|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
66.177.177.194
In Reply to: RE: update 2 posted by vinnie2 on February 17, 2019 at 03:59:41
With all that free time, connect a scope, let it clip into an 8ohm load and read the voltage.
Inquiring minds would like to know!
Follow Ups:
Why is it everyone assumes that anyone who is retired has nothing but free time? I have more than when I wasn't retired, but it is not unlimited. I still have all my chores around the place and doing things to help my with her art work and studio among others.Anyway, you need to be more specific. Connect a scope to where? How do I know when it is clipping? Anything else that pertains?
Edits: 02/17/19
1khz sine wave at input
Connect scope to output with an 8ohm load(load resistor best)
Turn up input until sine wave just starts to flatten at the peaks and read the voltage, its subtle on a tube amp but obvious.
Its crude but gets you close to output wattage. I did it this way for 30 years until i invested in an analyzer.
Well I gave that a try but I must have done something wrong because I came up with a voltage of 48vdc. If I were to use the formula P = V*V/R I would get a ridiculously high figure. Is there another way to calculate it that would give a much lower value? If not then I am doing the measurement incorrectly.
Edits: 02/18/19
"Well I gave that a try but I must have done something wrong because I came up with a voltage of 48vdc."First of all, the voltage you should be measuring across the 8 ohm load resistor is ac, not dc. (You can't get dc out of a transformer.) Trying to guess what you actually meant to type, one possibility is that you meant 48v pp (that is, 48 volts peak-to-peak of the 1KHz sine wave). That would correspond to 48/( 2 * sqrt2) volts RMS, which is about 17 VRMS. This would imply a power of 36 watts. That seems implausibly high, though.
However, there are too many uncertainties and guesses involved, with trying to estimate what you meant to type when you wrote 48vdc. Perhaps you could clarify?
Edits: 02/18/19
I meant peak to peak. In the morning I will see if I can set the scope to measure AC instead.
"In the morning I will see if I can set the scope to measure AC instead."
I shan't even try guessing what you mean by that.
There's nothing wrong with measuring peak-to-peak voltage of the sine wave; just divide by (2 * sqrt2) (which is about 2.83) to get the RMS voltage, like I said.
But the answer you got seems to be coming out to be much too large, implying 36 watts into 8 ohms, which just isn't believable for a single-ended amplifier like you have built. So there is evidently something wrong with your interpretation of what you are seeing on the scope. (Misreading the volts per division setting?) Or else something wrong with the scope?
You could practice by checking the output of your signal generator on the scope, and comparing the peak-to-peak voltage (divided by 2 * sqrt2) with the rms voltage you measure with a digital voltmeter. Do those two results agree, within expected experimental errors?
Then what exactly did you mean by the voltage I should be measuring is AC not DC? That would imply that either could be measured. Please add a little clarification.
By the way; you did have an 8 ohm dummy load connected across the OT secondary when you made the measurement, right? Because first of all, it could be bad for the output transformer and/or output tube if you run it without a load. And also, who knows what voltage the secondary would develop if you run the amplifier to clipping without a load. (And one really does not want to find out!)And, incidentally, if it really had been giving 48v peak-to-peak across an 8 ohm load, that resistor would have been getting very hot very quickly, with 36 watts being dissipated in it. I assume it wasn't getting very hot?
Edits: 02/19/19
"Then what exactly did you mean by the voltage I should be measuring is AC not DC? That would imply that either could be measured. Please add a little clarification."What I mean is that you cannot be getting a steady DC voltage from the output transformer secondary. You must have been seeing an AC voltage on your scope. And indeed, you then confirmed that when you had said 48vdc, you had meant to say 48v peak-to-peak AC voltage. Which would be fine, except that 48v peak-to-peak is way too high to be getting from that single-ended amplifier you have breadboarded. So something must be amiss with your measurement.
I am just suggesting that you try playing around with measuring the RMS output voltage from your signal generator using a dvm, and then comparing it with the peak-to-peak voltage that you can easily read off from the scope. The two should be related as I said before. If they are not, then it would provide concrete evidence that something has gone astray with your interpretation of the peak-to-peak reading from the scope. (Or that something is wrong with the scope.)
