The Tube Swapping Thread

Hi Jonny. The idea is that there are some of us who think that the amp has too much gain for our needs. It is designed to be able to cope with High Gain and Metal types as well as the Clean players so there is a lot of potential for overdrive there which means that those of us who only ever play with limited gain in use have to fiddle with our gain controls over the first cramped couple of notches. If you put a valve in the V1 slot which gives that stage lower gain, (that's not the same thing as a "lower gain valve" which is a misnomer), it means that every stage in the amp chain has less drive to it so you need to turn up the Gain control further to get the same original amount of overdrive. It simply spreads the gain we use over a larger section of the Gain control. Personally I would prefer to actually go into the stage and change out resistors to redefine it to a lower gain but that is more awkward.

If you are thinking this might suit your playing then my advice based on understanding circuit design, and that of a few others you may find if you dig around, is to NOT swap out your 12AX7 for any of the usual "equivalents", they plain aren't equivalents at all! 12AT7, 12AU7 and 12AY7 are in no way similar to 12AX7. Techs will generally insist they are. That is based on the number of times they have plugged them into a 12AX7 circuit and it has worked without blowing anything up. There are other parameters for valves which are more important than what people think of as "gain" which are wildly different and will make a mockery of the designer's careful work. The DC levels around the stage will change significantly which means the clipping p[oints will be different so distortion changes. The frequency response of the stage changes too as the loading alters massively. This may make it sound better to you of course but it is no longer a precision designed stage, it is now a random piece of good or bad luck. You pays your money and you takes your choice.

The valve you need to replace the 12AX7 is a 5751 which is very close to actually being a 12AX7 with lower gain, all other parameters are very close too and that matters. You can technically put a 5751 into any of the preamp stages, probably best not to in V3 which is the PI phase inverter stage which feeds the output valves. With 5751s in both the V1 and V2 slots you will drop your gain by about half which expands your Gain control range by a useful amount.

As you asked, you can quite happily put a standard 12AX7 in the V1 slot without any problem. As it is the first valve in the chain, any noise here will be amplified by the following stages so it is often of benefit to use a designated Low Noise 12AX7. That will be a little more expensive but be careful you don't get stung paying a silly "premium" price for something which is simply selected from the batch in a few seconds test. It also can help to have a "Balanced" 12AX7 in the V3 slot as each triode in the 12AX7 package drives one of the output valves, so balanced triodes mean a balanced signal on both sides of the output. That said the assymetric distortion that an unbalanced PI valve produces may sound better to you. Keep the metal valve cover on V1 too of course, it reduces noise pickup.

Thanks bordernbert. I've ordered a set of JJ EL84s for my GM40, plus three JJ 12AX7s (one being balanced). Might order another unbalanced 12AX7 too, now I think about it, just to see if I prefer its sound in V3.

Cheers for your help. Will report back once I've tried them out.

Ordering an unbalanced valve might be difficult, no one tests for that condition and what you get would be anything between an acceptable unbalanced limit and a perfectly balanced one. Without testing accurately you wouldn't know where your own valve sits. It might even make more sense to see if you can find someone who has a pretty poor "known unbalanced" one you could borrow.

I had the tubes changed on my GM 36 but not by choice...Sadly with only a few hours on the amp I turned on the standby and let it warm up and as soon as I turned it on, I heard the dreaded POP! No noise whatsoever after that...Contacted Hughes and Kettner and they said no worries we'll hook you up with a local shop to check it out. Turns out there wasn't one in Vegas...So I told them this and suggested a great shop here. They contacted the shop and set them up to do all warranty work for them. Just awesome customer service! As it turned out the tubes went So they sent the shop a complete set of new tubes and to my surprise, instead of sending the stock tubes the amp came with they sent Sovtek EL 84's and TAD 12 AX 7's! The amp sounds great and have had no issues with it since. Lot's of great products out there, but without great customer service it doesn't mean squat. Hughes and Kettner proved to me I can count on them if things go bad.

Here's something that H&K once said to make people think. They have always been criticised for their "cheap valve" policy. They do not sell their own amps, they pass them on through a distribution chain to final vendors. If they put in a set of much better valves at manufacture that puts the price up by let's say £20 as a ball park figure. This price is then inflated by a percentage at every stage between H&K making it and it getting to you. Insurers base their prices charged and customs and excise agencies base their prices levied on the sale price of the item. Distribution from point of manufacture to the main distributor country of sale does the same, and of course each individual sale outlet then calculates his price based on a percentage of what he pays which is the already inflated price. At 60% markup each time and let's say 3 distributor stages in the chain until you buy it, that would mean that £20 has become £81.92 to you the customer. It isn't as easy as saying just pay an extra £20 for the more expensive valves.

