EB-006-00-9 Driver issue

For E-blocks user to discuss using E-blocks and programming for them.

Moderators: Benj, Mods

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

bobingabout wrote:
Crenwick wrote:Do you know that when the CPC is idle there is a voltage of 37V over C18 ?
...
C18 is on my board warm...
Mine is the same, I just measured.

Considering this is a 25V rated capacitor, this is a problem that should be addressed, even if at the very least the capacitor is replaced with a 35V or 50V alternative in later builds.
+1 :wink:

Or place a Zener to limit to 20V for example (will test this asap).
Bernard

Flowcode 6 / EB006-V9

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Hello,

Looks like we may be able to switch off the charge pump using mLoader and only bring it back on when programming. This means that if you power up the board the charge pump will be active but after using Flowcode/mLoader to program it will be disabled leaving the cap free to sit at a much more pleasant 5V. Not sure when this mLoader mod will be made but I have asked for it to be done asap, if it works ok then I will post the update here.

I'll also change the BOM for any new boards so that we specify a higher voltage cap is used.

I'm told by a colleague that the cap getting warm is probably heat being transferred from the regulator via the copper on the board which is not strictly bad but far from ideal.

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

Benj wrote:Hello,

Looks like we may be able to switch off the charge pump using mLoader and only bring it back on when programming. This means that if you power up the board the charge pump will be active but after using Flowcode/mLoader to program it will be disabled leaving the cap free to sit at a much more pleasant 5V. Not sure when this mLoader mod will be made but I have asked for it to be done asap, if it works ok then I will post the update here.

I'll also change the BOM for any new boards so that we specify a higher voltage cap is used.

I'm told by a colleague that the cap getting warm is probably heat being transferred from the regulator via the copper on the board which is not strictly bad but far from ideal.
Hi Ben,

Yes, disabling the pump when not in use should be a very good idea ! Just add some delay or check 12VDET when ok before programming :wink:
Just to test I added a 47K resistor in parallel to C18 and C18 remains cold ! The 17.5V Rail remains then on 21V.
Bernard

Flowcode 6 / EB006-V9

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

Benj wrote:Looks like we may be able to switch off the charge pump using mLoader and only bring it back on when programming. This means that if you power up the board the charge pump will be active but after using Flowcode/mLoader to program it will be disabled leaving the cap free to sit at a much more pleasant 5V. Not sure when this mLoader mod will be made but I have asked for it to be done asap, if it works ok then I will post the update here.
If you can do this via mLoader, I presume the charge pump is controlled by the onboard PIC used for the USB interface and programmer. If this is the case, perhaps an alteration to the firmware in this onboard PIC could have it turned off on powerup also?
Benj wrote:I'm told by a colleague that the cap getting warm is probably heat being transferred from the regulator via the copper on the board which is not strictly bad but far from ideal.
Although that is a possiblity, being an Electronic engineer myself, I would sugest that putting 37V across a 25V Cap could also be generating excess heat.



On another note though, I've been looking in to this "Unknown chip type in data file" error...
I've basically been reverse engineering the HEX file to figure out just what data is stored in there.
I think it could be caused by the fact that my HEX file doesn't contain the chip type that it is for.

The HEX file is basically a text file that contains a string of HEX data. (I did post the file I'm disecting in a previous post)
The first line contains the program. (because my program is so small, it only takes one line in the file)
The second line contains the Config headder info.
The 3rd line... I can't identify what it is, but it doesn't contain the chip ID, my guess is an "End of file" marker.

The main problem i'm getting while trying to program things is... if you do not specify the chip type in the command line, drop down box in mLoader, or the data file, programming will always fail.

Therefore, another sugested edit to mLoader is that if no chip type is given, perhaps it should either assume it is for the chip in the programmer and attempt to program it, or ask if it should.

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

Hi Ben,

Another question about Mloader, does the "Receive File" (suppose read pic) works on your desk ?

I tried out all my pics and always receive the same error "Bad response From device" and after the error the connection is lost between mloader and board (Red light and no device detected).

For information, I added some 100nF decoupling caps on my ZIF socket to test programming 16F1937 but no success.
Bernard

Flowcode 6 / EB006-V9

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

I just tried it, I get the same problem.

This is the error box that pops up.
Bad Response From device.

