I'm a beginner and I've been doing my best to learn some stuff. I've read a couple of tube books. I've read about tube safety. However, there is something I'm missing (I feel stupid, but it confused my neighbor as well and he's an engineer). I was confused about my variac earlier yesterday. When I use my DMM to check for AC between the chassis and the earth on an outlet (before connecting a safety ground), I get 50V and my AC tester screams. I thought I had a weird short in the variac and my neighbor thought so as well. Then, I measured 128V between the PA amp and ground. I pull out my old Scott, and sure enough, I test 50V on it to earth ground. I suppose it makes sense, as the chassis is neutral. Somehow though, I'm not grasping how current doesn't run through one's body to the ground. Also, is their a good way to test for a live chassis, or if the fuse doesn't blow do you just assume you are fine? Obviously I've failed to grasp some basic concept.
pj
pj
pjanda1 said:
That happens. AC distribution is three phase. There would be no problems if every load was a three phase, balanced load. However, it's not, with a whole bunch of single phase loads connected and distributed across all three phases. That means that you get unbalanced loads, despite the best efforts of the power managers. If the load goes unbalanced, the AC neutral is no longer at ground potential. You can sometimes draw some big sparks between the neutral and ground. Test it again later and it'll be different.
There's really nothing you can do about this.
Thanks for your explanation; it makes plenty of sense. It sounds as if checking chassis to ground voltage has no purpose. My bench outlet has a GFI on it, and I assume that as long as I connect stuff with safety grounds it will identify any short. I also assume that live chassis conditions aren't that common.
pj
pj
Another possible explanation is that typical AC noise filters are formed by a capacitor from line to earth, another from neutral to earth, and a third one between line and neutral. There may be also some inductors, forming a low-pass filter with the capacitors. Now the capacitors to the earth (chassis) form a voltage divider so that the chassis is at half voltage between line and neutral, if the chassis itself has no direct reference to them. If the variac has an isolation tansformer then it could happen that the ground and the line/neutral are floating related to each other, and the line/neutral AC potential is fixed only by the above mentioned capacitive divider. So the chassis will be around half AC voltage.
In a US residential situation I'd never expect to find much difference between ground and neutral. They should be connected at the breaker box, and to a ground rod close by.
Whether or not you see small filter caps between hot and neutral or ground, there will always be stray capacitance. Transformers can have quite a lot. Since the chassis isn't connected to anything that provides a load or "drains it off", it doesn't take much stray capacitance to keep it at a surprisingly high voltage. Remember, AC will pass through a capacitor. Any out and out leakage resistance will also do the same thing. It's pretty typical to find 20-80VAC on the chassis of older non-grounded equipment. It's not a good thing! Always use a safety ground to keep chassis' at ground potential and to provide a safe path for fault currents, should something go amiss.
Remember that almost all Variacs have a single winding and provide no isolation. Keeping that knowledge in mind at all times can save your life.
Whether or not you see small filter caps between hot and neutral or ground, there will always be stray capacitance. Transformers can have quite a lot. Since the chassis isn't connected to anything that provides a load or "drains it off", it doesn't take much stray capacitance to keep it at a surprisingly high voltage. Remember, AC will pass through a capacitor. Any out and out leakage resistance will also do the same thing. It's pretty typical to find 20-80VAC on the chassis of older non-grounded equipment. It's not a good thing! Always use a safety ground to keep chassis' at ground potential and to provide a safe path for fault currents, should something go amiss.
Remember that almost all Variacs have a single winding and provide no isolation. Keeping that knowledge in mind at all times can save your life.
What really matter is CURRENT, not voltage. After all, you can build up 15KV by shuffling across the carpet, and you ain't dead, right? The safety limit for AC ground leakage current varies, depending on which standard is used, but most say 1 mA or less. If you can measure AC current, do that instead of voltage. Or use a 1K resistor and measure the voltage drop (1V drop = 1 mA).
Sometimes there's a capacitor from one side of the line to ground - in this case, the ground current will vary depending on which way the AC plug is flipped. If there are two caps, it will be the same either way. 1 mA is considered the "startle" threshold, where it could make you jump or drop something. If there's more than that, find the responsible capacitor and change it to .01 or .015 uF.
As mentioned above, a 3-wire cord is the best way to go, and would be required by safety standards for anything like this built today.
Sometimes there's a capacitor from one side of the line to ground - in this case, the ground current will vary depending on which way the AC plug is flipped. If there are two caps, it will be the same either way. 1 mA is considered the "startle" threshold, where it could make you jump or drop something. If there's more than that, find the responsible capacitor and change it to .01 or .015 uF.
As mentioned above, a 3-wire cord is the best way to go, and would be required by safety standards for anything like this built today.
It was the current reading that worried me too. Though I'm not sure I trust it in that range, my DMM was showing 10ma. I'll have to look for that capacitor. I will use a 3 prong cord. I know that ground loops can be an issue in PA/MI setups, so I'll probably use a resistor to keep the potential a little above earth. That is the only solution I found on my chip amp, as I had a loop with the cable connection to my TV.
pj
pj
Ground loops can be sorted by (presuming all the chassis and common earth points are linked) only grounding ONE of the apparatus via the plug, and connecting the rest of their metal chassiseseseses (?) together with ground strapping, etc. Thus it only returns to earth at one point and minimises/eradicates a ground loop from forming as all the devices are at the same potential
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.