There isn't really an issue of "setting the scope to measure AC or DC." It is just showing a trace of the waveform you feed into it. And you are feeding in an AC waveform from the output transformer. (It is true one can usually set an AC or DC input coupling on the scope, but that is not relevant here, since your signal you are feeding it will be pure AC, with no superimposed DC.)
But since the peak-to-peak voltage you are reporting from the amplifier seems to be implausibly high, the obvious first thing to try would be testing things out on a convenient AC voltage such as from the signal generator itself, and comparing with the RMS voltage measured with the dvm.
Edits: 02/19/19
Yes, I was using an 8 ohm resistor across the opt output leads, and no the resistor does not get warm, and yes I realize that voltage is way too high.
However;I went ahead and tried your test and got a 0.12 volts reading with the dvm on the signal generator output and 0.13 volts using the scope. I was very careful to follow the directions in the scope manual for measuring voltages. That seems like a pretty good match to me. HOWEVER, when I switched over to the amplifier everything went gunny sack. I did not change any of the settings, but I had to go clear up to 50 volts/division to get a sine wave that would stay within the screen of the scope, and that was with only 1/3 volume on the amp. I think the scope is working ok because it read the signal generator ok. I do not have a clue why the voltage is so high when I switch to the amp. I follow the instructions in the scope manual again, but obviously something is wrong somewhere. The final thing is the amp doesn't play any loader than my SET 2A3, so I know the voltage has to be wrong. I am just going to build the other channel for now and forget about trying to figure the watts unless someone has a eureka moment from reading these posts. Maybe deaf by khorns can add something if he checks back in. Thanks for the effort.
Edits: 02/19/19
Dont use the scope to read voltage
Clip in your RMS meter and set to AC
Turn up the volume until you see the sine wave start to flatten at its peaks and read the voltage.
E2/R= watts
I would say you want to adjust the level to where the amp is just below visible clipping.Turn up the volume until you see the sine wave start to flatten (clip) at it's peaks then turn down the volume just a little until the sine wave is not flattening at its peaks and read the voltage.
If clipping is visible on a scope, the distortion is quite high. That's not usable power.
"..."full power" for the purposes of measuring distortion in amplifiers is usually taken as a few percent below clipping..."
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 02/22/19
It works slick as a whistle when you use the right procedure. Now the only problem is deciding where the peaks start to flatten out. It can make quite a bit of difference in the voltage reading. Can you post a picture of what you consider it to be? Right now I can come up with anywhere between 2 and 4 watts output for the 845 depending on where I say the flattening starts.
It is usually pretty unambiguous when the clipping starts, in my experience. The scope trace looks like a nice sine wave, and as you crank up the level you reach a stage where the tops (and or bottoms) of the sine wave start to flatten off. The maximum level before clipping is the level just before any sign of flattening of the positive and/or negative peaks begins.
This is what 2.4% distortion looks like, you can see one half of the cycle is pretty flat. If you adjust the volume down just enough to make the flat part go away, your at about 1-1.5% distortion which is acceptable for a tube amp. This should get you pretty close without using an analyzer.
I think I have now tweaked the 845 about as much as I can. It sounds pretty good and the op points are close to target values.
Target values
plate volts 430 vdc
plate current 62mA
grid volts -51 vdc
As breadboarded now
plate volts 487 vdc
plate current 62.2 mA
grid volts -52.2 vdc
I like the sound of it in mono, so now it's time to breadboard the other channel and see how it sounds in stereo. After that I will take another shot at the 813. Wish I had some low voltage target op points for it, but I have not found any so far.
It may be my equipment or the operator, but I am having a hard time getting a sine wave like that. Here's what I have been able to get so far. It's kind of weird because it starts to flatten out at about 1/2 volume (2.3 vac) and the flattens more as you increase volume until you get up to almost full volume. That's the top pic (9.8 vac). The interesting part to me is I don't really hear any increasing distortion in the music until almost full volume. So what do you guys think?
Edits: 02/23/19 02/23/19 02/23/19 02/23/19 02/23/19
That top pic is pretty ugly, probably your driver giving up
The sine wave should be symmetrical, whats your driver tube?
