MEP 805B won’t stay running except In battle mode

[B8136sx2196]

Swapped around relays. Used dioxide. Switch Is good. I did try jumping across the starter to take the switch out of equation. Same thing happens. As soon as the starter engages the display goes blank and show to be rebooting. It will run 5-7 seconds and die. The generator makes voltage within a second of two of running. Checked between L2 and L3.

Now if I hit the battle short right after it starts to run and turn it off as soon as the display comes back up the engine will stay running. Then if I leave the unit running for an hour or two then turn it off it will restart like it should with NO use of the battle short. (normal cranking procedure).

Is the an internal battery somewhere that is charged when I let the unit run for a couple of hours and then discharges itself by the next day? A bad diode ?

all I know is if the display turns off during the cranking procedure it will die. If it doesn’t it will run just fine.

 

[nextalcupfan]

I'd look at the starter actually. It could be pulling way more amps than it should causing a voltage drop at initial crank.
I had this exact issue with an old car I had converted to Fuel Injection.
Whenever I hit the starter the voltage would drop low enough to reboot the computer in the fuel injection. This was even with a brand new Optima RedTop 1000 battery.
I had the starter rebuilt (they said it was totally worn out and literally everything needed replacing).
I gained an entire volt while cranking and it cranks 50-70 RPM faster.

 

[peapvp]

Swapped around relays. Used dioxide. Switch Is good. I did try jumping across the starter to take the switch out of equation. Same thing happens. As soon as the starter engages the display goes blank and show to be rebooting. It will run 5-7 seconds and die. The generator makes voltage within a second of two of running. Checked between L2 and L3.

Now if I hit the battle short right after it starts to run and turn it off as soon as the display comes back up the engine will stay running. Then if I leave the unit running for an hour or two then turn it off it will restart like it should with NO use of the battle short. (normal cranking procedure).

Is the an internal battery somewhere that is charged when I let the unit run for a couple of hours and then discharges itself by the next day? A bad diode ?

all I know is if the display turns off during the cranking procedure it will die. If it doesn’t it will run just fine.

there are several LED’s on the large backplane circuit board in the upper right hand corner. There is also a single red LED on the TCM-100

the best thing would be if you could video tape all LED’s without moving the camera when starting first normally until it dies and then whit the battle short switch start. The best thing would be to upload this video to YouTube since we can slow the video speed down on YouTube when watching the video

 

[peapvp]

I'd look at the starter actually. It could be pulling way more amps than it should causing a voltage drop at initial crank.
I had this exact issue with an old car I had converted to Fuel Injection.
Whenever I hit the starter the voltage would drop low enough to reboot the computer in the fuel injection. This was even with a brand new Optima RedTop 1000 battery.
I had the starter rebuilt (they said it was totally worn out and literally everything needed replacing).
I gained an entire volt while cranking and it cranks 50-70 RPM faster.

That would have been my next question, how low does the battery supply drop during cranking by measuring battery voltage with DMM

 

[Guyfang]

That would have been my next question, how low does the battery supply drop during cranking by measuring battery voltage with DMM

Look back to comments 10, 11, 12. I asked if you had load tested the batteries. You answered that the batteries are good. This is not the answer to my question.

 

[peapvp]

Batteries have been replaced and so has the starter within the past two weeks

Thanks Guy,
That’s the problem when these threads hang in the air for weeks and there is no movement meaning updates.

I am not in my twenties anymore where I could remember 15 different projects at a time plus this was during the times of no internet

So, LED Video it is

 

[peapvp]

Jay,

the Chinese are making these switches now. Ships from Shanghai.
US$ 7.23 + US$ 2.45 shipping

Switch EBay

Peter

so, the Chinese switch got here yesterday.

this is the smaller brother of the S1 switch.
All contacts are break before make (S1 is make before break)
Plastic is exactly as S1, contacts are cheap and may not carry the full current.

this switch is used in the AMMPS DCS. It definitely is the same plastic but the contacts in the DCS switch are gold plated

conclusion:
For the price they are excellent. Good to keep one on the shelve for an emergency when you have to wait for a correct spare part to come in.

This is a true domestic Chinese part and is usually not exported in this configuration



Peter

 

[B8136sx2196]

there are several LED’s on the large backplane circuit board in the upper right hand corner. There is also a single red LED on the TCM-100

the best thing would be if you could video tape all LED’s without moving the camera when starting first normally until it dies and then whit the battle short switch start. The best thing would be to upload this video to YouTube since we can slow the video speed down on YouTube when watching the video

Normal cranking. Display turns off unit dies.

battle short used to keep unit running

normal running after battle short used hours ago

 

[peapvp]

Normal cranking. Display turns off unit dies.

battle short used to keep unit running

normal running after battle short used hours ago

Jay,
I will check this tomorrow when I am back in my office.
As soon as the S1 (ECS) Switch hits start the CIM resets, even before the Start Relay activates the starter.
I slowed the video down to 0.25 speed and 1080 resolution
Then after start, the network failure light comes on with the corresponding LED on the backplane.
The TCM100 is in control of the network failure light and the heartbeat

I will post more info tomorrow when I have a chance to look at the manuals

Peter

cim reset before starter engages


Network failure light comes on and then engine shuts down but cim has not completed booting up

 

[peapvp]

Ok, problem is either CIM or connnection between CIM and TCM100


Check the wiring between CIM and TCM100
If ok, check the 100 PIN Connector on CIM

if both checkout, replace CIM, if problem persists then it’s TCM100 supplying the power to CIM

 

[peapvp]

This is the instructions Guy got for the 100 PIN Connector on CIM

 

[peapvp]

Ok Folks,
CIM arrived today and I was able to do some preliminary diagnostics.

