18Watt.com

A question to me from my bass player.

PA amp channel into an 8 ohm load is say 100 watts.
Same channel into a 4 ohm load is maybe 200 watts.

Why wouldn't the 18watt tube amp generate 36 watts into a 4 ohm load on the 8 ohm winding?

I started to give my expert tube amp guy answer (solid state vs tube/class type) and remembered I am only a wannabe expert.

Comments?

Looks to me like you already know the answer - the Marshall is a tube amp, not a solid state unit. Tube amps don't DO that!

EBE wrote:PA amp channel into an 8 ohm load is say 100 watts.
Same channel into a 4 ohm load is maybe 200 watts.

Why wouldn't the 18watt tube amp generate 36 watts into a 4 ohm load on the 8 ohm winding?

It's more about impedance matching. Most of our favorite tube circuits have a pretty high output impedance. You can drive high Z loads like peizo panel loudspeakers direct from push-pull output tubes, but you need a transformer to deal with relatively low Z loads like piston/cone speakers. With a multi-tap transformer, you're looking for the most efficient loading. If 8 ohms is right, then you'll yield less power at 4 ohms or 16 ohms.

Most SS amplifiers have lower output impedances than piston/cone speakers, so you get more power, the lower the speaker load is. However, you can't keep dividing by 2. Eventually you get too low and burn up the transistors, or if you're very lucky a fuse will blow first.

EBE wrote: PA amp channel into an 8 ohm load is say 100 watts.
Same channel into a 4 ohm load is maybe 200 watts.

Continue with 2 ohms...
you´ll at least have lots of fun..

jaysg already posted how it works.
It´s just so that many tube amps are pretty forgiving on load mismatch.
If you go one notch up or down, you won´t lose that much power... it´s more in the 10...20 percent range. Sometimes you don´t lose any power at all, might even go up. Just remember early Marshall´s with built in mismatch.. mismatch downwards and little happens except some tonal change... mismatch upwards and the output drops considerably.

Btw, a certain amount of mismatch can be favourable in combo amps with an external speaker jack. If the amp is a bit on the upper side with the internal speaker, and you connect the ext as well, it will just be a bit on the low side... almost the same power.

And, some guys drive their Marshall´s at extreme mismatch... just check the PP for EVH mod Well, almost anything works... if you have glass in the amp.

Adrian

wannatone, yes most tube amps are forgiving,but not Marshalls..
If you use a mismatch,your going to have trouble,or lose an OT.
they don't like the pull 2 tubes from a 100 watter thing either.
If your lucky,they only lose a tube or arc a socket,or lose a screen grid resistor.. To keep it healthy,always on a Marshall,match the output setting and speaker load.

Richie

Plexi wrote:wannatone, yes most tube amps are forgiving, but not Marshalls..
If you use a mismatch,your going to have trouble,or lose an OT.
they don't like the pull 2 tubes from a 100 watter thing either.
Richie

Richie, I've heard this before. Do you know the reason why? On one level, logic says that if you pull two tubes from a Marshall 100, you're only asking for 50W anyway, so how could it matter? The guys with the dead amps, however, tell a different story.

Phase maybe?

Any similar danger with el84's in an AC30 or Marshall 36W?

Those are good questions. Some i don't have the answer for,but i know by seeing,and changing many Marshall OTs.And knowing they used a mismatch. I'm sure there are other things that can Kill an OT..

Now as one of your questions,this is for 50 and 100 watt marshalls..I don't work on many old Vox amps,although i do get a few. And those new AC30s AC15s,And JTM 30 and 60 and 900s,the transformers are crap.
So i can't answer if the Old Vox suffer the same problems.
I don't think you see it as much,because they use the combo speakers in the cab. And its a bugger to remove tubes from those amps..lol
And cathode biased,you'd have to change the resistor for the amount of tubes.. I know the matchless run a mismatch on there OTs..
I wouldn't advise it to anyone.
I do know on Ampage the same question came up,and someone a while back did some tests,and found it would usually either take out a tube,or arc on the socket,using a mismatch on the marshalls. And i believe it was an amp set on 4 ohm,and load was 16..

