valve amp question

[Tony Raven]

I'm trapped (mostly willingly) in a rural area with a dearth of music stores. On the upside, this town has an old-fashioned repair shop. Ted is schooled, licensed, & can fix anything from a flat-screen to an antique radio. I've had him overhaul a couple of 1970s amps, & got great work at a quite reasonable price.

Though he's not a guitarist, he asked an interesting question: should a valve amp be biased (perhaps even some small mods) to the needs of the owner? For instance, a jazz player might prefer something crisp, while a rocker wants ready breakup.

I'm so entrenched in the plug-&-go mentality that this had never occurred to me, in decades of playing. If an amp suits my needs, it's in; if not, I move it along.

Is this something that's already common & I've just missed out?

[Jay Menon]

Whilst not an expert I have read that valve amps can indeed be biased according to the needs of the player. Bass players who traditionally want a clean sound with litle break up, tend to like their amps biased 'colder'. Rock guitarists on the other hand, would go for hotter biased amps, with a view to achieving power amp saturation that much easier, but at the expense of shorter valve life...

I'd be most interested to hear Dave's and Barry's (Loverocker) opinions...

[The Red Bladder]

I now have this vision of a valve amp that has been over-biased, ranting on about immigration, gay marriage and the Euro.

[ef37a]

Phew! This subject REALLY deserves that article by Merlin! I will however do my best but this is just MY take on the subject remember.
Take a pair of EL34s biased at 25mA per valve. For a 500V supply that is a total dissipation of 25watts, 12.5W per valve, very well within their rating and even if mains V goes up they will still be happy. VERY roughly those valves will stay in "class A"* up to about 15-20watts out so you have a very clean "hi-fi 15Wish amp. Put the bias up to 40mA per valve and the valve dissipation becomes 20W, a bit close to the 25W limit and certainly very little leeway for mains variation. The amps' power supply now has to deliver an extra 15W, not a lot you might think but all extra heat in the mains transformer and cooking the electrolytics. The class A power is now some 30watts perhaps but the ultimate maximimum power will be less.

So, IMHO it makes no sense to bias a valve amp hotter than the makers specification and that is 25mA per for my lot! Hotter biasing for 4xEL34 amp makes even less sense! You already have 30-40W on class A tap!

*This is a hot spud! People will tell you that the "classic" 4xEL84 cathode biased AC30 style amps do not run as "pure" class A. This is true but it is only when the amps are pushed to near their limit that they deviate much from A and a near flat out AC30 is a very loud beast indeed amd not something most of us can let rip with!...Wrapped.

[ef37a]

Colder biasing does not of course carry any heat penalty but many amp men would say that 25mA is pretty cold anyway!

The situation is further complicated (IMHO) by the use or not of negative feedback around the output stage. Many 50-100W fixed biased amps use feedback most cathode biased 15-20 watters don't.

The effect of feedback will be to "smooth" the transition from class A to class AB but also to cause the overload point to be much steeper. Put another way a NON NFB amp will get progresively more distorted as you approach clipping. With NFB the amp will have lower distortion up to near clipping then Woomp! Into harmonic mayhem. Amplifers with a lot of NFB, hi-fi valve and especially solid state have an even steeper ascent into gross distortion. I don't know if anyone here has ever tried a Quad ll as a guitar power amp but if so I bet they were not impressed!

So, only scratched the surface of a big subject (and a very subjective one at that!)I am now prepared to be shot at!

Dave.

[Tartaruga]

In my opinion,an amp can be ‘biased’ to be closer to it’s best performance.There’s no real point in ‘biasing’ it to suit the taste of the player.Tubes can change it’s performance,but it’s a kind of lotery…If you can find very good ‘old’ tubes(RCA,Sylvania),you probably will feel a difference against some of today’s brands.Usually,’Bias’ is used to suit the right fonctionning of each tube amp(equilibrium between tubes,as they compensate for differences)…But biasing it to suit the player’s needs...don’t know…maybe you mean,’finding the right tubes for your taste’…?
I have an old ‘all tubes’ amp,and the better tubes I put in it,the better the sound.’Very good tubes’ are hard to find,and quiet expensive,if you have the chance to find a few pairs...
Pedro

[zenguitar]

My tame amp tech is very old fashioned. He sees biasing as a technical issue rather than a tonal choice. And he doesn't measure voltage, he puts a sine wave test tone through the amp, a dummy load, and uses an oscilloscope. With a class A/B amp he pays close attention to the cross over point between the two phases aiming to make it as close as possible to a sine wave. Having watched this on a number of occasions it is very interesting. There can be some very interesting ripples and bumps in the wave form when you go too hot or cold. With a class A amp he biases to keep the 0V crossover in the centre of of the valves' range so it clips symmetrically.

