Firmware Uploading Problem: - Continuous Counting.

[LandRoverV8]

Hi Everyone,

I'm new to to MJLJ but am resonably experienced at building circuitry and have followed the 'Assembly Instructions' to the letter when building my kit. I've conducted all the tests suggested and get all the correct outputs.

I've come to upload the firmware onto the unit from my PC following the 'Flashing Firmware' instructions on the Autosport Labs Website. I'm confident that I've got the COM port configured correctly as shown on the site.

When I type in the command window:-

c:\hc08sprg.exe COM1 mjlJunior_V4.0.2.s19

I get:-

Waiting for HC08 reset ACK...received 0xfc (good)

as expected, but then this:-

Calibration break pulse sent. Count: - and the count keeps continuously increasing, if left, it will not stop!


Can anyone advise as to what the problem is here?


Thanks,
James

[brentp]

Hi James,

First of all, we want to emphasize the processor you received with your kit already has firmware loaded.

So- you don't need to perform a firmware upgrade unless you can confirm you need it- for example, you know 4.0.1 is already loaded and you want to go to 4.0.2.

Were you able to verify the unit communicates with the configuration software? This should be your first test after completing your kit assembly.

If the unit is communicating with the configuration software, you can check the firmware version by clicking "Help/About".


Other than that, here are two common problem areas when loading new firmware:

1. USB-Serial adapter: Are you using a USB-Serial adapter? If so, try performing the upgrade on a PC that has a real serial port.
2. Power interruption: After you apply power to the unit, make sure power isn't interrupted. Unsure if you're doing this in the vehicle or on a test bench- but it's worth mentioning.

Let us know what you find out-

[LandRoverV8]

Hi Brent,

Ok, I've tried to get the unit to talk to the Configuration Software. It returns a 'Timeout reading cylinder count'.

I've also tried to get a reponse from the unit using Hyperterminal and get nothing.

I'm using a home made Serial cabe which was built following the instructions on the site. I've tested this about 4 times with a meter to make sure there aren't any dodgy conections and that I've wired it up right.

I've tried it on 2 different PC's with real Serial ports. Neither get any further forward.

I've rechecked the board construction about 4 times and so have two other people with a 'fresh pair of eyes'!

I'm powering it on the bench with a 12V battery and don't currently have an EDIS equipped vehicle to test it with.

Is there anything else I can test? Any help is much appreciated.

Thanks,
James

[LandRoverV8]

As an additional note after some further playing:-

I've tested all voltages throughout the board, up and down current of all diodes, as you have suggested on other posts - all voltages are correct. I've tested all capacitors with a capacitance tester in such a way that proves the solder joints are sound (in the chip sockets etc).

I also stole my Dad's osciloscope and have confirmed that the oscilator is giving 16 of Her Majesties Finest Mega-Hertz!

So I'm am scratching my head as to what else might be wrong other than a PC Comunication problem somewhere?

Thanks again,
James

[brentp]

James,

What we can infer from your tests so far- specifically the information around the attempts to flash the firmware- is that the processor can *transmit* data to the PC but not receive data.

The reason why I believe this is true is because the boot-loader software on the processor sends a single character to the PC upon power up, signaling that it is ready to start boot-loader related activity. If it doesn't see a command, it jumps to the firmware code on the MJLJ.

In your case with the firmware program, it sends the "calibration pulse", yet never hears back from the processor. Evidently the processor doesn't hear the pulse from the PC.

It's good you have an oscilloscope! use this to monitor data on the rx and tx lines off of the PC serial port. Trace this from the PC, through the RS232 transceiver, and on through to the processor. Hopefully you will see a point where you *don't* see a signal where you should see one.

Related,

Are you at a point where you can hook the MJLJ up to a running engine set up with EDIS? If so, you can run the engine, and use a timing light to verify the ignition advance is being controlled by the MJLJ. This is another way to verify the unit is basically 'alive', and will narrow down your problem to just communications related issues.

Regards,

[dave9469]

Hi James

You could try taking the processor out (if it's in a socket) and making a link between the TX and RX with a thin piece of wire. Now in a hyperterminal window whatever you type will be sent straight back.