Other amps, fender etc seem to handle this..I've also changed alot of fender OTs And PTs,but those are mostly due to abuse..Ingo may know the whys ..hes the tranny expert..

Joe at OEI,his tranny i believe has yet to have these problems.
All i know is there is something to it,and i wouldn't try it.
And on an old Marshall.i wouldn't even think of trying it.

I know there are many debates about this,and some say it won't hurt a thing.
Kevin Oconnor talked of it in his books..here is a link..
and if you read this,at the bottom there is this paragraph..
He doesn't say why either..
Your scroll thing will be moved down about an inch to get to this last part..
I agree with this quote below, but the things he talked about above this paragraph, i don't agree on all of that.

http://www.londonpower.com/faq.htm

""In most amps, you can set the impedance selector to whatever sounds best. The one caveat is: NOT in English amps. Having replaced more Marshall OTs than anything else, I would advise that the impedance selector always be set to the rated load, or less. ""

Richie

Kevin O Connor wrote:In most amps, you can set the impedance selector to whatever sounds best. The one caveat is: NOT in English amps. Having replaced more Marshall OTs than anything else, I would advise that the impedance selector always be set to the rated load, or less. "

That is one statement of Kevin's that I've never been very happy with, and the he doesn't seem to really explain. They make an awful lot of very different amps in England too! And what about Scotland, Ireland or Wales. IS he saying they're safe? Another of his statements I have serious reservations about, is where he says that EL84s and EL34s are unreliable and to avoid using them, especially in upside down positions. That's wiped the 18W out right there!
Here's the deal. If you go into the theory and do load line analyses for tube power amps, it becomes clear that there are actually a range of impedances you can run them at. This is why for example the Matchless Spitfire gets away with running it's output stage with a 4k load, when everyone else uses 8k for PP EL84s. The problem is when you go to higher impedances on the output, you tend to get higher voltage swings on the anodes and OT. This *may* cause problems with some amps. There was a period when the bigger Marshall amps ran their EL34s at very high B+ voltages - I don't recall the exact model numbers off the top of my head. Eventually, Marshall lowered the voltages, due to the high numbers of failures in the field. Going to a higher impedance speaker load with one of these amps was a recipe for disaster, and it was this generation of amps that unfortunately gave Marshalls a bad name with some folks - or in Kevin O' Connor's case, to all "English amplifiers". And we see that Kevin does actually say "the rated load, or less". In fact I've seen high impedance load mismatches hurt some Fenders as well.

So if you pull a couple of tubes just be careful to also adjust your speaker impedance selector accordingly.

Phil,its hard to say, it could be tube type,like most think of english amps useing EL34s or KT tubes.. Maybe its the tube,instead of the OT. most amps that people say handle the mismatch use 6L6 tubes,as in fenders. yes there are a few different types of transformers used in English amps. I would assume they would be different.Partridge,Dagnall,Drake,RS.etc And marshall amps have had some design problems..one, alot of people see,that causes tube and OT failure is in some models,is the output selector. they melt,the amp will cut out from time to time,and not make good connection. Down to not working at all. Then one could be causeing the other,it would happen either way.So there are many things to think about,as why its in one amp and not another. All you know is..it does happen,don't do it..LOL

To stay out of trouble,on any {tube} amp i just like to use the correct tap for the correct speaker load.If it has one, Its put there for a reason.

I always wonder why the EL84 gets a bad reputaion. I guess its good for us,so they'll be more around for us to use..lol Maybe because they are in mostly combo amps,small ones,and get the heck shook out of them from the speaker vibration.
And they did make some tubes for mounting upside down,and sideways
Sometimes they will even show which is acceptable in the tube data sheet.
Just alot of things to think about..