So, there's a different approach to Dave's but the outcome is the same, not too hot, not too cold, just right. Thinking Bias? Think Goldilocks.

Andy

[ef37a]

Yes Andy.
That's the "hi-fi" way to do it! You can't set for symetrical clipping unless the amp has bias balance* and then only at DC/LF. "We" have this on all but the A100 but it is looked upon as more a way to reduce hum than anything to do with "tone" I think (tech' direc' is V hot on a good S/N ratio!).

Minumum crossover? Again a bit audiophile for guitar amps.They SPOSED to sound a bit rough!
Interestingly the A100 was originally spec'ed at 30mA per valve but I think nobody could tell the diff at 25mA so save the juice, the valves and the heat!

I can honestly say that in two years I never had a customer or even an "artist" ask me to "warm up" or "cool off" the biasing on any amp and I got some rite luloos of artist I can tell you!

*Well you can fork about till ungulates come to chez nous with a box of bottles!

Dave.

[IvanSC]

I have ONE fixed bias amp and it is my least favorite.

Subjectively, cathode bias is so much warmer sounding and somehow sags in a nicer way.
I would love to know WHY I have this bias (boom boom ) towards cathode biasing but every single amp I have fallen in love with was cathode biased.

Even the real early Watkins and Henri Selmer/Truvoice ones.

[ef37a]

Ah but Ive' what were the fixed biased amps you have tried?

There are some FB EL84 amps about, Laney for one, and some folks don't like them. (If I were a cynical type I would say it is a cheap way to get a 20W amp!).

You should give an HT-20 a fair go, that has both biasing regimes.
Then, cathode biasing tends to go with thermionic rectifiers (hateful things) thus the "sag". You might find a CB amp with silicon derived HT a bit different?

Dave.

[IvanSC]

Dave - I have been watching for a cheap HT for some time.

Loved both the HT5s I have played - both earlier ones though.

And I am a Laney hater, big time. uber boxy to my jaded ears.

Perhaps if I had played with a head only version into some decent speakers.....

Plus I don't do anything over 20 watts at all now.
Even my old MusicMan 65 stays slugged down to 25 watts.

And I LIKE thermionic rectifiers BECAUSE they sag (grin)

I don't even like going up to anything a 'powerful' as a GZ34 to be honest.....(more grinz)
Power supply droop is gorgeous if used correctly.

What we need is tiny, sweating transformers! Ouch!

[ef37a]

Ivan, mate!
HT-5; Fixed biased, does not sag worth a tiddle.

MusicMan 65* Fixed bias! And what do you mean "slugged down to 25W", power soak?

*At least the map I found shows so.

Dave.

[Folderol]

Is this a can of worm I see before me?

There is a whole lot more to it than fixed or automatic bias. For a start, with AB:

Is it a common cathode resistor?
If so does it have a decoupling capacitor?
If not, do the individual valves have decoupling capacitors?

For FB:
Is the bias regulated hard?
Is there some degree of compensation for HT variations?
Is it just a cheap resistor network? Yes I have seen that!
Is it low Z?

For both:
Is there grid current?
Are the drivers capacitively coupled ? (can temporarily bugger up bias if grid current is allowed)
Are there separate phase splitter and low Z drivers?

When looking at valve bias you absolutely must include the drive conditions. There is no point in having a precision setting regime if at the slightest hint of overload, rectification by grid current backs off the bias by 10V until the coupling capacitors and bleed resistors balance the charge.

