I was hinting in the last post that I might be using a .1 uF Soviet K40Y-9 paper-in-oil cap for the output cap. And so I am.
One thing I try to remember to do on components, especially ones I've had for a while, is to clean up their leads. On this cap I used some 0000 steel wool on the leads to get them nice and shiny. Usually there's a bit of oxidation on them, and they will solder easier and probably conduct better if they're clean.
Here's that cap installed. In an earlier post I mentioned that good wiring practice dictates that grid leads should be as short as possible. Which means mounting components as close to an input grid as you can.
You may recall that the stock wiring had the coupling cap way over by the preamp plate (output) and a single lead running to the grid of the power tube. Wiring this correctly will improve tone and cut down noise and hum.
If you're recapping an amp (or radio) or building one, this practice is easy to follow.
Here's more of that practice in use.
You can see the shielded leads I put in which run from the preamp tube (on the left) to the volume and tone controls.
Also note the coupling cap (an .02 uF K40Y-9) going right to the input side of the volume control, as well as the shielded lead going to that .1 uF cap I showed above.
Before I put the chassis in the cabinet, I wanted to check the bias on the output tube.
This is a single-ended Class A amp and had a 330 ohm resistor as the cathode bias resistor. I took some voltage readings and found that with the RCA 6V6 I had in the amp, the tube was biased to about 110% of its safe dissipation! So I decided to change the cathode resistor.
I used the bias calculator found on Rob Robinette's web site to determine the cathode resistor value. The calculator is super useful for calculating bias, and it even has the max dissipation numbers for many tube types built in. Sure, you could look them up in a tube manual, and do the maths yourself, but the calculator makes it even easier.
From the calculator, we know the max for a 6V6 (GT) is 12 watts. I already knew the stock 330 ohm resistor was too low a value, so I experimented with substituting 360 and 390 ohm resistors in place of the 330 ohm resistor.
The numbers you'll need for the calculations are the plate-to-cathode voltage and the voltage across the resistor. I wanted to get as close to 100% dissipation as I could. For a Class A amp, it's safe to run at 100% dissipation, unlike a Class AB amp where 70% is considered the maximum dissipation.
With a 360 ohm cathode resistor, I had 290 volts plate-to-cathode, and a voltage of 15.3 vdc across the resistor. That calculates out to 40.2 mA of current, and a plate dissipation of 97.5%! Pretty much right on the mark. With a 390 ohm resistor, the dissipation is about 89.2%.
I didn't have a single 360 ohm resistor on hand, so I just put a 330 ohm and a 30 ohm resistor in series. The resistors are rated for 5 watts. A 2 watt resistor would have been fine since there's a total of about 1.2 watts (290 volts x .0042 amps = 1.21 watts) going through the resistor, but 5 watt resistors are what I had on hand. They'll run nice and cool in use.
You can see the two resistors in place in the picture above.
The last thing I did, just for the heck of it, was to check out the lead dress to see if moving some things around changed the noise level.
I did this with the amp on, nothing plugged in, and the amp cranked. Then I did it with a guitar plugged in and the amp cranked.
You can see I use a chopstick to tap connections and move wiring around to see if the noise level changes.
The amp is so much quieter now that nothing changed at all. With nothing plugged in, and my ear next to the speaker (!) there is a very faint hum. I get some noise with a guitar (single-coil pickup) but it's still super quiet. I'm really happy with it. It's noticeably quieter than it was with the old (stock) wiring. I have fluorescent light fixtures over the workbench, and it picked up a lot of noise from those lights before, but now, turning the lights on makes no difference at all.
Now I'll clean the chassis up a bit and put it in the cabinet.
No comments:
Post a Comment