still wondering about redplating

[tedm]

I still wonder why some folks have reported redplating with a 130R, and then the redplating going away at 220R.

Seems like the opposite should occur, higher B+ and higher plate voltages reducing redplating??

[Shea]

The reduction in current more than makes up for the increase in plate voltage to bring down the amount of power that is dissipated.

Say, for example, with the 130-ohm resistor you measure 372 volts from plate to cathode, and a 9.8 volt drop across the cathode resistor. 9.8 volts / 130 ohms = 75 ma total current for both tubes. Multiply that by 372 volts, and you'll discover that the EL84s are dissipating 27.9 watts. That's almost 14 watts per tube.

This is close to the results we've been getting with the Hammond 272FX power transformer with a power supply of 120 volts or higher. Most Sovtek EL84s have been able to take it and deliver decent service, but some have started to redplate in a relatively short time (like yours, sorry).

With a 220-ohm resistor, less current is used. Therefore, there is less voltage drop across the EZ81 rectifier, which has an impedance of about 300 ohms. Less voltage drop across the rectifier means higher B+ voltage. With the same power transformer, we've been seeing plate-cathode voltages around 389. The voltage drop across the cathode resistor might be about 12.3 volts. 12.3 v / 220 ohms = 56 ma. See that? The idle current went down from 75 ma to 56. 56 ma x 389 volts = 21.8 watts, or almost 11 watts per tube. That's within the standard max rating for EL84s.

In the future, I will either use a 220-ohm cathode resistor or lower the B+. Sorry again about that redplated tube, but now you know one way to fix it. Your lightbulb device sounds like it might help too. Another way would be to use a series power resistor to bring down the B+, making sure it is rated to dissipate enough power and it is mounted in a good location.

Shea

[RossH]

Ted,

That is where I was going when I said this in the previous post:

But is isn't all about the plate voltage. It is also about total plate dissipation.

Total power (or dissipation) is equal to: ((PlateVoltage-CathodeVoltage) * PlateCurrent)...not plate voltage alone. Like Shea said, a bigger resistor lowers the cathode current, hence, bringing the total wattage down. The standard EL84 is set for a max plate dissipation (or power) of 12W.

There are a lot of ways to get the plate current. You can measure it off the cathode with the calculation or measure it directly. But the point is, redplating is a result of too much plate dissipation. The power rating, in watts, indicates the rate at which a device converts electrical energy into another form of energy, such as light, heat, or motion. If you give a plate more watts than it can handle than you get redplating.

HTH,
Ross

[tedm]

I understand the issue now. From now on I will take dissipation data as well as the plate voltage data.

If I were to calculate a 14W idle value, since the design is Push Pull AB, am I potentially exceeding the idle 14W when I strum hard?

e.g. After playing for awhile, getting the amp warmed up and caps/ps supply saturated, I can turn off the lights, and when hitting a dynamic power chord, actually see the tubes get slightly brighter. This is subtle, you almost need to be looking at the inner glow from a hole in the plate.

I don't think I mind 14W per tube if that were max, but < 13 might allow for more varieties of tubes and longer life while still getting great tone.

I will probably keep the 220R in there, which I'm guessing will (but will measure soon) is < 12W, but may open the amp this weekend to do some total power dissipation calculations, and may even try a 180R in there to see what the power #s are, and the tone is like.

Ted

[markh]

The complicated analysis is left as an exercise for the student (see Randall Aiken's site) , but in a nutshell, max dissipation occurs in clean territory, and when clipping hard you're actually dissipating less power. The rule of thumb is set the idle dissipation for 70% of max for AB1 designs.

--mark

[tedm]

I didn't get to my 18watt today, but spent several hours redoing my single EL84 AX84 and wanted to run my calculations past you guys:

I didn't have my current (mA) handy, so I only measured voltages from the plate, screen and grid.

After rewiring most of the AX84, and upping the bias resistor to 150 ohms, I'm getting 244 VDC on the plate, 248 on the screen, and 7.32 VDC of cathode voltage.

Are the following caluculations and assumptions correct?

bias current: 7.32 / 150 = 48.8ma ?

power dissipation: (7.32 / 150) * 244 = 11.91 watts?

Another odd thing is that both pentode and triode mode give 244 VDC on the plate, but triode mode gives 239 VDC on the screen (is that possible??), and a cathode voltage of 7.04VDC. However, triode mode sounds much less loud, like half as loud, but calculates power dissipation to 11.45 watts, or almost as much as pentode.

Could be I screwed up on that wiring. Have to check it tomorrow.

[RossH]

Are the following caluculations and assumptions correct?
bias current: 7.32 / 150 = 48.8ma ?
power dissipation: (7.32 / 150) * 244 = 11.91 watts?

The short answer is no.

Refer back to the equation in the earlier reply. You need to multiply the difference between Anode voltage and Cathode voltage by Plate Current. So, in this case that is 244-7.32V. Or, 236.68V. Now lets get the current.

If you are doing it by the cathode calculation you need to subtract the screen current as well (voltage drop across the screen resistor divided by screen resistor value). It is typically 4-5 mA. So:

bias current: 7.32 / 150 = 48.8ma - 5mA = 43.8mA
power dissipation: (236.7) * .0438 A = 10.37 watts

130 ohm cathode resistor will put you right about at 12W.

I don't know about the triode/pentode thing. I just wired mine up as a pentode. By the sounds of what you are saying I would say that it is a miswire...but I'm not positive.

HTH,
Ross

[Jason]

I still wonder why some folks have reported redplating with a 130R, and then the redplating going away at 220R.

Seems like the opposite should occur, higher B+ and higher plate voltages reducing redplating??

ted,

Actually, the tube doesn't care whether the grid sits at a negative voltage or whether the cathode sits at a positive voltage--that is the main difference between the fixed vs. cathode biased setup. The grid needs to be at less potential than the cathode no matter which setup you use. What happens in cathode bias is that current is drawn from the cathode to the plate. Since there is a resistance in the cathode circuit, a positive voltage is developed across this resistor because of the current drawn through this circuit. So what happens is that the cathode gains a positive potential instead of being at 0VDC. The varying AC signal voltage [music] on the grid varies in a cyclic manner. Because the signal is usually a sine wave, the grid voltage fluctuates in increasing or decreasing magnitude depending on the frequency and amplitude of the incoming AC voltage [music]. The tube acts like a valve which is controlled by this AC. The reason this works is that the grid when it is more negative than the cathode tends to keep current from flowing from the cathode to the plate. When the grid approaches 0V potential, the current is more easily drawn towards the plate. When the grid is more positive than the cathode, it has little effect on controlling the current flowing from the cathdoe to the plate and it may even draw it's own current. BTW, now we are touching on operating classes--that's a whole other can of worms!

Jason

[tedm]

Thanks for the extensive reply. I'm trying to understand the details you've provided and reconcile them with my other understandings from the O'Connor book and my own observations. I might have more questions coming, but let me try to continue digesting what I have. Thanks!

Ted