H&K prefer to keep the price cheaper knowing that you will probably make a decision to swap out the valves for better ones at some time in the future according to your own convenience. Some, (like me), prefer to wait until a valve change is called for. Others would swap out the H&K choice no matter what it was for some other esoteric flavour of the month type as a matter of course just to be able to say they have done it. It's basically a sound decision made as much for your benefit as their profits.

Their customer support is exceptional! As long as you know how to get their attention and then deal with them fairly and respectfully, just as you would expect them to deal with you. Glad to hear they have dealt with you in that way, like I said, I have always heard stories of their good dealing. And I can also say, they are also good to talk technical with too, they are open minded and listen and aren't too tight with their advice when there is a problem or question.

Anyone using JJ el844 in their amp? Thinking this may be a good swap so I can get the amp to breakup sooner at a slightly lower level. My main concern is if the lower power output messes up anything in the amp if the auto bias can’t cope. Or is all just sales bunkum?

"Or is all just sales bunkum?" We-e-e-e-e-ell, my own view is that it might lean that way. Apparently it's a 9W version of an 11W valve. Is that a really significant drop in power? Would you easily hear a drop in level of that amount? Current theory says not? A drop of 0.87dB (11W->9W) is not going to be particularly audible, especially in the context of a live band setting. It's almost impossible to discern -3dB after a wait of about 5 seconds.

There are a number of points which I would want to have cleared up before I would believe everything that is being said. How is that drop in power meant to apply within circuits? Does it mean that the valve has a lower anode dissipation, it can't handle as much power before overheating and red-plating? If so it can't sit straight in EL84 circuits. Does it mean that it needs to be driven from a lower line voltage? Again, it's a direct drop in into EL84 circuits so it would appear not. So if it is keeping the normal load impedances, (same output Tx and speakers), and line voltages then how is it going to work? The anode load impedance is set by the Tx/Speaker and that is crucial to the voltage/current curves and power output of the valve.

If the valve has the same shaped characteristics as its big brother then the breakup point is going to drop by only the same amount and will remain as difficult to hear, only -0.87dB remember. However, if the 844 has a different set of curves which start to bend significantly lower then it might give a different tonal balance. I'm mystified how some people think this sort of stuff works, (not you Madtone). I have just read someone saying: "They sound like el84 tubes to me. I was able to turn my amp up much more than usual for band practice which let me get the amps natural growl kickin...nice!!!" Now I don't know how this guy thinks that these things work but, if this valve is the same gain as a standard EL84, and there is nothing saying it isn't, only that it is a lower power, you wouldn't need to turn your amp up higher to get to the Crunch point. It should occur lower on your volume control. Isn't the valve recommended because it gets to it at lower signal levels?

Of course I could be missing something really obvious, it has happened before and no doubt will again. I'm looking for a datasheet at the moment as this intrigues me beyond marketing hype but it doesn't seem to be around yet.

Got the datasheet.  Here are some parameters from both the EL84 and the EL844.

Parameter	EL84	EL844
Ua (anode V max)	300V	300V
Ug2 (Screen Grid V max)	300V	300V
Wa (anode max dissipation)	12W	9W
Wg2 (max screen grid dissipation	2W	2W
Ik (cathode max current	65mA	60mA
Rg1 (for automatic bias)	1Meg	1Meg
Rg1 (for fixed bias)	300k	500k
gm (mutual conductance)	11.3mA/V	6.5mA/V
Ri (internal load resistance)	40k	100k

There is so much there which is exactly the same but so much which is very different and is significant, (look at gm and Ri which are both crucial design parameters), that I'm unsure yet how this valve can just be a drop in replacement for a standard EL84.

Excellent thanks for the response bordonbert, was hoping for a response with your usual engineering perspective. I know you only get marginal gains (no pun intended) with these things. I don’t know how valves work per se, but know enough to be cautious. No idea what gm and Ri do, but if these are important design parameters that suggest these are not a like for like replacement, bog standard JJ el84’s it is then.

I was having increasingly bad problems out of my GM40. Squealing, popping, massive volume drops, hissing. Tech guy from where i bought it suggested it was a power amp tube problem even tho the TSC was reading fine. Opened it up and sure enough the #2 tube was pretty funky looking. I got 4 JJ EL84 to replace with on recommendation of these forums, but instead went with "EL844" Low Power tubes. After about 12 hours playing time its been perfect. Maybe even better than before, maybe because im turning the master up higher with these tubes or maybe they just sound better. Maybe both. Fizziness is lessened ..i can actually use my guitar's tone pot now instead of having to leave it on 0-1. Thinking about preamp tubes now too since they are fairly cheap and now that i know how to swap them its quite easy.

Link

https://www.thetubestore.com/jj-el844-low-power-el84

My LP with '59 pickups is still an unplayable fizzy mess with this amp but thats for another thread.

