CW80 output stuck at 20V

That's almost surely a board failure in the CW-80, they're not as robust as you might like.

 

It's probably possible to fix this, but I've never actually tried.  I'm pretty sure they use common components in the CW-80 for the control circuit.

 

How are you testing voltage output?
How well does it control a train?
The CW80 should max out at 18 volts.

I used a half way decent multi meter, no load on the transformer. I got this from a friend several months ago and hooked it up to a new track - before final connection, I tested the output and the pot on the handle has no effect. Always 20v, maybe 18, this meter is analog with the multiple scales.

Gunrunnerjohn,

I wonder if they sell the whole pc board if that is the issue. Removing and replacing ics, scrs, whatever may be beyond my job classification!

Dale Manquen explains the transformer operation here

 

http://www.trainfacts.com/trainfacts/?p=34

 

Dale H

You need to put a load on the output to get an accurate voltage measurement.
Try operating a train with it.

Had to ask, which terminals are you testing?

A-U is track is variable track voltage

B-U is accessory voltage which is fixed and can be adjusted.

I am testing the right terminals. Per the details from Dale and flash, I will put a train on the track tonight and see if the load makes a difference.

A lighted car will work as a load.

Just in case you did not actually read the article.

 

"Each of the Triacs is bypassed by a resistor and capacitor in series to reduce transients across the Triacs. NOTE: These rather large 1 uF capacitors will cause misleading output voltage readings on a high-impedance digital voltmeter if there is little or no resistance connected from A or B to U at the output. The impedance of the capacitor, even with the Triac shut off, will be much lower than the input impedance of the digital voltmeter, causing virtually the entire available 18 VAC to appear across the meter. When testing this transformer, add a light bulb across the output being tested. The bulb will not only offer a low resistance shunt to help the meter read correctly, but it will also give a visual indication of what is happening with the output voltage as the lever is changed or the Accessory voltage is programmed.

 

   Don

I always test transformers with some kind of load, I usually use the bank of 8 ohm 75W resistors I have from when I worked on old hi-fi equipment.   Even a pair of automotive bulbs in series makes a good load, just enough to prevent false readings.

 

FWIW, with any of my meters, the CW-80 tests zero volts at a closed throttle setting open circuit.  If he's getting voltage of any magnitude at the zero throttle setting, I'd say something is probably wrong.  If you can get power under load from the transformer with the throttle closed, I'd take a stab at replacing the BTA32 triac on the track circuit first.

 

The schematic is available on that link Dale posted.

 

John,

Before you go replacing triacs, there are some adjustments that can be made.
First, there are two potentiometers. One big one, attached to the orange handle, and one small one in series with the big one. The small one can be adjusted to bring voltage down to zero, if necessary.
If that does not work, the magic light bulb trick will. Just add a 24v incandescent bulb in parallel with track output.

If the transformer is stuck on full output and he didn't tinker with any of the controls before, it's pretty unlikely that adjusting them will fix it.  The interior pot is just a "zero adjust", but being that far off sounds like something has failed.

 

Finally got some time a few minutes ago and put a loco on the track. Voltage went to zero and train ran normally normally as I cranked the handle up. This is one of the original cw-80 units, there is no date code on the bottom. The cw80 I have been using prior is one with the G designation and is a newer generation unit. That one functions like my 1033 and KW, meaning zero volts when handle is off and no load. Volts increase as speed handle is turned. The old cw80 units need the load as several have noted along with the article referenced and Don't act like the new ones in this situation. My lesson is to test transformers with a load and do not assume that old and new generations will act the same. Thanks to all for the help.

High value capacitors have some leakage. It is very likely that the capacitor/resistor across the triac is the reason you got the bad reading.

   If you had an old Simpson meter you probably would never have noticed the problem.  Don

Or a light bulb load.