Sinad Meter

[Pez]

Look what I found, thanks @Klondike Mike for making the video!

I just skimmed through the video, but it looks like his notch filter caps are bad. That's why at the end of his video, the meter won't adjust to the far left.

You need 2 things to calibrate these Sinadders properly:

1 - A very accurate tone (signal) generator, capable of 1KHz and 2KHz sinwave tones.

2 - "Matched" notch filter capacitors in the Sinadder that are still in spec.

When I did mine, I purchased 10 of the caps I needed. Then I tested them all on a capacitor tester to find the 2 x caps with the closest values. As soon as I replaced them, everything aligned perfectly.

FYI - In my Sinadder 3, the required notch filter caps were 10nF (.01uf) Polyester caps.

Great success.

I happen to have a picture of the offending notch filter caps - see the attached picture of a happy Sinadder 3 @12dB SINAD.

 

[loosecannon]

I saw Pez mention this but i'll go just a bit further.

it is SUPER important that the tone coming out of your 8656 sig gen and the frequency that the notch filter in the sinadder is tuned to be matched.

otherwise the sinadder cannot really do it's job and your results will suffer.

so, you are going to need a frequency counter of known calibration.

if you bought a lab quality one new within the last few years you can count on it being accurate, but if it's used/old/on the cheaper side, then you need to get it calibrated before you go to align the 8656 1k tone adjustment and the filter inside the sinadder.

if sending a freq counter out for calibration isn't really in the budget, there are ways to get pretty damn accurate results using a 10mhz rubidium standard that you can buy on ebay for less than 50 bucks.

You will also need to find instructions on-line about how to power it as they're not necessarily straight forward, but if you are willing to put in the research and effort you can have a very steady, accurate 10mhz standard you can use to calibrate equipment with.
LC

EDIT TO ADD: it is also very helpful to have a function generator that is also calibrated. these can be found pretty cheap for old ones and with a few new electrolytics and a calibration you have a very helpful test instrument.
i connect a speaker to mine and use it as a tone generator.

 

[Pez]

I saw Pez mention this but i'll go just a bit further.

it is SUPER important that the tone coming out of your 8656 sig gen and the frequency that the notch filter in the sinadder is tuned to be matched.

otherwise the sinadder cannot really do it's job and your results will suffer.

so, you are going to need a frequency counter of known calibration.

if you bought a lab quality one new within the last few years you can count on it being accurate, but if it's used/old/on the cheaper side, then you need to get it calibrated before you go to align the 8656 1k tone adjustment and the filter inside the sinadder.

if sending a freq counter out for calibration isn't really in the budget, there are ways to get pretty damn accurate results using a 10mhz rubidium standard that you can buy on ebay for less than 50 bucks.

You will also need to find instructions on-line about how to power it as they're not necessarily straight forward, but if you are willing to put in the research and effort you can have a very steady, accurate 10mhz standard you can use to calibrate equipment with.
LC

EDIT TO ADD: it is also very helpful to have a function generator that is also calibrated. these can be found pretty cheap for old ones and with a few new electrolytics and a calibration you have a very helpful test instrument.
i connect a speaker to mine and use it as a tone generator.

I can confirm, just as LC has said, a few Hz +/- either way on your generator WILL impact your calibration and your SINAD measurements.

FYI, I use a GPSDO for my 10MHz ref standard. Relatively cheap, and extremely easy to setup (plug and play).

73

 

[brandon7861]

I always assumed that if I were to adjust the sig gen freq for a minimum needle deflection, the signal would be in that notch and I wouldn't care where that notch is.

I don't have a sinadder, I use the sinad measurement on my oscope, so I've never encountered this issue in practice. So, now I am curious....

When capacitors in that notch filter drift, what is more a more prominent change, a change in notch frequency, or a change in notch depth? Do both caps tend to drift in a manner than maintains notch depth or is the procedure that I imagined of simply tuning the sig gen to match the new notch not enough? Do I really need to be Sherman Certified here or does the notch hold its shape as it moves?

 

[Pez]

I always assumed that if I were to adjust the sig gen freq for a minimum needle deflection, the signal would be in that notch and I wouldn't care where that notch is.

I don't have a sinadder, I use the sinad measurement on my oscope, so I've never encountered this issue in practice. So, now I am curious....

When capacitors in that notch filter drift, what is more a more prominent change, a change in notch frequency, or a change in notch depth? Do both caps tend to drift in a manner than maintains notch depth or is the procedure that I imagined of simply tuning the sig gen to match the new notch not enough? Do I really need to be Sherman Certified here or does the notch hold its shape as it moves?

In my unit I found it was the frequency that changed, not the depth. (I could get the meter to calibrate to the far left, but not at the required 1KHz (mine went high by around 20Hz from memory). However, depending on how much each capacitor in the matched pair drifts, and in what direction, would probably impact both freq and depth!
YMMV.

 

[Shadetree Mechanic]

Thank You everyone for your input, this has certainly given me a lot to think about. I hate to ask for help just to leave everyone hanging, but it will be several weeks before I can get back to this. And I am sure I have other threads with no conclusion too. Bench time for me is hard to get now days, but I will continue to learn about this all I can and will update here with my progress. I have a tone generator and a counter that I will have to check out too.

Here is a video that explains the inner workings: