Are solid-state amps subject to speaker load restrictions like tube amps?

[g1]

My question is (regardless of what % of SS amps use OT's): will a solid state amplifier using an output transformer be damaged if run without a load?

[J M Fahey]

will a solid state amplifier using an output transformer be damaged if run without a load?

Yes, definitely, even if said transformer is an externally connected autotransformer (very common in line distribution PAs in very large spaces, such as Hotels, Airports, Railroad or Bus stations or large Churches)

[Enzo]

REally? Let us forget for now large CV installations such as airports and just think about some 100 or 200 watt guitar solid state head with OT. COnsidering the circuit impedances and the low voltages, and the much lower turns ratio, would there really be any danger of a reflected transient arcing the primary?

[J M Fahey]

Short answer: yes, really, the SS amp would be damaged ; that was the question I answered.

Slightly longer one: who talked about arcing ? In a somewhat wider view: who talked about a damaged *transformer*?
In fact, I agree with you that the transformer would survive 100% of the time, under typical SS voltages.

I was talking "damaged SS amplifier" (what was asked), and failure mode would be second breakdown, overvoltage by inductive spikes or reverse polarization (as in having an NPN emitter more positive than both Base or Collector and viceversa on the PNP), or a deliciously destructive mix of all three.
In fact, much lower reflected of equivalent inductive impedance in heavy (as in heavy cone and voicecoil) woofers can and does destroy output transistors.
Clamping diodes from the "hot" output terminal to B+ and B- are added to take care of problems 2 and 3; second breakdown is not solved by them because they only act, by definition, when output voltage tries to exceed (positive or negative) peak voltage, AKA power rail voltage, while second breakdown can do all kind of nasty things *between* those two extremes.
That this damaging effect exists all day long and is not a Friday 13th phenomenon is shown in the elaborate curve bending applied to protection circuitry and the use of "unnecessary" extra devices, to extend the safe area for non-resistive loads , AKA speakers.

As an extra point, in the very few guitar amplifiers which use not autotransformers but "tube architecture" schematics, with center tapped output transformers , with CT going to B+ and extremes going to PP MOS sources, such as excellent KMG ones (widely discussed in SSGuitar) and my own humble, 12V powered, "Callejero" 60W , the danger is even larger, because you don't have a low internal impedance classic SS amplifier which would dampen spikes somewhat, but very high impedance Sources driving them "as tubes do".
There are also a couple Galliens with that archictecture that I know of; certainly there must exist some more, plus a bunch of 12V, "30W" cheap street PA amplifiers with 2N3055 outputs which I see everywhere.

[Enzo]

Fair enough. WHat kind of voltage spikes would occur? I am thinking that in a SS power amp, the output transformer would generally be a 1:1/1:2 ratio part, since the amp would/could normally drive one of the expected load impedances directly, before the OT was added.

[J M Fahey]

Yes, that's true, the usual ratio would be , say, 2:1 for a 4 to 16 ohms conversion.
In 70V or 100V lines it would be a higher 3.5 to 5x ratio, not *that* much either.

The problem for the transistors is the no_load flyback effect, which in theory has no limit, and in practice, can generate peaks 10 to 20X the original voltage.
You'll never see them that high, because diode clamps limit them to 1 V beyond B+ or B-, but if you don't add them, the clamp is the reverse biased transistor itself. BAD.
Worst part is that the breakdown occurs between its internal junctions, which is not expected and was not considered by the transistor makers.

Well, maybe they think about that in transistors meant to switch inductive loads, power converters or electronic ignition or, of course, in TV flyback drivers, but not on audio stuff.

As a practical example, think a typical SS power amp, say 50V rails, an NPN device on the upper half, driving a resistor.
No signal , (or on the sine wave zero crossing), the transistor emitter rests at 0V; the resistive load has one end grounded, the other one has those 0V applied.
No current runs through the load.
If the (let's say 50V peak ) sinewave starts to rise at the transistor base, the emitter starts to rise too, so does the "hot" load terminal.
Maximum reached?= 50V (let's ignore losses).
Can't get any higher than that, the transistor is saturated and can't offer more than B+ .

Now you have an output autotransformer, loaded with a resistor.
Voltage on the amp out terminal starts to rise as before, in this case this voltage starts to inject current in the transformer winding, which is absorbed by the load. It will show a reflected impedance, of course.

