Interesting discovery on my V18 (they screwed up!)

[Steve A.]

I just rewired the PI and output section of my V18 to Vox AC30 values and was checking for solder bridges and open solder joints when I found that they miswired the 470k grid loads on the EL84 output tubes (these resistors mislabeled as R71 and R98 on the schematic would be the bias splitters in a fixed bias amp).

The grid loads are normally wired at the junction of the coupling caps from the phase inverter and the grid stoppers going to the output tubes (as is shown on the schematic). However on the circuit board itself the grid loads are wired directly to the grids. So that the signal from the coupling caps goes first through the 10k grid stoppers (mislabeled on the schematic as R73 and R96) and then to the grids, with the grid loads wired to the grids.

Obviously in a fixed bias amp a mistake like this could be a total disaster but it kinda works in a cathode biased amp so evidently no one else has noticed it... The fix is simple- pull out the old grid loads and rewire the new ones so that lead nearest the grid stopper is connected to the *other* terminal of the grid stopper. There is plenty of room for that "tack joint" on the component side of the board.

Steve Ahola

P.S. If anyone here can enlighten me as to the effects of the miswiring the power section, it would be appreciated! I'm going to correct it right now because I'm trying to make this amp more like what we call "a real amp"...

Here's a link to the V18 schematics (I still need to add 12 corrections to the PI and power amp section):

http://www.blueguitar.org/new/schem/...d_drawings.pdf

P.P.S. I've attached a photo of the "fix" for this problem- note the blue heat shrink tubing on the extended leads from R45 and R46. BTW the identifying marks on the audio board are "OASIS 03V-0 E99661 99V-0 0726" (If your numbers are different your board might not have this problem.)

[tubeswell]

Maybe you should let the factory know? (or maybe not ;-)

[Steve A.]

Maybe you should let the factory know? (or maybe not ;-)

I guess that might void my factory warranty, wouldn't it? Then again I think I have already replaced 75% of the parts in the audio signal path so I guess that they could tell that I'd worked on the amp.

I think that Enzo is a factory service station for Crate so perhaps it would be better for him to explain it them...

Steve

[Enzo]

Customers carry more weight than service centers with most OEMs.

I don't see either way as a disaster. As long as the resistor is near the socket, it is in series with the signal and the tube capacitance, isn;t it?

[Steve A.]

Customers carry more weight than service centers with most OEMs.

I don't see either way as a disaster. As long as the resistor is near the socket, it is in series with the signal and the tube capacitance, isn;t it?

Well, if it was a non-adjustable fixed bias amp, would it not make a difference in the bias voltage that the output tubes were seeing?

In any case, I don't recall ever seeing a schematic drawn as this output section was actually wired up- if it didn't make a difference, I think that you'd see it both ways, a lot more often.

Actually I was more concerned about them using 470k grid loads since the higher signal seems to really slam the EL84's. Since rewiring the PI and output section closer to Vox AC30 spec the amp sounds a lot better to me. I did voice it for a Eminence Red, White and Blues speaker (which I put in another cab) and it was too bright when I tried two different varieties of Greenback. I have a Weber speaker that is like a copy of the old C12N and that does sound a lot like the Eminence RWB...

Now for a trip to the hardware store to pick up some 1" angle iron to reinforce the back of the chassis (there is one remaining rattle/buzz- this one at a low A and a low Bb. I think if I reinforce the large metal surface of the chassis that will solve the problem... knock on wood.

Thanks!

Steve

[Enzo]

Well, if it was a non-adjustable fixed bias amp, would it not make a difference in the bias voltage that the output tubes were seeing?

No. The DC bias voltage applies a potential to the grid, but no current flows. Only way to get voltage drop across a resistor si if current flows through it. Besides, if the bias voltage were coming through that 470k resistor, adding 10k or not amounts to about a 2% change in resistance anyway.


The stoppers want to be as close to the tube as possible. In the Fender layout we find so commonly, that means mount the resistor on the socket, while the grid resistors are back on the eyelet board. Much more difficult to mount the grid resistors to the socket on an old Fender chassis. Here, the sockets are right on the circuit board, so there is less layout pressure.

[Steve A.]

No. The DC bias voltage applies a potential to the grid, but no current flows. Only way to get voltage drop across a resistor si if current flows through it. Besides, if the bias voltage were coming through that 470k resistor, adding 10k or not amounts to about a 2% change in resistance anyway.


The stoppers want to be as close to the tube as possible. In the Fender layout we find so commonly, that means mount the resistor on the socket, while the grid resistors are back on the eyelet board. Much more difficult to mount the grid resistors to the socket on an old Fender chassis. Here, the sockets are right on the circuit board, so there is less layout pressure.

OK, I concede the point. However, if a similar error was found in other areas of the circuit, for example involving a voltage divider, it could make a big difference. But not here...

Thanks!

Steve Ahola