Repairing EmpirBus boards

Ian_Edwards

Ian_Edwards

Well-Known Member
Joined
9 Feb 2002
Messages
2,227
Location
Aberdeen Scotland
Visit site
Hi, my Southerly 46RS has an EmpirBus CLC system, which control almost all the power distribution and low power switching onboard.

The system is now about 8 years old and I'm starting to see some failures in the power switch boards.

These boards have 8 outputs, each capable of switching 8 amps at a nominal 12volts. Several outputs can be paralleled up to provide a high current output.

So far two output boards have failed, one provides power to the electric toilets, the other to the Eberspacher heating system.

I bypassed the Empirbus output for the two toilets with some wiring and appropriate fuses. The board powering the Eberspacer heater has only just failed and I'm going to work on that tomorrow. And I suspect the channels powering the Autopilot are on the way out.

Atlantis are the distributors for EmpirBus in the UK and they have replacement boards available, BUT they are an eye watering £635.60 inc vat each.

So I'm looking at options for repairing the boards. Since it's seems likely that it's just the power output stages which have failed, because on the board which powers the toilets, the other outputs (supplying much smaller loads) still work.

However, all the component numbers on the boards have been erased and I can't find any circuit diagrams online.

View attachment 55541

So questions:
1) Any ideas as to where I might get hold of a circuit diagram?
2) Suggestions for companies I could approach for a repair?
3) Ideas for where I might be able to get hold of second hand boards? (I look periodically on eBay and I've tried Google searches)
 
Re: Repairing EmpirBus boards - Update

Since I posted this, I've been in contact with the guys in Sweden.
We've agreed that, from the symptoms, that it more than likely that it's the output MOSFET's which have failed, and they have given me the part number, Infineon BTS-6143D. So I think it's worth having go at repairing the boards given the very high cost of replacement.
The chips are surface mounted and covered in a heavy conformal coating.
Assuming I can source the MOSFET's:
Is it easy to replace surface mounted components? They're the 8 square chips on the photo of the board.
How do I take the conformal coating off?
Or would I just be better to send the board to a company like Tinley and get them to do the repair?

They also suggested that I put a transient diode in parallel with the load, i.e. the DC motors, Any idea on what would be a suitable diode. The maximum load to be switched would be 40amps at a nominal 12 volts.

In previous incarnations I have designed and built circuit boards, both on Vero and PCB's, but that was in the days when the 7400 series of TTL devices were the new kids on the block!
 
Re: Repairing EmpirBus boards - Update

Ian_Edwards said:
Since I posted this, I've been in contact with the guys in Sweden.
We've agreed that, from the symptoms, that it more than likely that it's the output MOSFET's which have failed, and they have given me the part number, Infineon BTS-6143D. So I think it's worth having go at repairing the boards given the very high cost of replacement.
The chips are surface mounted and covered in a heavy conformal coating.
Assuming I can source the MOSFET's:
Is it easy to replace surface mounted components? They're the 8 square chips on the photo of the board.
How do I take the conformal coating off?
Or would I just be better to send the board to a company like Tinley and get them to do the repair?

They also suggested that I put a transient diode in parallel with the load, i.e. the DC motors, Any idea on what would be a suitable diode. The maximum load to be switched would be 40amps at a nominal 12 volts.

In previous incarnations I have designed and built circuit boards, both on Vero and PCB's, but that was in the days when the 7400 series of TTL devices were the new kids on the block!
Click to expand...

The output devices are a bit more than simple MOSFETs. They are high side "smart" switches. They also contain over current and over temperature protection. So the suggestion of protecting new devices with some form of transient suppression is a good one since it is unlikely they died of over temperature caused by over current.

The coating will probably either scrape off or burn. (I've only ever come across one coating which wouldn't, it's rare and expensive.)

If you want to try this yourself; it's at your own risk..........

Removal - by cutting the device legs you are more likely to preserve the PCB.

