(Slow) Build log for Vox JMI-era AC15 OA-031 circuit

[dchang0]

Nice! Thanks for the link and tips!

[dchang0]

This may be an easier-to-read drawing of the B+ on PT center tap fuse and diodes-on-rectifier mods:

https://robrobinette.com/5F6A_Modifi...rotection_Mods

A question: Merlin's recommendation on B+ PT center tap fusing is to put TWO fuses in, one on each half, rather than one on the center tap that they share like in the last schematic on this page:

The Valve Wizard

From what little I have read so far about the difference, there seems to be some situations where the one fuse in the center tap doesn't protect against certain failure situations where two fuses on the legs does. Is this correct? Is it worth it to put two fuses as Merlin suggests? (Probably not, in my case, as finding a place to put the two fuses where they are accessible is probably too challenging. I notice that modern-day JMI Amplification puts one B+ fuse in a new fifth hole drilled between the two chassis halves. Pretty clever location.)

[pdf64]

The fuse on the CT doesn't protect the B+ winding if the rectifier shorts plate to plate.
Adding the diodes between the winding and tube rectifier plates should prevent such a failure damaging the PT.

[dchang0]

If I understand Merlin's first diagram on the aforementioned page (with the unhappy face) correctly, then if BOTH diodes fail AND the rectifier fails, the center-tap-fuse would not protect the PT. Is that correct? (An ultra-rare failure event, to be sure. This hypothetical is posed out of curiousity.)

If it helps others, here are the two diagrams from Merlin's website:

Fuse on each half/leg:


Fuse on center tap:

[dchang0]

Had to drill two holes at the edge of the steel chassis for the B+ fuse block, to the right of the PT. This fuse will be visible and accessible from the side if the slider board is pulled out far enough.

[pdf64]

If ... BOTH diodes fail AND the rectifier fails, the center-tap-fuse would not protect the PT. Is that correct? (An ultra-rare failure event, to be sure. This hypothetical is posed out of curiousity.

I think that if either Si diode and the tube rectifier failed, then the CT fuse wouldn't help; so a slightly higher likelihood but still hopefully low
So it would be a good idea to use Si diodes with a sufficient voltage rating; for this amp, 1kV should be fine.
Though if the PT B+ winding voltage may exceed 350-0-350 (consider unloaded at power up, high mains) then a higher voltage rating is required.

[dchang0]

Ah! Good point--thanks for the precise explanation. I will be ordering UF5408 diodes rated at 1kV and 3A. I heard that the 1400V-rated diodes Fender uses are quite "noisy in audio applications." More important, they are hard to source and cost ~$5 each, ten times as much as the UF5408.

[pdf64]

As the Si diodes are in series with the tube diodes in this application, we don't need to be concerned about switching noise, ultra fast types etc; the smooth switching characteristics of the tube will control the current flow.
A single 1N4007 in series with each plate would be fine, make it a series pair on each plate if you really want to have belt, braces and another belt.

[dchang0]

Cool. I'm not at that level of caution, but two diodes per leg is still ridiculously cheap insurance...

[dchang0]

Made a little progress on the steel chassis. Added the passive components in roughly the same manner as the vintage layout researched and drawn by Turret over at Amp Garage forum (sort of point-to-point). It was a pain getting the Z-axis arrangement down--I probably will follow the current-day JMI-15 approach next time (using longer terminal strips), although it is pretty clear that current-day JMI has their chassis drilled in much different locations than the vintage holes.

I didn't go crazy and install 12W wirewound resistors, though. Just used a 3W 22.1K metal film for R33, 5W 100R wirewound for R24 and R25, and 5W 130R wirewound for R22. R21 and R23 are the usual 0.5W 1.5K.

NOTE: On V4, I had to use a 4-40 0.25" screw on the side of the socket under the criss-crossed resistors to minimize the chance of shorting. The other tube socket screws are all 3/8" long.





Not shown: the extra potentiometer flat-nuts arrived, so I was able to get the proper height on all the pots.

[Tom Phillips]

A suggestion:
You should allow for some physical spacing between the power amp cathode resistor and the cathode bypass cap. The goal is to minimize heating of the cathode bypass electrolytic cap by the heat given off by the resistor. Less heating = longer cap life.

You are referencing designation numbers for the components in your narrative. It would be helpful if you would post the version of the schematic that uses those exact designation numbers.

[dchang0]

Thanks, Tom, for the good advice. I pried them apart a bit--maybe 1/8". Each component is now airgapped in all directions.

The schematic is attached here. It is the original OA-031 that I relettered a few months ago with help from members of this forum.



This is the layout that Turret over at Amp Garage drew up. I can't find the original he posted (turretboard.org is gone), so I am attaching it here. It is drawn from an actual vintage JMI-era AC-15 and is true to the schematic except for R38 (the amp he examined uses a 2k2 resistor, while the schematic says to use a 1k5).

voxac15layoutrev4_112.pdf

[dchang0]

Question about using tube socket pins for the "insurance" diodes...

Hi, guys--

I remember reading a long time ago an exchange between an amp builder and an known-expert builder. The first builder was using unused pins on a tube socket as a convenient terminal to solder other wires/components, and the expert said this was a very bad idea because a guitarist could accidentally plug in the wrong tube into the wrong socket and blow something up. I had a hard time envisioning exactly what was being argued, but the gist of it was that inserting the wrong tube would cause a new circuit path to appear, which would then burn some valuable thing up.

So, I'm asking for your help in laying out how to do the double-diode or single-diode "insurance policy" on the EZ81 rectifier as described by pdf64.

I've attached two diagrams with my proposed diode layout--are these dangerous/bad if a wrong tube (most likely an EL84 but also possibly some other 9-pin rectifier) is inserted? I am assuming yes, since I'm unidirectionally-shorting an awful lot of pins out. If they are, what I'll do is fall back to the safe method of not putting the diodes across any pins at all (just go with single diodes soldered inline with the wire with heat shrink to protect the solder joint between the wire and diode).

Of course, it may be very bad to insert an EL84 in the EZ81 socket even without any diodes in place, in which case the answer becomes, "DON'T INSERT AN EL84 IN THE EZ81 SOCKET!!!" But the diodes probably make things worse.

Fuse on center tap of PT secondary is not shown.





Thanks in advance!

[Leo_Gnardo]

Having "insurance" rectifiers is an excellent idea. My practice is to attach them to a separate tie strip, not the tube socket. I'm more worried about hazard of an arc with components crowded so closely together on a socket. Breakdown voltage of air is 1000V across 1/8 inch or 3mm. Remember that's instantaneous voltage, and under "ideal" conditions, 0% humidity at sea level pressure, & no dust accumulation on your parts. Give your hi voltage parts some elbow room, more cheap insurance, that's my tuppence worth.

[dchang0]

Advice taken--I'll forgo using the socket pins. I don't particularly feel like drilling another hole in the chassis for attaching a terminal strip, so I'll use the inline approach and single diodes. Thanks very much!