I'm attempting to repair my friend's "DermaWand". It's supposedly a skin-improving device sold on late night TV.

Anyway, the device produces a glowing spark inside a glass tube very much like those toys you see in the toy and hobby shops that look like lightning inside a glass ball (Tesla's plasma ball). The "lightning" is attracted to your skin when you touch the glass. The DermaWand's output is contained in a small glass cylinder about 1/2" in diameter, with a flat end that is applied directly to the skin.

The one I'm working on does glow, but my friend complains that the intensity is too low and not nearly as much as it used to be. Turning the knob has little or no effect. I've verified that the pot is fine and working, the resistor values match their color codes, and the diodes have about 650mV across them in the right direction. D2 and D3 show open in the wrong direction, and D1 shows about 2V but the schematic shows why.

I'm thinking the triac, as the only active component, is suspect. But I don't have much confidence in being able to diagnose this simple circuit. The coil and tube are pretty much a black box (literally).

Any ideas?

 

▼ This circuit was redraw from yours.

▼ This circuit was changed the wire that I think more match the circuit theory.

Figure 2. Simplified AC Power Control Circuit using a Triac
http://www.allaboutcircuits.com/vol_3/chpt_7/6.html

http://www.learnabout-electronics.org/diodes_07.php

 

Other than layout, it looks like you've changed 2 things; the direction of all 3 diodes and the pot wiring. I'll check the diodes again and it wouldn't surprise me if I got them backwards. The pot I'm sure about - it uses only one pole and the wiper, leaving the 3rd pole open.

Any thoughts on whether it's worth trying replacing the triac? The fact that the light is glowing, although weakly, makes me think the triac might be working.

If the problem is the coil and/or bulb, I think I'm done. Can't seem to find the bulb anywhere.

 

Here's a circuit to test a Triac. Maybe this will help. Are you testing
D2 and D3 with the coil/tube connected? I've repaired light dimmers using triacs. As you can see a voltage applied to the gate should turn the triac on. With AC it's triggered each cycle.

 

The diode directions in the original drawing are accurate.

 

The diode directions in the original drawing are accurate.

I've seen triacs fail 'weak' but still work marginally.

 

I've seen triacs fail 'weak' but still work marginally.

OK, I might as well pull it.

 

How old is it? It could be the gas in the tube part has slowly leaked or diffused with time.

 

That's definitely a concern, and if it's true I'm cooked. I think the parts are proprietary and can't easily be replaced. I gather it quit working suddenly rather than gradually, so maybe there's hope for trying a new triac.

 

Ugh. I pulled the triac and tested it as shown above by tubeguy. It works. So my only hope is that is has failed "weak" as mentioned by kc5tpa. I gather it is not uncommon for a triac to fail at higher voltages but appear OK at test voltages.

In both configurations, the light comes on brightly and stays on when the gate is activated by a momentary connection thru a 1K resistor. In one configuration, the gate resistor must be touched to +12V to turn on. In the reverse configuration, it must be touched to 0V.

 

What if you reverse the main terminal triac connections, leave the gate where it is. Because in most triac circuits the load and control are between the gate and MT1 as in the test circuit or Scott's links. This circuit has them between gate and MT2.
I know it doesn't make sense if the circuit was originally working, but might be worth a try.

 

Have you tried checking or replacing C1 and C2? They will often fail opencircuit with no visible signs.

If either of those have failed (C2 is my guess) you could get a semi-working result but greatly reduced tube power.

 

Have you tried checking or replacing C1 and C2?

Good idea, I will now.

Lacking a meter to test capacitance, the easiest test I can think of is to see if these caps can couple an audio signal more or less like another cap of the same capacitance. Open to better ideas.

 

OK, I did an even faster test. I took them off the board and tapped them to the poles of a 13VDC supply, with my DMM attached, and watched the decay of the voltage on the meter when I removed them from the supply.

With no capacitor, the meter display re-zeroes in 3 seconds. The 474 and 105 caps prolonged this to 6 and 8 seconds (counted out by me, very approximate).

I conclude that neither is shorted nor completely open.

I don't have any handy replacements to try because of their high voltage ratings, so I reckon I'll just put them back and hope the new triac will do magic and get this thing zapping again.

 

Ugh. I pulled the triac and tested it as shown above by tubeguy. It works. So my only hope is that is has failed "weak" as mentioned by kc5tpa. I gather it is not uncommon for a triac to fail at higher voltages but appear OK at test voltages.

In both configurations, the light comes on brightly and stays on when the gate is activated by a momentary connection thru a 1K resistor. In one configuration, the gate resistor must be touched to +12V to turn on. In the reverse configuration, it must be touched to 0V.

When you say reverse configuration do you mean swapping MT1 and MT2?

 

Yes, I just swapped them.

 

Good idea, I will now.

Lacking a meter to test capacitance, the easiest test I can think of is to see if these caps can couple an audio signal more or less like another cap of the same capacitance. Open to better ideas.

As a really rough test you could charge one by touching on a battery or PSU, then touch it on a LED, it should give a little flash.

As a better suggestion you should keep an eye out for a cheap multimeter with cap and inductance meter built in, they are not expensive (cheap brands) and some even have a crude 3 or 4 digit freq meter function as well.

If you are after a nice cap meter I did a "high resolution" cap meter project which measures down to fraction of a pF, here;
http://www.romanblack.com/onesec/CapMeter.htm
you can build oyur own if you have a PIC 16F628A and an LCD.

OK after the caps it's probably wise to just replace those 3 diodes, assuming they are all typical small HV diodes like 1N4007. Are you sure none are a DIAC?

 

Are you sure none are a DIAC?

Pretty sure. They look like and are marked like normal black silicon diodes but I couldn't read the numbers before.

Now with those big caps removed I can see that they are all the same and labelled N4007. There's also a script "HB" logo.

They all test good using the meter diode test; all at about 630mV ∆V one way and open the other.

 

Pretty sure. They look like and are marked like normal black silicon diodes but I couldn't read the numbers before.

Now with those big caps removed I can see that they are all the same and labelled N4007. There's also a script "HB" logo.

They all test good using the meter diode test; all at about 630mV ∆V one way and open the other.

Looking more like the triac or the bulb.

 

Pretty sure. They look like and are marked like normal black silicon diodes but I couldn't read the numbers before.

Now with those big caps removed I can see that they are all the same and labelled N4007. There's also a script "HB" logo.

They all test good using the meter diode test; all at about 630mV ∆V one way and open the other.

Yeah it's standard to use 1N4007. They might still be breaking down at high voltages even if they test good for Vf. They are very common and only a few cents each, if you do decide to replace them. I keep stocks on hand and would just have replaced them with the TRIAC.