Simple Phaser – Choosing Components

In the previous post I covered some details of the circuit similar to MXR Phase 45. A relatively simple circuit … but still with a lots of components😊. Ultimately it has only a couple of circuit blocks/components that we really need to look into.

This time around I’ll be looking into choosing JFETs and tweaking biasing and LFO to get them to work the best. I did cover how to measure JFET parameters in an earlier post. For this post I used a slightly more advanced measuring rig, but ultimately it’s the same thing.

As usual, if you don’t want to read just skip to the video, you can always come back for diagrams here.

Choosing JFETs

Few days ago, in preparation for this, I prepared a measuring rig. I’ll post a schematic of it at some stage. I thought, if I need to measure JFETs again, I better have it permanent and not on a breadboard.

I did a few batches of JFETs I had lying around: J111, J112, J113, J109 and BF256B. Not the most exciting bunch, but as we’ve seen in the previous post, I’m sure I can make them work.

The goal of this exercise is to identify several pairs that are relatively closely matched so I can try them out together to see if I can get better sound.

Here are the measurements from the video in a handy bunch of tables. I might need to refer to these measurements at some stage again. Last time I did it I took notes in a notebook. I’m unsure which notebook and where it is … I’m not making that mistake again 😁.

J113

You can notice in the photo above that I marked all my transistors so I know which one is which … makes sense. If you have loads of them, marking doesn’t really work. In that case, you could put them in separate bins that have similar parameter values and marking the range of values in the group.

J113	|VP| (V)	IDSS (mA)	Notes
I	1.39	7.7
II	1.9	19.2	Pair 2
III	1.41	10.9	Pair 1
IV	1.95	19.9	Pair 2
V	1.4	10.3	Pair 1

All 5 transistors would work in the circuit because I can bias them properly. I can’t really pair I with anything, so I chose the closest two pairs.

At this point, it’s important to check whether these values match what we expect from datasheet:

And this matches what we expect. V_GS(off) is expected to be between -0.5 and -3.0 volts, and in my case above I measured between -1.39 and -1.95 volts. I used absolute values above, I could not be bothered with minus sign.

I_DSS in datasheet indicates that it should be over 2.0 mA and all are over that value, so we’re good.

J112

J112	|VP| (V)	IDSS (mA)	Notes
I	3.86	50
II	3.78	52
III	3.87	52
IV	3.92	57
V	3.99	56
VI	3.78	50	Pair 5
VII	3.57	46
8	3.79	50	Pair 5
IX	3.63	47

Looking at cut-off voltage, this is at limits of how much I can get from biasing without modifying the circuit a bit. That sounds dramatic but what it really is in the end is just swapping a resistor or two with more appropriately valued ones. Should be easy.

At this stage it’s worth checking how much current your power supply can provide. Well, I should’ve checked that before, I had thought it was around 50mA so I was skeptical about the accuracy of this measurement, but turns out my power supply is rated for 120mA. All good so far.

J111

J111	|VP| (V)	IDSS (mA)	Notes
I	8.05	109
II	7.98	110
III
IV

I gave up measuring after the first two. I can’t get them biased properly. The whole point of biasing JFETs around their cut-off voltage is to be able to use that increased resistance around that point. The resistance goes down very quickly and if you recall from the previous post/video, low resistance gives us notch at high frequencies only.

We need this resistance to go both high and low. At the biasing point that I can achieve, the JFET will be mostly ON and with low resistance. Looking at the datasheet, this was expected. Anyway, moving on.

J109

Moving on to J109. While J111, J112 and J113 are the same JFET family, J109 is different. It is sourced from process 58 and according to application note AN-6609 Selecting the Best JFET for Your Application by Fairchild, it should be perfect for Analog and Digital Switching.

The transistor would be a secondary choice for low noise amp … there’s a saving grace. However, characteristic of P58 transistors is low r_DS … so expectations are low 😁.

BTW, J111/2/3 are sourced from process 51 – primary choice for analog/digital switching, oh wait … and voltage variable resistor – BINGO! Exactly what I need.

Now, the application note is from 1977 … yes, you read that right. I bet it is still 100% relevant (except that lots of the transistors in it are obsolete now). I haven’t got a clue if I read this application note correctly or not if I’m perfectly honest, so let’s reserve judgement for the actually trying them out.

J109	|VP| (V)	IDSS (mA)	Notes
I	3.82	210
II	3.78	212	Pair 6
III	3.90	214
IV	3.77	212	Pair 6
V	3.99	217

Looking at cut-off voltage, pretty close to J112, so the same consideration applies. I am very dubious about I_DSS measurement, but that’s the current that did happen. I’m guessing my power supply can source more current than what’s written on the box.

I’m guessing 120mA rate is when all 5 outputs are used … on the second though I should’ve probably been more careful not to exceed 120mA in the first place. If you do this at home … just be careful and gentle to your power supply 😄.

BF256B

BF256B is one of the transistors that are easiest to source in Europe (from a general component supplier).

The transistor is sourced from process 50 … looking back at AN-6609 – that should be good for voltage variable resistor! Great! Btw, do note that the pinout is different from the rest (G-S-D the rest have D-S-G pinout, just flip the transistor).

BF256B	|VP| (V)	IDSS (mA)	Notes
I	2.86	8.8	Pair 4
II	2.84	8.6
III	2.88	8.8	Pair 4
IV	2.64	7.7	Pair 3
V	2.63	7.6	Pair 3
VI	3.17	10.0
VII	2.72	7.8
8	3.06	9.4
IX	3.12	9.5
X	2.61	7.5

Cut-off voltage looks good, and transistors sourced from P50 should have generally higher ON resistance than ones sourced from P51, so this looks promising.

Choosing the Best JFET Pair

Once I chose 6 pairs to try out it’s just a matter of plugging them in and figuring out which one I love the best. It’s really just a matter of plugging them in and adjusting bias. For J112 and J109 I had to adjust R15 to a higher resistance value to bring the biasing point down a bit.

In the end, I felt I had to change R18 to a lower value so the voltage going out of LFO has a wider sweep. This was required to counter R15 increase that consequently decreased the swing. Also, consequently the maximum frequency dropped, so I changed R16 to a lower value.

Finally, I chose BF256B unsurprisingly, here’s the final version I used:

Experimenting in Action

The hardest part of doing this video is stopping playing 😎. I love those pedals that just keep you going. I had to try out 6 different pairs, so that went on for a while. There is a pulsating quality (on higher speed setting) but at any position there is this moving, driving quality.

It just makes you want go on and the frequency of the LFO makes me wanting to play it differently at different settings. I hope you enjoy this as much as I did 🤩.