Which guitar are you using when it sounds fine Jthom?  If it is a lowish output single coil type then it sounds like it could be down to hitting the front end too hard with the humbuckers on the LP.  I keep trying to spread the word that the input stage of these amps includes a solid state buffer which has 3 modes.  Low input signals are passed through with a small amount of amplification absolutely transparently to set the level for the valve stages.  Medium level signals get clipped on one side adding even harmonic distortion which is very musical.  Higher signals get a full Tubescreamer type clipping.  All of this is outside of the amp's controls other than the Lead/Ultra channel selection which ups the gain of that stage a little and depends absolutely entirely on the guitar or last pedal level control.

It's a mistake to push too much signal level into it if you are getting your distortion from in front of the amp and/or the later amp stages.  If you weren't already aware of this I really suggest you try turning the signal from the LP fully down and try using it again increasing it a little at a time and balancing it against only the Master Volume control until you get the level of distortion you hear as right for you.  These amps sound very rough when they are pushed into distortion with pedals, the input stage, the Gain stage, the preamp valve stages and finally the power amp.  That's definitely overkill for all except the most hardened extreme metal cases.

I used my own Les Paul Traditional set at around 4 for rhythm work and about 6-7 for lead and I had plenty of Gain available for overdrive/distortion.  That was without pedals too.

The guitar that I feel sounds really good with this amp is a Yamaha RS420. After a few hours of fiddling and pondering I feel like i made progress with the LP. The relationship between the different tone controls is a completely different world compared to other amps I have played. Turn one knob and everything else responds so differently. I guess i just need to get good
Thanks for the replies, I'll keep working with this.

I just order new JJ EL84 tubes, unfortunately the numbers on the box aren't the same. Can someone confirm if I can safely mount them in my GM36.
1. 29.1 -7280
2. 29.1 - 7270
3. 29.1 - 7150
4. 29.3 -7140 (doesn't  look like matched for me but I am not an expert so I ask you guys)

Another think cut my attention is in amp documentation the Bias[V] range is 10 to 24.4 so I am also no sure if replacing to 29.1 won't destroy my amp?

Thanks for all good people...

Voltaire wrote:le mieux est l'ennemi du bien.
best is the enemy of good.

Matching is always done only to within a specific percentage.  It is unrealistic to expect suppliers to look for perfect matches in all respects, to how many decimal places would that have to be?  Usually this will be to within 5%.

29.3 - 29.1 = 0.2 and this is only 0.7% difference.  That's a great match!
7280 - 7140 = 140 and this is only 2%.  That's another good match.

It isn't necessary to have perfection despite the current thinking in the music gear field.  There is a growing feeling that if 2% is good then 0.2% must be better.  This is wrong.  These figures will vary over time no matter how accurately they were matched at the start and may even be very different from day to day. And the difference in 2% and 0.2% will almost certainly be absolutely undetectable with your ears.  We all need to be realistic with our expectations and accept that it's a real world out there and most real life good stuff lies within the twenty percent range not at the perfection pinnacle.  This is the very reason that systems like H&K's TSC are adopted.  Bias balance can be performed and maintained electronically even when valves drift away from each other as they age.  There are limits to that of course but with reasonable valves it is a great process.

Put these valves in place and try them with the "plectrum" bias measuring procedure on the back of the amp with its flashing LEDs.  Let us know what the results of that are and we'll see if there is any problem.

...should have said also.  Your bias of 29.1V has been measured at what current?  We don't know do we?  The bias voltage is for a specific current and anode voltage and without knowing that H&K and your supplier have decided on the same test conditions comparison is pretty much meaningless.  Don't worry, run it up and test it with the plectrum then you can be sure it's working as it should or if there is a problem.

Not sure if anyone is still reading this, looks like it's been a few years, but worth a shot!

I love the idea of more headroom at lower gain settings, as it does seem like it ramps up awfully quick with the stock tubes, but I also love me some metal. I was looking at the SRV option on Eurotubes but am wondering how much impact this will have on what I'm hearing with the gain dimed on the Ultra channel?

Hi Rob.  Yes, you're right, you won't get too many flaming arguments here nowadays.  If it helps here is my controversial 5 penn'orth, anyone else disagreeing can quite rightly argue the toss with me.  I hope people would have factual evidence to provide to back up any claims they make, that's the only way forwards with issues to do with electronics.  I'm sorry if this is more than you wanted but it may contain information that you are not aware of and it may actually be of interest, I don't know.  It is here again for anyone who may be interested to read through and think about.

