Speaker impedance question

Cabicular

This has no doubt been discussed before but why don't amp manufactures auto detect impedance and set appropriately?
Can't be impossible can it?

ecc83

Not impossible but very expensive to do and a shade pointless! Then you need to differentiate between vlave OP stages and transistor. For the latter the load impedance (Z) only matters to the extent that it should not be too low otherwise the amp will go into protection mode or even pop a fuse (NO modern sstate amp should be able to be damaged by any sensible load!)

Connecting loads HIGHER than optimum, usually 4R, will just result in less power.

For valves I suppose you could have a CPU sensing, motor driven switch? But in practice valve amps are nowhere near as bothered by the "wrong" load Z as most peeps think, so again, pointless?

Yes, if you slam a 100watter set to 4 R into a 16 R load at full welly it is liable to arc over. If you set it to 16R and drive the bllx off it into 4R you are liable to blow a fuse or even a valve or two but with well designed amplifiers it IS quite hard to break them! (I HAVE tried!)

Dave.

ICBM

Or why more don't take the Leo Fender approach and make the amp robust enough to drive any normal impedance without blowing up, and not bother with that impederance selecterer thing on the back.

Leo knew about musicians...

ecc83

ICBM said:

Or why more don't take the Leo Fender approach and make the amp robust enough to drive any normal impedance without blowing up, and not bother with that impederance selecterer thing on the back.

Leo knew about musicians..

Heh! I don't know if modern Fenders allow such liberties?

"We" all know that no decently made amp will be harmed by an impedance "mismatch" unless as I mentioned you load* it and drive it to extremes. The problem is that we live in a "sue 'em for everything you can get" society. So whilst amplifier IN PRACTICE might be more robust and tolerant than generally thought YOU CAN'T SAY SO IN THE BOOK!  It is also a very competitive industry now with many more players and margins are very tight.

*Not "auto sensing of Z" but I think protection in valve amps could be improved? There does not seem to me to be a huge problem in detecting an O/C or near O/C load and shutting down the drive?

Fairly easy to do at the amp end of the jack (as you know IC) bit trickier to work out that the "musician" has not plugged in the other end! Not however impossible IMHO.

Maybe when EL84s are a nifty each and you need a mortgage to revalve  a 100 watter we shall see such protection?...Don't laugh! It will come.

Dave.

jpfamps

I have admit the idea of putting some sort auto detection / switching on the output of a valve amp does not inspire confidence.

Fenders tend to be combos so attaching the amp to an inappropriate impedance is less likely.

From my experience repairing amps in the real world, then the major issue in valve amps with inappropriate loading are usually arcing, either externally or in the output transformer, due to too high loading.

External arcing is difficult to prevent, although putting a small piece in insulating heat shrink on pin 3 seems to help.

Internal arcing in the mains transformer is simply down to the insulation in the transformer being compromised.

Fender still use transformers wound in the old fashion way on bobbins, and this method employs a layer of insulation between each layer of the windings, so there is more insulation in this style of transformer than a transformer that's wound wire-on-wire (the transformer wire has a insulating coating as well obviously!).

This method dates from a time when the insulation on the transformer wire was not as robust as it is now. 

Modern transformer wire has better insulation, so in theory you don't need to have a layer of insulation between each winding on the transformer, and of course makes the transformer cheaper to wind, as a) not adding insulation between each layer save a lot time in the winding processes. b) you can use a smaller stack as using less insulation means you have more space for wire.

For what it's worth from the mid 60's onwards it's very rare for me to find a Fender amp with a damaged output transformer (and I've worked on 100s of Fender amps), so they must be doing something right.

ICBM

jpfamps said:

For what it's worth from the mid 60's onwards it's very rare for me to find a Fender amp with a damaged output transformer (and I've worked on 100s of Fender amps), so they must be doing something right.

Same here. The only two I can remember off the top of my head were a Bandmaster Reverb head that had been run cranked into a 16-ohm Marshall 4x12", and a Champ that had probably been used as a preamp with the speaker disconnected.

