Adjusting BIAS on EL34s - Wooo-Hooo!!! I'm Still Alive!!!

[MrBump]

I know that this has been asked before, and I've trawled the web looking for articles, but I thought I'd stick something here anyway.

I've just bought a digital BIAS meter, that will display the cathode current of a valve.  However, I'm not sure how to get from that to practical reading of plate dissipation, which is what I believe I need.  From what I've read, this cathode current needs to be x by plate voltage in order to determine plate siddipation in watts.

Am I missing something?  Can BIAS bi adjusted simply using cathode current?

Mark.

[MrBump]

Actually, I've just found this:

http://www.marshallforum.com/workbench/10890-biasing-70-vs-90-vs-plate-current-vs-cathode-current.html

Very interesting read, and explains the how things got confused with the the introduction of "dumbed down" bias meters (for idiots like me...)

[jpfamps]

I've looked through that article, and there a lot of twaddle in there, which I suspect is designed to scare people into paying an amp tech to bias their amp for them (bear in mind I charge people to bias their amps too).

A few points.

1) Stating that a 1 ohm resistor "could be 20% off so your reading could be 20% off" is nonsense. You would be hard pushed to buy a resistor with 20% tolerence. 1% 1 ohm resistors are very cheap. Some people think that say a 100k resistor will have better tolerance than a 1 ohm resistor, but this is nonsense.

2) Same is true with a "cheap" multimeters. Even cheap multimeters are now very accurate.

3) Adding the 8W plate draw to the static plate dissipation is daft. Static plate current is at best a few mA. I know I've measured it, so it will introduce minimal error to your reading. Regardless the error will be on the safe side anyway.

4) The transformer shunt method, recommended in the post linked to, is DANGEROUS, can be just as "inaccurate" as the cathode current method, and for amps that have more than one valve per side won't tell you the current draw through individual valves, so you could have one valve drawing no current and the other valve drawing twice the recomended amount.

Look at Randall Aiken's web-site for details on the 70% rule. This is a good rule of thumb for most EL34 amps running so 450-500 on the plates, ie like most Marshalls. Ultimately the 70% rule is quite a conservative limit so a 5% error either way will be fine.

[MrBump]

Thanks for that.

I still can't see how you can adjust the bias on just the cathode current - surely you need to know what the plate voltage is (i.e. 25 watts of an EL34 x 0.7, divided by plate voltage should = maximum allowable cathode current, the figure that the digital bias meter is measuring)?  Or am I missing something?

Can one assume plate voltage?

Mark.

[jpfamps]

You are correct, you also need to know the plate voltage.

The more sophisticated bias probes will also allow you to measure the plate voltage as well as cathode current.

To measure the plate voltage accurately you really need to measure the voltage whilst you are adjusting the bias of the amp. This is because as you increase the current draw of the amp then the plate voltage will drop. In an amp with a solid state rectifier this effect will be much less than in an amp with a valve rectifier.

If you don't have a bias probe that allows you to measure plate voltage, then to measure plate voltage with the amp under load you will need to have the amp open and on, which I'm sure you will appreciate is potientally very hazardous, so I don't recommend this.

You have two further options.

1) Guestimate the plate voltage, by say looking at a schematic, asking someone who works on amps, making a educated guess. Generally most modern amps with EL34s use about 450-500 VDC on the plates.

2) Measuring the unloaded voltage with a DMM. To do this remove the power valves. Attach the negative probe of your DMM to the amp chassis and check you have a good ground connection. Insert the red probe of the DMM into pin 3 of a power valve socket. Check that there is not continuity between the DMM probes. Set your DMM to VDC and fire up the amp. This will give you the unloaded plate voltage, and I would use this to calculate plate dissipation.

The 70% "rule" is just that. It's not a target, and is a fairly conservative. The datasheets for most valves in Class AB recommend biasing at 100% of dissipation! I've seen a ton of amps where the valves have been biased very hot and they seem to be fine, although I don't recommend this.

