- 2025
- Apr
- 7
I’m not going to “just ground it!” because the Internet says so.
Recently, I had someone on a forum quite emphatically indicate to me that I was doing it wrong, and that I should ground (earth) everything - also indicating that I should just use IEC connectors and cut holes in irreplaceable chassis units and simply “Make some paint.” I then got the reddit response of “Gulp!” and “Who told you that?” when I said I just use isolation and let things float, and that no holes shall be cut.
Well, thank you - but I’m not going to do that. One, I don’t want to destroy equipment. If you want to after I’m dead, that’s your decision.
But also, I’m not going to ground everything. You don’t ground everything because you can. This was in relation to the Heathkit AF-1 that’s I’m currently rebuilding. It has some quirks that don’t really allow that to happen. And, I learned a different way. Just because the internet tells you that every capacitor is bad, and that every device has to be earthed, doesn’t mean you do it.
Who told me that you don’t ground everything? Two people, actually.
The first gentleman was a rusty crusty old military man that spent his second career working for Rockwell before they shut his shop down in the 90s. Mr. Lehman was the maintenance guy for electronic controls, and had seen it all - everything from tubes to modern (for the time) DCS. I got involved with him because he was a friends father, and I had an amplifier that was floating 70VAC above ground. You’re probably familiar with this, it’s either because the grounding is improper or something is leaking. When I took this amp to him in the early 80s, he poked at it for a moment and said “This stuff wasn’t necessarily built well. There were rules, but they didn’t always follow them because they didn’t have to. You’ve either got something in here leaking or it’s just built that way. The stuff it was supposed to be connected to didn’t care, that’s why it’s all painted with plastic knobs.” We went over the device, he related how to check things to make sure that you COULD properly earth this device - in ths case, the transformer wasn’t leaking, it had a transformer, and you weren’t going to cause ground loops because earth and input were at differing potentials. He taught me about making sure that nothing in there is going to leak a lot of current if you ground it, and in this case it was just that the grounding in the device was crap. He said that in this case you can and should ground it, but you always need to make sure that you’re doing it for the right reasons - not just because you have an earth/safety ground.
Second, was my first electronics instructor. Mr. Norman had been around in electronics almost as long as there was electronics to be in. He too was a military guy that did his WWII term and went back into consumer electronics when he was discharged, having been in that same field beforehand. He was the local guy that RCA, GE, and the others gave the warranty projects to. He saw the peak of radio and the rise of television, and had more interesting stories than you could remember. One of the first things we did in his class was build an AA5 from one of those kits you used to be able to buy. Of course, they had been used and abused over the years, so we had to be careful - and this was before you had an Intenet source for tubes so all of the 35W4 had been replaced by diodes and a big honkin’ stack of resistors. (He said don’t do this because you strip cathodes, but for this we don’t care - you’re the last class that’s going to use these.) But those could be hot chassis sets, and they didn’t have polarized cords. I didn’t understand that at first, but it was because he was teaching us good practices. He showed us what happened if you accidentally touched or grounded a chassis that could be hot, and said you’ll not always have the ability to replace cords on customer devices with polarized cords. (This was when we still repaired TVs and stuff.) Best practice would be to put a polarized cord on that, but you don’t always have that option. (All of the stuff I get has a polarized cord replacement, even if the old one is still good.) He showed us what happened if you got differing potentials on things, and the damage you can cause if you’re trying to talk to something that doesn’t like being earthed. He taught us that you should ground things when you know you need it, and in some cases you’ll need to lift the ground. Again, he (in much more colorful terms because it was all guys) told us that you should always make sure you know WHY you’re grounding, not just because you can.
I learned about lifted grounds the first time I burnt out a very expensive isolated transceiver loop in a device I’d acquired.
That brings us to this chassis. Let’s say I ground it to earth.
In the manual, Heath warns us that the signal return is actually chassis. This is specifically called out as a note. It doesn’t say why, but it’s important. You’re expected to know why at this point, because this was considered to be an advanced kit and didn’t even include assembly instructions, just some guidelines. Why did they warn us? Because you could potentially see AC on the chassis. You could see high voltage DC on the chassis. Not from this device itself, but from the UUT. This device is transformer isolated.
DC return is also chassis. That doesn’t really matter for this case.
If I chassis ground this device (and no, it does not have the ground hole in the back and should not be grounded) then I run the risk of shunting something that doesn’t like to be grounded directly to ground because signal return is chassis. That’s the takeaway here. This device was not designed to be operated like this, and could be dangerous doing so.
The isolation transformer I use for all my testing is there so you don’t leak back to line.
I understand that some are not comfortable working with ungrounded equipment, and that’s ok. We all have the methods we use. My methods were taught to me in a world where you are absolutely going to run across things that leak, have hot chassis, or worse, and you may not be able to do anything about them. I was taught how to mitigate that, and I’d rather have all of my work floating hot on an isolated line than run the risk of accidentally connecting signal to ground and destroying things because I grounded something that wasn’t designed to be grounded.
That’s where I get my methods.
But also, some of the equipment I have could be grounded, if needed. Look on the back of a IG-72 signal generator. See that little hole by the cord, how it looks just the size of a Heathkit screw? That’s for a ground - IF YOU NEED IT, not just because you can - and you have to understand that it’s also using chassis for signal return. That’s the takeaway here. Don’t do it just because you can, make sure you’re doing it for necessary reasons.
The long and short here is this chassis is a good example of why you don’t just do the one-size-fits-all thing. In this case, grounding a chassis signal return could provide you with some sparks. Always be aware of WHAT you’re doing, and WHY you’re doing it.