Mind you, we are talking guitarists here, so maybe that's a form of distortion and 'shunting' effect they want

[ef37a]

Hi Will...Ok!
I challenge you to find a c biased guitar amp with a cathode resistor per valve and moreover one without a decoupling cap!
Regulated bias? Waste of time since the time honoured method compensates well enough for HT variations.
Cheap resistor network? Do you mean a cap and R off the HT winding? Yes some "famous" names did that and "we" paid for the results!
Low Z? How low do you mean?

PI to op valves is almost universally 100k anode load, 22nF ish and 220k grid leaks. The A100 is like this (but a proper bias winding)no one ever complained! But yes, HT and Series One amps all have low Z PI/drivers, C:10k. This is more to keep grid leak resistors low because of the very high, very variable grid current found in modern valves.

But yes Will, if you read the various texts about blocking distortion virtually all guitar amps should sound pants. Somehow they don't!

Dave.

[Folderol]

Hi Will...Ok!
I challenge you to find a c biased guitar amp with a cathode resistor per valve and moreover one without a decoupling cap!

Oh dear! I think my HiFi slip is showing

I keep forgetting guitarists don't want quality
{ducks and runs, weaving erratically}

Incidentally, did you know that if you have really chunky drivers and independent screen ballast resistors. using caps to cross feed the drive to the screens gives you an interesting compression cum saturation effect.

Not only do not a lot of people know that, it's likely that not a lot would want to know it!

[ef37a]

Quality? Well not so much but "clean" is desired but not TOO clean (some early Fenders had 12dB of overall NFB, about as much as you could use with the grott traffs!).

Cross coupled screen grids? Not seen that Will, not sure what 'appen, got a map? I do know that if you leave G2 undecoupled in say an EF86(a much maligned valve) stage you get compression effects and by choosing the cap value it can be made frequency dependant. One of the 1001 things I want to investigate!
BTW there are two schools of thought re G2 resistor ratings. Make them very feeble so they burn out if the valve is abused. Screen grid overload is said by many to be the prime cause of valve failure in guitar amp op stages. Or, big jobs that can stand being pulled to chassis for 100mSecs or so and therefore do not need replacement. "We" go for the latter course of action and now that these days it seems good valve techs are like hens' teeth maybe the best option?

Lastly; did anyone catch the Vox Story on BBC4 recently?

Dave.

[Folderol]

Quality? Well not so much but "clean" is desired but not TOO clean (some early Fenders had 12dB of overall NFB, about as much as you could use with the grott traffs!).

Did you ever come across that WW article suggesting centre taped speaker winding (to ground) and the sides going to the OP cathodes. Supposed to give better phase stability and more resistance to off-load prolems, but needs a lot more drive. Never tried it myself - always wanted to.

Cross coupled screen grids? Not seen that Will, not sure what 'appen, got a map? I do know that if you leave G2 undecoupled in say an EF86(a much maligned valve) stage you get compression effects and by choosing the cap value it can be made frequency dependant. One of the 1001 things I want to investigate!

Not seen it commercially. It was an idea that I picked up from somewhere (as a callow yoof) and tried out myself. All I did was stick cathode followers on the screens and fed them via a standard RC network from the phase splitter. It was all rather hairy, and I've no idea what it did to valve life, but a guitarist friend at the time liked the effect. I used a pair of EL34s, mostly because (if you remember) they were classed as 'industrial' valves and carried no purchase tax so were actually cheaper than 6V6s.

BTW there are two schools of thought re G2 resistor ratings. Make them very feeble so they burn out if the valve is abused. Screen grid overload is said by many to be the prime cause of valve failure in guitar amp op stages. Or, big jobs that can stand being pulled to chassis for 100mSecs or so and therefore do not need replacement. "We" go for the latter course of action and now that these days it seems good valve techs are like hens' teeth maybe the best option?

I preferred the 'Philips' method of drop-off resistors. If they get hot enough to melt the solder they {ahem} disconnected
I'd love to get back into valve stuff again, but time, logistics and lack of a proper home workshop all conspire against me

Lastly; did anyone catch the Vox Story on BBC4 recently?

Dave.

Now he tells me!

[ef37a]

"Now he tells me!"

If you can't get it on iPlayer I can send you a burn.

Dave.

[Tony Raven]

Now y'see, that's why I love this site. Heavy-duty info presented in a lively (& generally friendly) manner.

