8ohm or 16ohm loading

Hi.  Running my TM36 with a pair of 1x12s carrying 8ohm Celestion Vintage 70s.  So according to H&K I can use either a single one or the two in series giving 16ohm.

I wondered what the impedance was for the quoted 36W.  If you don't know about these things it matters!  You cannot get the same 36W output power into 8ohm and 16ohm without using a separate tap on the output transformer.  From the same amp setup 8ohm should ideally give you twice the power out of the amp to 16ohm, (that's ideally but it will draw more, output stage allowing).  So I asked H&K for advice, I asked if, considering that the H&K 2x12 is 16ohm, an 8ohm cab would affect output power. This is what I got from their technical department:  "Hi Albert. If you have an 8 ohm cab it will indeed increase the volume. Kind Regards". We're all apparently too stupid to take in anything deeper than that, (despite in my case, a lifetime in the electronics industry designing stuff like this). I wasn't too impressed though I will grant them that I could have asked more specific and probing questions.

The subject of the "8-16ohm output impedance" needs a little clarification from H&K I think. My assumption is that they imply the output stage is beefy enough to work with a mismatch within that range, but I could be wrong! I would also assume that, if the output power is quoted at 16ohms, then 8ohms will push it much much harder, (more about the results of that in a few lines). If 36W is quoted at 8ohm load, then 16ohm loading will draw significantly less power, even though it still represents a mismatch and once again puts the output stage under more stress.

So to my reason for seeking other opinion. I have seen a horrid scary thing! I was running the amp into a single 1x12 8ohm cab with the power soak set on 1W. Great sound but you tend to push the amp harder of course. Turning it up to get enough power out of the 1W setting and tone from hard driving it means that one pair of output valves is actually running at full output power internally into the power soak load. It cut out completely. I stopped playing of course as you would and after 4 or 5 secs it kicked back in again. I've experienced this same effect 3 times now.  I've taken to practicing at home on the 5W setting and not so high on the MV to keep internal dissipation down.

Even more scarily, once I was playing at 1W and high settings into 8ohm and the thing dropped out and the whole internal structure of the left EL34 started to glow red hot. Now this is a very bad thing for valves. It is known to happen usually due to low bias, but should not be accepted. I immediately kicked it into standby and it stopped trying to make the room look cosy and romantic. After that event the amp had dropped into diagnostic mode with that pair of valves turned off. I shut it down, waited a minute or so, then ran it up again convinced I had blown something. It was fine. I checked the bias settings, with my "bias setting" pick, LEDs flashed, I counted, and everything checked out ok.

I reported this to H&K asking for advice a few days ago now. The reply I have is "Hi Albert and thanks for your email. As the UK distributors for Hughes and Kettner your email has been forwarded to ourselves. I've copied our technical department into this email and they should be in touch shortly. If I can be of further assistance please do not hesitate to contact me. Best regards". It's only been a couple of days so they haven't replied yet. Has anyone else seen this sort of behaviour, or does any other tech savvy guy have any info on the output impedance question and whether it is linked in any way to what I'm describing?

And incidentally, is this site monitored by H&K themselves? Hi guys, are you out there? Is there anybody listening? Phone homeeeee........

Well, you're lucky.  State side I've received zero answers to any questions.

I'm going to make like an ass and nit-pick - at 16 ohms it may dray have the current, but with their dynamic auto-magic bias circuit, they could change the bias - oh wait, that might be a bad idea.  Hmmm...

Anyway, it does make one wonder how they handle a 2x range of impedance on a tube amp.

So, how hot did the gm get at the moment of failure?   Heat isn't a huge deal for tubes, but it is quite a big deal for semiconductors, which this amp has...  I'm having a hard time buying the shtick that this thing isn't bothered by the heat it generates.  I'm in data storage, and people make a living from removing heat from semiconductors.  Many commercial chips are made to operate up to 70C core temp, but that value is easily reached.  So both my practice and gig rigs have active cooling on the GM.  And also, because I have an external active power soak, I only use the 18 and 36W modes.  I don't feel comfortable cooking those internal 20W resistors on the bottom of the amp...

It is a little surprising that tube didn't get damaged while glowing - I've never seen a tube glow, but from what I've read it ain't good.

Later,
-Bob

I'd love to have a look at that auto-bias setup.  It's a clever idea and there are a few ways I can imagine they would approach it, but I think it must be simpler than my flights of fancy might suggest. Do they really adjust it dynamically in use, or do they simply have a way of adjusting it statically and accepting that it will drift when being driven.  A standard output setup will be set correctly at tickover, vary dynamically under drive and drift overall over time.  Maybe the H&K way is to accept it isn't always perfect dynamically but it at least always varies around the correct static spot?  Or to vary it dependent on some not quite so long time constant, (sort of integrating or averaging it over time and adjusting as you go).  I can't believe it would adjust itself accurately over a very short time period.

In any case the aim is a good one and, for my money, I will accept however they do it as long as the amp remains reliable and sounds good doing it.  If being out of perfect bias doesn't damage it or spoil the sound then why worry?  Distortions around the crossover point tend to have more influence as the signal levels go down so when we are driving hard and loud, an increase of that type of distortion may well not be audible, or may even sound better!

