News:

Forum may be experiencing issues.

Main Menu

J113 for phasers

Started by Aentons, April 27, 2020, 04:42:30 AM

Previous topic - Next topic

Aentons

All the talk of upcoming phasers reminded me of an old phaser I have that sounds really really good, really thick and watery and drippy, distinctly different from a phase 90, and I realized I'd never actually looked inside. I had always assumed a straight clone but it turns out that's not exactly the case.

It's a DOD Phasor 401 and it uses J113's and TL022's.

Anybody have any experience with using the J113 in a phaser?


Max

#1
I have no experience with phasers in general, but I have a bunch of J113 unused, so this is interesting...
I would say that the use of the transistors in a phaser is very different from a distortion pedal, so in terms of sound is not so critical, given that they are properly matched.
Please correct me if this is just gibberish  ;D
Actually, I was planning to use the J113 in the switching part of an echo project, they shouldn't really make any difference there, but I have enough for a little phaser ;)

EDIT: I just checked the schematic with a phase45 and they are really close, there's only a couple of resistors different (and the fets and op-amp of course).

Aentons


Aentons

Sorry folks, I'm learning out loud... :)

After looking at the schematic a little closer, it looks like the 401 is using more gain at each stage and doesn't use transistor for any gain before mixing at the output. Could that account for more of the sonic differences than the J113's?

Also, when comparing the 401 to the MXR, I noticed in the Electrosmash Phase 90 analysis that they mention the following about the transistor in the output stage: "Later circuit revisions (Block Logo), included a capacitor C11 from base to collector. It is a Miller Capacitor which shapes the high-frequency response (roll-off the highs) providing a dominant pole compensation preventing oscillation." I have a 90's Phase90 and the high end roll-off is very noticeable but in a very pleasant way. I had asked a question a while back about how this was achieved so I guess this answers it. ;D

Scruffie

The J-FET is used as a variable resistor, it's not really there for its 'sound'.

The way they're used however, is different to the phase 90 and more like the phase 45.

I haven't compared the phase stage values, perhaps they're different too, that will affect where in the spectrum the phasing is occurring.

Works at Lectric-FX

Aentons

#5
Quote from: Scruffie on April 28, 2020, 11:03:21 PM
The J-FET is used as a variable resistor, it's not really there for its 'sound'.

The way they're used however, is different to the phase 90 and more like the phase 45.

I haven't compared the phase stage values, perhaps they're different too, that will affect where in the spectrum the phasing is occurring.

Thanks!
Wow, I've never looked at a 45 schematic and did not realize that the 45 had differences other than the 2 vs 4 stages. It then appears that the dod 401 is actually a dod 201 with 2 extra stages rather than a straight 90 clone. That makes more sense now.

also, sorry, I did not word my question about the J113 very well. I was more referring to the way that they may, or may not effect the sweep and not sure if they have any impact on where the notches are. The 401 seems to have a very smooth sine sweep and more low mids.


Edit: I did find this statement on another forum post:
"There are subtle differences between the 45 and the 90. The 90 is not just a 4 stage 45, the 45 is not just half a 90.
The 45 has a JFET gate-drain feedback circuit which cuts the JFET distortion a lot, making it sound much smoother,
whereas the arrangement in a 90 is much simpler."

Aentons

#6
Quote from: Aentons on April 28, 2020, 05:04:24 PM
Also, when comparing the 401 to the MXR, I noticed in the Electrosmash Phase 90 analysis that they mention the following about the transistor in the output stage: "Later circuit revisions (Block Logo), included a capacitor C11 from base to collector. It is a Miller Capacitor which shapes the high-frequency response (roll-off the highs) providing a dominant pole compensation preventing oscillation." I have a 90's Phase90 and the high end roll-off is very noticeable but in a very pleasant way. I had asked a question a while back about how this was achieved so I guess this answers it. ;D

Just a bit more about the C11 cap in the output stage. I found a schematic that has the value. 0.0068u.


Aentons

Now back to some semblance of the original question... put a different way.  ;D

It seems like the J113's may be fine to use in place of the 5952/5457's, so why aren't those ever recommend in their place or why don't you really see anyone using them for this purpose?

