Biasing AC-30 amps?

[Mark]

Hi

I’m working on a 70’s Vox AC-30TBR amp and the original 47 ohm resistor was in the amp. The voltage on the anode is 350VDC and the current through the valve is 55mA that calculates as a 12 watt device dissipating 18.7 watts. I have found in the past an 82 ohm resistor or greater provides the best bias.

What puzzles me is why wasn’t the resistor changed out years ago. We don’t have a problem removing resistors used to bias fixed bias amps but cathode bias resistors don’t get changed to a more appropriate value.

I’m wondering if I’m missing something by biasing the amp correctly?
Does the amp sound better if the output valves are running beyond spec?

I have included the link as it’s good info on the 70’s AC-30TBR though it has some mistakes. The reverb on these amps is very sensitive to noise.

https://ampgarage.com/forum/viewtopic.php?t=38103

[nuke]

Because that's what Vox put there. Vox always ran the snot out of the output tubes. It's part of the sound, the aura, the mythos that is Vox.

At the cathode, you'll be seeing the sum of the plate current and screen current, times 4. And of course, subtract out the cathode voltage (~10-12v) from the plate voltage (330v-ish). Figure about 16ma (4ma each) for the screens, the rest is plate current. Still a lot.

The earliest drawn schematic I know of did show an 80-ohm cathode resistor. The AC15's usually had a 130-ohm cathode resistor, with somewhat reduced plate supply.

But by the 60's, the tubes were likely better than the old original 6BQ5/EL84 12-watts of plate dissipation.

And tubes were cheaper and easy to find then.

[Mark]

Good point about the screen current, it typically is higher than other valves types. I can determine the screen current from the voltage across the 100 screen grid resistors.

[pdf64]

I think the issue with these amps is that the HT voltage is too high for the design of the output stage. My approach is therefore to add a sag resistor, in this case between HT winding and the FWB rectifier. As I recall, 100R 50W aluminium clad bolted down to the chassis worked well last time, it got the HT down from about 370 to 325V DC, and the idling anode dissipation became reasonable.
Dunno if the end result is more or less saggy than a real vintage arrangement but it sounded great, still incredibly loud and not particularly saggy
Yes the reverb is noisy. I found that the biggest contributor was the footswitch cable for the reverb; disconnecting that at the amp end tagboard reduces buzzy hum a lot.
The owner was fine with that.
It would probably be worth trying the amp with Greenbacks or alnico blues, as I suspected the Vox badged blue cover ceramic magnet speakers were probably too efficient with a nasally tone (maybe Fanes).

[nuke]

Just for fun, put a voltmeter on the heaters and use a variac to dial in exactly 6.3VAC. Note the resulting mains voltage to get the heaters exactly at 6.3vac, with all tubes in and operating at idle.

I kind of wonder if the UK power windings are more conservative than the US transformer windings.

My original 5F6A Tweed Bassman likes about 110vac, and runs pretty high voltage on modern 120-125 vac. I built a voltage bucker to use with it and keeps everything a little more sane.

[pdf64]

... What puzzles me is why wasn’t the resistor changed out years ago. We don’t have a problem removing resistors used to bias fixed bias amps but cathode bias resistors don’t get changed to a more appropriate value. ...

The problem with increasing the value of cathode bias resistor (in a AB output stage) is that the bias voltage across it can increase significantly as signal increases, perhaps doubling when cranked.
That has various consequences, I can't recall any amps with a good rep that operate like that.

That's why if an AB output stage design requires a colder bias, eg due to high voltage on anode and screen grid, fixed bias is almost invariably used.

[Littlewyan]

That HT is a bit high for an AC30, I think the older ones ran around 320V. Let's hope you have a good amount of screen current to explain why the bias reading is so high, otherwise I'd be looking to install some sag resistors seeing as it's using a bridge rectifier instead of valve.

I often find manufacturers use high voltages for EL84s and I really don't understand why.

[Dr Tony Balls]

I often build EL84 based amps and cant always find a power transformer that suits my needs. If that's the case i'll often shoot a little high on the secondary voltages and then drop the B+ with a MOSFET B+ reducer to get it in a range i'm happy with.