First let's make sure I have you right in what you are saying.  Eurotubes SRV option seems to be a selection of valves which they say is in keeping with Stevie's preferred sound.  It includes a 5751 in the first slot for a cleaner signal path through the preamp section, then hitting the power amp section harder with the Volume control full up.  It also contains a very common piece of factually incorrect information which has real consequences in the real world.  The idea that a 5751 is "in-between an ECC81 and an ECC83S in gain" is a VERY VERY BADLY WORDED piece of advice.  Firstly, it pre-supposes that the ECC81 is an option for this type of amp stage and it isn't, or let's say it shouldn't be.  The ECC81 is the original UK designation for the US 12AT7 and the correctly specified ECC83 for the 12AX7.  They are the same valve just redesignated for each side of the Atlantic.  The ECC83S is simply a genuine ECC83 with a slightly different physical structure inside it, it has exactly the same electrical parameters.  And now for the controversial bit.

The 12A*7 "family" is NOT a family of similar valves with increasing "gain".  First of all valves do not have a "gain", they have a series of interrelated parameters which create the gain of the stage they are put in.  You can easily design a circuit based around a high gain 12AX7 which has a stage gain of x3 say and then simply change a few resistor values and have a stage with gain of 30x.  So how can this mean the valve has a "gain"?  Valves do not amplify voltages according to a "this many times as big" criteria.  In very simple terms, a valve has the ability to read a voltage on its grid pin and to use that to alter the current passing into its anode.  The grid voltage increases so the anode current increases and vice versa.  (It uses the cathode as a shared reference point to do this.)  The anode output voltage is then produced by selecting a load resistor for the anode to pass this current through.  It is the resistor which defines how big that voltage will be, not the valve.  We can also affect how much gain is in the stage (not in the valve) by adding a resistor between the valve's cathode and the circuit ground which introduces feedback and reduces the gain.  You can see, at no time is the valve changed in any way but we have methods of using resistors to control the size of the output voltage.  Up to a certain level of stage gain the valve has very little effect and good circuit design reduces its effect more.  That is what we aim for in design so as to prevent any differences in the valves which may be used from making differences in the effect of the circuit.  Designers actually try to design out the need to "tune" your valves by selecting "golden" ones.

Now, one of the valve parameters which is quoted for any model is "μ", (that's a Greek letter pronounced 'mew').  It is the valve's "amplification factor", sometimes incorrectly thought of its gain, and no they are NOT the same thing.  For a 12AX7 this is typically about 100 and for a 12AT7 this is about 60.  That seems to be significant but with a cathode resistor in place, which is absolutely normal in guitar preamp circuits, feedback is introduced to the stage which reduces its gain and makes the effect of the valve's characteristics much less prominent.  You will not halve the gain of the stage by changing to a valve with half the value of μ.  What you will do is to upset other aspects of the circuit which you know nothing about because of the wildly different other parameters which no one looks into.  For example, the anode bias current, (it's resting state current), will certainly change.  If a valve change increases this, (12AX7 to 12AT7 or 12AU7 say), then you are increasing the heat dissipated in the anode resistor and this can easily get to the point where the resistor can burn out.  That depends on an individual amp's resistor specification of course but many are close to the limit for long term safety and no one even thinks of aspects like that.  Most techs are not even aware of that side of things, (note I say MOST - I did not say ALL).  The standard tech is an expert on fixing and maintaining a huge variety of models of amp and of processes of troubleshooting them.  He works from his broad specialist experience of those models and of the processes of troubleshooting them.  He is not an expert on design!  He has no need to be, it is not what he does for you.  Now, some are interested enough to go into this side of things but that does not mean they will all be aware of aspects like these, especially less experienced ones.

My advice is, do NOT take any advice from anyone that is "fine" to swap an ECC83/12AX7 for any other model in the "family", 12A-T/U/X/Y-7.  These valves are all different in all their parameters, not just the misquoted "gain".  They were originally designed to perform different tasks and have nothing in common with each other, other than their sockets.  However, you have the answer in your original post.  If you look at the datasheet and specifications for the 5751 you will find that that model is virtually identical to the 12AX7 in every important way.  The 5751 IS a safe and sensible option for a valve swap where the 'μ' of the valve is not an important parameter.  I stress, halving the valve μ does NOT halve the stage gain in any practical design I have come across, it reduces it only slightly.  Feedback built into the stage prevents it from having too much effect but it has a little.  I use 5751 valves myself in just the same way as Eurotubes describe, (one of the factual things about their "advertising speak").  I use a 5751 in the V1 position to limit gain in the early stages and to prevent overdrive too early on.  It can't hurt.  I am aware however that, on the other hand, I can simply turn down the volume on my guitar to achieve much the same effect.

Beware the advertising man Rob.  They love an old wives tale or a piece of "internet common knowledge".  It's a sure hook into your brain and then on into your wallet.  Always look for factual engineering data to back up any claim.  If you can genuinely hear it you can definitely measure it.  In our amplifiers, if you can't measure it - it doesn't exist!