Engineering q...

[jbeckva]

Ed or Joe,

I've been messing around with HPIB enabled components - HP8903A analyzer and a HP3455A meter. I have an IEE-488 USB adapter - acts like an old fashioned COM port so it's pretty easy to use to communicate with both.

What I would like to do is a routine that will inject a preamp level signal into an amp, and measuring the output "calibrate" to a specific wattage as required. I do have that working somewhat now, but not quite refined. It's like you'd first want to do a good "estimate" on what input voltage to apply then based on an optimum increment (hmmm.. even a stored "cal factor" but getting off point here) really dial it in.

Also would like the same type of "cal" routine but based on stopping the increase of input where an exact (or pretty close) measurement of distortion limit would be met.

So I'm thinking.. how to actively measure and calculate the gain of an amp, then based on that figure calculate the optimal voltage and step increment settings to use when performing the tests above - specific wattage and a distortion limit test where it checks at what wattage the amp is putting out when the desired limit is met. And perhaps calculate an estimated input voltage to apply when after the initial measurement is met and the gain predetermined, i.e. the difference btwn that and the desired measurement in the range of the input needed.

So first it would perhaps start out at a specific (but low.. don't want to blow it up) voltage input, and measure the output. I am assuming multiple measurements need to be made when increasing the input voltage, right? Then from that, or am I off base?

Thanks!

 

[WOPL Sniffer]

HUH?

 

[orange]

He says caps don't smoke well, even with good rolling papers.

 

[oldphaser]

Jerry,

I sent you a PM.

Ed

 

[jbeckva]

Thanks Ed, replied to the PM.. Basically as I stated, let's say I proceed as follows.. Bear in mind that the amp or any amp under test would have it set to maximum gain if there are level controls involved:

1. Start with an input voltage low enough to as to NOT swing the amp full scale.. Let's say for argument's sake 100mV on the input.
2. Measure the amps output, save it as measurement "A"
3. Increase the input to 200mV and re-measure the output, measurement "B"

Now, based on A and B and the difference between them, would I have enough to figure out the overall gain of the amp? Or would I need to take a few more measurements to see?

Then, based on the gain, say I wanted to have the amp output right around voltage "C". What would be the formula to "estimate" the input voltage I need for that?

Is it as simple as the ratio output to input, or do I need to deal in dB's somewhere in my calculations?

 

[oldphaser]

In terms of dB, there are a variety of db units of measurement. In most cases, I'm using dBV for most of my measurements and performing the necessary calculations, in order to obtain certain figures with respect to full output (0dB), -3dB, -6dB, -10dB or referenced to 1 watt. However, in your case you are interested in the gain. So for the following example dbW may be more appropriate. 1 watt is equal to 0dbW. A Phase Linear 700 series II amplifier is rated at 360 watts or approximately 25.56 dBW.

This does not completely answer you question. For that there are many other variables to consider. And I don't have all day to type them all down. I'd much prefer phone calls because it is so much quicker for me and I can get back to work building and testing amplifiers. I have discovered many interesting things during the course of my testing. It seems like every week I discover something new that I didn't know before and I've been working on these things for several decades.

There are variety of db calculators on the internet:
http://www.sengpielaudio.com/calculator-db-volt.htm
https://www.rapidtables.com/electric/dBW.html


Ed

 

[jbeckva]

Ok, another way of asking the question.. Is the *voltage* gain of an amp linear i.e. if I have 100mV on the input and 2V out, when I increase the input to 200mV will I have 4V out? If not, then what would it be? And would you need more measurements or info to determine the answer - given that at time of measurement the gain of the amp is "unknown"?

Not (at this point) concerned about power output.. just straight up voltage. From there and given the impedance I would think I can calculate the output power.. ?

 

[Gepetto]

Ok, another way of asking the question.. Is the *voltage* gain of an amp linear i.e. if I have 100mV on the input and 2V out, when I increase the input to 200mV will I have 4V out? If not, then what would it be? And would you need more measurements or info to determine the answer - given that at time of measurement the gain of the amp is "unknown"?

Not (at this point) concerned about power output.. just straight up voltage. From there and given the impedance I would think I can calculate the output power.. ?

Yes until you hit clipping of various elements in the circuit (like SOA protection clipping in) or clipping due to the limitations of the bulk supply.

A*K = B

K is the voltage gain. The global feedback path ensures this.

 

[jbeckva]

Thanks Joe, exactly what I needed. So this ought to be interesting.. The concept is to first measure the ratio input to output, probably a few times to get an average, then based on the result I'll see about dialing the output in by the same ratio.

Gotta go pull an old Rotel I have still.. amps kind of scarce here at the moment LOL. Results forthcoming..

 

[oldphaser]

Some food-for-thought.
I could be wrong, but the variables might well include the following:

Single channel or both channels driven.
Amplifier loaded or unloaded.
Constant voltage source (i.e servo/motor driven Variac) or a source that sags.
Temperature
Input (signal) source (flatness throughout its range, i.e. DC to 200KHz +/-0.1dB)
Duration of test. (i.e. 30 second to 5 minutes)
Input impedance.

Input sensitivities for rated output can also change dependent upon frequency. Most commonly, input sensitivity is rated at 1 kHz.

I have some results of (3) Dual 500's and (2) 700 series II's taken recently with a variety of input levels in order to generate an output at rated output (0dB), -3dB, -6dB and -10db.

If my memory serves me correct, the amplifiers certified to THX's (home theater) standards are all required to have a certain amount of gain for a given input. As an example; Input sensitivity: 1V for 28.28V output, THX Reference Level. To some degree, I suspect this might well dictate to designers how much feedback they have to use. Perhaps Joe can respond on this one?

Joe, Please forgive my ignorance, but could you verify the other terms in A*K = B where K is the voltage gain. Is "A" for amplification and "B" gain?

Ed

 

[Gepetto]

Some food-for-thought.
I could be wrong, but the variables might well include the following:

Single channel or both channels driven.
Amplifier loaded or unloaded.
Constant voltage source (i.e servo/motor driven Variac) or a source that sags.
Temperature
Input (signal) source (flatness throughout its range, i.e. DC to 200KHz +/-0.1dB)
Duration of test. (i.e. 30 second to 5 minutes)
Input impedance.

Input sensitivities for rated output can also change dependent upon frequency. Most commonly, input sensitivity is rated at 1 kHz.

I have some results of (3) Dual 500's and (2) 700 series II's taken recently with a variety of input levels in order to generate an output at rated output (0dB), -3dB, -6dB and -10db.

If my memory serves me correct, the amplifiers certified to THX's (home theater) standards are all required to have a certain amount of gain for a given input. As an example; Input sensitivity: 1V for 28.28V output, THX Reference Level. To some degree, I suspect this might well dictate to designers how much feedback they have to use. Perhaps Joe can respond on this one?

Joe, Please forgive my ignorance, but could you verify the other terms in A*K = B where K is the voltage gain. Is "A" for amplification and "B" gain?

Ed

Hi Ed
A is voltage in and B is voltage out.

 

[laatsch55]

Lively discussion this am, coolness!

 

[oldphaser]

Hi Ed
A is voltage in and B is voltage out.

Joe,

I see where I made my mistake.

Thanks!
Ed