| Author |
Topic: output impedance? |
Bobby Burns
From:
Tennessee, USA
|
Posted 23 Jan 2010 6:24 pm |
|
| Is there anyway to know what impedance speaker load to use, if it is not listed on the amp, and you don't have a schematic? Would this differ from solid state to tubes? |
|
|
 |
Danny Hall
From:
Nevada, USA
|
Posted 23 Jan 2010 6:53 pm Re: output impedance?
|
|
| Bobby Burns wrote: |
| Is there anyway to know what impedance speaker load to use, if it is not listed on the amp, and you don't have a schematic? |
Yes, there is a way. Your tech needs to put it on the bench and apply a volt or two of AC to the speaker jack. Then he reads the resulting voltage between the plates of the two output tubes.(tubes yanked) Simple division gives him the turns ratio. One then squares the turns ratio and multiplies the resultant by (say) 8 ohms speaker impedance. By looking at the factory specifications and output curve of the tubes in "push pull" he determines what the ideal speaker load would be.
| Bobby Burns wrote: |
| Would this differ from solid state to tubes? |
Not really, but the numbers are completely different. A transistor is inherently a low impedance device while a tube is a high impedance device.
_________________
The Last of the World's Great Human Beings. Ok, well maybe one of the last. Oh alright then, a perfectly ordinary slacker. |
|
|
|
J Fletcher
From:
London,Ont,Canada
|
Posted 26 Jan 2010 5:22 am
|
|
| If it's tube amp, I would think that the load impedance that gives the most clean output power would be the ideal speaker load. You would need a signal generator, oscilloscope, dummy load(s), and ac voltmeter to measure this. And a calculator...Jerry |
|
|
|
Danny Hall
From:
Nevada, USA
|
Posted 26 Jan 2010 11:12 am
|
|
| J Fletcher wrote: |
| If it's tube amp, I would think that the load impedance that gives the most clean output power would be the ideal speaker load. You would need a signal generator, oscilloscope, dummy load(s), and ac voltmeter to measure this. And a calculator...Jerry |
Dummy loads don't react to changes in frequency. Literally. There are lots of schemes available for speaker load emulators but none on them do it perfectly; and different speakers have different reactance curves.
It's best to make the tube work in a straght part of it's load curve. With tubes like the 6CA7/EL-34 that line is straight over a wide impedance range. Marshall runs EL-34s at 3400 ohms per push pull pair, while Trainwreck, and most HI-FI designers shoot for around 6600 ohms. 6L6GC have a load center of about 5000 ohms while 6V6 types load at 8800 ohms or so. Generally, tubes will create more headroom and "clean" power the higher in the load range you go. But with musical instrument amps, a sterile approach using static load resistors a scope and a signal generator won't yield a "musical" amplifier. The "magic" of some legendary amps is where they break the rules, not where they adhere to formulas.
Leo, as a rule, ran his power tubes well above the recommended voltage limits. While it's true that he used common topologies out of reference materials available at the time, he experimented using his ears and his soldering iron to get the honk and twang he liked for western swing. Jim Marshall couldn't have cared less about western swing but he was no genius either, so he copied Leo's Bassman amp and once again, using his ears, found the legendary sound of British Rock by changing the transformer turns ratio, input tube type and biasing arrangement. All these changes hardly constituted "the most clean output power"
Truly great equipment comes from solid engineering and an artists hearing and sensibilities and on extraordinary accidents.
_________________
The Last of the World's Great Human Beings. Ok, well maybe one of the last. Oh alright then, a perfectly ordinary slacker. |
|
|
|
J Fletcher
From:
London,Ont,Canada
|
Posted 26 Jan 2010 1:17 pm
|
|
Hi Danny
That's interesting, but I would say for the most part, if you have a tube amp, and want to determine the correct speaker load, the method I describe will get you there.
What would be an example of a situation that would yield an incorrect result using the method I've described?
"A sterile approach using static load resistors a scope and a signal gwenerator won't yield a "musical" amplifier." ...I'm not describing how to design a good sounding amp, I'm describing a method to determine the optimum speaker load, for best power transfer from the amp to the speaker.
Determining the turns ratio of the output transformer can be useful, and I've done it many times, for many years.
As far as 6L6GC's go, a push pull pair with 450 volts on the screen and plate, fixed bias, seems to like about 4000 ohms plate to plate impedance. At least in Fender amps.
Which 6L6 amps have a primary impedance of 5000 ohms?
