Zener value for MPF102

[rasco22862]

Hi i´m building the MXR phase 90 from tonepad, i´m going to use de MPF102 as the JFET. I read that the zener diode has to be relative to the JFET used. So what Zener diode i should use with the MPF102. Right now, i have a 5.0 v zener diode.

Thanks

[R.G.]

You're kind of stuck.

You only have a 9V supply, so you can't go much higher than 5V without running out of headroom on the top side. But the MPF102 may have a Vgsoff of as much as 8V. You're going to have to select MPF102s to work in the circuit. It is NOT certain that any MPF102, even new, undamaged, fully specified ones, will work in the circuit.

Even if you get JFETs that will work in the circuit, you will still have to match them to get good phasing.

[rasco22862]

1)What kind of MPF102 would work in the circuit, and what zener diodes should i use them?? I already built the Improved JFET Matcher.
2)Idem for 2n5485
Thanks

[rasco22862]

nothing?

[Aharon]

Check the Build Reports at Tonepad,full of comments about diferent Jfets used for the 90.

http://www.tonepad.com/project.asp?id=7

bottom of the page.
Aharon

[rasco22862]

thanks, but nothing about the zener diode and the mpf102

[mdh]

Can you tell us what voltages you're reading for your particular batch of MPF102s in the JFET matcher?  I can't claim to fully understand the meaning of that measurement, but I think it's supposed to indicate Vgsoff.  My guess is that if you're getting readings around 2-4 (or -2 to -4, depending on how you're placing the probes), then you'll be fine with a zener around 5V.

If I'm wrong, someone please correct me!

[rasco22862]

Ok, thanks.

What do i want to know is waht reading has to be on the DMM on the Improved JFET Matcher for the MPF102, because i know a friend in an electronic store that let me choose the transistors. So i have to be prepared 

[mdh]

I seem to recall R.G. saying that the ideal Vgsoff value (which, as I said above, is what I believe the matcher is supposed to measure) for JFET phasers running on 9V is about -3V.  I've build a Phase 90 and a Phase 45 with 2N5485 FETs that measured close to -2V in the matcher.  You can get an idea of what kinds of values to expect by looking at the range specified for Vgsoff in the data sheet.  I think the main thing is to stay away from Vgsoff values that are too large in magnitude.  I don't have the data sheet in front of me, but I think the range for MPF102 is something like -2 to -8V.  There are a lot of threads on this, so if you search for FET and Vgsoff with R.G. as the author (use the advanced search), you should find a lot of info.

That said, I'm thinking that if you look for FETs that measure close to -3V in the matcher, you'll be in the right ballpark.  Is there any reason you're going with MPF102s?  Could you get 2N5485s or 2N5952s?

[rasco22862]

i could get 2n5485 but is very expensive compared to the MPF102. Is it so bad the mpf102??

[mdh]

It's not that they're bad, per se, it's just that you may have to sort through a lot of them to get four matched parts with appropriate values of Vgsoff, and there's really no guarantee that you'll get lucky.  Seriously, take a look at the data sheets for the following part numbers:  MPF102, 2N5485, 2N5952.  Concentrate on the ranges given for Vgsoff.  The MPF102 probably doesn't indicate a minimum, and indicates -8V for the max.  2N5485 (and I believe 2N5952) should show a narrower range.  It might be worth your while to find out what other N-channel FETs they have at the store where your friend works, and take a look at those data sheets before you go in, to identify other potential parts.  But if you go in and find 4 MPF102s that all measure somewhere around -2 to -4V and are within a couple mV of each other, by all means, buy them and try them out.  If, on the other hand, they measure between -4 and -8V, you should probably keep looking for something more appropriate.

Is that clear?

[rasco22862]

yes, very clear!  thank you very much.

but, that about the zener diode? wich one should i use??

[mdh]

If you can get FETs with characteristics like I described, then you'll probably be fine with the 5V zener.  You might also do a search here or look at the build reports on Tonepad to see what experiences others have had.  It frankly sounds like you have pre-build jitters, so you should just proceed to build with the best parts available, and see how it works out.

Good luck!

