JFET Matching for Phasers 

Love how this showed the breadboard. If I don’t see a picture of it I have trouble still…

Would love to see more videos describing how these work! Really like how you showed the theory and then gave us a taste of what it sounded like matched/unmatched. Thanks.

Nicely done. In theory, pretty much ANY set of JFETs of the same part number (i.e., 2N5457, MPF102, 2N5952, SK30a, BF245, etc.) will yield audible phasing, as you nicely demonstrate, and as a brief scan of the universe of phaser schematics will corroborate. The question is whether the movement, and audible result, will be *consistent* throughout the entire range of sweep. Imagine we have 4 rowers in a boat, in one of those Cambridge-vs-Oxford or Harvard-vs-Yale competitions. Two are always rowing in sync and with equal force. One of them always joins each stroke a little late, and the 4th one always lifts their oar out of the water a little earlier than the others. The boat is unlikely to go straight all the time, but for that brief moment when all 4 oars are pulling together and in sync, it does. So it is with JFETs in phasers. Ideally, all JFETs used will be changing their drain-source resistance at the same time, in response to what the LFO is telling them to do. With larger differences between them, some will require a higher turn-on or turn-off point. If one was to restrict the sweep of the circuit to ONLY the range where all the JFETs overlap in characteristics, it would sound as good as one might expect. The problem is, it may not be the width/range of sweep you want. As you aptly note, the MXR Phase 90 (which I assume you're working from) only had a speed control, and fixed the sweep width (i.e., how high and low it sweeps) to some sort of compromise that would sound decent for the slowest and fastest sweep rates. The internal bias trimmer would set the "starting point" so that all pedals sounded the same. Overall, the necessity for matching is a function of how much variation in sweep you want. If one has fairly modest requirements (e.g., you generally use only medium-fast "spongy" speeds with little feedback and only moderate sweep width) then matching does not need to be stringent. If you're aiming for wide sweep, especially at very slow rates, and you want to include variable offset (i.e., where in the spectrum the sweep "begins"), then better matching is called for, such that the responsiveness of the JFETs to the LFO is similar across the entire range of what you want them to do for you.

Enjoyed the video and explanation, I’m also in the process of matching a bunch of J113 for a phase 90 build, using the same process. Looking forward to building the PCB when you guys have it in stock!

This looks like a great sit for guitar pedal builders. Just subscribed.

OMG thank you for showing the unmatched set difference, I was looking for this and didn't even know it.

Erik, another great video mixing theory with real-world sound. The matched set sounded like a very solid multi-stage phaser while the unmatched set has a univibe quality. Not bad to my ears, just different. Guess I need to build two pedals or switch between them in a single enclosure. Thanks, E! You rock (and nice socks).

Great video! I also had the impression the difference would be worse, but better have them match. Thanks for this video, I'm just working on it and had this question.

very instructive videos! keep it up! ;)

Very helpful. Thanks

Yeah I liked the unmatched set it had more of a present sound and the matched set was more subtle and seemed to have a shallower sweep

if a set the multimeter on gate and souce of the transistor it gives me the same result by placing it on the gate and the third pin of the Tl072 is it correct?

May i ask u something sir, when i saw the circuit, i found something dont understand LFO, where i put input n output LFO? Far as i can see is only vcc n bias vcc . Tx u sir 🙏

Hey. I was hoping you could help me. I finished building my phase 90 from effects layouts. I have good signal on and off. But no swoosh. Tried turning the trimpots still nothing. Checked my 741's theyre good. But i used this cheap 5457's ...

I decided to build the breadboard tester unfortunately it doesn't seem to show where the wires go that you use for the probes and I'm afraid I'm not an electronics expert, any help would be appreciated cheers 😊

So as I understand it, when matching JFETs, one looks to the VGS (off) value only?? Funny, these values are in the -0.5 to -1.5 range, when in the RG Keen article, they are in the 1.5 to 2.8ish range (thou he says the spec is -0.5 to -4.0!! So, maybe lower because different JFET?!? I picked up a Peak Atlas DCA75 meter that does all of this! My readings for a a handful of 2N5952s were -2.46 -2.15 -2.17 -2.33V I have a bag full of them, so just wanted to confirm VGS is what I want to match up and if so, will do just that with closest values! I read somewhere, but forgot, what is the maximum spread between values (say 1% or 5%) before you can judge them as "matched"?!??

Yep. That's how you match JFETs. I was surprised that I didn't hear $20 worth of difference between the two sets.

I would like to build this breadboard matcher.

Hey! Great video, wondering if you can direct me to where I can find Fritzing parts for guitar pedals? Thanks!

What phaser circuit - mxr phase 90?

Thanks so much for this God bless you sir but my question is Can experiment this circuit with a bi polar ic and transistor possibly 4558 and 2n222 or 3904. Etc to achieve the phase effect Cause Tl072 is hard to fine sir . waiting for your reply sir thanks alot

Not gonna lie, the unmatched set still sounded good. The matched had a more consistent sweep.

Hi, what program do you use to make the patterns? Thanks so much.

??? Can you make a an Echo Pedal as i would be really interested in making 1 if thats possible thanks

If I want to use this circuit to match p channel JFETs would I just leave the breadboard layout as you have it and swap power source polarity?

When is your phase pedal being released 🤩?

Not sure…so I gotta ask: will J113 work with the same circuit, same cap and resistor values?

Wheredo u keep multimeter prob? To test it. Is it in gate qnd drain?

Can i use lm358 op amp for this ?? Or it has to be tl072?

3:40 all you need now is for the cat to jump up on your piece of paper... Edit: thanks for covering the LED...

Ada kah scema penuh layout untuk pedal ini?? Terima kasih

The bad/mismatched set sounds more natural to my ear

So where are the test points?

I like the bad set to my ears

Hello Kevin, Informative video, good explanation. Did you get Jason Johnson's email address? I asked but never heard back of it was ok to give out, I heard from him but no mention of that. He has Facebook and Instagram @ Fuzzlord Effects. In a recent video, Fuzzlord Q&A, from May 15, 2020, may answer your question, it is gmail. I have to respect his information, I'm sure you understand. All the best from Oregon, C.

0.69 0.72 0.75 0.74 Will this be a matched set?