AC30 reissue (PCB) blown Vscreen/preamp fuse

[DonMoose]

Gentlemen -

I have on my bench a PCB version of a Vox AC30 with a blown fuse (500mA) between rectifier and choke. The grid (series) resistors on two of the EL84s are cooked, but read 1K instead of the expected 1k5.

I get solid connections from the cut circuit to the respective grid resistors.
Solid Cathode connections the bias circuit, which reads 50R0 to GND.
Good screen connections, good plate connections.
No unexpectedly low ohmic readings to GND from anywhere in the power amp.

The only oddness I find with power off and a DVM (20M scale, tubes in place) is about 2M across the PI coupling caps - but they read the same.

Apart from a suddenly bad tube, what could cook that Rg like that? There's a charred ring around both middles.

I have to source a fuse - looks like 4x20mm glass - but want to keep the new tubes and replacement components safe.

Thanks!

[TheGimp]

You could be measuring through the grid resistors attached to the PI caps, thorugh the PS, back thorugh the HT PS through the anode resistor of the PI and back to the other side of the coupling cap.

I'd expect both caps to read the same.

[DonMoose]

You could be measuring through the grid resistors attached to the PI caps, thorugh the PS, back thorugh the HT PS through the anode resistor of the PI and back to the other side of the coupling cap.

I'd expect both caps to read the same.

Something like that path, but these are series Rg, not shunt Rg - it's a cathode-biased amp, so there shouldn't be a meaningful DC path from EL84 grid to PI plate unless the EL84 is shorted.

EDIT
*** no - you were right - I found the DC path - both sides of the CUT control have pulldowns to GND, then thru the PSU. THANKS!***
EDIT