[TwoSheds]

Hi,
I was wondering where you are? If you are local to another Megajolter you could see if their laptop can communicate with your unit perhaps? or vice versa..

[brentp]

Hi James

You could try taking the processor out (if it's in a socket) and making a link between the TX and RX with a thin piece of wire. Now in a hyperterminal window whatever you type will be sent straight back.

That is an excellent idea.

[LandRoverV8]

Hello All,

Thanks for the suggestions. I've tested the Tx and Rx lines as suggested to help trace the fault.

I found that I could use Hyperterminal to return all characters entered when using a wire jumper between Tx and Rx on just the lead, when unplugged from the board. As soon as it was plugged into the board, with U2 removed (mine had an additional chip holder) and a jumper inserted in the chip socket as suggested I get nothing. Testing with a meter I found that Pin 13 'R1 IN' has a direct connection fault to +5V! So held in this 'High' state all the time, no signals can get through.

So I've looked at all the solder pads under a magnifier with a bright light to check there's nothing shorting that I've soldered and everything is ok. I've removed the U2 chip holder, in case that was causing a problem. Still the same. I really is a solid short too, just lead resistance, not like its a bit of dirt or something.

I really don't want to blame the board itself because I know they have such a good reputation, but I really cannot work out what has caused this!

Has anyone else got any ideas?

I'm not able to test the rest of the board on an EDIS engine yet because mine isn't finished. Ironically the old clock work distributor points died the other day and I had to replace them, it's calling out to be 'Jolted'!

Thanks,
James

[dave9469]

Hi James

I see you've taken U2 out, the MAX232, and you have 5v on pin 13. Is this also the case with and without your PC lead connected?

What pin out have you used for your home made cable? Are you 100% sure you read the correct pin numbers on each end of the male/female 9 pin plug? sorry if this is sucking eggs but it can be the simplest of faults sometimes.

What version of board do you have? V3 or V4?

I've got a V3 in front of me and it's very difficult to see how pin 13 could have 5v on it because it's only a very sort piece of track to the 9 pin socket. I'd assume the V4 will be the same but I'd need a good quality picture of the underside of the board, possible Brent could add one?

[pr_race]

James
I have the same problem on a v4 board and can't get to the bottom of it either. Going to have another look at it tomorrow ans see if I can find anything.

Peter

[brentp]

I believe James has a V4 board, correct?

James-

The 5V plane is the bottom of the board, (The 5V plane receives it's feed right between D1 and R9)

With nothing connected to the board (no serial cable connected, no power), and with U2 unplugged, You're saying that you are measuring zero ohms between pin 13 of U2 and the 5V plane?

Realize there is only one short trace between U2 pin 13 and X1 on the board- if you do have a short that's the area where you want to focus on. That trace is on the backside- the front side of the board has no trace connecting to pin 13 of U2.

[LandRoverV8]

Hi Guys,

Thanks again for the support.

Yeah it's a V4 board.

Before I wrote my most recent post, I was staring for some time, at that short trace from U2 Pin 13 to X1, only being 30mm long at most, thinking I must have missed some other link. You've confirmed I haven't at least!

I'm reading 0 ohms between U2 pin 13 and +5v plane, with no serial cable, no power applied or U2 fitted and can't see where the short is at all. Several of us have looked at the back side of the board with a magnifier + brightlight without being able to spot any stray solder etc.

What's best to do now?

Thanks,
James

[brentp]

Hi James,

Seems odd, doesn't it? Although the answer is almost always simple and obvious after the fact.

* Without U2 installed, and the unit powered up- do you actually measure 5v on pin 13?
* Can you measure any other shorts from U2 pin 13 to any other areas on the board?
* Check around on the solder side for X1. make sure there are no solder bridges from the pads to the 5v plane.
* scrape around the trace to make sure there are no solder whiskers.
* I know it sounds bonkers, but can you run any appreciable amount of current between pin 13 and the 5V plane? can you power an LED?
* What about other pins on the U2 socket? Can you measure any other shorts as you probe around?