5E3 Voltages in General

[Delmar Evans]

I've built several of the 5E3 Tweed Deluxe amps and have a question about lowering voltages. Im Using all NOS tubes, (RCA) in the correct locations. 12AY7 12AX7 6V6 and 5Y3 Also I changed the Cathode resistor to 300 ohm. There is always discussion about lowering voltages. I wondered why. U use a 330-0-330 PT and a Hammond OT in my build. My amp sounds great. I thought I would share my readings and see if anyone has any comments .
Line at 122.5 B+ 385 vdc 352 vdc 258 vdc respectively at power filters
V5 pin 4 and 6 349 AC Pin 2 and 8 390 vdc
V4 Pin 3 382 vdc Pin 4 353 vdc Pin 8 22.85 vdc
V3 Pin 3 383 vdc Pin 4 351 vdc Pin 8 22.72 vdc
V2 Pin 1 173 vdc Pin 6 209 vdc
V1 Pin 1 136 vdc Pin 6 138 vdc
Actual MA draw on pin 3 of V3 35.5 (shunt method)
Actual MA draw on Pin 3 of V4 35.3 (shunt method)


The amplifier sounds great , but the math indicates 31.82 ma is 100 percent plate dissipation. The bias calculator indicates I am at 107 percent at 12.9 watts dissipation. The amp sounds great, and I am having no issues.

Back to the voltage scaling question. I have a variac and performed all of the same voltage readings at 110 VAC. With that said, here you go

Line at 110.3.5 B+ 340 vdc 309 vdc 228 vdc respectively at power filters
V5 pin 4 and 6 310 AC Pin 2 and 8 341 vdc
V4 Pin 3 336 vdc Pin 4 311 vdc Pin 8 19 vdc
V3 Pin 3 310 vdc Pin 4 310 vdc Pin 8 19 vdc
V2 Pin 1 157 vdc Pin 6 187 vdc
V1 Pin 1 124 vdc Pin 6 125 vdc
Actual MA draw on pin 3 of V3 29.8 (shunt method)
Actual MA draw on Pin 3 of V4 30.9 (shunt method)

So doing the math again , would the lower voltage be running the amp way under bias at 29 ma? That would be less than 80 percent. I guess the question I have is why bother shaving off 12 VAC from the circuit

Delmar

[Firestorm]

There's no good substitute for running at the "ideal" plate voltage, but in cathode biased amps boosting the cathode R even farther is useful. You can also put a voltage divider on the screens to keep them down as that will also control plate current. Oddly, 6V6 tubes seem to run hot with not much ill effects. Especially NOS.

[pdf64]

The shunt method seems needlessly hazardous and pointlessly inaccurate.
The resistance of your meter when set as an ammeter will likely be significant in comparison to the OT primary, and so the actual current will be higher than the measured current
Just measure the volt drop across the legs of the primary and divide it by the relevant primary leg resistance.
Your voltages seem higher than ideal but if you like it then that's what matters.

[hitchcaster]

put the amp in the dark and see if the plates glow red at all.. after its been on a bit... the 250ohm will bring the voltage down a little, but be harder on tubes... if they aren't red plating you should be ok... depends what tubes your using 🕵👍🏼

[Delmar Evans]

Thanks to all for the replies. As indicated in my post Im Using all NOS tubes, (RCA) in the correct locations. 12AY7 12AX7 6V6 and 5Y3 Also I changed the Cathode resistor to 300 ohm. I changed the Cathode because of high dissipation. This helped , but still a little high. No redplating is occurring. Also , Im using a Fluke 88 in all of my measurements. The comment that the voltages "seem higher than ideal" which ones ? Firestorm commented about bumping up the cathode resistor. I have several to choose from , I could bump to 350 and give it a try. Im still puzzled why everyone speaks of lowering the B+ through various methods, yet when I brought it down 12 volts with a variac, it biased it way too cool. If you lower the B+ , then would you have to change the Cathode resistor ?

[Firestorm]

I disagree with pdf64 that the shunt method is inaccurate, especially with a Fluke. I have three Fluke meters and two Radio Shack and they all read close. When I want a "real" measurement on a single tube, I stick in an SE amp with known cathode resistance. But that tube will perform differently in push-pull. It's hard to force the bias down in cathode biased amps. The tubes will fight you. They want to run where they want to run. Drop the screens. Otherwise, enjoy the sound of the amp and expect to replace the outputs regularly.

[Delmar Evans]

Thanks Firestorm. The screens are already at 30 vdc or so below the plates Should I add screen resistors to bring the voltage down lower on the screens ?

What do you think about the voltages the amp is currently running at ? Should I bump up the Cathode resistor to 350 ohms or so ?