Scruffie

I can think of several reasons;

1) DIYers can be stubborn when it comes to 'special' part choices.
2) Up until relatively recently, those were very easy to acquire J-FET's.
3) There's plenty of J-FET's good for phasers, but the specs may not get the best out of a specific design.
4) You can still get the ones mentioned in SMD form and people migrated to that.
5) Engineers/Designers/Manufacturers/People moved on from J-FET's as they're a complete pain in the ass.
Works at Lectric-FX

Aentons

#9
Quote from: Scruffie on April 29, 2020, 10:51:16 AM
I can think of several reasons;

1) DIYers can be stubborn when it comes to 'special' part choices.
2) Up until relatively recently, those were very easy to acquire J-FET's.
3) There's plenty of J-FET's good for phasers, but the specs may not get the best out of a specific design.
4) You can still get the ones mentioned in SMD form and people migrated to that.
5) Engineers/Designers/Manufacturers/People moved on from J-FET's as they're a complete pain in the ass.
Yep, I get it. You are totally right.

I was just curious if there is anything in particular that is not suitable or would need to be compensated for with these for that application since they are about the cheapest thruhole jfet that Mouser has right now.

BTW, I'm gearing up for a VFE tractor beam build and am a bit nervous about it

sharpan.sv

I used j113 in Phase 90 clone, works just fine.

jimilee

Quote from: sharpan.sv on May 09, 2020, 02:41:59 PM
I used j113 in Phase 90 clone, works just fine.
Just reading the tractor beam doc, it was built around the recommended jfets more or less.


Sent from my iPhone using Tapatalk
Pedal building is like the opposite of sex.  All the fun stuff happens before you get in the box.

Aentons

#12
Ok, so I got some J113's and put a matched set in my tractor beam, biased it as best as I could by ear, and it's distorting. I chose a set matched at -1.39

I see the VFE site has this to say:

"The input stage buffers the signal to strengthen it for the signal split coming next. Technically, there is a small volume cut in this stage to improve the headroom of the phaser, but later stages makeup any volume loss in this buffer."

"The Tractor Beam uses JFETs to change the center of the phase shift. In this configuration, JFETs have a natural clean headroom limit. If the phaser section is distorting, try reducing the input volume."

Is that what is happening? Do I need to adjust the volume cut in the input section to accommodate the J113s? It looks like that is done by increasing 470k R1 R22. Does that sound correct?

Aentons

To clarify... I'm playing my build right next to my production Enterprise v2.5. I'm using the same guitar(s), amp, and PedalPower2+. When I turn the guitar volume down, the distortion goes away. I have a strong lfo and I can hear both turnarounds clearly, but the distortion happens mainly in the middle of the sweep. The prod vfe build is not distorted hardly at all.

Here are my voltages:
OPA2134
1 = 0 mV
2 = 0.1 mV
3 = 0.1 mV
4 = -7.53 V
5 = 0.1 mV
6 = 0 mV
7 = 19.6 mV
8 = 8.25 V

LM358
1 = ---
2 = ---
3 = ---
4 = -7.53 V
5 = ---
6 = ---
7 = ---
8 = 8.26 V

TLE2074
1 = -6.1 mV
2 = -1.7 mV
3 = 0.1 mV
4 = 8.25 V
5 = 0.1 mV
6 = 0.3 mV
7 = 1.5 mV
8 = -15.4 mV
9 = -2 mV
10 = 0 mV
11 = -7.53 mv
12 = 0.1 mV
13 = 2.6 mV
14 = 11.3 mV

J113 = G,S,D (removed)
Q1 = ---,0.1mv,0.1mv
Q2 = ---,0.1mv,0.1mv
Q3 = ---,0.1mv,0.1mv
Q4 = ---,0.1mv,0.1mv

Aentons

#14
Power from the source measures at 9.42V

Here are the voltages for the VFE prod build for comparison:
OPA2134
1 = 0.5 mV
2 = 0.5 mV
3 = 0.2 mV
4 = -8.15 V
5 = 0.2 mV
6 = 0.7 mV
7 = 8.8 mV
8 = 8.85 V

LM358
1 = ---
2 = ---
3 = ---
4 = -8.15 V
5 = ---
6 = ---
7 = ---
8 = 8.86 V

TLE2074
1 = -3.0 mV
2 = -0.3 mV
3 = 0.1 mV
4 = 8.85 V
5 = 0.5 mV
6 = 0.8 mV
7 = 2.2 mV
8 = -4.0 mV
9 = 0.3 mV
10 = 0.4 mV
11 = -8.15 mv
12 = 0.5 mV
13 = 0.7 mV
14 = 5.7 mV

2n5952 = G,S,D (not removed) 
Q1 = ---,0.4mv,0.4mv
I couldnt get a good read on some of these



----------------------------------

Also, for the input gain, R22(470k) and R23(390k) look to be the same in both builds

---------------------------

And a screenshot to compare
Left is VFE prod and right is mine