Author Topic: Help with the simulation of Phase 90 on LTSpice!!! JFET's giving 0.3 Volt sweep  (Read 11267 times)

Rob Strand

Quote
That's what the current owners say. They weren't there. The likely partner is Cornell Electric Manufacturing Corp., New York, Manufacturer of voltage supplies (battery eliminators). Not a major technology but maybe knew the right people in the right stores.
It's hard to know what the truth is.    I couldn't find any announcements for the name change or mergers.

It seem the Dubilier name remained in the UK after 1933,  and didn't change until 1935 to 1938.


I did find this odd document,

https://ethw.org/w/images/9/90/Dubilier_-_Radio_Catalogue_and_Dealers_Guide%2C_1925-1926.pdf

"Dubilier Condensor" and "Radio Corporation" (NY)

Presumably Radio Corporation is RCA before they split from GE.

From 1928 to 1933 there was apparently a court case between RCA and Dubilier so I can't imaging they would merge.  Yet it is odd it's about the same period.

After that history skips forward to 1938-1939 with a new name.   
The front section shows promise but then says nothing!

https://ethw.org/w/images/5/55/Dubilier_-_Catalog_No._161,_1938-39.pdf
Plopping around the pot since an early age.

PRR

> Presumably Radio Corporation is RCA before they split from GE.

No way. RCA was eat-only: nobody survived after RCA (until Sarnoff left this world).

Everybody was a "radio corporation".

Dubilier was mica-caps, then added electrolytics and other types. The Cornell I linked was selling battery eliminators, meaning they bought lots of caps. Or made them. But certainly sold lots of cap-filled products. I could see them joining forces to dominate the cap market. Or one or the other falling behind on his rent and a forced merger. Those catalogs you link are from William's own library, but marked Florida. Several businessmen who visited FLA in that period just didn't come back north again. Dubilier may have sold to Cornell and retired to Florida.
  • SUPPORTER

Rob Strand

Quote
Everybody was a "radio corporation".
Yes, that's the way to read it.

Quote
Dubilier was mica-caps, then added electrolytics and other types. The Cornell I linked was selling battery eliminators, meaning they bought lots of caps. Or made them. But certainly sold lots of cap-filled products. I could see them joining forces to dominate the cap market. Or one or the other falling behind on his rent and a forced merger. Those catalogs you link are from William's own library, but marked Florida. Several businessmen who visited FLA in that period just didn't come back north again. Dubilier may have sold to Cornell and retired to Florida.
I didn't think about checking he might have got out of the company, or even checking when.


I did find something stuff narrowing the name change down to late 1933 to early 1934.  I guess that agrees with the 1933 merger claim.


Dubilier
Filed Jan 1933,
https://patentimages.storage.googleapis.com/aa/e9/a5/e4b6ee339113cb/US2130532.pdf
Filed April 1933
https://patentimages.storage.googleapis.com/39/61/8e/935dbe8a5a3696/US1970776.pdf
Filed Aug 1933
https://patentimages.storage.googleapis.com/d8/75/b8/ecc51bc5278979/US2070435.pdf


Cornell-Dubilier
Filed Aug 1934
https://patentimages.storage.googleapis.com/58/2d/ca/7c69983bee960d/US2088693.pdf
Plopping around the pot since an early age.

savethewhales

https://www.cde.com/resources/catalogs/AEappGUIDE.pdf

This is what I quoted from. It's under the name of Cornell Dubillier Electronics (or smth like that). It has few information about who actually wrote it and specific things like dates.

What I know is that seems well written and has valuable information.

https://linearaudio.nl/sites/linearaudio.net/files/Bateman EW 01 2003 mar2003 10 to 100uF caps and 100 Hz measurements_0.pdf

This is the Bateman, C. document I quoted.

Either way I don't know too much of history around caps but I was told that CDE was a know manufacturer of caps..


Rob Strand

Quote
This is the Bateman, C. document I quoted.

Either way I don't know too much of history around caps but I was told that CDE was a know manufacturer of caps..
That info is all fine.
Plopping around the pot since an early age.

PRR

https://patentimages.storage.googleapis.com/58/2d/ca/7c69983bee960d/US2088693.pdf

Yum, nitrobenzol! Better than Halowax! Now called Nitrobenzene, and produced in the millions of tons, Nitrobenzene is a notorious poison.
  • SUPPORTER

Rob Strand

Quote
Yum, nitrobenzol! Better than Halowax! Now called Nitrobenzene, and produced in the millions of tons, Nitrobenzene is a notorious poison.
The last 100 years has been quite frightening chemical wise.    Old capacitors had all sorts of evil stuff in them.  Even upto the 70's we were pretty stupid.    There's still a lot of cracks.  The large hardware outlets still sell non-animal friendly snail bait   There's even deceptively marked animal friendly versions which say poisonous if eaten by pets.   Plenty of poisonous chemical in paints for dumb consumers to wash down their kitchen sinks.   You can't trust people with anything evil in the MDS,  they have no idea what it means.