I will say that since the "incident" I have had nothing but good results from using it.  The speaker change from Celestion Vintage 30s open back to closed back gave me a little more depth to the tonal range which improved things.  The Clean/Crunch/Lead options are really one of the best I've heard.  If I had a criticism I would have to say that really old smooth classic rock tones are sometimes a little difficult to achieve and I find it leans towards a more modern overdrive tone too easily.  That will be down to me and really depends on the guitar I use too.  I'm a Les Paul man using a Gib Traditional for day to day work, (I like my weight training too   ), with a USA Strat for lighter tones and a Gib Junior Special Double Cut when it needs that P90 growl.  I also play some slide on an Epi 335 Dot and a Dillion LPJ DC.  The latter is a really cheap and nasty unit I ended up with by accident but found it has an astounding tone for slide.  I'll use it until it falls apart which it surely will one day!  I mainly play 70s/80s/early 90s rock, and it can be great for things like Bad Company and Thin Lizzy.  (No one seems to want to play Humble Pie stuff which ticks me off no end!  Mind you where is a Marriott when you need one?    )

I think you're right to question the ability to run over a 8-16ohm range though.  I can only assume the output stage is Victorianly over engineered somehow, or made to match somewhere in the middle which means a lesser mismatch at either end.  In truth most speakers have such a varying impedance curve that there is always a degree of mismatch and power lost into the transformer.  If that's a good quality item then a lot of misuse can be weathered.

I also agree that those power soak resistors must take a hammering too.  Like you, I design my own power soaks.  Nothing clever, just simple switched passive units, and why pay loadsamoney for a commercial unit when it's so easy.  Personally I always go for aluminium cased 25W or even 50W resistors with that Victorian approach.  Why not use bigger components when they're so cheap and having a slightly bigger box is no problem?  Maybe with a real music signal this is not the issue that running with sig gen waveforms would be, but it seems to be an area where the one eyed man like you and I would become curious and maybe uncomfortable.  Soldering them down across a straight gap in a PCB in a straight line like that may be asking for the PCB to grumble at the thermal expansion side of things in the long term.  I would have preferred to at least see them with a kink or loop at each end to soak up that potential movement.  But all that said, the amp sees a much less reactive load with a power soak in circuit.

Anyway, in the long term the proof of the pudding is in the eating as they say.  We'll have to wait and see if the amps remain reliable and sounding as good as they do when just played in.  I've always got my old Vox Valvetronix AD120VT which I also love waiting in the wings for the star to sprain an ankle!

bordonbert wrote:(No one seems to want to play Humble Pie stuff which ticks me off no end!  Mind you where is a Marriott when you need one?    )

I was just reading this thread and came across this!

I have NEVER managed to get any band to play ANY Pie material. Why I do not know. I saw them 16 times & still listen to their music very often.

Do you live near me?

bordonbert wrote: I wondered what the impedance was for the quoted 36W.  

On the H&K Blog the official H&K account replied to a comment asking about the impedance of the GMD with "The impedance match of the GrandMeister is 12 ohms, so both 8 and 16 ohms are fine. With 4 ohms, you’ve got to be much more careful." I bet the TMD impedance is is the same. Seems kind of silly to me to use a 12 ohm output though. No one has 12 ohm cabs and did they really save that much money by giving us 1 12 ohm output instead of 16 and 8 ohm outputs?

I think the point is that with a 12ohm output neither 8 or 16 is a problem for the amp, these impedances are always very approximate and a compromise anyway even in amps which offer multitaps.  If you have separate 8 and 16 ohm outputs there is a cost and complexity penalty though it isn't great, that rearboard is pretty stuffed with components already, and if you plug into the wrong one it is further off the mark that the 12ohm mismatch would be.  It's a pretty good compromise in production engineering terms, (I'm not a production engineer, things are generally purist for me as a R&D guy ), and it does mean simplicity in use for the "uninformed" guitarist.  (There are actually a lot out there though they generally don't see it themselves. )







(EDIT: KernelSanders! Just clocked it and took it in. I'm going to change my name to Ken Tuckee to match.)

Ken Tuckee, I like it!

I'm really wondering if it was a good compromise to just use one 12 ohm vs separate taps, or even just 1 x 8 ohm tap or just 1 x 16 ohm tap. I noticed going through a 16 ohm cab sounds tighter/ less fizzy with my GMD in 40w mode. I've played around with resistance amp & cab mismatches with other amps so I'm familiar with what a mismatch sounds like. Makes me wonder what the GMD would sound like with a proper impedance match. I'm also wondering about 40w vs 20w mode. Normally when you pull two power tubes on an amp you half the cab resistance as well. I doubt they put anything in the meister amps to compensate for this so if running in 20w or less, maybe we really should be using a 6ohm cab? I'll have to try my 2 x 8 ohm cabs out now that I think of it.

Tried 8//8 + 16//16 = 12ohms?  I'm still not really convinced that it will make too much difference.  The speaker impedance wanders about wildly across the frequency spectrum and is only close in a narrow band in the middle so most of it is very rough approximation anyway.  But that said, there is always the chance that you could stumble across something useful.

Oh, and I can confirm that there is no compensation in any way in the GM36 for when you kick in the 18W mode.