A cautionary tale re Exactoscale RSU

User avatar
Paul Townsend
Posts: 775
Joined: Mon Sep 14, 2009 6:09 pm

A cautionary tale re Exactoscale RSU

Postby Paul Townsend » Tue Sep 11, 2018 6:01 am

Exactoscale were one of the earliest suppliers of kit for resistance soldering. Bernard Weller sold two power supply versions, one cased and ready to go or just the bare bones transformer. I bought the naked transformer from Bernard in late 80s, built it into a steel box with switches, connectors etc and have used it from time to time ever since.
Over the last couple of years it has had extensive activity in building my broad gauge track.

I was happily using it this week until it suddenly stopped. The fuse could not have blown as the output still had 2 to 6 volts but could not supply the high current required.

This implied a high resistance connection inside. Investigation showed that the transformer outputs were connected by the manufacturer to a large nylon 4 way "Choc Block". 7 of the the screw terminals felt tight but the nylon on one pair was melted into a blob and the screws were inaccessible.

The manufacturer had made the classic mistake of solder filling the thick flexible wires before clamping them into the screw terminals. All electrical users, especially manufacturers should know that over years the solder migrates under pressure thus loosening the joint and causing heating.....the screw tip was burnt hence my high resistance connection. My repair removed the choc block and I soldered my leads to the transformer short tails, the joints now encased in heatshrink insulation.

I know people have been tempted to do this solder filling to avoid stray strands. DON'T DO IT!

Anyone using this otherwise excellent product should check their connections for a similar problem or risk a fire.

martin goodall
Posts: 993
Joined: Fri Oct 02, 2009 6:20 pm

Re: A cautionary tale re Exactoscale RSU

Postby martin goodall » Tue Sep 11, 2018 3:18 pm

Only people who really know what they are doing should delve inside mains equipment of this sort.

Paul definitely does know what he is doing, having spent his career in electrical/electronic engineering. He could no doubt build his own RSU from scratch if he were so minded!

My own RSU is the London Road Models version, designed by the late Mike Grey (another very experienced professional who was a university lecturer in electronics). So I hope and believe Mike got it right when designing the LRM RSU.

Tony Wilkins
Posts: 316
Joined: Tue Mar 20, 2012 3:57 pm

Re: A cautionary tale re Exactoscale RSU

Postby Tony Wilkins » Tue Sep 11, 2018 3:44 pm

martin goodall wrote:Only people who really know what they are doing should delve inside mains equipment of this sort.

Paul definitely does know what he is doing, having spent his career in electrical/electronic engineering. He could no doubt build his own RSU from scratch if he were so minded!

My own RSU is the London Road Models version, designed by the late Mike Grey (another very experienced professional who was a university lecturer in electronics). So I hope and believe Mike got it right when designing the LRM RSU.

I would second that.
Mike issued the specification for the transforms for the LRM RSU and had them specially made.
Paul's problem though was not with the transformer, but with the low voltage / high current output where any significant resistance can cause local heat generation.
The general perception is that it is volts that are dangerous, but low voltages with high currents can be just as bad. Look at the recent problems with cars etc catching fire due to electrical problems, something I fear we may see more of with the growth of electric vehicles.
Tony.

User avatar
Rod Cameron
Posts: 650
Joined: Fri Jul 25, 2008 12:01 pm

Re: A cautionary tale re Exactoscale RSU

Postby Rod Cameron » Tue Sep 11, 2018 3:50 pm

Paul Townsend wrote:The manufacturer had made the classic mistake of solder filling the thick flexible wires before clamping them into the screw terminals. All electrical users, especially manufacturers should know that over years the solder migrates under pressure thus loosening the joint and causing heating.....the screw tip was burnt hence my high resistance connection. My repair removed the choc block and I soldered my leads to the transformer short tails, the joints now encased in heatshrink insulation.

I know people have been tempted to do this solder filling to avoid stray strands. DON'T DO IT!


Would you apply this rule to layout hook-up wire as well, Paul?
Rod

User avatar
Hardwicke
Posts: 652
Joined: Sun Sep 14, 2008 10:25 pm

Re: A cautionary tale re Exactoscale RSU

Postby Hardwicke » Tue Sep 11, 2018 7:43 pm

Would that be a problem in DC or DCC? I've never had an issue in DC but not wired it up for DCC yet.
Builder of Forge Mill Sidings, Kirkcliffe Coking Plant, Swanage and Heaby. Still trying to "Keep the Balance".

martin goodall
Posts: 993
Joined: Fri Oct 02, 2009 6:20 pm

Re: A cautionary tale re Exactoscale RSU

Postby martin goodall » Tue Sep 11, 2018 8:19 pm

Tony Wilkins wrote: The general perception is that it is volts that are dangerous, but low voltages with high currents can be just as bad.


I seem to remember an old saw - "It's volts that jolts, but mils that kills." ("Mils" being milliamps or more likely whole amps).

User avatar
Paul Townsend
Posts: 775
Joined: Mon Sep 14, 2009 6:09 pm

Re: A cautionary tale re Exactoscale RSU

Postby Paul Townsend » Wed Sep 12, 2018 6:51 am

Rod Cameron wrote:
Paul Townsend wrote:The manufacturer had made the classic mistake of solder filling the thick flexible wires before clamping them into the screw terminals. All electrical users, especially manufacturers should know that over years the solder migrates under pressure thus loosening the joint and causing heating.....the screw tip was burnt hence my high resistance connection. My repair removed the choc block and I soldered my leads to the transformer short tails, the joints now encased in heatshrink insulation.