1. Use a scalpel or Stanley knife to cut the small legs. Do this close to the body of the devices by pressing the blade down towards the PCB. Cut one leg at a time.
2. Heat the "tab" with a soldering iron. You will need to apply a reasonably powerful iron for a surprising long time. Apply more solder to the iron and tab to ensure a good heat flow. Use a large tip if you have one but not so large as to damage other parts on the PCB. It's likely the tab will be soldered to a large area of copper and that will take sometime to heat up.
3. While heating the tab, gently push the device with a screwdriver or similar (not your finger) until you feel it loosen. Lift it off with a small pair of pliers.
4. Heat and remove the cut legs.
5. Use copper braid to remove the excess solder.

The hard part of replacement is making a good joint with the big pad under the device body.

6. Tin the large pad so that you have a thin layer of solder as evenly spread as possible. After cleaning with braid that might be what you have.
7. Put the new device in position and solder 1 of the small legs to hold it in place.
8. Place your larger iron tip so that it touches the tab and pad.
9. Feed in some solder to make a good thermal contact between the iron, pad and tab.
10. Gently press down on the centre of the device (not with your finger) until you feel it move down as the solder under the tab melts.
11. Remove the iron and wait for the solder to solidify.
12. Solder the remaining legs.

You are unlikely to achieve a joint as good as one done on a proper rework station. Heat dissipation though the main pad to the PCB depends on the quality of the joint. But if it over-heats the device will protect itself, although it will do that by turning off.

If you struggle to hold the PCB still on the bench, try a large lump of Blue-tak.

Good luck and have fun:)
 
Last edited:
Re: Repairing EmpirBus boards - Update

John,
Thanks for the detailed instructions.
On reflection, at least for a start, I'll ask Tinley for a quote.
On looking up the data sheet for the Infineon BTS-6143D, it is indeed a smart device, and it also includes diodes to clamp spikes from inductive loads, so I guess the original designers didn't think it necessary to include additional diodes.
 
Re: Repairing EmpirBus boards - Update

Ian_Edwards said:
John,
Thanks for the detailed instructions.
On reflection, at least for a start, I'll ask Tinley for a quote.
On looking up the data sheet for the Infineon BTS-6143D, it is indeed a smart device, and it also includes diodes to clamp spikes from inductive loads, so I guess the original designers didn't think it necessary to include additional diodes.
Click to expand...

It depends what the designers specified the outputs to be capable of and how carefully the installer/specifier studied the data sheet. There is a limit to the spike protection and as your motors wear there will be more and bigger spikes. I'd add a Transil to each output, particularly if you're going to the expense of having them repaired.
 
Re: Repairing EmpirBus boards - Update

Just a thought..........

Are you sure the devices have failed?

Test them with a small load to be sure.

They might just be overheating and shutting down.
 
Re: Repairing EmpirBus boards - Update

I'm fairly sure they've failed, when I connect the load directly to the power supply via fuses and relays, the loads work OK and they don't blow the fuses, which are rated just above the nominal working current.
When I look at the spec' for the MOSET's they are working well within their rated output, they are supposed to be fused at 8amps per channel and the chips themselves are rated at 40amps. if they were overheating, I'd expect them to operate briefly from cold, but there's no sign of life.
The EmpirBus nodes are mounted in very tight spaces on board, so you have to be a contortionist just to get at them and you can only get one hand to the board, so it's difficult to test thing.
I plan to take at least one of the fault board home after my trip to the boat this weekend, and I may be able to do some bench testing, now I know what the Chip is and have the data sheet. Pulling pin 2 to ground should enable the output and I can rig a dummy load and measure the current drawn.
Once again thanks for your input, it's really valuable when you're digging around with very limited information.
 
Re: Repairing EmpirBus boards - Update

Given the tests you have done and the fuse values, I think you are probably correct and the devices have been damaged by over-voltage spikes.