Cathode biased, with 400 volts on the plate, and say 360 volts on the screen, 6600 ohms.
Deluxe Reverbs with 410 volts on the plate and screens, 6600 ohms to 8000 ohms.
Very early Marshalls, 6600 ohms.
I've heard all that stuff about breaking rules and musical amplifiers,many times. It's interesting...Jerry |
|
|
|
Pete Burak
From:
Portland, OR USA
|
Posted 26 Jan 2010 1:56 pm
|
|
| What is the Brand/Model of the Amp? |
|
|
|
J Fletcher
From:
London,Ont,Canada
|
Posted 26 Jan 2010 2:21 pm
|
|
| What is the Brand/Model of the Amp?....yeah,you could go that route!...Jerry |
|
|
|
Danny Hall
From:
Nevada, USA
|
Posted 26 Jan 2010 8:37 pm
|
|
Jerry:
No fight intended. What do you look for on the scope? Maximum output power without clipping? In which stage? What if the input to the finals is already clipping? How do you tell when the output tubes enter crossover if the signal they receive at the grid is already clipping hard?
Optimum output from the power tubes can be discovered by feeding their grids with a clean sine wave at 80/90/100Volts whatever, PtoP and looking for crossover clipping to start. But what good is that going to do you if the amp was not designed for that parameter? All I'm saying is Leo didn't design that way nor did John Marshall or the folks at Vox or Hiwatt or Supro. They generally employed load impedances much lower than optimum then drove the grids like mad.
In transmitter design and Hi-Fi things are done differently.
_________________
The Last of the World's Great Human Beings. Ok, well maybe one of the last. Oh alright then, a perfectly ordinary slacker. |
|
|
|
J Fletcher
From:
London,Ont,Canada
|
Posted 27 Jan 2010 5:40 am
|
|
Danny, if somebody showed up at my house and wanted to find out the correct speaker load for their amp, assuming that there was no indication on the amp, or on the scematic I'd probably do something like this:
First I would look at the amp. Is it a combo amp with one speaker made in the USA or Canada? Probably 8 ohm is the correct speaker Z, unless it's a single ended 6V6 amp, where the Z is usually 4 ohms. Two speaker combo amp? Probably 4 ohms. And so on.
Made in Europe? Different story usually, maybe 16 ohms for a single speaker, 8 ohms for two speakers.
Is it a head only? Pretty much follow the same procedure.
I would then connect the speaker output jack to a 8 or 4 or 16 ohm (determined by the above method) dummy load, connected to a scope, put a 400 hz 50 mv signal into the input, and adjust all amp controls for the cleanest undistorted power output at the speaker jack.
By looking at the tube complement, and size of the transformers, I would already be expecting a certain amount of power. Say somebody showed up with an amp with two 6L6's and hefty transformers, I would be looking for 40 watts or so.
I calculate how much clean power the amp produces into 8 ohms (for example), then switch to a 4 ohm load and do the same calculation. If it's a British or European amp, best to check at 16 ohms too.
Anyhow, it boils down to the correct speaker Z is either 16, 8, 4 ,or 2 ohms in 99% of the amps out there. I know there are exceptions.
If there are multiple speaker jacks, best to determine if they are paralleled or come from different output transformer secondary taps, using the same method described.
I know you catch my drift here.
"Jerry:
No fight intended. What do you look for on the scope? Maximum output power without clipping? In which stage?"
Yes, at the output jack.
"What if the input to the finals is already clipping?
How do you tell when the output tubes enter crossover if the signal they receive at the grid is already clipping hard?"
Crossover distortion looks different than clipping on the scope. If the amp is starting to clip at 15 watts and I'm looking for 40 watts, then I'm either connecting the wrong load or there is a fault in the amp.
"Optimum output from the power tubes can be discovered by feeding their grids with a clean sine wave at 80/90/100Volts whatever, PtoP and looking for crossover clipping to start"
Well sure, but it's usually not necessary, if there's a fault somewhere perhaps. 30 volts should do it..Jerry
"All I'm saying is Leo didn't design that way nor did John Marshall or the folks at Vox or Hiwatt or Supro. They generally employed load impedances much lower than optimum then drove the grids like mad."
I'm not designing, I'm trying to determine the correct speaker Z for an existing amp. If you want to discuss design, I'm interested, but that's a different thing. Ken Bran at Marshall didn't so much design as take an existing design and run with it. Who knows what Leo Fender was thinking? We can guess by looking at the evolution of his amps from the late '40's to the mid '60's though...Jerry |
|
|
|