I understand that these things run at 100+ plate dissipation. Just wondering about all of the hype on lowering the voltage VIA scaling, using the VVRunits or diodes on the center tap. Is this typically an issue when using a transformer that is too high of secondary voltage? I just wondered why when I pulled it down to 110 , as all the talk is about 1950's voltage supply, that it is cold on running.

[Firestorm]

These amps were not designed to run Class A. Close. But not really. The other manipulative thing you can do is drop the 220K grid loads to 150k - 180K. If you like it, fine. Otherwise, if the tubes aren't cooking, you're fine.

[martin manning]

Line at 122.5 B+ 385 vdc 352 vdc 258 vdc respectively at power filters
The bias calculator indicates I am at 107 percent at 12.9 watts dissipation...
The amp sounds great, and I am having no issues...

Line at 110.3.5 B+ 340 vdc 309 vdc 228 vdc respectively at power filters...
So doing the math again , would the lower voltage be running the amp way ...under bias at 29 ma? That would be less than 80 percent.

I guess the question I have is why bother shaving off 12 VAC from the circuit?

You would want to lower the cathode resistor value to get back up to ~100% dissipation at idle with the input voltage reduced to 110VAC.

The difference is, regardless of where you set the idle dissipation, the maximum dissipation along the load line will be higher with the input voltage at 122.5, 137% vs. 107% at 110VAC (assuming an 8k load). There will be a difference in output power, tonal differences, and reduced tube life at the higher voltage, but it will probably work either way.

Pete, the series resistance of a DMM in current measuring mode is very low. Even if it were zero ohms, the difference in plate voltage when measuring idle current by shunting the OT would only be a few volts, so the accuracy is good. There is added risk for sure, but it is quick!

[pdf64]

put the amp in the dark and see if the plates glow red at all.. after its been on a bit... the 250ohm will bring the voltage down a little, but be harder on tubes... if they aren't red plating you should be ok

Wow!
I'm not aware that red plating is a controlled tube characteristic, ie it doesn't happen at 150% or 200% etc.
Rather red plating just indicates that the limiting plate dissipation is likely being significantly exceeded.
The absence of red plating does not indicate that the average plate dissipation is within its limiting value.

Pete, the series resistance of a DMM in current measuring mode is very low. Even if it were zero ohms, the difference in plate voltage when measuring idle current by shunting the OT would only be a few volts, so the accuracy is good. There is added risk for sure, but it is quick!

The resistance of a DMM when in current mode can be derived from its burden voltage spec. http://www.fluke.com/fluke/uses/comunid ... denvoltage
My Fluke 189 is 1.8mV/mA on the 400mA range; the checks out well with its resistance when measured with my Fluke 89, of 2.2 ohms.
So, the meter's resistance will likely be in the range 1 to 10 ohms.
OT primary CT to plate resistance may be in the range 15 ohms (Selmer 100 watt), 45 ohms (BF SR), to 150 (5E3 replacement type) ohms.
Yes any change in plate voltage will likely be insignificant, and being pentode / tetrode, probably immaterial.
The issue with regard to measurement accuracy is the parallel current paths presented by the meter and the OT primary.
Such a shunt method could only be valid if the parallel path was insignificant; that does not seem to be the case here.

To me, it seems irrational / poor practice to choose a hazardous, inaccurate method, when less hazardous and more accurate methods are just as practicable.

[martin manning]

DC resistance of a Hammond Tweed Deluxe OT is listed at 150 ohms a-a, so 75 ohms each side. No problem there. Could be an issue in some other cases, but the error is not significant, IMO.

[pdf64]

Sorry Martin, I was editing to add such info.
My 5E3 type OT was from AES; yes, nearly 300 ohms p-p!

To me, the point of getting the VB+ closer to 350 than 400 is that plate dissipation becomes more reasonable, without having to deal with the increased squish that raising the cathode resistor value brings.
Subjectively, tone seems more pleasant too, may be smoother.

[martin manning]

I certainly concur that using primary DCR is the best and safest way to go to measure anode current directly (not cathode current), but to snatch a quick reading the shunt method is ok, as long as you are confident in your technique!

[Delmar Evans]

Thanks to all for your input. Yes I am confident using my Fluke on the Shunt. I have not blown myself up yet. When I brought the line down 12vac to 110 range we ran in the 50's. the voltages throughout dropped 13 percent. (337 on the plates) This gave me bias readings that were way cool, telling me that I would have to change my 300 ohm cathode back to 250 ohm. If I am reading this correctly. I would still like to know, what voltages are optimal at pin 3 and pin 4 of V3 and V4 and coming through the filters , what I call B+ 1,2 abd 3. also , yes the OT is 8 K

Regards.

Delmar