Really, when I try to keep up on these discussions, I truly regret mis-spending my youth studying computer science by day & playing in rock bands by night -- lord, if I'd only gone to trade school, I'd still have something both current & useful.

What with all the debates of tube-swapping & such, seeking for The Tone, if there were something that could be done in the biasing to further that quest, then wotthehell

So. Consensus appears to be something like: Yeah, it can be done, but probably not worth it. Set the circuit to factory spec & stop messing with it, yeh? Any disagreement? Willing to listen!!

[ef37a]

First off Tony NO learning is ever wasted if it is true so no regrets!
How I wish I had done a night class in computing in my 30's or enrolled in the OU. Maybe I would be able to fix the niggles in my systems like the infernal wrap that is likely (only happens in SoS and studio-central?).

Next. Mate, IMHO 90% of what you read about guitar amps is bllx. I have been an observer (and sometime Ptype builder) with a team of top engineer, gifted player with some electronics nonce and 2 or 3 periferal people steeped in the industry with fantastic ears (don't ask me how they have preserved them!)The process is, get the engineering right first, i.e. properly rated components, make as sure as you can that the inevitable valve failure causes as little consequential damage as possible. Check that allowable excess mains volts are tolerated. And LBNL, try to break it! Such things as bias current values are (obviously it would seem) very low down the list of "tone" tweaks so set bias at the best value for best valve life.

Then comes hours of cut/play/fag/tea/solder/play/cut/sub'/tea/swear..to get each stage working against the next to give the correct drive/overdrive and frequency response. When THAT is all done a Test Specification is done (sometime by me) where every voltage is noted and the response checked at spot frequencies. For evem a "simple" amp such a the HT-5 this can run to several sides of A4. This is the "bible" that ensures every amp made sounds the same as the first lab specimen...grr...

[ef37a]

Grr cont'...
You will read ont'web of all sorts of super tweaks by swopping valves,sometime type for (better?)type, sometimes a different type, a practice I deplore. But no matter how effusive the modder, you will NEVER get a before/after voltage table or an A/B recording.

I was tasked to make a triode switching device. This allowed both sections of two ECC83s to be changed over in any given circuit at the tap of a foot. The result was that yes. 2 very experienced players, one of which was a top design bod, COULD tell certain valves apart, but!This was in very controlled conditions and the differences were small and confined to the overdive sound alone. In practice they did not think a player would ever know.
NB. This was done with available current production valves not "NOS" or other exotica, nobody can base a commercial design on such but even so, those claiming great things for NOS have never done such an in depth a study and until they do the claims stay with the fairies as far as I am concerned!

Dave. (Tony; get Tube Preamps; Merlin Blencowe. He has promised us a power amp book but could be 4years!)

[4TrackMadman]

One case in point for biasing amps hotter - Eddie Van Halen. He was cooking his Marshall Plexis in order to get more saturation and his early tone is heavily based on that. He also used to blow up amps like crazy...so, for each his own.

[ef37a]

Morning 4 track,
I have done a bit of Googling re Eddies amps and I cannot find any mention of "hot" biasing?

The techniques that seem to be most commonly mentioned are the use of a Variac (but to reduce mains to 90V ref 115 NOT boost it!) a power soak system and feeding the resultant power stage signal into a big MOSFET PA.

The bottom line tho' is of course that Rock Gods have godlike resources and can do as they like. Most of us have to run amps without blowing valves or traffs and no manufacturer could afford to market such an unreliable product!

Indeed, "we" found evidence of "red plating" with a certain brand of EL84 and tweaked the bias R to avoid it. Listening tests proved that the overhot bias was contributing nothing to the sound and certainly not one punter ever complained about his amps' sound post mod (when they came in for other problems).
Dave.

[Folderol]

... of course, the problem with using a variac on the supply is that you will also lower the heater voltage, which is not good news for the cathode.

I would think the only simple, non-valve destroying, way to lower the saturation point would be to lower the screen volts.

This might be a problem if the O/P valve is a 6080

[4TrackMadman]

I think the variac came in later, first he was biasing them ultra hot. Remember seeing that somewhere in a video explaining how to get his tone, some nerdy mid-80s guitarist on a vhs tape, wish my memory would serve me better as to who it was.