Plopping around the pot since an early age.

Rob Strand

Found this one, Sept 1933, adverts for Cornell-Dubilier,
https://worldradiohistory.com/Archive-Service-Magazine/30s/Service-1933-09.pdf

So I guess that narrows the name change to Aug/Sept 1933.

Previous ad in the same mag was in, Feb 1933, under Dubilier Condensors,
https://worldradiohistory.com/Archive-Service-Magazine/30s/Service-1933-02.pdf



For the name I got a similar name to what you found previously,

"In 1933, Dubilier's company merged with the Cornell Electric Company to form the Cornell-Dubilier Electric Corp."

https://www.electronicdesign.com/technologies/components/article/21795312/william-dubilier

And

http://ai.eecs.umich.edu/people/conway/Awards/ElectronicDesign/ED%20Hall%20of%20Fame%202002.pdf
« Last Edit: October 30, 2020, 03:26:07 AM by Rob Strand »
Plopping around the pot since an early age.

savethewhales

Hey!! Do you guys know what's the normal output voltage of a guitar signal? I've read so many different things: -10 dBV, -20 dBV, 100 mV RMS, 3V RMS.. I can't really understand what would be correct..

I'm just asking this because tomorrow I'll have to do tests on my pedal but I need them to be at a reasonable level for guitar outputs.

Thank you very much for your attention and help,

Cheers

savethewhales

Hey guys!!

I really need some help, I've already done some tests on my pedal on an Audio Analyser (APX), and I was having this response (with the circuit INSIDE the pedal box):



Different trimmer position:



BUT Now after some time without testing it (because I had to go to work these days), I went to test it again and this is what I was having:





Does somebody know what might be going on here?

There's some things I did:
- Tested continuity in all of the circuit (except the LFO, which I seen on the oscilloscope and was behaving ok) and everything seemed to be connected (but of course there's some problem here).
- Changed cables (this helped me to avoid the -5 dB ceiling, otherwise getting 0)
- Seeing the frequency response at the op amp outputs. The 1st op amp of the phase shifting stage was alright, the 2nd was alright, but the 3rd was like this:

(when I continued, the slope was the same).

Knowing that the response should supposedly be of 0 gain throughout the whole spectrum (because in the ouput of the 2nd op amp of the phase shifting stage it was like this), I concluded it was happening a problem of high pass filter somewhere around the 3rd op amp (I don't even know where exactly could this be happening...)
So I started checking frequency response on the terminals of components near that area.
For the sakes of me trying to explain myself and you guys understanding what the hell am I saying, I put down here the schematic on LTSpice and schematic on EasyEda (which is the definitive final one, in terms of values).

LTSpice (imgur link to see better https://imgur.com/X0W84Dg):


EasyEda (I know it's a mess.. if you guys want I can try to make it to 2 different images for better visualization):


Talking EasyEda, I got (Note that I refer to "bad" as the high pass unwanted freq response):
- Left leg of R18: good
- Right leg of R18: bad
- Before R15: good
- Between R15 and R14: bad
- Out1 of U5 (TL072): good
- Pin 6 of U5: bad
- Pin 5 of U5: bad
- Pin 1 of Q3: bad
- R13 on the upperside: bad
- Right leg of cap U7: bad
- Left leg of cap U7:

- Both sides of R14: bad
- One of the legs of cap U9: bad;
Other leg of U9:
I got these 2 little different responses from 2 different testings:






These were the ones I could take note of, but I will be testing the rest of the points on college on monday.

I'd really appreciate any kind of help that you guys could give me.. I am actually lost.
« Last Edit: November 15, 2020, 02:41:25 PM by savethewhales »

Rob Strand

We have to wait for the image service to come back on-line.

Some conversations are going to get difficult without images.
« Last Edit: November 15, 2020, 07:20:56 AM by Rob Strand »
Plopping around the pot since an early age.

bluebunny

Plenty of places you can have your images hosted - imgur.com for one.
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

savethewhales

Plenty of places you can have your images hosted - imgur.com for one.

