Recent Posts

91

Do-more CPUs and Do-more Designer Software / Re: MQTTPUB Gets Stuck

« Last post by Bolt on May 11, 2026, 10:45:43 PM »

I've been commissioning a different system using the same MQTTPUB logic, and this one has a much longer scan time (~14ms) and have had to bump up the timeout with a DEVWRITE to get it be able to publish data on time. It takes over 50 seconds. That's not a huge problem per-say, but it keeps returning a lower word in the error code of 22, but the data is going through. What does 22 mean? How can I further diagnose these issues?

92

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by BobO on May 10, 2026, 03:34:16 PM »

The FPGA doesn't directly connect to the bus. All bus connections are through buffers. It is certainly possible a buffer is damaged, but 405s are notorious for bent pins.

We're happy to look at the unit.

93

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by rlp122 on May 09, 2026, 06:50:01 PM »

<pushes chips into center of table>

I'm all in on it being a bent or retracted pin.  This is the most common issue with this PLC.  I wouldn't place much faith in ChaptGPT, it hallucinates a lot of stuff.

Retracted pins are very hard to spot.  I couldn't even begin to count the number that I have seen in my career.  It doesn't take much, even 0.5 mm is enough to cause it to not connect fully.  Get out a magnifying glass and look very closley.

94

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by Angelo on May 08, 2026, 07:39:42 PM »

Ah.... ok.

But do you think the problem is the Altera chip, even with the multimeter showing a short between pin 3 and the GND pin on the LJ245 IC?

95

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by franji1 on May 08, 2026, 08:36:04 AM »

The Altera part

96

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by Angelo on May 07, 2026, 09:45:31 PM »

That's an FPGA, which is not field replaceable (it must also be programmed).

Contact support at hosteng.com about RMA

What ? The LJ245A ?

97

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by franji1 on May 07, 2026, 07:45:44 AM »

That's an FPGA, which is not field replaceable (it must also be programmed).

Contact support at hosteng.com about RMA

98

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by Angelo on May 06, 2026, 10:38:28 PM »

I have checked. All pins are perfect aligned.

The system was running fine for the last 8 years, nobody removed the ECOM100 from the bus since then. So the bent pins hypothesis  would have 0% to be the case.


I forgot to ulpload the photo.

99

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by franji1 on May 06, 2026, 12:07:20 PM »

It may be a bent connector pin.  405 modules have that issue since they are rotated/levered into position and the connector sometimes doesn't line up perfectly and ends up bending pin(s).

100

ECOMs and ECOM100s / Re: ECOM100 blocking the bus

« Last post by Angelo on May 06, 2026, 12:01:06 PM »

Just removed the board to make some measurements with voltimeter.

What I´ve found:

 the IC (U13) which is a LJ245A, see photo below, there is a short between ping 13 (GND) to pin 3 (A1).

See that the U15 and U14 are also the same LJ245A

The LJ245A is the marking of Texas Intrument SN74LVC4245A. This CI is Octal Bus Transceiver and 3.3-V to 5-V Shifter
With 3-State Outputs.

Using ChatGPT I got a response like this:


"9. Definitive Test

The best method is to:

remove the IC,
test whether the bus communication returns,
or replace it directly.

Since you already work with PLCs and industrial buses:

if removing the IC causes the system to “come back to life” or restores communication,
then the IC is most likely internally locking up the bus.

This is extremely common."


Which seems to be the case as it cites  "lockin up the bus"


I will try to find the IC to purchase and then replace it.