Richie

Whichever way, there's no risk in going to a lower speaker impedance with a tube amp, even going by what Kevin O'Connor says. The worst that can happen is you loose some power or it doesn't sound as good. But sometimes you get lucky and it ends up sounding even better. Increasing the impedance is another story and carries some risk...

""... I would advise that the impedance selector always be set to the rated load, or less. .... EL84s and EL34s are unreliable...""

Sounds like plain BS to me. I don´t know the guy who has published this opinion (presented as facts...), so I can´t judge if he understood anything at all about tube amps. But, I guess he already regrets that the ink never gets back into the bottle...

Well, I might be wrong, but I see at least two causes why certain Marshall models might have been some less reliable than the rest of the world.
Excessive screen voltage and, I´ve heard some came even w/o screen resistors?
High primary impedance. Just remember the earlier amps run two tubes into up to 8 k ohms (4 tubes into 4k). So, if I follow the questionable advice to plug the impedance _selector_ below the rated impedance... our poor 2 tube primary gets to run at 16k... fortunately, those amps don´t have 4 ohm taps as well... to get into 32k ohms
No, I wouldn´t try this...

But, to do this guy justice, I guess he just accidentally misspelled what he really meant.

Btw, EL84´s had been in almost any european tube radio for decades, and I can just tell you that the old western europe stuff had been very reliable. And, most EL34´s that I pulled out of old amps still tested like new, even after 30 years...

Adrian

Ha! Kevin O' Connor is recognised authority on valve amps, and has written many books on the subject. And he does actually know his stuff. But one or two of his opinions like that really make me splutter!

You know it ain't easy being a Tube amp Guru as it seems every guru wanna be is trying to knock you off the hill or catch you making a mistake.

I had some dealings with Kevin a couple of years back on a Thanksgiving day when I called his number to order some books. Kevin himself answered the phone (Canada no do the Thanksgiving thing) and he was not just helpful in taking my order, but spoke to me for almost an hour about tube amps, answering all my questions and even seemed excited to talk to me about his favorite subject, "Tubes!". He ended the call by saying "..call or email anytime you've got a question and I'll be glad to help." He is just a righteous guy and tops in my book.

Gerald Weber catches more grief than any other Tube Guru out there, but, he is just as helpful and willing to share or just talk any ol' time. Doug Hoffman, Bruce of Mission amps and Randall Aiken are more of these giving individuals in the tube world. In fact, there are a few of you right in this forum and other forums as well. The only thing I can figure is going on when I see someone slam one of these guys (and I'm not seeing that here) is that it's just the passionate mentality of us tube amp fanatics. Our passion seems to get the best of us sometimes.

I tell you one thing, I for one am glad these guys are sharing their wealth of knowledge in this field, cause if I had to get it from the library I wouldn't be half as far on my learning journey as I am now. Long live tube amp Gurus I say!

Hello

perhaps a small calculation will show the problem by mismachting the OT.

If you are using a secondary load of 16 Ohms and put into the 16 Ohms tap of the transformer, the reflected load as example for a 50 Watt Marshall with two EL34s will be 3400 Ohms and at 50 Watts the voltage swing at the anode is 412 volts rms or 582 Volts (peak to peak). Using now the 16 Ohms load into the 4 Ohm tap of the transformer will give a reflected load to the primary of 13200 Ohms. Now at 50 Watts the voltage swing at the anode is 812 Volts rms or 1148 Volts (peak to peak). This mismatching will cause very high AC voltages in the tube and in the primary windings of the transformers.

Comparing Marshall transformers to other British companies like Partridge for example, these have not an isolation between each wounded layer and the possibility for a short between the layers of the primary winding or a short between primary or secondary is much higher. Also the chance to get spark-overs in the used tubes will rise.

