Magnatone 107 / How do I make this 'widowmaker' safe? continued

[poligow]

sorry for the delay, i was on holidays !

i followed your instructions, everything looks ok ...

i made a quick video to show you the amp working, and my weird noise : http://we.tl/A4f8pRDrzb

If the isolation transformer is now in place, test your work by using your ohmmeter and testing resistance between the plug-in terminals of the (unplugged!!!) AC cord and the chassis. You should show > 1M between both line and neutral terminals and the chassis YES , and the safety ground terminal should show zero ohms to the chassis. YES

It this is true, use your ohmmeter again and test the resistance between the (-) terminals of the power filter caps and the chassis. This should read the value of the 170K resistor that remains after you pulled out the 0.05uF capacitor. it reads 270k (looks more 270k than 170k on the schematic) Also test the resistance between the input jack bushing/sleeve and the chassis (should be zero YES ) and also to the (-) terminal of the filter caps, which will likely be 170k YES .

If all that checks out, you can either (1) put the 0.05uF cap back in; it's not a death cap any more or (2) put a wire where it was. I would probably put the 0.0 5uF cap back in and see if it now works OK.

What was bothering me about this was that the input stage was connected to the chassis, but the filter caps had only the path through the 170K resistor to the chassis ground and that was the DC path for the current in the input tube. This still bothers me, but if you have confirmed that the chassis connects to AC power line safety ground, and that neither of the AC power cord power prongs connects to the chassis or to the negative side of the DC supply, then at least you can't get shocked if you've done the wiring right and have the isolation transformer in the unit as you've described.

Still, it bothers me. Do NOT trust this thing until you've made the ohmmeter tests, and then also powered it up and tested for AC volts between the input jack bushing/sleeve and the chassis of some other plugged-in and safety grounded equipment. This should show nearly 0V AC. YES

Something still bothers me.

[poligow]

ok tried this morning to replace the removed cap by a wire; like this :

the big hum and the feedback are now gone.

i tested for AC between the input jack sleeve and a grounded equipment : 0V

if i'm right, i just made a direct connection for one side of the AC line to the chassis :

wich is the same as found is this post

can this little one can now be considered as safe ?

[J M Fahey]

That was not a >death cap<

Why worry when the whole amp is a death trap?

As in: ground is straight one leg of the wall plug, and +V is the other.

To boot, old 110V flat blade plugs could be connected both ways .

Oh well.

So best they could do was to make just the first triode have its own little floating ground, to which are referred both input grid and cathode (no amplification possible otherwise) and float/separate it from actual ground through 170k (for DC coupling, although you can easily find some 30 or 40V DC dropping across it) and the .05uF cap, to allow some grounding at audio frequencies.

Oh well [2]

So in this very particular case, it's fine to replace it by a wire, AFTER HAVING FULLY ISOLATED THE CIRCUIT WITH AN ISOLATION TRANSFORMER.

[poligow]

thanks for the feedback and the explanations JM

to finish the work should i also remove the 270k resistor or leave it as is :


by the way here's a better view of the schematic :

[eschertron]

If you've replaced the cap with a wire (short) then the 270k resistor is doing nothing. Leave or remove. No difference.

It is weird that the signal ground for stage one (in the original) was the chassis, separated by the cap+270k resistor from signal ground for the remaining stages and rectifier negative. I think I've seen low-valued resistors (10R - 50R or so) used like this, but not 270k. I'd like to see the rationale for that value.

[poligow]

sounds good !

thanks a lot to everyone for helping me in this project !

[J M Fahey]

If you've replaced the cap with a wire (short) then the 270k resistor is doing nothing. Leave or remove. No difference.

It is weird that the signal ground for stage one (in the original) was the chassis, separated by the cap+270k resistor from signal ground for the remaining stages and rectifier negative. I think I've seen low-valued resistors (10R - 50R or so) used like this, but not 270k. I'd like to see the rationale for that value.

The modern 10-50 ohm resistors are in modern , properly fed and grounded amps, to avoid a few millivolts of hum which can appear in a ground loop.
We are talking signal here, maybe annoying but harmless.

The widowmaker era high value (270 k here, usually >100k) was meant to avoid it killing you if you got the amp ground connected to live mains ... which would happen 50% of the time.

Statistically playing Russian Roulette with three bullets in the cylinder every time you plug your amp and I am not kidding here.

So they made a (very poor) "floating" ground just for the chassis, first grid and cathode.

I bet they were so used to it that they didn't even notice the hum.