was deactivated on all pre-production modules, but they still carry some of the components of that subcircuit. modules from the production batch no longer have these parts on board. power cunsumption figures in the manual refer to a -512 module from the production batch.

hans
Find all posts by this user
Quote this message in a reply
06-09-2015, 06:26 PM
Post: #132
RE: [FRAM71] Pre-Production Batch
(06-09-2015 05:03 PM)Dave Frederickson Wrote:  Despite this, has connecting pins 6 and 7 on the FlashPro4 end of the cable been tried?
no, didn't try that, as i don't have the FlashPro4. could be a viable, though.

hans
Find all posts by this user
Quote this message in a reply
06-09-2015, 06:53 PM
Post: #133
RE: [FRAM71] Pre-Production Batch
(06-05-2015 11:23 AM)Sylvain Cote Wrote:  
(06-05-2015 09:44 AM)Oulan Wrote:  What model of programmer could be used to re-program the module ?

I have not tried it yet because I still need to make the adapter cable, but I bought the Microsemi SoC FLASHPRO5 programmer at DigiKey (1100-1179-ND) for this.
From what I read it should work, but Hans, Dave, Robert or Jean-Fran├žois are better placed to answer this.
Sylvain

It's likely that the FlashPro5, like the FlashPro4, doesn't provide VJTAG to the FRAM71. You might try connecting pins 6 and 7 on the FlashPro end of the cable and letting us know if that works.

Dave
Find all posts by this user
Quote this message in a reply
06-09-2015, 10:55 PM (This post was last modified: 06-09-2015 11:27 PM by Dave Frederickson.)
Post: #134
RE: [FRAM71] Pre-Production Batch
I removed the zero ohm resistors, R10 and R11, on my FRAM71 that connect VJTAG and VPUMP to 3.3V. Made a new cable that connects pins 6 and 7 on the FlashPro4. This connects the 3.3V VPUMP generated on the FlashPro4 to the VPUMP and VJTAG on the FPGA and VJTAG on the FlashPro4.

Flashed V502 then V510 again. No problems. The warning message went away, too. Smile

Note: The LT1521-3.3 regulator in the FlashPro4 can provide 300mA.

Dave
Find all posts by this user
Quote this message in a reply
06-10-2015, 04:04 AM (This post was last modified: 06-10-2015 05:16 AM by Dave Frederickson.)
Post: #135
RE: [FRAM71] Pre-Production Batch
(06-08-2015 07:13 AM)Oulan Wrote:  Yes I would be interested with pics for modding a module, but first I need to get a spare module ...

Here's one way to make a "serviceable" OD module.

1. Obtain a donor module.

2. Remove the module cover. I haven't found an easy way to do this, but I use a hobby knife and a combination of cutting and scraping along the cover edges at about a 45 degree angle. Remove the circuit board. That's the hard part.

Click to zoom
[Image: osor-JRUPyJAMD0r_G5ZH7yOpiKmJgUHN4gmuy4D...32-h774-no]

3. Apply Kapton tape to both sides of the pins. This will support them during soldering.

[Image: m2PW4uJIl1QFPTxeM5ci3mEV84ekQir1Lds780Sz...32-h774-no]

4. Flip the board over and solder a wire between pins 1 (VDD) and 11 (OD). Without the tape the pins, which are unsupported, could become misaligned or fall off when soldered.

[Image: Udko9Yzz0s4jYctgmq9zAMez22t0ibtLRPZ9AztB...32-h774-no]

5. Flip the board back over and cut the Kapton tape with a hobby knife along the top of the pins and in between pins 1 & 2 and 5 & 6. Remove the excess tape leaving pins 2 - 5 insulated by the tape. This isolates the BUS[0:3] data lines.

[Image: DIOkP4e4R0dTXxdIgMOgw0xJ-McN4vVrgKyaY61N...32-h774-no]

6. Put the whole thing back together and close with Kapton tape. Make sure the tape goes around the middle of the module and goes all the way around as this is where the pins apply significant pressure.

[Image: sSKkXaEoNasHhfsEocok_5ECCPXjTPbFvKYg4_2F...32-h774-no]

When the module is inserted into Port 1 it pulls the Output Disable (OD) signal TRUE, disabling SYSROM and allowing FRAM71's SYSRAM to take over. Since the BUS signals are prevented from making connection, the module has no effect on FRAM71's memory configuration. The module is "serviceable" in that it can be opened up easily if the tape needs to be replaced or reverted back to it's original function by removing the tape, cutting a wire, and taping the module back together.

Comments and suggestions are appreciated. Smile

Dave
Find all posts by this user
Quote this message in a reply
06-10-2015, 07:59 AM
Post: #136
RE: [FRAM71] Pre-Production Batch
Hi Hans & Dave,

Thank you so much for identifying (so quickly) the FlashPro4 VJTAG support programming issue with the production version of FRAM71. Sounds like you already have some possible cable modifications that will help us enthusiasts.

Kind regards,

Michael
Find all posts by this user
Quote this message in a reply
06-10-2015, 02:26 PM
Post: #137
RE: [FRAM71] Pre-Production Batch
(06-10-2015 07:59 AM)Michael Lopez Wrote:  Thank you so much for identifying (so quickly) the FlashPro4 VJTAG support programming issue with the production version of FRAM71. Sounds like you already have some possible cable modifications that will help us enthusiasts.

The LCPS was never intended to be used without the development board. The dev board provides VPUMP and VJTAG to the LCPS. The blue wire on the programmer in the manual connects 3.3V developed on the LCPS to VPUMP and VJTAG which provides the programming voltage to the FPGA and powers the JTAG block. It wasn't designed to power anything other than the onboard circuitry.

FlashPro 4 and 5 are designed to provide VPUMP to one external device. By connecting pins 6 and 7 on the FlashPro this is doing exactly what is done with the LCPS, except with a beefer 3.3V regulator.

Ideally VPUMP and VJTAG should be provided separately. Installing R11 allows the FRAM71 to provide VJTAG, but at the expense of battery life.

Dave
Find all posts by this user
Quote this message in a reply
06-10-2015, 08:33 PM
Post: #138
RE: [FRAM71] Pre-Production Batch
(06-10-2015 04:04 AM)Dave Frederickson Wrote:  Here's one way to make a "serviceable" OD module....

Great explanation Dave, thanks for taking the time to help other FRAM owners make an OD module. I was thinking I will need to make a 2nd module for my second machine, but realized it's not needed since that one is already a 2CDCC ROM.

--Bob Prosperi
Find all posts by this user
Quote this message in a reply
06-10-2015, 09:43 PM
Post: #139
RE: [FRAM71] Pre-Production Batch
The instructions for updating the firmware specify that FlashPro v11.2 be used. v11.2 is no longer available so I have to use v11.4 or v11.5. Both of those versions give this warning:

Code:
Warning: One or more programming or configuration files have changed since generation.
         To view details and update a PDB, you may click on Modify PDB button for the target device.
         To update a STAPL file, you may reload the file or reload the project.

Is this of concern? Are there specific instructions needed to updae the PDB file?

Dave
Find all posts by this user
Quote this message in a reply
06-11-2015, 07:58 AM
Post: #140
RE: [FRAM71] Pre-Production Batch
Dave,

I noticed the same warning when attempting to use FlashPro V11.5 (version presently available for download) but due to the VJTAG error I didn't get far enough to know whether it would impede the flash process or not. Also, thought this warning may be related to the file locations on my PC (initially they weren't in the C: root directory).

Cheers,

Michael
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 2 Guest(s)