Now some other words to the EL34s in Marshall amplifiers. They have rea very high screen grid voltage, which is more than the max. ratings in the data sheets. I know three versions of the EL34, which never have had problems, these were original Mullards, Philips or Valvo. The East-German version of the EL34 was very critically, first the high voltage of the screen grid and second, the angle of the systems inside were different to all other EL34s. So the hottest points of the tubes were facing each other. But don´t´forget, Marshalls are really the most selled amplifiers, so it is in the nature, that they have the most failure rates.

In my eyes mismatching is always dangerous, because the transformer isn´t built for higher voltages or higher currents inside the primary winding, but it will be no problem to use a 4000 Ohm transformer for a PP Amplifiers with only two EL84s, if it is designed for.

Kind regards
Ingo

zaphod_phil wrote:Ha! Kevin O' Connor is recognised authority on valve amps, and has written many books on the subject. And he does actually know his stuff. But one or two of his opinions like that really make me splutter!

Phil,

just not to get misunderstood.. I highly respect the guys who have written amp books... and I appreciate their stuff.

Just don´t take every printed word as if it was written in stone...

Well, at some point I got tired of the 100th pictured step by step "how to blackface a Twin Reverb" receipt... w/o any word about what it actually does. I haven´t even found any word about one of the most important tube amp design topics, i.e. how to calculate the plate load. Everybody appears just to know what is required... guess it´s just like in the old days when they picked it from transformer catalogs and tube data sheets....

Maybe I should have bought one of Kevin´s books. I appear just to have accidentally started with the wrong stuff?

I would appreciate any recommendations on books.

Peace!

Adrian

Adrian,
If you want math regarding tubes the book "Valve Amplifiers" by Morgan Jones is full of it. ISBN 0-7506-4425-7

Also the RCA Receiving Tube Manual RC-30 has a lot of tube operational characteristics in the first 100 pages.

As far as Kevin's books are concerned, if you want operational calcs and theory, Principles of Power is a good one.
ISBN 0-9698-6081-1
I would buy Kevin's book first as it is better written than Jones' book.
Barry

I agree with Barry...O'Connor is the real deal. He provides (arguably,of course) the best published resources for amp junkies out there. Expensive but well worth the money. Unfortunately for him (or any other tech writer) he can't cover every nuance in 300 or so pages. That's what makes the BBs so important...they fill in the the thin spots. I imagine if it was possible to edit and publish the AX84 and the 18watt.com archives a new standard would be (easily) be set. Phil's threads at AX84 are some of the most substantial "real world" explanations I've seen; not to mention Paul, Mark and, well, just too many of you guys to count... Imagine 400 or so pages of just the meat in those archives. What a resource that would be. Adrian, you can add 30 or so pages on simplifying the calculation of load lines. I'm going to need at least that many. Make the drawings really colorful and put in lots of pictures of pretty girls or I'll never get through it.

Good old Fleabay. Just got an ' Audio Power Meter ' pic attached. I am going to try it on a solid state junker before I try it on the 18 watters etc.
Anybody used anything like it before ? any tips ?
Steve UK

bnwitt & dualopamp,

Thanks! Sounds much like it´s exactly what I´m looking for!
I had some introductory tube lessons back then, but the stuff was already going to be obsolete and everybody was too excited about solid state Glad to hear that much of the old knowledge could be saved from decline.

stevesuk,

4 and 8 ohms, stereo... are you going to build a stereo 18 watt? Jusst kidding. Maybe a bit hard to take readings from that scale? The meter looks more like it was intended as a hifi decorative device. Well, a more precise indicator could be added..

Adrian

Ingo wrote:perhaps a small calculation will show the problem by mismachting the OT. Using now the 16 Ohms load into the 4 Ohm tap of the transformer ...

What you are saying is absolutely correct for running a higher impedance speaker load - it is risky. However, as you also know, it does not apply when going lower in impedance loading.

Your comments about the Marshall transformers as well as the EL34s are also very interesting.