Transistor substitions for a phase 90

[svstee]

I really need a phaser, and a saw that there was already a vero layout made up for the phase 90 so I think I'll take a whack at it. I was wondering if I could sub out the 2N5952's in favor of J201's, and the 2N4125 for a 2n5087, is this OK? I'll be careful about pinouts, but will these subs be OK?

Layout:
http://www.aronnelson.com/gallery/main.php/v/Grooveyard/mxr_phase_90.gif.html?g2_imageViewsIndex=1

Datasheets:
http://download.theicstock.com/2N5/891/2N5952_638891.pdf
http://www.fairchildsemi.com/ds/MM/MMBFJ201.pdf


http://www.synthdiy.com/files/2007/2n4125.pdf
http://www.datasheetcatalog.org/datasheet/fairchild/2N5087.pdf

[mdh]

J201s can work in a Phase 90, but they're not ideal, and IIRC, you could get a batch that won't work at all, due to the (again, IIRC) wide Vgs_off spec of that part.  I think you would be much better off getting a pile of 2N5485s.  You should consider buying 50 or more to yield a high probability of getting a well-matched quartet.  Also, search the forum on this issue, there has been a lot of discussion of this issue in the past.

[R.G.]

When I originally did the work on cloning the P90 and getting the JFETs right, the 2N5292 was hard to find. Since then, that's gotten easier. Mouser sells the 2N5292 now, and cheaply. I did about a week of comparing JFET spec sheets and came up with two alternatives, the 2SK30 and the 2N5485. Both of these have Vgsoff voltages compatible with the P90 LFO, and have worked well in practice.

The J201 is an odd JFET. It's not all that good in a P90, although it can be made to work. I personally would use any of the other three before going to the J201. Of course, you're asking this because you have J201s in hand, right?  In that case, Keen's fourth law applies - use what you have to get it running, find the perfect thing when you can.

[Mark Hammer]

2SK30s are commonly used in many Japanese-made and Chinese-made phasers.  The only thing is that there are several types of 2SK30s, like the GR and  A / AY.  It is my recollection that the GR are preferred, but the trouble is that they are less available than the other type.

Anybody care to chime in on this?

[R.G.]

2SK30s are commonly used in many Japanese-made and Chinese-made phasers.  The only thing is that there are several types of 2SK30s, like the GR and  A / AY.  It is my recollection that the GR are preferred, but the trouble is that they are less available than the other type.

Anybody care to chime in on this?

The Japanese often (in the past, when they still made transistors) make a part number and then add a color code (black, yellow, green, etc.) to it to indicate selected groups within the part. For the 2SK30 and most Japanese JFETs, these are Idss groups, which has the side effect of selecting them into Vdsoff groups as well. The GR(een) group happens to have the best Vgsoff for the P90. In the USA, the makers tended to make several different part numbers. The 2N5484/5485/5486 group share the same datasheet, like the 2SK30 B/G/Y does, and the 5485 happens to have the most advantageous Vgsoff of that group for the P90.

Both US and Japanese systems tag on -A and other suffixes for reasons know primarily by the maker. In the USA, this often denoted a lower gain version in bipolar transistors; don't know if there is a trend there in later transistor years, or for the Japanese.

In any case, even getting 2N5485 or 2SK30A/GR, you'll have to match them, and at that point you'll know from the design of my matcher whether the bias voltage in the P90 (3.1 in the stock unit, up to 5.0 works OK) is bigger than the Vgs at 10K DC resistance that the matcher selects for.

[shadowmaster]

It has become a habit for my case to socket every transistor or IC in my builds. This way, part substitution would be less of a hassle and can even experiment on some of my components lying around unused for years and find for them a suitable use.

For my Phase45 and Phase90, I both used 2N5457 for the fets. These sound really good and clean to my ears. I used matched BF245B initially in my Phase90 for debugging purposes and they do phase OK but not good sounding in my opinion.

Another suggestion for your Phase90, if you opt to have sockets in your fets and transistor, I suggest you do the same with the zener diode as they also need some tweaking to get phasing if turning the trim pot doesn't help with different fets.

[Martien]

a bit of topic but could you use monolithic dual jfet packages like a 2n6905with good results they oath to be matched right?