[jbrew73]

That HT is a bit high for an AC30, I think the older ones ran around 320V. Let's hope you have a good amount of screen current to explain why the bias reading is so high, otherwise I'd be looking to install some sag resistors seeing as it's using a bridge rectifier instead of valve.

I often find manufacturers use high voltages for EL84s and I really don't understand why.

What would be a good starting point for sag resistors? Would you put them on the a/c or dc side?

[Littlewyan]

Depends on what's possible with the room you have or the layout.

AC side - Gives the diodes some protection if there's an inrush of current and you can split the resistance in half.

DC side - Only one resistor required but you'll need twice the wattage.

Ball park figure from the EL84 datasheet is 228mA total for 4 x EL84s (Screen and anode current max dissipation). If we use 100Ohms resistance that drops the HT by 22.8V and dissipates 5.2W into the resistor. If we use resistors on the AC side we can get away with 2 x 47Ohm 5W resistors. If we go with the DC side we'll need a 15W 100Ohm resistor and unless we can panel mount it, we're getting big!

I also haven't bothered to include the preamp valves in my calculations so we want a bit of wiggle room which is why I've gone with those values. We could just increase the cathode resistor but as the bias gets colder the HT may increase due to less current draw and the amp may not sound as good. The issue in guitar amps (Particularly with an overdriven power amp) is as you drive the amp harder, because of how cathode bias works it biases the valves colder. In hifi world or clean amp world, not so much of an issue but in distortion land we push beyond those limits and end up with an even colder biased power amp.

[Dr Tony Balls]

not to harp on it, but here's a recent implementation of a B+ dropper (on the left wall). Tiny, cheap and effective.

[Mark]

not to harp on it, but here's a recent implementation of a B+ dropper (on the left wall). Tiny, cheap and effective.

That is small which is good as there is a lot of real estate in some amps. I typically worry about heatsinks for such devices, I’m not sure the chassis can adequately keep the mosfet cool.

[Mark]

Depends on what's possible with the room you have or the layout.

AC side - Gives the diodes some protection if there's an inrush of current and you can split the resistance in half.

DC side - Only one resistor required but you'll need twice the wattage.

Ball park figure from the EL84 datasheet is 228mA total for 4 x EL84s (Screen and anode current max dissipation). If we use 100Ohms resistance that drops the HT by 22.8V and dissipates 5.2W into the resistor. If we use resistors on the AC side we can get away with 2 x 47Ohm 5W resistors. If we go with the DC side we'll need a 15W 100Ohm resistor and unless we can panel mount it, we're getting big!

I also haven't bothered to include the preamp valves in my calculations so we want a bit of wiggle room which is why I've gone with those values. We could just increase the cathode resistor but as the bias gets colder the HT may increase due to less current draw and the amp may not sound as good. The issue in guitar amps (Particularly with an overdriven power amp) is as you drive the amp harder, because of how cathode bias works it biases the valves colder. In hifi world or clean amp world, not so much of an issue but in distortion land we push beyond those limits and end up with an even colder biased power amp.

Thanks for the advice, how does the resistors affect the performance of the amp?

I ask as I put a 100 ohm resistor between the valve rectifier (5AR4) and the output stage of an 5F1 Champ clone and the amp didn’t sound so good when the amp distorted. I imagine it could be worse with an amp with greater wattage.

[Littlewyan]

It recreates the sag of a valve rectifier. Some newer Voxes use it, like the AC15C1, instead of a valve rectifier. So it’ll give a softer response which may be desirable in your case as the original AC30s had valve rectifiers.

Not sure about the Champ as Class A doesn’t sag. If you’re worried about the effect you could start smaller, 47ohm total instead of 100Ohm.

Did you measure the screen current?

[Dr Tony Balls]

not to harp on it, but here's a recent implementation of a B+ dropper (on the left wall). Tiny, cheap and effective.

That is small which is good as there is a lot of real estate in some amps. I typically worry about heatsinks for such devices, I’m not sure the chassis can adequately keep the mosfet cool.

You can add a small heatsink and paste if so, but depending on the application the chassis itself is often enough to keep the temperature down.