Spin Semiconductor

Hi

I am building a board based on the spin refrence schematic. For some reason, I get distortion on my output when it is over .5v peak-to peak. Any ideas where to begin? I am using the refrence schematic with the output tied into no load other than the 1k resistors or the ouput tied into a buffer, I get the exact same results. Any ideas where to start? This one has got me stumped!

Just tried it with a straight in/out block. I get clipping at 1v peak to peak. Hmmmm

Here is the schematic. Distorts even without no load after R3

And yes....voltages are fine

Is the clip indicator pin on the FV-1 indicating a clip is happening inside the FV-1?

frank wrote:Is the clip indicator pin on the FV-1 indicating a clip is happening inside the FV-1?

Yes it is, I accidentally had the LED reversed...

Im thinking my oscillioscope may be having some issues? Im not sure what else could be wrong.

Since it is clipping with a through program you are overdriving the input, check the levels around the input buffer. Also check the probe and scope setting, recently had a person using a 10x probe but scope set for a 1x probe so all the readings were off by a factor of 10.

Okay, so I was finally able to check again. Everything on the scope seems fine.

I have noticed that their is high frequency distortion before the onset of clipping. Im wondering if it might be a decoupling or grounding issue?

OKay, so If I feed it more then 300mv (with just an in-out program), I get strange distortion. I can clearly see where the Fv-1 clips )square waves) and that problem seems fine now. I did another board revision. I will report back with a video.

Clock signal seems fine, no modulation. All my decoupling caps are okay too.

Im going to run a signal sweep and see if maybe my filter isnt working, might have too much bandwidth on the output? The weird distortion frequency is pretty low!


https://www.youtube.com/watch?v=hV5pOI2 ... ata_player

What does the output look like at chip pin and at output buffer? And power rails, are they clean? What are the voltages to the opamp power pins? Do you have caps on the opamp power pins?

frank wrote:What does the output look like at chip pin and at output buffer? And power rails, are they clean? What are the voltages to the opamp power pins? Do you have caps on the opamp power pins?

Chip pin/output buffer is totally clean. If I tap the signal off directly, we have no issues. Yep, buffer has .1uf bypass cap. Proper voltages check out too. Doing a 9v single supply, and biased at 4.5

There is about 5mv of ripple on the 3.3v volt rail. Im using a 78l33 regulator. I have tried a few caps from 1uf to 330uf as an additional output filter, and while the ripple is cut in half, I detect no audible effect on the noise. I have also used a more advantage capacitance multiplier filter earlier in the prototyping process and coundnt hear any difference even though performance was increased down to about 1mv of ripple.

Just discovered another interesting fluke related to this problem. If I sweep my signal generator up to 60khz-100khz only the noise appears on the output. Its the strongest between 80khz-100khz

Here is a video of the distortion. It appears only on the negative swing of the signal.

https://www.youtube.com/watch?v=DZlkcl3 ... ata_player

Here is a video of that noise when Im sweeping through with the generator. I can cause it to stop with a cap across the input, but at lower frequencies, it is still seeping in, even with a cap across the input ( I used .1 just to be dramatic. )


https://www.youtube.com/watch?v=B7VZ5uO ... ata_player

I also just tried some alternative power supplies and the performance is pretty much exactly the same. I dont really have anything with ripple below 3mv-ish

EDIT:
Whats the maximum output swing? I believe the output is clipping. This is really confusing me.

In the YT video it sounds like you are getting hiss not distortion, if that is the case you might need to do some signal filtering which should get rid of any aliasing noise.

Why are you sweeping from 20KHz to 100KHz? The chip is designed for 20Hz to 1/2 Fs (16KHz if a 32KHz xtal is used) Please repeat the sweep in band only.

Okay, so fun little update. APparently this issue only exists on 1 of my bioards. Im going to try building some more powerful low pass filters

ice-nine wrote:In the YT video it sounds like you are getting hiss not distortion, if that is the case you might need to do some signal filtering which should get rid of any aliasing noise.

Thanks for the tip, this is what Im assuming the noise must be as for now. The issues I was describing earlier were partially a fluke. Im still having issues getting reasonable headroom out of the chip, but I believe that the high frequency noise may be aliasing related. I will check back.

frank wrote:Why are you sweeping from 20KHz to 100KHz? The chip is designed for 20Hz to 1/2 Fs (16KHz if a 32KHz xtal is used) Please repeat the sweep in band only.

I did quite a few in band sweeps too. Just though it was interesting how some components of those signals were getting demodulated into the output.