But given the expense involved in repairing the boards, I'd definitely do the check you've outlined to make sure they are faulty. It's just possible that your motor loads have started to draw enough current at start up to trigger the protection rapidly. But on balance I think "dead devices" is the most likely explanation.

These devices are not rated at 40A unless they have an external heatsink. IE a heatsink other than the PCB.

Your picture suggests there is just the PCB so the limit is 8A per device. The 8A limit assumes a PCB area for heatsinking of 50mm x 50mm for each device with the PCB mounted vertically and a free air flow over it. This will be significantly improved if there is blown air cooling. It also assumes that there is thermal isolation from one device to the next. It is clear from the picture that there isn't 50mm x 50mm PCB area per device.

This is confirmed by the EmpirBus datasheet which states up to 8A for each channel and up to 40A for all 8 channels (5A each on all channels).
 
Re: Repairing EmpirBus boards - Update

Once again, thanks for your input.

When I was thinking about the 40amp rating, I wan't thinking about thermal dissipation, more about the ability to take a transient current when starting an inductive load.

My plan is to get the boards repaired (if I can) and then use the output from EmpirBus to drive the coil of a relay which will then switch the inductive loads. This will still allow me to use the distributed coms of the EmpirBus, whilst protecting the output devices from the transients resulting from switching the inductive loads, and ensure that the EmpirBus boards aren't operating near their maximum current rating.

I've also been looking at Transient Voltage Suppression diodes, and it looks like the 1N6277ARL4G, would do the job. I has a 1500watt peak power rating and what seems to me to be about the correct voltage ratings, although I have to admit that I'm a little confused about all the different voltage ratings!
Any comment about the choice of the 1N6277ARL4G, or suggestion for a better alternative?

I'd fit the diodes on the 12volt power lines as close as I can to the electric motors, so that the spikes are clamped before they get into the DC supply.
 
Re: Repairing EmpirBus boards - Update

Ian_Edwards said:
Once again, thanks for your input.

When I was thinking about the 40amp rating, I wan't thinking about thermal dissipation, more about the ability to take a transient current when starting an inductive load.

My plan is to get the boards repaired (if I can) and then use the output from EmpirBus to drive the coil of a relay which will then switch the inductive loads. This will still allow me to use the distributed coms of the EmpirBus, whilst protecting the output devices from the transients resulting from switching the inductive loads, and ensure that the EmpirBus boards aren't operating near their maximum current rating.

I've also been looking at Transient Voltage Suppression diodes, and it looks like the 1N6277ARL4G, would do the job. I has a 1500watt peak power rating and what seems to me to be about the correct voltage ratings, although I have to admit that I'm a little confused about all the different voltage ratings!
Any comment about the choice of the 1N6277ARL4G, or suggestion for a better alternative?

I'd fit the diodes on the 12volt power lines as close as I can to the electric motors, so that the spikes are clamped before they get into the DC supply.
Click to expand...

Motor start shouldn't be a problem since the devices also have a soft start function.

The 1N6277ARL4G should be fine because;
- It will not begin to breakdown until 17.1V which is comfortabley above the maximum charging voltage in a 12V system.
- The voltage across its terminals will be no more than 25.2V while conducting 29.5A (temperature and tolerance will effect these figures somewhat).
- 25.2V is comfortably less than the maximum operating limit specified on the EmpirBus datasheet (32V).

Relays are also inductive loads. The 1N6277ARL4G is over the top protection when driving the relays although it will work. A rectifier diode would be fine for that.

As an aside........
There are a number of old threads in which people have suffered a burnt out switch in a circuit with an inductive load. Usually they do two things in addition to replacing the switch. Add some form of suppression to the inductive load (a diode is best) and add a relay. What gets missed is the need to add a diode across the relay coil as well. A while later the switch burns out again. There then ensues a discussion on MOSFETs, smart switches and diodes verses RC.

On a brighter note (for me at least) the software fix I've been testing for the last couple of days is holding up.
 
You must log in or register to reply here.

Share this page

Top