Thank you soo much! I just updated the post I had done, with images from imgur. Hope you guys can see it good.
« Last Edit: November 15, 2020, 12:07:20 PM by savethewhales »

bluebunny

When you upload your picture to imgur, hover over your picture and click on the "..." on the top right, then click on "Get share links".  Now click on "Copy Link" next to the "BBCode (Forums)" option:



You can paste this directly into your post, where your picture become visible in-line.  Then we don't have to go chasing your links.  (And it's exactly what I did to show the picture you see above these words.)
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

savethewhales

When you upload your picture to imgur, hover over your picture and click on the "..." on the top right, then click on "Get share links".  Now click on "Copy Link" next to the "BBCode (Forums)" option:



You can paste this directly into your post, where your picture become visible in-line.  Then we don't have to go chasing your links.  (And it's exactly what I did to show the picture you see above these words.)

Thank you very much! unfortunately, some images are too big now.

UPDATE: Now everything is ok at the post!!!
« Last Edit: November 15, 2020, 02:41:53 PM by savethewhales »

bluebunny

some images are too big now.

Apart from resizing your pictures before you upload them, you can also make them appear smaller by including HEIGHT=nnn (or WIDTH=nnn) inside the [IMG] tag.  Like this:

   [img height=20]

Or this:

   [img height=120]

And the reader can click on the image to see it full-size if they wish.
  • SUPPORTER
Ohm's Law - much like Coles Law, but with less cabbage...

savethewhales

some images are too big now.

Apart from resizing your pictures before you upload them, you can also make them appear smaller by including HEIGHT=nnn (or WIDTH=nnn) inside the [IMG] tag.  Like this:

   [img height=20]

Or this:

   [img height=120]

And the reader can click on the image to see it full-size if they wish.

Thanks! But I guess i'll let these ones stay like this, and if someone asks me for resizing or smth, maybe i'll do it!

PRR

....unfortunately, some images are too big now....

The "WIDTH=" or "HEIGHT=" technique changes the display size but not the size in storage or over the wire.

IrfanView will resize images off-line. Free, Windows-only. https://www.irfanview.com
  • SUPPORTER

savethewhales

Does somebody know where could I be getting that high pass filter in my circuit?

I already have the tests on almost every node of the circuit, so I can show you guys if you want!

savethewhales

This post is a definitive testing of my circuit, meaning it's better documented than the first one.
It's therefore an ask for help to all of you who might know anything about circuits and guitar pedals.
I'm very lost and lacking of time right now, so I would thank ANY help!

Soo I was able to test every point that I wanted today.

Remembering, this is the schematic on LTSpice (for easier visualization):
part 1 -
part 2 -

And this is the official schematic (with real values) on EasyEda:


Now, as I referred in that previous post, I was getting a very strange frequency response after some time not using/testing the Phaser Pedal (before this time everything was ok). My conclusion is that there is a high pass filter throughout the circuit (that will be proven by images show below), that I cannot understand where comes from.

Before that, let me show the frequency response I once had with my pedal:
1 -

2 - With the trimmer in different pos:


As I referred, after some time without using the pedal, the freq response got kinda crazy:
1 -
2 -

That being said, note that "Good"="flat" and "Bad" = High Pass filter non sense (you will see below).

Output of the 2nd op amp (first of the phase shifting) is ok (zoomed in)


However the inverting and non-inverting inputs of the same op amp have this same response (bad):


In my head this isn't even possible..

U3, which is the capacitor, has a good freq response on the left leg, but the leg connected to the JFET has a bad freq response.. This I can't see why also.

The pos and negative inputs of the 3rd op amp are "bad" but it's output is good, just like the 2nd op amp, like (???)

Now, between U6 and R10, the freq response is good, but on the other leg of U6 it's already bad.

As for R15, which is connected to the 3rd and the 4th op amps, has a good freq response on the left side and a bad one on the right side.

Same goes for U7.

As we can conclude, the 4th op amp has a bad freq response on the output:


The inverting and non inverting inputs of the 4th op amp have the same response..

In this matter, both legs of R14 are bad.

R8 (Vref side) and every resistance that is making the all pass filter in parallel with the JFET, has a req response looked like the output of the pedal, but attenuated. The other leg is bad.

As for the 5th op-amp, it has a bad freq resp on the non inverting input.
The output and the inverting input are also bad

Both legs of U9 Cap are bad too (which is different from the other caps).

The 6th op-amp has at the non-inverting the freq response of the output of the pedal but attenuated:


Now as for the JFET's, in every drain point of every JFET, the response is bad. In the last FET the response is stabilized before the other ones:



In the source and the gate the freq resp of every FET are equal. They're the same as the output of pedal but attenuated:


If someone got to here and thinks he/she can help me, please say something! I'm desperate! Thank you very much