Blowing CDU's!!!!

[Fr Ted Crilly]

Hi.

 

I'm testing a newly built layout using the ages old method of Seep motors and push to make switches, and for a few days the points all worked fine, then the Gaugemaster CDU blew! Subsequently I've managed to blow a further two units very quickly, despite checking the wiring for shorts etc.

 

They seem to last for four or five point changes then blow! I've never had any problems on any other layout I've built over the years!

 

I'd be very grateful for any thoughts and suggestions as my head is getting raw with scratching!

 

Many thanks.

[Ray H]

Could it be that one of the push to make switches is making but taking far too long to break the contact?

[Dungrange]

or are the push button switches the latching type - ie push it to make the connection and then push again to break?

[SHMD]

Are these the CDUs with a big transistor built in or are they completely passive?

How many points (SEEPs) are you firing with each press?

 

 

Kev.

[Titan]

1 hour ago, Ray H said:

Could it be that one of the push to make switches is making but taking far too long to break the contact?

 

1 hour ago, Dungrange said:

or are the push button switches the latching type - ie push it to make the connection and then push again to break?

 

These won't have the slightest effect on the CDU, let alone cause it to blow up!  Indeed one of the features of a CDU, and definitely the Gaugemaster one is that the current drops off very quickly, so even if a switch gets stuck, neither point motor or CDU will come to any harm as the current falls to a safe level.

 

It will be worth checking the supply voltage to the CDU, especially if you are using a power supply that you have not used before. From what I can recall the max voltage of the capacitors used does not leave much margin, and if you are supplying them with AC or unsmoothed DC the peak voltage is approx. 1.4 times higher than the RMS voltage, so if you are using a power supply voltage that is close to the maximum of the capacitors rating, you could find that the peak voltage is actually more than they can handle.

Edited January 3, 2022 by Titan

[Ray H]

I only mentioned the sticky switch because we've had a couple on layouts that I've been involved with and blown CDUs. Another way to blow CDUs is to keep pressing the switch over and over again with minimal time between each push - we've had that happen too although in fairness it was with stud & probe and the probe was inadvertently left touching the stud.

[meil]

As has been said by Titan you can't blow-up a CDU by sticking switches. Two guesses: providing too high a voltage to the CDU or the Capacitors have been put in backwards. The polarity of Electrolytic capacitors is VERY important. Also their max operating voltage is also important.

Edited January 4, 2022 by meil

[Brian]

According to the images on the Gaugemaster web site Link to GM CDU the two capacitors are rated at 40 volt and each is 2200uF   Therefore the maximum AC or unsmoothed DC input should not exceed 28 volts which is very high anyway IMO!  

 

Things that I would be checking...  Input volts and the type - AC or DC or regulated DC.    Disconnect a known working CDUs output and read DC output voltage.   Ohm test each point motors coils possibly to be found at around 4 - 6 Ohms per coil.   Check for a short occurring somewhere on a particular motor or its wiring.  If using diodes as Bemf suppression ensure none of these have gone on short circuit. and none are fitted the wrong way around!

 

 

 

[Crosland]

1 hour ago, meil said:

As has been said by Titan you can't blow-up a CDU by sticking switches. 

 

You could if it is poorly designed.

 

E.g. with a transistor current limit that does not cut off when discharged may stress the transistor if the switch remains closed longer than anticipated by the designer.

 

I have seen such designs.

[Titan]

1 hour ago, Crosland said:

 

You could if it is poorly designed.

 

E.g. with a transistor current limit that does not cut off when discharged may stress the transistor if the switch remains closed longer than anticipated by the designer.

 

I have seen such designs.

 

That may be true, but we are specifically talking about a Gaugemaster unit which is well designed and has an appropriate current limit that does not stress the transistor, so it is possible to eliminate sticky switches and shorts as a possible cause.  However, as we don't know for sure that the voltage applied is not exceeding 28V, and if it does it could cause the problem described, then that needs to be assessed and confirmed/discounted before other causes are considered, otherwise another CDU may end up as toast.

[Fr Ted Crilly]

Thank you so much one and all for your comments and advices, they are all very much appreciated.

 

The CDU came with its own 16v AC Gaugemaster power supply. The switches which are now of the toggle variety which you momentarily push from the central off rest position to either side have all been replaced. I did initially try Cobalt S lever switches as I have a great interest in mechanical signaling , but in view of the failure, and as this is to be an exhibition layout they were to large and unwieldy to be practical.

 

I am only firing one point at a time with approx 5 seconds delay in between, there are 7 motors in total. One switch does fire two motors at the same time as they operate a cross-over.

 

One thought that occurs to me is that as all motors are within 6 feet of the switches, do I really need a CDU?

 

Any further thoghts and observations are most welcome, thanks.

[SamThomas]

Just a thought - have you run this past the technical guys at Gaugemaster ?

It can take a bit a patience to get through to them though.

[SHMD]

2 hours ago, Fr Ted Crilly said:

Thank you so much one and all for your comments and advices, they are all very much appreciated.

 

The CDU came with its own 16v AC Gaugemaster power supply. The switches which are now of the toggle variety which you momentarily push from the central off rest position to either side have all been replaced. I did initially try Cobalt S lever switches as I have a great interest in mechanical signaling , but in view of the failure, and as this is to be an exhibition layout they were to large and unwieldy to be practical.