I dont like the idea of putting switching on a speaker output either - in fact, it's one of the main stupid design flaws with the Marshall JCM2000s (both the switch in the 16-ohm jack and the failure to parallel the two halves of the impedance selector switch) - let alone those horrible 4x12" input panels. Any switching introduces a potential contact failure point which could produce an open circuit, so eliminating it where not essential is a good thing.

jpfamps

ICBM said:

jpfamps said:

For what it's worth from the mid 60's onwards it's very rare for me to find a Fender amp with a damaged output transformer (and I've worked on 100s of Fender amps), so they must be doing something right.

Same here. The only two I can remember off the top of my head were a Bandmaster Reverb head that had been run cranked into a 16-ohm Marshall 4x12", and a Champ that had probably been used as a preamp with the speaker disconnected.

I dont like the idea of putting switching on a speaker output either - in fact, it's one of the main stupid design flaws with the Marshall JCM2000s (both the switch in the 16-ohm jack and the failure to parallel the two halves of the impedance selector switch) - let alone those horrible 4x12" input panels. Any switching introduces a potential contact failure point which could produce an open circuit, so eliminating it where not essential is a good thing.

Off the top of my head of post 1963 Fenders I've changed the OT in a Twin that with two 16 ohm speakers in it so had been running at 8 ohms (for several years), a Bandmaster where the one of the wires in the OT had arced over to the end bell and a Bassman 135.

My guess is that the later 2 had been run into the wrong load as well, so essentially all three had been used incorrectly.

Pre 1963 I've changed loads. 

ICBM

ecc83 said:

*Not "auto sensing of Z" but I think protection in valve amps could be improved? There does not seem to me to be a huge problem in detecting an O/C or near O/C load and shutting down the drive?

Fairly easy to do at the amp end of the jack (as you know IC) bit trickier to work out that the "musician" has not plugged in the other end! Not however impossible IMHO.

The early Ampeg SVTs had a 4-pin cable to connect the amp to the cab, with two wires for the speaker current and the other two jumpered on the connector in the cab and used to operate an extra standby circuit - the HT to the driver valve! So it was impossible to operate the amp without the cab connected, even if the cable was plugged into the amp.

They dropped this fairly quickly though, presumably because it was 'slightly' dangerous feeding 350VDC down a speaker cable .

It could be done in a much more sensible way if you wanted to though (in the same way as the Blackstars with switching in the jacks for example) and with a Speakon cable it's now easy to get four connections to the cab.

Cabicular

So what I'm hearing is
Load of shagging hassle
Not much reward ?

ICBM

Cabicular said:

So what I'm hearing is
Load of shagging hassle
Not much reward ?

Correct .

Probably 90% of owners never need to worry about it anyway, they either use combos or always use the 'proper' cab for their head. The ones that do need to worry about it because they mix and match stuff should educate thesselves as to how to connect them up correctly, it's not hard .

A very simple rule that's useful if you're faced with the usual nondescript unmarked cab at a venue or a rehearsal room, and you don't have a multimeter: assume the cab is 8 ohms. An amp set to 8 ohms is at less risk from a 4-ohm or 16-ohm cab than an amp set to 4 is from a 16-ohm cab or vice versa.

ecc83

Yes "If in doubt go eight Ohms" is a good plan as is not driving the H out of the amp if going into an unknown cab (this is where ICBMs liking for Big Mother amps pays dividends!).

Further on the protection front... It would not seem very hard to have some form of current sense? Just detecting over voltage at the speaker jack would, it seems to me, be a recipe for spurious protection triggering but if you have a speaker voltage signal and no or very low current that is pretty conclusive of no load.

But this is where we came in? The extra circuitry would cost. Partial protection, i.e. no plug inserted can be and is done in at least one range of amps (and they ALL have another protective measure which I cannot mention, not foolproof but helps) . People can and should be "educated".

And even when gross loading errors occur, modern amps seem very rugged. Last of all. RTFM!

Dave.