advertisement


New Regulator

Thanks I'll go for as short a scope lead as possible.

Just reverted back to a standard scope lead and the amplitude is much greater so the soldered on lead seems more immune to pickup, this implies it is pickup as does the same amplitude when all is powered off.

Sounds OK but I have a downer on the regulator right now so not listening as normal, hard to be positive about a subjective issue.

ETA
The DAC, only one at the moment, is at the far end about three boards below the DUT.
 
Use two probes, tie thecearths together, put the two probe tips on you measuring points, earth and volt and subtract the two waveforns. When looking for small signals, you really need to go differential.
 
russel said:
Use two probes, tie thecearths together, put the two probe tips on you measuring points, earth and volt and subtract the two waveforns. When looking for small signals, you really need to go differential.
Click to expand...

Will give it a go, thanks.
 
Just looking at the frequency on the above shots.
I see 100MHz and 7.7MHz.
7.7MHz is quite roughly estimated.

I have a 100MHz clock but that must be a red herring as the clock was powered down in one of the shots.
7.7MHz is close to my original oscillation problem.
 
OK, using two scope leads, channel 1 & 2 set to add with channel 2 inverted.

I guess this has just become expensive.

All shots are regulator output.
1st All powered down, also same when both leads are shorted.
2nd Powered on not playing.
3rd Powered on and playing.
4th Implementation.

5mV and 0.5uS per division.
IMG_2512_zpsiajck5jn.jpg


IMG_2513_zps9uo8quf9.jpg


IMG_2514_zpszcx7brk8.jpg


IMG_2516_zps3s8rjmmr.jpg
 
Tony,

From the sequence of photos, it appears that:
1. There are two frequencies coming from the unpowered setup: a low level signal of about 10 MHz (.1us/cycle) and a similar amplitude one of about 285kHz (3.5us/cycle) (photo1). This is RFI that must be coming from something else since all your circuits are powered off.
2. One of your circuits enhances the lower frequency noise with a larger glitch about every 3.5us (photos 2 and 3) - RFI coming from one of your circuits.

It seems likely that the lower frequency is related to the ESS operation but you'll have to study the clock frequencies. I can't figure out the ESS data sheet with regard to clock frequencies to save my life but maybe you can.

Last thought is that none of the interference or noise seems to be in the audio band or the result of a circuit malfunction. So, if you can't hear it and the system sounds great, then maybe this is what it's supposed to look like!

Kit
 
RyanSoundLab said:
Tony,

From the sequence of photos, it appears that:
1. There are two frequencies coming from the unpowered setup: a low level signal of about 10 MHz (.1us/cycle) and a similar amplitude one of about 285kHz (3.5us/cycle) (photo1). This is RFI that must be coming from something else since all your circuits are powered off.
2. One of your circuits enhances the lower frequency noise with a larger glitch about every 3.5us (photos 2 and 3) - RFI coming from one of your circuits.

It seems likely that the lower frequency is related to the ESS operation but you'll have to study the clock frequencies. I can't figure out the ESS data sheet with regard to clock frequencies to save my life but maybe you can.

Last thought is that none of the interference or noise seems to be in the audio band or the result of a circuit malfunction. So, if you can't hear it and the system sounds great, then maybe this is what it's supposed to look like!

Kit
Click to expand...


Ooops I goofed.
Was wondering where you got those frequency’s from.
So first off the time base in all these shots is 0.05uS NOT 0.5uS
I see 100MHz and 7.7MHz, may be 100MHz and 15.4MHz.

This is RFI that must be coming from something else since all your circuits are powered off.
Click to expand...
Maybe the scope has just had its day.
Shot 1 is also when the scope leads are just shorted, so that looks like its the scopes noise.

The burst noise that appearers when playing a stream I am happy with.

ESS clock is 100MHz which would have had me excited had it not been powered down!
LOts of other clocks in that lot, Arduino is connected on I2C, Rpi connected on I2S, LCD screen seems to have a clock and IS noisy, ESS internally up-samples, when running it had a 44.1kHz I2S input to 192kHz internally.

There are 14 regulators running (I switch from normal dual mono to straight stereo to simplify the procedure) the DUT is the quietest of the lot.

To my ears this single DAC setup sounds better than my previous dual mono but hoping "themysticals" golden ears will visit and confirm.

Taking shot 1 as a control measurement there does not (sadly) seem much point in proceeding further. As you say its all well beyond audio.
 
Those frequencies are even less of an issue so you're good to go now. Don't blame your scope. Earlier shots you took show even less noise than these. My newer Tek digital at 5mv has at least that much noise, too, but looks a little more random. Time for that glass of wine and a good CD.....

Kit
 
Well Kit you do make sense but the scope is driving me madly.
Either the scope or regulator is not the same from day to day, as you said I’ve had better shots than today although they were with a PD style scope lead.
Russells suggestion today of differential measurement made sense so today I reverted back to conventional leads.
Perhaps I should try differential with a pair of PD leads...

Perhaps using a knocked about scope from nearly 50 years ago doesn’t cut the mustard.
Its a frustrating machine with dodgy switches, never sure what’s happening.
I have four leads with only one that seems reliable.

I just started a new thread here

Sadly no listening tonight as we have to go and listen to a live band or two :D
 
Somewhat related to the project is having the PCB's made. I just came across this site in a magazine that instantly compares the cost of getting boards from up to 25 different vendors. I don't know if this works in the UK, but when I tried it, I found I was not using the least expensive factory, particularly for getting stencils made (needed for SMD parts). Just thought you'd be interested.

http://pcbshopper.com/

Kit
 
Thanks for the useful link kit and apologies for my tardy response.
Been busy putting LMS on Rpi2 with TC7 this week.
 
You must log in or register to reply here.


advertisement


Back
Top