[rasco22862]

another simple pair of questions

1)Wich one is of the layout is the 22k feedback resistor? If i matched the transistors, do i need it? So what is for? Do i pút a jumper?
2)If i buy a 500k linear pot and wire a 100k resistor like this (the middle one):

Do i get a 500k rev-log pot?
3)Also i used 0,047uF caps instead of the 0,05 uF caps. Is it ok right?
4)For the 15 uf electrolytic cap, i put in parallel one of 10uf, and other of 4,7uF. Is this ok?

[rasco22862]

5)I found the 2SK-136 and the 2SK-104

The 2Sk-136 has Vgoff: MIN:0,3 MAX: 2,5
The 2Sk-104 has Vgoff: MIN:-0.25 MAX: -4.25

Is there a better one?

[mdh]

1. The feedback resistor is the one that connects pin 7 of IC3 to pin 6 of IC2.  In other words, looking at the Tonepad schematic, it's the 22k resistor closest to the word "Schematic" You might as well put it in, and if you get a bad ticking sound that you don't like, remove it, or try increasing the value. Mine ticked with the feedback resistor in place, so I removed it. I have been meaning to try a larger value there, but I haven't gotten around to it. In any case, don't put a jumper there.

2. Short version: no.  Long version: read "The Secret Life of Pots" at GEO.  Then read it again.  Lather, rinse, repeat. Such is the way of GEO.

Less glib version: the maximum resistance of the pot in parallel with any tapering resistor still needs to be 500k. If Rx = 100k and the maximum resistance of the pot is 500k, then the maximum resistance you'll get from this arrangement is 500k * 100k / (500k + 100k) = 83.3k. I would (and did, as a matter of fact) just use a 500k linear pot here.  If you happen to come across a 500k reverse log pot, then go for it. Otherwise, use what you have, and don't worry too much about the taper.

3. Yes, that will be fine.

4. That's cool as well. FWIW, I originally had a 22uF cap there, and later swapped it for a 10uF. I liked the 22uF better, but the 10uF increased the maximum rate to something approximating a genuine Phase 90 that I played in a music store.

5. I think the 2SK104 sounds like a better prospect, but if your friend is going to let you hang out and measure transistors, then take your time and find a matched quartet with Vgsoff close to -3V. BTW, be sure to check the pinout of the transistor that you choose, keeping in mind that it may not be the same as the one indicated in the Tonepad layout.

[BJF]

Hi,

Try for this circuit to match transistors with an Ugs at about 1,4 to 2V5's and a set matching within a few millivolts.

It will not matter what the case of the transistor says: what you want to do is operate the FET in it's resistive range, also known as triode range.

If you look at the circuit you will see that the transistor works as a parallell shunt to a resistor.
To get each allpassfilter to sweep in the same way for deepest effect controlrange need to be the same for each transistor used

You could find a quartet with the MPF's as the Ugs is poorly defined but you may need a bag of them to find this: they would be allover the place likely

You might also want to try 2N5487 with a better likelyhood of finding a matching set.

If you use transistors with a very low Ugs, like 0V6, you might want to change the zener to say 2V7 but distortion through circuit would be worse......examples of such transistors would be 2SK30A.

You can ofcourse then try linearization circuits to each FET so as to drive them midways.
I would not recommend this even so it has been done as in for instance the Guyatone Rollophase

However best results you'd likely get with a matched set within range.

Just a few thoughts

At your service
BJ

[rasco22862]

I found 4 mpf102 that has -2.75. are they gonna work right?
PD: the jfet matcher reads 4,9 v with anything plugged. is this correct?

[mdh]

Sounds about right. Have fun!

[rasco22862]

Ok, here are some voltages:

2n3906 E:5.10  B:4.36 C:2.20
JFETs MPF102 from left to right on tonepad´s layout:
1) G :0.92 S:5.09 D: 5.09
2) G :0.92 S:5.09 D: 5.09
3) G :0.92 S:5.09 D: 5.09
4) G :0.92 S:5.09 D: 5.09

IC1:
1 5.10
2 5.10
3  3.47
4 0
5 4.47
6 5.08
7 5.07
8 9.23

IC2:
1 5.10
2 5.10
3 5.07
4 0
5 5.09
6 5.10
7 5.10
8 9.21

IC3
1 5.10
2 5.52
3 5.06
4 0
5 5.10
6 5.55
7 5.10
8 9.21

I matched de mpf102 with -2.75 with the improved jfet matched. I dont know if its right but, the DMM reads 4,09v when nothing is plugged

Thanks