I would suggest running a signal through the bad ones for 100 hours and then retesting them. However I found the Xicon polystyrene to be in general poor performers.
If you really have time try desiccating the bad ones. Store I. A closed container with silica gel for a few months!
I put some bad Xicon examples on the hot water pipe for a few days, but, generally speaking, it is only a game. I think, that these parts are unwanted.
Fair enough. Xicon is a lower cost part.
The other issue was spreading of the FFT at the base of the test frequency. More of this on the film/foil capacitors than the ceramic ones. My suspicion is this is from vibration. The test would be to have the plates parallel to the earth and then redo the test with them perpendicular.
Unfortunately I think that this test is impossible with the Wien bridge oscillator. The spreading of the FFT probably is the result of the oscillator's opamp noise and the relatively low depth of the frequency rejection near the base frequency.
I would expect the base spreading to be the same for both film and ceramic capacitors if the issue was elsewhere. Just try the oscillator flat and on edge.
I think there are enough things in the display to question its usefulness on the primary freq picture. The asymmetry of the skirts is often a result of slight mistuning of the notch or of a speed/windowing issue. For the purpose used the fft was fine, but as the fundamental was only 3pixels wide, resolution is limited.
To really see a good picture of the skirts would be a different picture/resolution. And using a steep notch makes it doubly complicated.
Cheers,
Alan
Many thanks for the measurements
To really see a good picture of the skirts would be a different picture/resolution. And using a steep notch makes it doubly complicated.
Cheers,
Alan
Many thanks for the measurements
Last edited:
Not sure what you are trying to tell me......
I grabbed a miniDSP of the shelf and set it up as 48db/oct 1KHz low pass filter and get same results..... cutting out the harmonics of the generator gave lowest thd and a more pure sine wave resultant for further use/app needs. A low noise filter is needed if doing thd+n... the miniDSP is a bit noisy compared to analog filter in analyzer. But still makes the generator useful for a -100dB app. Can be applied with general purpose generators to improve them?
THx-RNMarsh
Last edited:
In regards to distortion in caps or resistors with attached lead wires........... disregarding the alloy make up of the resistor....... the 2H is usually from the material... this is esp true with caps... and the 3H would be a termination/connection issue between part and lead wire.
THx-RNMarsh
THx-RNMarsh
Okay I see what your doing. You improving the generator rather than measuring the generator.
we need a low noise sharp cutoff filter design to see how it improved ultra low thd generators, as well.
THx-RNMarsh
My measurements years ago implied the spreading was primarily noise sidebands from the effective Q of the circuit. When I completely eliminated non-linearity in the amplitude control loop by using a precision wire-wound Tel-Labs PTAT resistor as a thermistor (the oscillator would only stabilize briefly and slowly wander away) I noticed the spectrum would snap in briefly with dramatically low sidebands.
BTW LIGO uses WIMA caps at the most critical point (filtering the amplitude control on the laser) and this is not included in the ultra-vibration isolation loop. At this point -180dB is required.
Okay, so for less money, we still can use the HP339A which if I recall,
"The THD is -105dB (re .7v/1Khz) or approx. .0005%" and can do multiple
frequencies and levels.
And how/what is the affect with the combination active-BK1607?
Bummer of a deal the Xicon Polystryene are...I guess we are stuck
with good ol' RTX...what say you Richard?
Oh speaking of which, Richard, how do those cheapo Chinese caps you like to use
test out?
"The THD is -105dB (re .7v/1Khz) or approx. .0005%" and can do multiple
frequencies and levels.
And how/what is the affect with the combination active-BK1607?
Bummer of a deal the Xicon Polystryene are...I guess we are stuck
with good ol' RTX...what say you Richard?
Oh speaking of which, Richard, how do those cheapo Chinese caps you like to use
test out?
Last edited:
I'm having a hard time y'all, I'm having to remove everything from the bench
area and then figure out a way to put it all back in in some organized manner.
The bench has become just an ongoing battle. My bench is a combination of
JW and Bob Pease...cept, I'm not so learned. The latter of which I'm taking
steps to rectify. Love the pun too. Back in College learning Electronic
Engineering Technology along with Manufacturing with the new lab at
my community college and will eventually get associates degrees in each
which I think is better than just certificates. The good news is because
I have undergrad degree and master's (in non-related fields) I don't have
to waste my time with Speech, Psych, Math, Chem, English, etc.
And I get to use and learn at least manufacturing wise with state
of the art Mills, Lathes, various CNC Machines etc.
Question, which software should I be particularly interested
in getting as a student? I can get LabView, I have AutoCAD,
and I hear that AutoCAD just bought Eagle (the boar layout
people). ARTA has the full version for students don't they?
Where you can set up and save parameters?
Cheers,
Sync
area and then figure out a way to put it all back in in some organized manner.
The bench has become just an ongoing battle. My bench is a combination of
JW and Bob Pease...cept, I'm not so learned. The latter of which I'm taking
steps to rectify. Love the pun too. Back in College learning Electronic
Engineering Technology along with Manufacturing with the new lab at
my community college and will eventually get associates degrees in each
which I think is better than just certificates. The good news is because
I have undergrad degree and master's (in non-related fields) I don't have
to waste my time with Speech, Psych, Math, Chem, English, etc.
And I get to use and learn at least manufacturing wise with state
of the art Mills, Lathes, various CNC Machines etc.
Question, which software should I be particularly interested
in getting as a student? I can get LabView, I have AutoCAD,
and I hear that AutoCAD just bought Eagle (the boar layout
people). ARTA has the full version for students don't they?
Where you can set up and save parameters?
Cheers,
Sync
I found the polystyrene caps to be worse for EMF pick up than the poly polypropylene.
The distortion was about the same as the Panasonic polypropylene. Wima came a close second. The Panasonic are bulky though 400 - 600V but they are very consistent in quality.
Haven't had a bad one yet.
The distortion was about the same as the Panasonic polypropylene. Wima came a close second. The Panasonic are bulky though 400 - 600V but they are very consistent in quality.
Haven't had a bad one yet.
You mean the epoxy coated brown ones? Use them too, next to WIMA. Pretty good on pricing too.
There are very few ploystyrene film makers remaining for making film capacitors. There is only one left now for very high quality ps caps. Strangely it is in China. They intend to stay in the business..... They have the highest quality film...... no pin holes etc.
REL uses them. RTX.
THx-Richard
Last edited: