Speaker ohms???

[Canucker]

I'm sure the answer will probably be over my head....or so simple I'll wonder why I didn't see it (although I'll bet money its the first one)...but what decides on how many OHMS a speaker should be in an amplifier design?

I have a nice little vintage amp that I want to upgrade the speaker for as it is so NOT the original....the amp is a Maestro GA 1RT and somebody threw in a Realistic brand speaker....which makes me think it was just lying around and fit so why not? So I'm not trusting that just matching the 8 ohms of that speaker is the way to go....and I can't find the right info ANYWHERE online. I do have the schematic though but there is no other technical info around out there......

[ashcat_lt]

I'm guessing the answer is quite a bit more complex than I could possibly start to describe...

My understanding is that for maximum power transfer you want the load impedance to be as close to the source impedance as possible.  Since this is pretty much the only instance in our work where we actually want maximum power transfer, I think that's maybe the place to start.  Without info on the OT, I don't think we can be completely sure what that might be from the schematic (though I wouldn't be surprised if somebody around actually just knows), and I don't know how to tell you to measure it...

That said, it's a safe bet that the "rated" speaker impedance there is either 4, 8, or 16, and I've always heard that most amps can handle being off by a factor of 2 in either direction.  So 8 is very probably safe either way, right?

[CodeMonk]

Don't have an answer for your question, but I did find this:
http://shop.guitarpoint.de/media/images/popup/Maestro61ReverbGA1RT_108406_8.jpg

Looks like you can see the model number on there.
Maybe you could contact Jensen and ask them.

Sorry, thats all I got

[Transmogrifox]

In an output configuration like this, the speaker impedance is reflected to the plate of the tube as (speaker_impedance)*N^2, where N is the turns ratio (plate_side/speaker_side).

If the speaker impedance is too large, then you will drive the output into saturation (clipping) before you get much power output.  

If the speaker impedance is too low, you risk either damaging the tube, or damaging the transformer, or both.

I would say an 8 ohm speaker is the safest guess.  If you feel the amp can safely put out more power before clipping, then change to a 4-ohm, or 2 8-ohm in parallel.

If you had an oscilloscope, then measuring turns ratio would be a trivial matter.  If you know the transformer turns ratio and high voltage connected to the transformer, then you can determine the power output for any speaker impedance.

The next thing would be looking at the tube datasheet for power rating on its plate.  From there one could deduce the speaker impedance design target, and also what power rating you should expect your speaker to have.  

[R.G.]

How hard are you willing to work?    

The 6BM8 output tube is specified for a nominal recommended plate load of 8K with ~270V on the plate. Most amps of this time simply used the recommended amplifier conditions of the tube makers, so it is a very good bet that the plate load it wants to see is 8K.

The question is then reduced to "what speaker load makes 8K appear on the plate?". As it happens, this is simple, if a bit laborious to find out.

The way you do it is to remember from your transformers and motors course as a sophomore EE that the impedance ratio of a transformer is the square of the turns ratio, and that the turns ratio is the same as the unloaded voltage ratio. So you measure the voltage ratio, square it, and that is the impedance ratio. With that ratio in hand, you assume the plate is seeing 8K, and divide by the impedance ratio to get the necessary speaker ohms.

This involves unplugging the amp and removing the output tube to get at the primary with no power or tube attached to it. You also make sure that there is no load on the speaker lines, disconnecting any internal resistors and so on that might mess up your measurements. Then you put an AC voltage at some middle frequency - 500Hz to 1kHz is good - on the secondary, momentarily misusing it as an input, not an output. You carefully measure the voltage on the secondary, setting it to some easy to remember and easy to measure value, like 1Vac on your meter.

Then you measure the AC voltage on the primary. This will be bigger. Maybe a lot. So with the voltage on the primary and secondary measured, you calculate
Vratio = Vprimary/Vsecondary, and get some number. Then square the number. This is the impedance ratio.

If the transformer happened to be designed for 8 ohms, working backwards we would expect an impedance ratio of 8000/8 = 1000. This is the square of the voltage ratio, so we take the square root to get 31.62. That says that the transformer will make a voltage on the secondary that is 1/31.62 times the AC voltage on the primary. So if you used a 1Vac input to the secondary, you'd see about 31 - 32Vac on the primary.

If the trannie was designed for 16 ohms, the impedance ratio would be 8000/16 = 500, and the voltage ratio would be 22.3. If it was designed for 32 ohms (yep, some of these were used way back when), you'd have an impedance ratio of 8000/32 = 250 and a voltage ratio of 15.8.

[Canucker]

mind blown!!!  My assumption that it wasn't that simple was right on the money. I have so much left to learn. I full nights sleep and a thorough reread is required. Thank you for letting me pick your brains yet again....your huge huge brains! 

[Canucker]

Don't have an answer for your question, but I did find this:
http://shop.guitarpoint.de/media/images/popup/Maestro61ReverbGA1RT_108406_8.jpg

Looks like you can see the model number on there.
Maybe you could contact Jensen and ask them.

Sorry, thats all I got

Nice....I enlarged the photo and stared for a minute or two and saw that there are actually two lines to the code....saved the image....opened it in microsoft pain and then inverted the colours to reveal the full code..... i have a pretty limited knowledge of electronics (comparatively) but I have creative problem solving ideas sometimes!

[PRR]

> somebody threw in a Realistic brand speaker

So we do NOT know the original Ohms.

In older gear the speakers were almost universally "3.2 ohms" (what we now call 4 Ohms). The low impedance is the cheapest construction. The low impedance is not a problem when the speaker wires are short (probably under 1 foot?).

Somehow around 1970 the most-common impedance became 8 Ohms. This may relate to the longer wires in home hi-fi, the current limitations of early transistor amps, or simple change of fashion.

I would *ASS*ume 4 ohms is most likely.

As R.G. says, you can measure the OT, assume a likely primary impedance (I get 7K for these voltages, but 8K is all the same), and do math.

However if you find a plot of power output versus load impedance for a power pentode you will see that the "optimum" is very broad. 2:1 either way is not a real change of power.

This is complicated by the fact that most of these curves over-simplfy the case. Most over-drive the amp at low impedance. The even/odd distortion balance changes with impedance. Also with drive level (3rd comes up MUCH faster than 2nd).

Now if you were playing the Stadium and knew you needed a 200 Watt amplifier, you would want to get pretty darn close to whatever "optimum".

But at-most this is a three and a half Watt amplifier. Do you have venues where 3.5W works but say 2.2W will get you laughed out of the club? No.

If you must replace the speaker (the Realistics were not total junk), get an 8 Ohm because you might end up using it in something else, and today 8Ohm is most common.

You will NOT melt *this* amp with a too-low load (you also won't get more power). Single-ended tube amps will suck-in a certain amount of DC power, and divert none some or much of it to the speaker. Going too high OR too low simply reduces the most that can be diverted to the speaker.

You *may* hurt it with a way-too-high load and LOUD playing. This is the "American Woman" Herzog. A Fender Champ intended for 3 Ohms but mis-loaded into the 1Meg (1,000,000 Ohms) of another guitar amp. When LOUD, the tube plate cut-off KICKS the OT to super-high voltages and punctures the insulation. You must always have "some" reasonably low load. Herzogs often used 5 and 10 Ohm resistors to save their OTs. If no resistor and the speaker might get un-plugged, the safest thing for *this* type amplifier is to SHORT the output. (Quite the opposite of what you do for transistor amps and the very biggest tube amps.)