Board Type: EB006 v9
IsOpen : no
ID from Board: -1
Chip CmdLine:
USB Hardware: USB\Vid_12bf&Pid_0021&Rev_0000
USB Description: EB006 Multiprogrammer
ADCval: 0.000000
Location: mLoaderv3Dlg.cpp @ 1598 (CmLoaderv3Dlg::OnBnClickedBtnReceiveFile)

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Hello,

Right we have now made the mod to the mLoader software to switch off the charge pump after programming, working well here for me but want to test it a bit before releasing on the main site.

I've also asked for a build number/date to be shown in the diagnosis window and for the chip read functionality to be fixed but these haven't gone in as yet.

Have a play and let me know how you get on. To add the new mLoader to Flowcode simply extract the files here: "C:\Program Files (x86)\Flowcode 6\tools\mLoader"
Attachments
mLoader.zip
(5.92 MiB) Downloaded 380 times

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

Hi Ben,

Just tested with 16F887 and 16F1937 without ZIF and it doesn't program.
I receive "1:Flash memory not verified-may be due to code protect"

I checked also Vpp and it's start when programming and just after falls down to 5V. What I see on my scope is that there is some time needed before reaching 9V !

Let me know if you want something other
Bernard

Flowcode 6 / EB006-V9

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Hello Bernard,

We added a 1s delay after switching on the charge pump, I'll try upping this to 2s and do some more testing here to see if this makes any difference.

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

Ben,

As test purpose I just disconnected L1 and put a stable 15V over C18. Programming is then OK, so it has for sure to do with pwm out and time to raise.

In my last post I said that the Vpp raised to 9V so I also tried with 9V over C18 and the error comes back (9V seems to be to low).

Maybe worth to check if the pwm values in mloader has also be changed ?
Bernard

Flowcode 6 / EB006-V9

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

Okay, I've tested it.

it gives me the "Unknown chip type in data file" error every single time I tried it, which apears to be a step backwards.

I'm starting to think it throws up this error even if that isn't the actual problem, since it doesn't matter what chip is in the data file, I'm telling it that it is a 16F84A in the command line.

If the only change is that it turns off the charge pump when not needed, perhaps the issue is that it is failing to reach the required programming voltage?

So I did a test... probing the voltage on the MCLR pin (which needs to raise to at least 11V when programming).

First test from board boot, it reached 12V and programmed correctly.
Second attempt, it topped out at 7.7V, then brought up the unknown chip error. after pressing continue, it reached 10V and succeeded.
Third attempt, same as second, except it peaked at 11.5V.
Fourth also brought up the error, but I pressed it right away, seeing what would happen without bothing to take readings, and once again it failed to program. (the program doesn't run, and it brings up the "Flash memory not varified" error)

So I tried a different program... a longer program to see what happens over more time, and to my suprise it worked first time, but I was able to see that the VPP/MCLR line took a second or so to stabilise, worryingly so at only 9V, but it programmed successfully anyway.

I'm guessing the issues I'm having are in fact hardware related to an extent. There's not enough current going through the charge pump, which makes it take a while to stabilise, and difficult to maintain a stable 12V, which might be causing a few issues...


I took a look at the V8 and V9 data sheets, to compare the charge pump circuit (I havn't done any tests on a V8 board, but it seems to work flawlessly every time) The circuit given looks the same (except the resistors in the 12V monitoring circuit are different values, but the same drop ratio), However, it only shows what comes after CP_PWM, I wonder if it's something before R25 that might be different, providing less current?


EDIT: I should have checked for posts before posting.

Yes, a bigger delay, to give it time to charge and stabilise would most likely help, perhaps a delay after turning on the VPP line, before sending the data to give the 12V(9V) time to stabilise too?

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Hello Bernard,

Your right switching the charge pump off and back on is effecting the PWM duty!! :(

Looks like the best option is for you to return your board to us making sure you include your name and address in the package and we will reflash the board with the modded firmware.

If you have access to a PICkit 3 then we can send across the new firmware to you so you can reflash the board yourself.

We will go through all of our existing stock and reflash them and also issue out a notice to anyone who has purchased the board letting them know the specifics of the issue and asking if they want it to be reflashed.

The new mLoader is working fine for me on the 1937 and a 877A even with the VPP C18 voltage at around 9V so even with the new mLoader it looks like it's a tolerance issue.

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

A couple of questions first...

1. There's an update bootloader button in mLoader, I assumed that was to update the firmware, if it isn't, what does that do?

2. what would be involved in re-flashing the board myself? (I don't think I have the equipement to do it myself, I'm just curious)

3. What if someone tried to solder something to the EBlocks programmer? Like Crenwick soldered a resistor across C18.

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

bobingabout wrote:A couple of questions first...

