Impedance measurement question

[tedsorvino]

It has worked ok for about 3 years with 2 tung sol tubes. So...even if it eventually burns them...I guess it would be about time to change tubes. I mean...changing 2 tubes every 4-5 years it's not the end of the world.

[Leo_Gnardo]

I'm quite hesitant with some things because in the past I had bad experiences with ss amps. I've used a 4 ohm speaker to a switched impedance ss amp (with the switch at 8 ohm) and I blew a fuse.

We're wandering a little but some SS amps, notably certain Peavey models, had output autoformers where you could select the optimum winding for the impedance you intend to drive. A mismatch could well result in drawing extra current through output transistors and roasting the fuse, hopefully before the transistors fry. A joke from decades ago, transistors are electronic devices whose function it is to short out and save having to change fuses.

Also, some of the larger Fender amps, mid-70's onward, had a special jack on the ext speaker function, that had a switch which selected the appropriate output tap. Some ran the speakers in parallel, others in series. Twins, Bassman 100 or 135, Super Reverbs among others. Look inside & see what you have Fender owners, repair persons, and others, also consult the appropriate schematic. Some of the more modern Fender HotRod & Blues series combos have a similar setup, all those that I've seen with a switching jack on the ext speaker output are parallel wired.

[nevetslab]

Back in 2018, I had set up an automatic output impedance plotting system, amp under test was a Mesa Simulclass 2:90 Power Amp. The basic concept was feeding a constant current into the Output of the Tube Amp. Since the output impedance changes with frequency, a compressor loop was added into the measurement system to maintain the constant output current fed into the output of the amp under test. It was presented in this MEF post: Tube Amp Output into Resistive vs Speaker Loads. - Music Electronics Forum (music-electronics-forum.com)

You'll have to wade thru the first batch of posts to get to the test set-up and resultant plots.

[J M Fahey]

That.

What the power of the test resistor has to be in WATTS?

In principle, higher than expected measurement, so 60 to 100W.
That said, if you will not leave amp full power for a long time (say burning it in) but a couple minutes for measurement, you may use a smaller o e, say 25W
Stop as soon as you see smoke he he.
Personally I use a 20/25W one and drop it in a bucket of water, not kidding.
Easily stands 100W or more.
Clean tap water is not conductive.

And what happens when you connect a speaker to the external speaker out of an old fender tube amp impedance-wise (is it in parallel or in series to the internal speaker)?

As mentioned above, 99% of classic Fender amps used parallel output jacks, no tap switching.
In fact, no extra tap was available, period.

Personal idea, and I may be wrong, is that Fender did not actually expect people adding extra cabinets, and statistics must have proven this.
The very very few who did, which we all know, were famous bands playing to huge audiences, think the Beach Boys, Chuck Berry, Ted Nugent, maybe Jefferson Airplane or the Grateful Dead, all from the heroic era of absolutely inadequate PA systems, BUT: they used the best, most expensive amps, typically *JBL loaded* Dual Showman or Twin Reverb, adding an expensive but very loud extra Dual Showman 2 x D130 cabinet to each head or combo.

Unavoidable slight power loss because of mismatching was more than compensated by the added cone surface , focusing and plain "acoustic gain" offered by those advanced speakers.

Very much doubt a musician bothered to boost, say, his Super Reverb by adding more cabinets, he´d simply upgrade to a Twin Reverb.

If a plain Twin was not enough, he´d get the JBL upgraded one.

Only if *this* one was not enough, he might sit it on top of a Dual Showman cabinet.

Just thinking aloud and remembering old movies and pictures.

[Helmholtz]

Back in 2018, I had set up an automatic output impedance plotting system, ....

You may measure amp source impedance with a good LCR meter that displays AC resistance. But that's not what you want here.
(A tube amp would require some base load for stability).
Or it can be calculated from the output voltage difference between different load resistors at some medium output.

You may find values between 1 Ohm and 100 Ohm depending on tube type and the amount of NFB used in a 2 tube PP amp.

Amp source impedance does not equal rated load. It's only useful to determine the damping factor.

[Chuck H]

First, it's not uncommon for "other brand Fender copies" to use different values for the transformers and/or different ohm loads for the factory speakers. It's usually a matter of economics and other manufacturing logistics. As long as the "schematic" is copied (didn't often include transformer impedance) the maker sells it as an alternative for affordability and/or geographic availability. If a different primary impedance from the original Twin series amps is the case that may actually be to your benefit with the kt66's. Like Helmholtz said, 4 to 8k for those tubes. Since most Twin series amps ran around 2k (IIRC). Pull a pair of tubes and plug it into an 8 ohm load (if the amp was rated for 4 ohms like a twin) and you have a 4k load. Low-ish. Especially since this is a Twin copy and may run a similar high-ish plate voltage. You can hope your particular transformer is more in the generic 6.6k range (they often are for copies) and if so then you'd be close enough to ideal for an 8 ohm load (so, safe to mismatch up or down). If you're actually in the 4k range I would avoid mismatching down. But...

You don't know if you don't measure it

[Helmholtz]

Clean tap water is not conductive.

It actually is. But in the 1M range (at least over here). So it won't disturb.