The culprit is the internal Powersupply in the CIM.
I will trouble shoot the cim tomorrow

This video shows the erratic behavior b8136sx2196.107357 actually experienced during the start and is amplified here by using current limiting on my bench power supply
Link to YouTube Video - this video is a YouTube short and cannot be embedded here

and here is the video after the CIM has booted up and when the panel light switch is activated (here the two wires from the switch)

the display brightness of the backlight varies. Not sure if this is clearly visible in the video. When panel lights / backlight is turned off for a few seconds and then turned back on then the backlight is noticeable dimmer. Turning it on / off quickly then results in full backlight.
This is not normal and indicates a problem in the High Voltage part of the power supply. Most likely bad capacitors and / or DC DC cinverter

I will post more detailed troubleshooting results here tomorrow

If you have the same issue, leave the panel lights off for about 2 minutes after turning master switch on - then you should be able to cycle to the higher brightness like in my video and then you should be able to start your generator without using the battleshort switch

Peter

 

[peapvp]

Ok folks,

two problems found:
1. One ground wire was loose on its crimped connection on the connector which connects PSU and LCD (low voltage)
This crimp was bad from day 1

2. Both electrolytic caps are out 100uF and 330uF
The 330uF stabilizes the 5 Volt supply for the SBC ( Single Board Computer )

next update will be after changing caps and fixing connector

 

[peapvp]

Ok folks,

we found the actual cause of the resetting of the CIM:

the 330uF / 10V capacitor has no solder on its plus pin, they are SMD and the bottom of the pin solders to the top of the PCB solder pad.

The Plus Pin of the Cap made intermittent contact with the PCB

This cap stabilizes the 5 VDC Supply for the Single Board Computer
A bad contact can reset the SBC or in the worst case damage the SBC completely.
It’s like a power surge on your home computer

The cap was only attached by a tiny amount of epoxy next to the Minus Pin.

We replaced both caps on PSU with new vibration resistant caps.
The 100uF / 50V cap was soldered correctly.

Peter

 

[peapvp]

Update:

I have crawled now into the CPU and SSD and I am just sharing a couple of pictures.
Stay tuned and don’t assume anything

 

[peapvp]

Update:

Since this is the 3rd hardware version with the latest software installed in terms of operating system, I was able to get the following information:

1. The windows CE version installed is WINCE .net (basically WIN CE 4.0) with last driver's updated in February 2008 to accommodate winsystem GX500 SBC with USB
I was able to boot the system by creating a bootable DOS 6.22 USB Stick and retrieve the NK.BIN file, which contains the image of the created WIN CE Operating System

2. I was able to boot the NK.BIN File by making the flash disk bootable with WIN CE 6.0 ( 2 revisions past WIN CE .net)

3. There is only USB support / driver for Keyboard included but no USB Mouse support in NK.BIN

4. In our case here the EEPROM BIOS got corrupted because of all the failed start attempts

5. The Log records I found on the flash disk indicate over 8000 aborted starts since the Genset and CIM were built in 2009

6. The actual recorded runtime accumulative on this Genset and as recorded in Software is 138 Hours

7. The only way to move the pointer is with the Genset Keypad which is attached to COM 3 RS232 because of a lack of usb mouse driver in NK.bin

8. The flash disk is 128 MByte and 5 MByte are bad (bad sectors)

9. The BIOS corruption caused now a mismatch between the application software MCIICIM.exe and the 4 COM Ports
as a result, the application software MCIICIM.exe still transmits on RS232 COM 4 (connected to TCM100 J20) but does not process incoming data.

10. This is the same problem on COM 3 for Serial Mouse (Keypad)

11. The respective Baud rates on COM 3 and COM 4 are correct for the connections

This gets us to the following conclusion:

The CIM should be shutdown at least once a month through the CIM Software
The Logfiles should be read on screen, transfered to USB stick and then deleted so they don't get to large.
The CIM software will create log files up to a certain size, then create a new one

WIN CE .net does not have any maintenance utilities like defrag, scandsk etc included, so the flash disk can not be properly maintained like a hard drive in a Windows PC

The same is true for the WIN CE registry files - they cannot be cleaned - no tools included

The MBR can get corrupted when Generator shuts down with a heavy load attached in a fault condition (E-Stop, Low Oil, Over Speed, Over Temperature etc.)

This can also corrupt the EEPROM Bios

Winsystem incorporated a Brownout detection utility on this SBC, which will reboot the SBC if the 5 Volt Supply drops momentarily to prevent the above issues, but with a full load of 30KW or 60KW and the associated interference when emergency stopping, it becomes a gamble

The Windows CE operating system fails in some ways like a Android device (if you are familiar with them)

I will see if I can create a new NK.BIN in Windows CE 6.0 and add utilities and full usb support

I will keep you posted

Peter

 

[peapvp]

Update:

Keypad is up and running
If you remember the Windows 3.1 days with like setting up dial up modems with address, IRQ etc…..

 

[peapvp]

Update and conclusion:

CIM fixed.
CIM Software is basically written to run in Windows 3.1 (Everything 16 Bit)
If I ever felt old and grey then it was today.............
The TCM100/102 I am using here on the bench is not connected to rest of Genset, hence your getting these random numbers.
However, communication works fully when you change DIP Switch Settings for Frequency (60Hz / 400Hz) and KW, ask me how I know
If you need any help PM me
Peter

P.S. So who is your Daddy now????