1. There's an update bootloader button in mLoader, I assumed that was to update the firmware, if it isn't, what does that do? Good question !

2. what would be involved in re-flashing the board myself? (I don't think I have the equipement to do it myself, I'm just curious) ?

3. What if someone tried to solder something to the EBlocks programmer? Like Crenwick soldered a resistor across C18. Don't worry for me, I have a PICkit :)
Bernard

Flowcode 6 / EB006-V9

medelec35
Matrix Staff
Posts: 9520
Joined: Sat May 05, 2007 2:27 pm
Location: Northamptonshire, UK
Has thanked: 2585 times
Been thanked: 3815 times
Contact:

Re: EB-006-00-9 Driver issue

Post by medelec35 »

Something odd.
My board was working with the latest mLoader software.
Since yesterday just keep getting:

Code: Select all

File loaded from...
=> EEpromTest1.hex
File sending!
......................................................
File sent with 1 warnings:
  1: Flash memory not verified - may be due to code protect


FINISHED
Benj wrote:If you have access to a PICkit 3 then we can send across the new firmware to you so you can reflash the board yourself.
Perhaps I could reflashed and see if it get's things working again?
Ben would you mind sending me the hex file please?

Maritn
Martin

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

medelec35 wrote:Something odd.
My board was working with the latest mLoader software.
Since yesterday just keep getting:

Code: Select all

File loaded from...
=> EEpromTest1.hex
File sending!
......................................................
File sent with 1 warnings:
  1: Flash memory not verified - may be due to code protect
Yeah, try measuring the Voltage of VPP (Pin 1 of the 40 pin PIC socket) during programming. anything below 9V and it will give that message, which is what this firmware fix is supposed to solve.



There are a few other things going though my mind too...

Will this be the only fix, or might there be more later?

I also soldered a headder into J5, so I could use my breakout board for our old equipment on it. This isn't much of a problem, but I later tried to de-solder it using my crappy desolding equipment, and ended up ripping out a bunch of the VIAs on the pads for the pins of J5, which caused a couple of pins on the D9 connectors to stop working, as a fix I soldered it back in, on both sides of the board, but to do that i had to remove the plastic shell, and it looks a bit of a mess... functions as well as it did before, but looks a mess.
That thing is definitely "Out of warrenty".

Since it isn't actually mine personally, I'm just prototyping it for the college, I'd just order a new one, but due to... things... nobody is allowed to order anything untill August.


Problem is, I don't think I have the tools to flash it myself, Could I send it back anyway?

Going home now, I'll check for an answer on monday.

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

I've sent the new firmware complete with field upgradable bootloader to Crenwick and Martin.

Lets see how they get on with it over the weekend and if it's good then I will roll it out to anyone that wants it or accept returns and reflash any boards that need reflashing.

Note that the latest firmware I have given you guys will only work with the new mLoader which switches on the charge pump and waits a second befomre doing anything else.

Also note that the new firmware has a bootloader so it takes about 5 seconds for the board to power up into normal operation. The Prog LED will flash slowly while in bootloader mode.

medelec35
Matrix Staff
Posts: 9520
Joined: Sat May 05, 2007 2:27 pm
Location: Northamptonshire, UK
Has thanked: 2585 times
Been thanked: 3815 times
Contact:

Re: EB-006-00-9 Driver issue

Post by medelec35 »

Hi Ben,
After reflashing here are my observations.
1)
Benj wrote:it takes about 5 seconds for the board to power up into normal operation. The Prog LED will flash slowly while in bootloader mode.
That does not happen.
Prog LED onlt lights during programming stage.
2)
medelec35 wrote:File sent with 1 warnings:
  1: Flash memory not verified - may be due to code protect
Error is still present.
3) If the cancel is pressed when sending to target, VPP stays at 9.8V

4)VPP is higher than before.
With the attached flowchart, on file sent to programmer VPP does the following:
VPP1.jpg
VPP1.jpg (91.48 KiB) Viewed 14840 times
(time frame 0 to 1.6 seconds)
After that voltage stays steady between 9.5 and 9.9V.
A different programming session with longer time delay:
VPP2 Diff prog session.jpg
VPP2 Diff prog session.jpg (94.04 KiB) Viewed 14840 times
Looks like mLoader should wait for at least 2 seconds before establishing communications.

programming continues for 3mins & 45 seconds Which I thought is a bit long considering the size of program (unless extra time is taken for programming EEPROM)?