If I were you, I would guestimate the plate voltate and shoot for 60% dissipation which will give a you a good margin for error.

[MrBump]

Thanks, appreciate that.

I think that you're last option is the best one for me, currently - my Orange was revalved recently, so I'm kinda hoping that the plate current is in the ball park of what it should be with the new valves. 

Mark.

[Keven]

sorry to hijack, but if you were to set the bias on an amp that has 2 tubes pulled, do you need to measure at all 4 sockets or just set the optimum cathode current for the two valves that are running and be done with it?

[jpfamps]

You ideally you should measure the current through each valve anyway, so if you remove a couple of valves then just use the current draw in the remaining valves to calculate plate dissipation.

Obviously any socket without a valve in won't be drawing any current!

[HTH AMPS]

I've looked through that article, and there a lot of twaddle in there, which I suspect is designed to scare people into paying an amp tech to bias their amp for them (bear in mind I charge people to bias their amps too).

I've seen that exact post on the Metro Amps forum too - that guy really gets himself wound up about biasing.  Anyway, I've found the 1ohm resistor method to be perfectly workable and accurate enough for guitar amps - a few mA either way ain't gonna kill anything.  If you're drawing too much current you would soon know when the plates start glowing red.  Seems to me that a fairly simple procedure is trying to be made much more complex (why?).

[jpfamps]

Just to demonstrate that the cathode current method is not that inaccurate, I measured the screen current in an amp I was biasing today.

In a 6L6 loaded amp with 470 on the plate and 34 mA of current draw there was about 520 mV drop across the 470 ohm screen resistors, which works out at about 1.1 mA, ie an error of around 3%! This will be inconsequential.

These are real measurements taken in a real amp, ie not someone's opinion, and are typical of what I have found in past amps.

[MrBump]

All good stuff.  Thanks for the replies.

Mark.

[MrBump]

Sooo!

I'm still alive, for which I'm very thankful.

I decided to put E34Ls in my amp instead of EL34s... I'd heard that they might be a little brighter and more Fender-y than the EL34s.

My word, what a difference!

Now back with 4 valves rather than 2.  Everything has so much more clarity and sparkle.

I had to adjust the bias right down - with the EL34s the cathode current read 36/37, but with the E34Ls it was coming in at nearly 60 - sounded great, but I suspect that the valves wouldn't have lasted long...

So I took the cathode current reading down to 35 - assuming 500 plate voltage (was too much of a wuss to measure this, so took a guess, given that the old cathode current was 36/37), a new cathode current of 35 would give a plate disipaton of 17.5 watts, which seemed to my ears to be bang on the money.  Old Oranges apparently sound better biased a little more cool than more modern amps, and I couldn't agree more...

Running it now with the back off to make sure that there's no red plating, but so far so good...

It's been fun - I'm still terrified of putting my hands inside the old beast, but slightly less so now.  Good learning experience.

Mark.

[Keven]

Hooray, glad to hear it went well. thanks for the rundown on e34l's as well! i am running with 2 valves here too because i thought the amp was quite bottom heavy (and had a redplating valve.. )

did you think your amp with 4 el34's were overly boomy and couldn't get a tone with a decent hi-mid ,IE, felt too scooped even with the mid control cranked up?

[MrBump]

Hooray, glad to hear it went well. thanks for the rundown on e34l's as well! i am running with 2 valves here too because i thought the amp was quite bottom heavy (and had a redplating valve.. )

did you think your amp with 4 el34's were overly boomy and couldn't get a tone with a decent hi-mid ,IE, felt too scooped even with the mid control cranked up?

I thought that they were missing something - I posted in another thread about getting another amp, as my Orange was very thick sounding, particualy for the poppy/folky music that I'm playing at the moment.  The EL34s were great with loads of distortion, but were a wee bit too boomy for sparkling cleans.

The E34Ls definitely sound sharper and brighter to my ears.  Much more what I'm after.

Mark.