I know people have been tempted to do this solder filling to avoid stray strands. DON'T DO IT!


Would you apply this rule to layout hook-up wire as well, Paul?


Yes. Wires into screw terminals should not have soldered ends.
If you don’t want to insert bare wire, crimped ferrules are available

User avatar
Paul Townsend
Posts: 775
Joined: Mon Sep 14, 2009 6:09 pm

Re: A cautionary tale re Exactoscale RSU

Postby Paul Townsend » Wed Sep 12, 2018 6:52 am

Hardwicke wrote:Would that be a problem in DC or DCC? I've never had an issue in DC but not wired it up for DCC yet.


Yes

Tony Wilkins
Posts: 316
Joined: Tue Mar 20, 2012 3:57 pm

Re: A cautionary tale re Exactoscale RSU

Postby Tony Wilkins » Wed Sep 12, 2018 8:34 am

I am not a great fan of screw terminal connectors in general. They may be all right for short term testing, but as a long term or permanent arrangement, no.
The biggest problem is that people don't do the screws up tightly enough to begin with and wires have a habit of coming loose causing failures or shorts, usually at the most inconvenient times. Solder tags with properly soldered joints are a much better method.
Regards
Tony.

User avatar
Le Corbusier
Posts: 1285
Joined: Wed Feb 11, 2015 3:39 pm

Re: A cautionary tale re Exactoscale RSU

Postby Le Corbusier » Wed Sep 12, 2018 10:00 am

Tony Wilkins wrote:I am not a great fan of screw terminal connectors in general. They may be all right for short term testing, but as a long term or permanent arrangement, no.
The biggest problem is that people don't do the screws up tightly enough to begin with and wires have a habit of coming loose causing failures or shorts, usually at the most inconvenient times. Solder tags with properly soldered joints are a much better method.
Regards
Tony.

Why do these not suffer from the same problem of heat build up and solder migration?
Tim Lee

User avatar
Rod Cameron
Posts: 650
Joined: Fri Jul 25, 2008 12:01 pm

Re: A cautionary tale re Exactoscale RSU

Postby Rod Cameron » Wed Sep 12, 2018 10:20 am

Tony Wilkins wrote:I am not a great fan of screw terminal connectors in general. They may be all right for short term testing, but as a long term or permanent arrangement, no.
The biggest problem is that people don't do the screws up tightly enough to begin with and wires have a habit of coming loose causing failures or shorts, usually at the most inconvenient times. Solder tags with properly soldered joints are a much better method.
Regards
Tony.


Going a bit off-topic now

I would agree with that for underboard wiring, but when it comes to linking looms together from board to board I've always used mains-rated choc-block connectors with screw terminals. A bit OTT perhaps, but when you are putting the layout up and connecting boards it's easy to see if any wire has come loose (and it happens very rarely in my experience). D-connectors etc may be smaller, but involve more work in soldering and fault-finding - why go miniature when you don't need to?
Rod

User avatar
grovenor-2685
Posts: 3161
Joined: Sun Jun 29, 2008 8:02 pm

Re: A cautionary tale re Exactoscale RSU

Postby grovenor-2685 » Wed Sep 12, 2018 11:42 am

Why do these not suffer from the same problem of heat build up and solder migration?

The heat produced is a result of the power dissipated in the joint
W = I2R
With an RSU the current is much higher than for any normal model railway operating current, potentially well above 10amps
So even if the joint resistance is just half an ohm you get 50 or more watts in the joint that will get it very hot.

The solder migration is a result of the constant clamping force of the screw connector, there is no such clamping force on a solder joint.
Regards

Tony Wilkins
Posts: 316
Joined: Tue Mar 20, 2012 3:57 pm

Re: A cautionary tale re Exactoscale RSU

Postby Tony Wilkins » Wed Sep 12, 2018 11:46 am

Le Corbusier wrote:
Tony Wilkins wrote:I am not a great fan of screw terminal connectors in general. They may be all right for short term testing, but as a long term or permanent arrangement, no.
The biggest problem is that people don't do the screws up tightly enough to begin with and wires have a habit of coming loose causing failures or shorts, usually at the most inconvenient times. Solder tags with properly soldered joints are a much better method.
Regards
Tony.

Why do these not suffer from the same problem of heat build up and solder migration?

Hi Tim.
If I read your question correctly, the problem with soldered wire ends under the pressure of the screw causes the solder, which is a very soft metal, to creep and over time produces a loose connection resulting in a high resistance joint. A soldered joint between a solder tag and copper wire will not be under any external pressure and therefore not move. Also a soldered joint effectively becomes one solid unit. True, soldered joints can fail, but it is not very common, dry joints being the main cause.
Tony.

User avatar
Le Corbusier
Posts: 1285
Joined: Wed Feb 11, 2015 3:39 pm

Re: A cautionary tale re Exactoscale RSU

Postby Le Corbusier » Wed Sep 12, 2018 11:59 am

Tony Wilkins wrote:Hi Tim.
If I read your question correctly, the problem with soldered wire ends under the pressure of the screw causes the solder, which is a very soft metal, to creep and over time produces a loose connection resulting in a high resistance joint. A soldered joint between a solder tag and copper wire will not be under any external pressure and therefore not move. Also a soldered joint effectively becomes one solid unit. True, soldered joints can fail, but it is not very common, dry joints being the main cause.
Tony.


Thanks
Tim Lee


Return to “Paul Townsend”

Who is online

Users browsing this forum: No registered users and 1 guest