[tedsorvino]

First, it's not uncommon for "other brand Fender copies" to use different values for the transformers and/or different ohm loads for the factory speakers. It's usually a matter of economics and other manufacturing logistics. As long as the "schematic" is copied (didn't often include transformer impedance) the maker sells it as an alternative for affordability and/or geographic availability. If a different primary impedance from the original Twin series amps is the case that may actually be to your benefit with the kt66's. Like Helmholtz said, 4 to 8k for those tubes. Since most Twin series amps ran around 2k (IIRC). Pull a pair of tubes and plug it into an 8 ohm load (if the amp was rated for 4 ohms like a twin) and you have a 4k load. Low-ish. Especially since this is a Twin copy and may run a similar high-ish plate voltage. You can hope your particular transformer is more in the generic 6.6k range (they often are for copies) and if so then you'd be close enough to ideal for an 8 ohm load (so, safe to mismatch up or down). If you're actually in the 4k range I would avoid mismatching down. But...

You don't know if you don't measure it

Thanks once again for the reply. The transformers are not similar to original fenders. I remember I had checked it in the past. But they may measure similar. Hadn't measured them back then. The reason I chose 8ohm speakers for its main speaker ( I'm using one speaker - either 10'' or 12'') is because I bet 8 ohm is on the safe side either way (I bet it's an 8ohm amp, but even if it's a 4ohm one, it works well with 8ohms). But having worked with a 4ohm speaker sometimes the result was not bad.

[mozz]

"But you won't damage the amp running 8 ohms where 4 ought to be or some such."
Many a Fender Champ has it's output transformer blown by someone running a 8 ohm speaker where it should be 4.

[g1]

I think marshalls are more generally sensitive to the matching that way too. But I also think both the Champ and marshalls are more likely to be driven much harder than classic fender clean style amps.

[J M Fahey]

It actually is. But in the 1M range (at least over here). So it won't disturb.

Nitpicking.

We are talking 2-4-8-16 ohm range here, so 1M is irrelevant.

Many a Fender Champ has it's output transformer blown by someone running a 8 ohm speaker where it should be 4.

Really?
Leo must have used real cheap stuff in those OTs.

and/or geographic availability.

Can give you one example.

Very common Fender output impedance was 4 ohm, based on using 2 8 ohm speakers on the "large" ones (Bassman 50 - Bandmaster - Twin - Dual Showman) which where available in USA (jensen or JBL).

Our "go to" speaker was an excellent 20 - 25W 8 ohm one, 4 were needed for 100W amps, so 8 ohm total was the usual choice.

I was the exception to the rule, because I started making my own 8 ohm 50W speakers, quite a feat way back then, so my bread and butter 100W amps were 4 ohm, same as my 2 x 12" cabinets.

Being SS and having limited rail voltage by 2N3055 (only game in town) , the scale was: 100 W RMS into 4 ohm , 200W RMS into 2 ohm (4 x 12" in parallel) and 400W RMS into 1 ohm , using 2 x 2 ohm cabinbets, 4 x 12" each.
Those amps and cabinets DID move some air!!!!

[Leo_Gnardo]

Leo must have used real cheap stuff in those OTs.

Leo Fender, that is. Champs were intended as cheap practice amps for beginners, to be played at polite volumes. (HA! Imagine that!) Those who wish to spend the extra cash, go ahead and replace your thumbnail sized OT with a more hi fi one, and while you're at it substitute a more robust speaker.

On occasion I have demonstrated for customers a Champ or similar, running into a pokey 4x12 with say Celestions or my favorite Altec 417's. "Omigod, you can do that???" Followed quickly by "Holey Moley, that sounds fantastic!"

Can't say I've run across any Champs whose OT's or 6V6's were obviously wrecked by running an 8 ohm speaker. Champs, for such a simple amp they have so many ways to break down. I'll save time & not list them all but you repair guys know what I mean, right? For an amp that's supposed to be run at polite volumes, theoretically anyway, the whole output section is under some stress even with no signal passing through, no wonder they croak fairly often.

[Chuck H]

For an amp that's supposed to be run at polite volumes, theoretically anyway, the whole output section is under some stress even with no signal passing through, no wonder they croak fairly often.

This^^^^^^^^^^^^^^^^^^^^^^^^

Most Champs ran hot and later Champ models had stupid plate voltages with the same grandfathered in cathode resistor. And that OT that was about the size of a 9V battery. When the Princeton evolved to two power tubes the Champ and Princeton still used the same PT. This put the Champ in the 400+Vp range and biased at something like 130% diss with the stock cathode resistor. Crank 'em up and things pop. Though it's usually a 6V6 and not the OT. I rebuilt a silver face not long ago with a 10" 8 ohm speaker, an EL34, bigger OT matched to the new load and some circuit modification. Dumble-esque TS values and a diode to clip the asymmetric end of the swing just post clipping at the power tube grid. This makes the clipped tone more push/pull sounding and was an "improvement" over the rhaspy cranked Champ tone (IMHO). It put out about 7.5W and sounded REALLY good. It's the high Vp and small OT that are the lynch pins in those models (and the tiny speaker I suppose). The rest of the amp is pretty solid with typical Fender build and an over rated PT. I'd recommend the above mods to anyone with a SF Champ.

Sorry for the continued off topic.