Questions Regarding Proper/Safe use of Isolation Transformer and Current Limiter

[dankinzelman]

Hi everyone,

beginner questions here, but maybe this will be useful to someone besides myself. I am going to state what I understand to be safe practices in a number of cases. BECAUSE I'M A BEGINNER, DO NOT TAKE THE FOLLOWING AS ADVICE, BUT RATHER AS AN INVITATION TO DISCUSS AND LEARN.

Please correct my errors or tell me if I am getting this right:

0. Valve amps have very high and potentially deadly voltages inside. Always ensure capacitors are safely discharged before working on the amp, if possible through a resistor to avoid arcs. Never touch the inside of the amp with both hands at the same time unless you are absolutely sure there are no dangerous charges left in any of the circuitry.

1. Dim bulb current limiter should be used when powering up untested devices, newly modified devices, devices which have blown fuses, or in situations where there is suspicion or risk of internal shorts. Once it's clear that the device is free from short circuits and doesn't seem to be consuming excessive current, it is no longer needed and should not be used, because it will reduce line voltage available to the device, falsifying internal voltages and oscilloscope readings.

2. When using oscilloscope, DUT should be powered by ungrounded isolation transformer with no current limiter. Probe ground clip should be the first connection made to DUT, and that connection becomes ground reference for DUT. First ground clip can be safely connected ANYWHERE on DUT. For subsequent probes, ground clips may be left unconnected. If used, they should be connected ONLY to a position with identical voltage potential to that of first ground clip (ie, in the exact same connection as the first ground clip, or in a very nearby position with perfect continuity to first ground clip). All oscilloscope connections should be made with DUT powered off and caps discharged.

3. When using a mains-powered signal generator, the signal generator should also be powered from the same isolation transformer as the DUT. Both devices will receive ground reference only through first oscilloscope probe ground clip (see point 2).

Is all this correct? Can I actually safely connect the ground clip to (for example) the B+ on an amp if it is floating?

Also, if the DUT is known to be in good condition and I am certain I won't get my probes near high tension areas, can I safely use the oscilloscope without floating the DUT? For example, can I connect across a dummy load to measure distortion, or look at line-level output of a preamp without the iso transformer? Or should it always and invariably be used?

Thanks in advance.

[Ralph5474]

Dan;
A couple of important safety tips.

First thing about working with tube amplifiers (and higher powered solid state amplifiers), is you are playing around with high voltages. If you touch the wrong spot in the amplifier circuitry, at best, you will receive a nasty jolt and at worst, DEATH!

First thing, is to make sure the chassis is at earth ground potential voltage wise when energizing the device. In so many words, make sure the ground pin is intact and properly connected when connecting the amplifier to power.

Do not, I repeat ***DO NOT*** connect the ground clips of your oscilloscope, signal generator, analyzer or other device to anything that is not chassis potential. Connecting the ground clip to B+ would be potential suicide to both equipment and yourself.

The "use only one hand while standing on a dry wooden floor" rule is very helpful in preventing injury and possible death. You do not want electrical energy flowing through your chest (and heart).

BEWARE! - Some vintage electronic devices actually connect the chassis to line potential.

Those safety statements apply whether or not you are using an isolation transformer.

[Merlin]

2. When using oscilloscope, DUT should be powered by ungrounded isolation transformer with no current limiter.

You don't *have* to power the DUT from an isolation transformer; just make sure the DUT chassis is connected to wall ground, then you can safely connect the scope ground clips to chassis. This will be true for any modern amp with 3-prong power cord.
It is only with unmodded vintage amps with 2-pin power cords where you have to engage brain and either mod them to have a grounded chassis, or use an isolation transformer.

Is all this correct? Can I actually safely connect the ground clip to (for example) the B+ on an amp if it is floating?

Yes.

Also, if the DUT is known to be in good condition and I am certain I won't get my probes near high tension areas, can I safely use the oscilloscope without floating the DUT?

Yes. If DUT chassis is connected to wall ground, then the scope probe clip should also go to chassis and you don't need an isolation transformer.

[acheld]

What is a DUT?

[Dave]

Device Under Test

[acheld]

2. When using oscilloscope, DUT should be powered by ungrounded isolation transformer . . .

I dunno.   I've heard that before, but honestly, while I do have an isolation transformer sitting above my bench, I haven't used it in ages.   My 'scope has not complained.   Yet. 

My bet is that most members of this forum do not use an iso transformer much, if at all . . .  Not to say there is not a place for them on the workbench.

[dankinzelman]

Thanks everyone, I appreciate the tips. I am going to integrate the one-hand suggestion into my original post, as well as a statement about discharging caps before touching anything.

I have an amp which runs B+ at ~740VDC. Since my scope probes are rated to 600VDC I am hesitant to connect them to the plates with the ground probe connected to the chassis. Based on my understanding, it occurred to me it might be possible to float the amp, then connect oscilloscope ground clip to the screen grid (+340VDC) before probing the output primary winding. This way the probe would only see the difference between plate and screen grid (~400VDC), so it should be ok, right?

But after thinking about it a little more, it seems to me that the screen grid then becomes ground referenced, so it would mean I should regard the chassis as potentially lethal since it will be at -340VDC relative to ground, right? In this sense, my initial idea seems pretty stupid, so I don't actually plan to do this. But I want to know if my understanding is theoretically correct, or if I have misunderstood how this works.

[pdf64]

I have an amp which runs B+ at ~740VDC. Since my scope probes are rated to 600VDC I am hesitant to connect them to the plates with the ground probe connected to the chassis. ...

Scoping output valve anodes is a bad idea, they swing +/-kV.

If absolutely necessary (present your case), scale things down, feed the probe via a high voltage potential divider.

The resistors in the potential divider will need to have suitably enormous voltage rating.
eg several resistors rated for 1kV each in series.

[dankinzelman]

Thanks pdf64.

My use case is that I'm working on an amp which sounds like crap and plays really quiet. The signal at the power tube grids looks good to me, but the output across my dummy load looks really strange, with distortion that looks nothing like clipping to my (admittedly untrained) eye. This is with a new pair of matched EL34s biased at ~17W, but they can't manage to make anything that looks remotely like a sine wave at more than a few watts into 8 ohms.

So I was wondering if looking at the primary side of the output transformer would be instructive.

[mresistor]

What is this mystery amplifier?

[dankinzelman]

Davoli C3 (Italian). I will start another thread to discuss, for now schematic is attached.

[HotBluePlates]

... The signal at the power tube grids looks good to me, but the output across my dummy load looks really strange, with distortion ...

The schematic shows a bias voltage of -33v.
   1. First step is measured dc volts at the output tube grids to determine what bias voltage you have present for the idle current you have set.
   2. Second step is feeding a sine wave to the amp, but measuring the peak volts at the output tube grids.
       - Say you have -33v bias, you cannot drive with more than about 32v peak & expect clean output from the amp.
       - If your DMM measures RMS Volts for AC and doesn't have a peak-reading function, convert:  32v peak x 0.7071 = 22.6v RMS
   3. When setting up a test signal, check that it's not too hot at the input jacks (probably <100mV peak), then adjust Volume.  The schematic shows a shared master volume that doesn't appear present in the photos I've seen of a C3.

That amp has feedback from the speaker to the driver stage ahead of the output transformer.  So if driven to distortion, it might look strange just due to the feedback.  If you have to eliminate this as a variable, unsolder one side of the series 33kΩ feedback resistor.

[dankinzelman]

Thanks HotBluePlates. I'll start another thread to discuss this since I don't want to derail the safety discussion.