For reference using Hashcalc, CRC of mLoader = da7fa9ef
Posted that since there are a few different versions of mLoader about that state they are all V3.0

Martin
Attachments
EEpromTest1.fcfx
(6.57 KiB) Downloaded 259 times
Martin

User avatar
Crenwick
Posts: 127
Joined: Sun Aug 07, 2011 1:40 pm
Location: Belgium
Has thanked: 7 times
Been thanked: 21 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Crenwick »

Hi Martin,

I can't flash this weekend because I lend my PICkit to a friend who is out until tuesday :cry:

The latest mloader I have is also CRC : da7fa9ef and the version before was CRC : 0136ad14

@Ben , like Martin remarked it should be a lot easier if the version on display was modified at every change :wink:
Bernard

Flowcode 6 / EB006-V9

medelec35
Matrix Staff
Posts: 9520
Joined: Sat May 05, 2007 2:27 pm
Location: Northamptonshire, UK
Has thanked: 2585 times
Been thanked: 3815 times
Contact:

Re: EB-006-00-9 Driver issue

Post by medelec35 »

Hi Bernard,
Crenwick wrote:I can't flash this weekend because I lend my PICkit to a friend who is out until tuesday
No worries thats just the way it goes sometimes.
Crenwick wrote:@Ben , like Martin remarked it should be a lot easier if the version on display was modified at every change
I did mention this to Ben when I was at matrix HQ, so I'm sure that is on the to-do list. :wink:

Martin
Martin

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Cheers guys,

Here's a version of mLoader with a 2 second delay. Let me know how you get on.
Attachments
mLoader 2s.zip
(5.92 MiB) Downloaded 298 times

bobingabout
Posts: 45
Joined: Thu Dec 06, 2012 3:41 pm
Has thanked: 4 times
Been thanked: 8 times
Contact:

Re: EB-006-00-9 Driver issue

Post by bobingabout »

I've tried programming my chip with the 2s delay version, I couldn't get it to program once.

the only one I seem to have any luck with is the one that doesn't turn off the charge pump.

User avatar
Benj
Matrix Staff
Posts: 15312
Joined: Mon Oct 16, 2006 10:48 am
Location: Matrix TS Ltd
Has thanked: 4803 times
Been thanked: 4314 times
Contact:

Re: EB-006-00-9 Driver issue

Post by Benj »

Hello,
I've tried programming my chip with the 2s delay version, I couldn't get it to program once.
Sounds like you will need to return the board to us so we can test it out and return a working board to you.

medelec35
Matrix Staff
Posts: 9520
Joined: Sat May 05, 2007 2:27 pm
Location: Northamptonshire, UK
Has thanked: 2585 times
Been thanked: 3815 times
Contact:

Re: EB-006-00-9 Driver issue

Post by medelec35 »

Hi Ben,
Reflashed as per instructions.
Took screen shots of each stage.
I will PM to you just in case there may be something that should not be public, so you can see the process.

Now the Prog LED not lighting at all.
When programming with mLoader 2sec version the following pops up:
After Reflash1.png
(31.65 KiB) Downloaded 8681 times
Same pop up if programming directly with mLoader GUI or via Flowcode.

I will take the programmer into work tomorrow as they maybe something that requires testing.

Martin
Martin

User avatar
JohnCrow
Valued Contributor
Valued Contributor
Posts: 1367
Joined: Wed Sep 19, 2007 1:21 pm
Location: Lincolnshire
Has thanked: 364 times
Been thanked: 716 times
Contact:

Re: EB-006-00-9 Driver issue

Post by JohnCrow »

Hi Ben

Just been reviewing this thread. Not had chance to do anything much for the last few months, but did get chance start a new project over the weekend so take the opportunity to test out the new V9 Board.

Started with the 16F1937 supplied with the board and mloader (the last release before the one that turns of the charge pump)
The board was programming all ok.

I then had to move to a bigger PIC as I ran out of memory in the 16F1937. Using the 18F4550

I initially got the Flash memory not verified - may be due to code protect error and the program didn't work.

But i found it was due to an error in my config settings for the new device. After correcting this all worked fine again.

Not had any problems over the 2 days of fairly intensive usage.

I've just put the latest version of mLoader (2 sec delay) and tried programming the last version of my flowchart, and again its programmed all ok.

If the new firmware is ready I'll give it a go if you can send a copy over please.

I'm using a Matrix PSU set to 9V to power the board
1 in 10 people understand binary, the other one doesn't !

Post Reply