1/2 of a 6sn7
Edits: 02/24/19
According to my simulation, your 845 is starting to pull grid current at about 3 volts output (about 1 watt) using your operating points. Your 6SN7 driver is not going to do its job alone. Your sine wave isn't clipping but it sure is producing harmonics (on paper). It would be interesting to see what the real harmonics would be using your OPT. Depending on your music source, your probably not picking up on the harmonics but they are definitely there according to your scope. What is your plate and cathode resistor values.
You should look into a UMC204HD and the free ARTA Software. You get a scope, THD, FFT along with many other features. Its easy to implement and setup, I set it up in an hour. It's not too bad and reads very close to my HP Distortion analyzer.
Edits: 02/24/19
What would happen if I used both halves of the 6sn7 in srpp?
I'm not an expert in SRPP but I don't think so. It may help the output impedance and distortion but it wont provide grid current.
An IT transformer or a mosfet (grid drive) would be better suited for this or run the voltage up a little higher.
With your current driver and voltage, not sure you will get more than a couple watts on a low voltage 845. Those transmitting tubes start pulling grid current early.
You can do better using a 300B and its easier to drive. I would use some DHT's to drive it as well.
Actually, I can live with a couple of good watts with my Altec's. I am playing one channel in mono right now, and I do believe it would play more than loud enough for my tastes with a stereo pair. Sooooooo, before I spend any more effort on this project I am going to breadboard the other channel and see how it compares to my SET 2A3. That is the amp it has to beat to stay in my permanent setup. I already determined a while back that I like the SET 2A3 better than the 300B model 91 clone it replaced.
Edits: 02/25/19
If your are referring to the WE 91 you probably didn't like it because of the pentode drivers.
That sounds like a good candidate for another breadboard project one of these days. Would you happen to have a schematic of a good one you would be willing to share?
You don't really need much of a schematic, simply run any DHT at its op point and couple it with a cap.
Here's one I did back in the day, i got about 6-8 watts depending on transformer
nt
.
Have Fun and Enjoy the Music
"Still Working the Problem"
I got the following values with the amp turned on, no signal.
plate volts pin 2 to pin 3 103.7 vdc
grid volts pin 1 to pin 3 1.31 vdc
bias volts pin 3 to pin 1 minus pin 1 to grnd 1.31 vdc
how much plate current?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
It's got a 1.5vdc nmh battery from the cathode to ground. How do you figure the plate current with that arrangement?
Measure the voltage across the plate resistor.
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Well there is part of my problem, I thought the flattening would occur on both the positive and negative sections of the sine wave. Thanks, I will try it again.
"I thought the flattening would occur on both the positive and negative sections of the sine wave"
It would if the tube was 1. very linear and 2. biased to idle right at the half way point between cutoff and saturation.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"The maximum level before clipping is the level just before any sign of flattening of the positive and/or negative peaks begins."
I agree. Once clipping is reached, the HD is high and that power is not of much use.
It's a available power BEFORE clipping that vinnie wants to know.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Yeah, that is what def said too, but I am finding very difficult to find a spot where it seems to be clear cut. I mean there is quite a bit difference in voltage between where it is a rounded top and where it is a flattened top. I was hoping to get a better estimate of the power than just 2 to 4 watts.
Edits: 02/22/19
Just noticed that formula for power is different than what I saw else where. Do you mean E*2/R? The other one I saw was V*V/R where V was the rms voltage.
Edits: 02/21/19
You measure 5 volts out when it begins to clip5*5/8ohms=3.1 watts
If you want to check freq response doing manual plots, you can do a frequency sweep starting at 5hz ending at 20K and plot it on paper.
Your meter needs to have a good freq response to do this. Simply plot the voltage as it drops across the sweep. You could simply sweep 5hz, 10hz, 20hz, 30hz, 40hz, and then start at around 12KHZ up to 20KHZ
Edits: 02/21/19
He means the square of the RMS voltage, divided by the resistance.
Ohhhhhhhhhhhhh! Now that sounds much different. I will give it a try on the 845 I have breadboard now in the morning. I know that it is supposed to be close to 4 watts. If that works I will try the 813 again. Thanks!
That sounds like something even I might be able to do. I will give it a try and report back when done. Actually, I am kind of curious too.
Edits: 02/18/19
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: