[daveone23]

Can someone help me figure out whats wrong with this signal generator. I recapped it, tubes are good but when i hook ground to the scope it sparks. None of the knobs on the sig gen change the scope pattern.

[MrFixr55]

Re the sparks, verify the condition of C33 and C34, the power line RFI filter caps on the generator. Verify the filters on the scope also if the scope is not grounded. If the caps in the generator are original, replace them with type Y safety caps or remove them. verify that there is no leakage from the power transformer or miswiring, solder splashes, grounding of one leg of the power cord, etc. Consider grounding the generator using a 3 wire cord. Use a 3 prong - 2 prong "Gozinto" (this gozinto that) if you must float the generator.

What is the frequency and voltage of the sine wave being displayed? is it 60 Hz, 400 HZ or something else? If it is 60 Hz and of high voltage, then it is from leakage and may be totally overriding the generator waveform.

[daveone23]

Here are the .05 line caps i replaced (yellow). How
Do i check for a leak at the transformer?

[daveone23]

The .05 caps have no continuity, the line cord has no continuity to ground. The ground probe has full continuity to chassis and the signal probe to ground probe is 8ohms.

[MrFixr55]

Hi Dave,

Unfortunately, the pic is not completely "painting" on my computer.

How big was the spark? I assume that your o'scope is grounded and your sig gen is not. If they are both grounded by 3 prong plugs there should be no spark.

Start out by ohming between the chassis and each pin on the power plug. the results should either be an open circuit or a high ohm reading, climbing to an open circuit (effect of the EMI caps charging).

Any steady Ohm reading would be bad; the lower the reading, the more the leakage. A very low reading indicates a short or miswire.

If this test passes, then:

Obtain a 1K 10 watt resistor. Connect 1 lead to the chassis, the other lead to ground. Connect a DVM set to 200VAC across the resistor. The leakage current = voltage / 1000, and should be less than 500uA (0.5mA). A Class A (residential) GFCI should trip at 5mA.

If the leakage is less than 1mA, then ground both the scope and the gen, then repeat testing the gen for output and continue troubleshooting.

hope that this helps.

[MrFixr55]

OOps, Sorry, Dave, the pic finally painted and your last comment posted while I was still writing.

Definitely do the ground leakage current test that I just described. It is always possible for something to ohm open at a low voltage but to have leakage at power line voltage, causing sparks. You have to have the scope probe grounded to the generator in order for any reading to make sense. Sparks, especially big sparks are bad. Best way to protect the scope and gen is for both to be grounded to the same outlet.

The other thing that worries me is the 8Ohm resistance between the probe output and ground. Ohm that again while changing the "output multiplier". If the resistance does not increase with the multiplier value then something is wrong with that end of the circuit.

[radiorich]

Hello Dave,
I also would use safety capacitors !
Sincerely Richard

[daveone23]

Isnt the chassis and ground the same on this machine? My other 288x that works is an identical machine and it does not have this issue and does not have a 3 prong to ground.
Also when testing the transformer, one side of the 6.3 volts goes to chassis. Shows on schematic.

[462ron]

Just a thought here, those caps C33-C34 are .05mfd. That high of a value can possibly pass enough 60hz current to cause a slight arc when connecting to a grounded piece of test equipment. This would be considered normal. I’ve seen this myself in similar situations. Replace them with “Y” safety caps possibly of lower mfd!

Ron

[daveone23]

They worked for 85 years as .05's and my other machine which operated perfectly has them. Doubt thats the issue. What type of safety caps to you guys use for your line inputs. The only one I could find for my application is the one shown below. Y rating is the one to use because they blow open, right?

[daveone23]

One problem i noticed is there is no real voltage on the 31 caps, 20uf elec.
On the probes, good machine
X1=8ohms
X10= 31 ohms
X100= 600 ohms
Bad machine
X1= 27 ohms
X10= 50 ohms
X100= 1.2k ohms