*Now* we repeat the experiment with no load applied.
Voltage will rise as before, *even without load* a "magnetizing current" will circulate in the transformer primary; when transistor saturates that inductor will try to keep that magnetizing current in its last known value and will develop a voltage pulse on that hot terminal, "whatever's necessary to keep that current at the same value"; that's the flyback effect.
That voltage will go beyond the 50V available; it does not depend on the rail voltage but on the magnetic energy stored in that core.
(That's exactly how those tiny step-up converters work, which get 12V out of 5V and similar wonders)

What's the transistor situation?
Collector: +50V (it was always there)
Base: +50V (the driver transistor has saturated and can't supply any more)
Emitter: that unloaded inductor keeps rising voltage there *way* beyond 50V; the Emitter to base is reversely polarized until breakdown= BAD.
It does not seem that much, even on a scope, because that breakdown voltage is around 5 to 7 V.
Will that transistor like (or survive) the output current it delivered before, now being shoveled through its throat, the wrong way?
Don't think so.

[Steve Conner]

Well, this is all true, but any SS amp with BJT output devices should have the catch diodes, because otherwise even the reactive load of a woofer voice coil can damage it.

They are the exact same thing as the diodes sometimes added from plate to ground in tube amps, except there isn't the issue of the OPT leakage inductance, so they clamp better. KMG's power amps use MOSFETs, which have those diodes built in, so they have a "catching" mechanism too.

What the diodes don't do in a tube amp, is save the screen grids from burnout on an open load. But transistors have no screen grids, so they don't die that way. Hey, maybe we can improve reliability in tube amps by putting a clamp diode from the plate to the screen. A 100V TVS or neon lamp or whatever, so the plate can still get below the screen, but not by enough that the screen draws lethal amounts of current.

Also, in the case of a SS amp driving an open-circuited transformer, the load might be reactive, but it's a high impedance, so current draw and dissipation will be low. I don't see any risk of second breakdown, and I conclude that a good SS amp should be unharmed by an open circuit, even if it has an output transformer.

Now, if the transformer secondary was shorted, or the transformer was saturated by subsonic garbage, that could get nasty, but a properly designed protection circuit (when did you last see one of those? ) should deal with it.

[g1]

Thanks for the info guys. What I was getting at with my question was whether this was a big enough issue to over rule the caveat: "don't hook up a solid state amp to a load until it's repaired" (or at least running proper without a load).
Seems to me the risk is not significant enough to worry about.
My personal feeling is that any amp that needs a load at all times should be labeled as such (or noted in user manual) or have a shorting jack/resistor at the output.
Unfortunately that leads me to another question: can a shorting jack be used on the output of a solid state amp that has an output transformer?

[J M Fahey]

I think a short is always bad for an SS amp.
Anyway, *if* you want something equivalent without shorting anything, you just need to design some simple circuit that mutes the power amp with no plug connected.
Anyway I find this a non solution for a non problem.

[petetexas]

Hi,
Just to add my little bit, I recently purchased an Ampeg SVT 450H Bass head (All tranny)
I e-mailed Ampeg with the question "can I run the amp using the DI out for recording, and NOT connection any speakers"
The reply came back " We as a company do not recommend the amplifier being run without the correct speakers attatched , but being that the Amp is a Transitor Amp, no harm will occur " - Covering every situation eh !
Pete

[gcbruce]

Has anyone tried their Fender Mustang V head, to see if it works without speakers connected ? Wouldn't that be the real way to find out ? From what I am reading, no damage whatsoever should occur with this particular amp.

[gpar7]

Hi , I have a mustang V head. I've been gigging with it and using it at home with the puter , fuse and playin along with u tube . i have played it plenty with no speakers or phones using the usb to puter only . it doesnt seem to bother the amp at all to play right to the puter.( its got latency issues though with my set up) I have also played it thru a crate 4*12 in stereo parallel @8 per side , then only one side thru 2 16 om @ 8 ohm , sounded great btw , did two gigs this way. I now have it running thru a vox 2 12 vt100 cab , it sounds awesome this way . ( thats 16 ohms per side , stereo sounds great)

[gcbruce]

Mustang

I tried my Mustang V head for approximately 1/2 hour with no speakers plugged in. Listened through headphones. No damage or heating or anything . Worked like a charm. Also then removed the headphones from the jack. Still no problem. I am convinced that this head needs no load to still operate perfectly.