 

I am only firing one point at a time with approx 5 seconds delay in between, there are 7 motors in total. One switch does fire two motors at the same time as they operate a cross-over.

 

One thought that occurs to me is that as all motors are within 6 feet of the switches, do I really need a CDU?

 

Any further thoghts and observations are most welcome, thanks.

 

Do you have a multi-meter and can you measure the "no-load" ac voltage coming out of the Gaugemaster power supppy?

(This would answer many questions...)

 

 

Kev.

 

[DCB]

Its not clear from the post if the CDU stopped working or if it went bang and stopped working.  I have successfully blown up capacitors by (a) connecting them backwards. and (B) connecting them correctly as per the instructions due to the fact that they were rubbish made by Joseph Lucas.  They go "Pop" quite loudly and shoot out liquid, which stains wood quite a delicate shade of yellow.  

If they go bang with the correct gauge master CDU and  power supply it has to be a manufacturing issue.   If the CDU has failed due to the transistor or something on the PCB then it could be incorrect use such as a sticky switch leaving the output connected.    I would return the components to Gaugemaster as I believe they have a lifetime warranty.   I would also check for continuity between the CDU output and return,  When none of the point switches are activated there should be no continuity, when one or more is activated obviously there should be continuity. Most people would use a digital multi tester, I prefer a buzzer or analogue meter which I can hear.  

 

[Pete the Elaner]

13 hours ago, Fr Ted Crilly said:

 

One thought that occurs to me is that as all motors are within 6 feet of the switches, do I really need a CDU?

 

 

The CDU is not there to overcome volt drop over a longer wire (although it does help with this).

 

A solenoid is an inductor, which opposes change of current. A transformer is also an inductor, so while it supplies the voltage, it also tries to oppose the change of current. This causes it to throw the motor a little softly.

A capacitor does the opposite: it discharges as fast as possible. This is what gives CDUs their kick rather than a more tired throw.

Another benefit of the CDU is that after the initial kick, the residual current is very low. This not only prevents damage to the point motor if the switch is left closed, but it reduces the arc created when the switch is opened after use, so if you are using small switches, it makes them last a lot longer. If you are using stud & probe, they stay cleaner.

The arcing is created by the rapidly collapsing magnetic field around the motor, which generates a voltage.

[SamThomas]

I

7 hours ago, DCB said:

 Most people would use a digital multi tester, I prefer a buzzer or analogue meter which I can hear. 

I have a couple of DMM's that I use, but find the old analogue one very useful to measure something that is fluctuating.

 

I also have a 24v vehicle test lamp - also very useful & for faulfinding a set of test leads in series with a 55w bulb (with a buzzer in parellel with the bulb) and decent PSU - all 12v.

[Brian]

Hi

Can you confirm your layout Seep PM? motors, CDU and switches are wired like this?  

 

[Fr Ted Crilly]

Once again many thanks for all your help and comments etc. Apologies for my slow replies however I haven't been feeling to good in recent days! It's great to know so many people want to try and help.

 

I confirm the wiring is as per Brian's diagram.

 

As I'm not used to using them, after reading the multi meter instructions carefully (LOL) I confirm the out put of the transformer is 18.6/18.7 V AC, its labelled as 16V.

 

The caps haven't blown, there hasn't been any explosions etc, I suspect it's the transistor on the CDU that has gone.

 

I hope I haven't missed anything, and still suspect its my error in some way or other. Anywat, I look forward to all your further responses.

[Brian]

As you say your wiring is as per the drawing I supplied, then my next step would be to return the CDU and its power supply to Gaugemaster for them to examine and see what is wrong and hopefully repair or replace the items.  (Note, many GM returns are repaired FOC under their Lifetime warranty!)  Contact their Service dept. first to obtain any returns info. they may need.  01903 884488

[SamThomas]

1 hour ago, Brian said:

As you say your wiring is as per the drawing I supplied, then my next step would be to return the CDU and its power supply to Gaugemaster for them to examine and see what is wrong and hopefully repair or replace the items.  (Note, many GM returns are repaired FOC under their Lifetime warranty!)  Contact their Service dept. first to obtain any returns info. they may need.  01903 884488

Make sure you get a retuns number then it will be easier to trace in the event of a query.

 

Not too sure of the GM Lifetime Warranty applies to CDU's like it does with the analogue controllers.

[John Besley]

I am not blowing up CDU's but need some help, what terminals on the in coming and out going should be pos and neg, can't see anything on the instruction sheet that came with my CDU

 

I am also a bit surprised that the circuit board doesn't have anything marked on it for in and out untill you turn it over .... Which doesn't help if you've screwed it down

 

 

I have marked it up for my own sanity

[Lurch]

The arrow points to the negative end.

 

Usually the stripe denotes which leg is negative on radial capacitors (where both legs are on the same end). The arrow points to the negative end on the radial capacitors. The stripe usually has the minus sign on it.

[cliff park]

I suspect the OP is referring to the polarity of the supply, rather than the capacitors. but isn't this fed with AC.  The output does not matter, since the point motor coil will work regardless of polarity.

[Brian]

Cliff Park is correct,  its like the the below

You can feed the CDU with AC typically 16v AC or DC at around 19 - 21 volts.   The CDU has a diode in the input so getting a DC input the wrong way around result in it not working 

 

 

[RAF96]

Feed it AC and meter the output.