Low Cost Alternatives To Comonly Used JFET's

[Pavlos]

Hi, I'm new to the site though I have been lurking for a couple of months, reading, learning, getting rather baffled, learning some more. So far I've not actually built anything, but I've been doing a bit of bread board prototyping on some of the simpler circuits to decide on what I fancy building first. And so far, barring the odd oversight or two on my part it has been a total success.....and I have plans.....many plans lol. But enough waffle and now to get to the point........

I have noticed that most of the circuits that use JFET's (Especially the ones intended to emulate particular amps) seem to use either j201, 2n5457, and occasionally MPF102 at the core of their circuits, which if I'm reading things correctly from the comments (and a lot of googling) though not direct, identical replacements for each other, are largely compatable, albeit with slightly different characteristics. The thing is, these JFET's appear to becoming rather more difficult to reliably source and are also becoming quite expensive (In UK, if you can find them they vary from about £0.60 - over £3 each depending on where you look) Now I'm aware that there are surface mount versions available at low cost per item, and adapter pcb's to mount them on which will allow them to be used as direct plug in replacements for the existing 3 pin packages, but to me this seems like it could be quite a bit of extra, and rather fiddley work, especially when some of the amp emulating pedals call for anything from 3 to 10 JFET's each. So from both a simple interest and slightly lazy standpoint the leading question is....

Has anyone had any success in using other easily available and low cost JFET's than these three, and if so how close were they sound-wise to the original, and what circuit modifications (Resistors,caps etc) were required if any?

Oh and by the way, great site, lots of excellent information and very helpful people on here, many thanks in advance :o)

[Chris60601]

So - here's a really short answer to your question with a somewhat longer after.
What I do (and I'm not suggesting you should try this, this works for me), I have started to buy SMD versions of some trannies (J201 for example) and create my layouts (PCB) to use them. They are waaaay cheaper and much easier to find but the caveat to that is; you need you be patient and further develop your soldering skills. Now that's not to say you cant find adapter boards for SMD to work with Vero - you can!
There are options on OSH and a few others sell them (www.diyguitarpedals.com.au for example).
And it seems you covered that and may not be an option for you.

The other way is to review the datasheets and find subs/alternatives. I use this site:  http://alltransistors.com/ to find them. Additionally, if you either use sockets on your Vero, that is a fast way to kinda plug-n-play or if you fancy breadboarding, that will work just as well and from your post, the latter 2 options seem to be the way you want to go.

The thing about this hobby; its not to try to create the EXACT stomp that you might normally pay a poop-load of money for, it's all about taking that device and making your own. Tuning it to sound the way YOU want it to sound.

Believe me, I have tried going the way of, "I just want to build the device I would buy, nothing more". It will never happen my friend :)

Well, it might (for a build or three) but soon enough, you'll be like, "Hey, this guy/girl built it this way but this other guy/girl did it this way with some added features". That's when you'll have turned and become hooked. You'll have late nights, little snips of wire in the carpet, you'll find ways to validate spending an extra 50.00 on parts from Tayda. By then, it's too late. You have become.... A Stomp-Zombie  and you'll never be the same.

Ask Rocket :)

[Frank_NH]

Yes, SMD is the way to go these days, but... you can use other JFETs.  You just need to look at the datasheet for the proposed JFET.  The two parameters to focus on are Idss (idle drain current) and Vgs(off) (the "pinch off" voltage).  For example, in place of 2N5457s you can often use 2N5458, J202, and J230 becuase Idss and Vgs(ff) are in the same range.  The main thing which will change for different JFETs is the gain and the drain resistance needed to bias the JFET.  Also, if the circuit uses a mu-amp configuration (e.g. Sabra Cadabra, Formula #5, BSIAB2), then you can use a range of JFETs and not have to worry much about biasing.  Note that the circuits may SOUND different (due to reduced gain or slight misbiasing) but that may be a GOOD thing.  I made a BSIAB2 without J201s and it sounded great to my ears.  Use sockets for the transistors and experiment.

[Pavlos]

Hi Chris, thanks for the quick response, I'm planning on sticking with the stripboard approach for the time being (Re-wiring a guitar is a tiny bit simpler than this ;o) and as per quite a bit of advice I've seen, I'm definitely going to use sockets and experiment a little bit with trying different transistors, makes a lot of sense for the meagre cost of the sockets and the extra versatility it allows in fine tuning a build, not to mention the lack of any chance of excessive heat on those transistors. I did stumble upon http://alltransistors.com/ (Mind blowing for novice like myself but looks like a very useful resource) a few weeks ago actually, and have been looking at some of the data sheets too. Though I have started to get a little bit of understanding, I'm still a very long way off even beginning to think I have a clue where the boundries of common sense would break down on what would or would not work. Some of the values that spring up there on the compatibles seem rather extreme by a number of factors so I'm hoping to at least find out where the ball park is so I can play safely :o) one thing that does look relatively safe is replacing within 'families' such as 2n5457,58,59 but as I'm thinking along the lines of availablity and cost as well, I'm wondering how far you can go before it just won't work, or you need to adjust the circuit to accomodate a different component. And also interested in how anyone has found the replacements, is it vaguely analagous to swapping 12AX7 for 12AU7/12AT7, or going from EL84 - EL-34 - 6V6 in a valve amp? was it higher/lower gain, more responsive, tighter/flabbier/more grainy in character? yep a thousand an one questions lol

My soldering skills aren't too bad at the moment but it certainly wouldn't harm improve and maybe have a go at the surface mounts sometime in the future. All help, advice and information gratefully received :o)

[Pavlos]

Thanks Frank, think I've seen the mention of Idss (idle drain current) and Vgs(off) (the "pinch off" voltage) before, one question on that though is how far off those values can you reasonably go and still expect the circuit to work? or will any of the alternatives suggested on  http://alltransistors.com/ be ok to use? Sounding different can be good sometimes, nothing wrong with twaeking with a purpose or even just a bit of of good old serendipity :o) As for the gain and drain biasing, that is where I run out of knowledge at the moment, I know some circuits have trim pots, but for those that don't how much of a change in resistor values is usually needed and also which resistors? (I'm on a pretty steep learning curve at the moment)

Formula ~5 is one of the builds I'm interested in actually, and the DLS mk3, and a couple of the Wampler pedalds and.......oh dear, I think I'll need a lot of coffee!

[Pavlos]

Oh and by the way, I'm starting with build it like the original for now........well until I understand things better (Hence the breadboard) but I already have ideas of things I'd like to change.....just not sure which components I need to change yet lol..........I thnk that breadboard is going to get a good workout though :o)

[Beaker]

Hi Pavlos,

I really don't have any counters to you points about SMD versus through hole JFETs, as your arguements and logic are  perfectly sound.

What I would say though, is that once you get over the initial shock and horror of just how tiny SMD transistors are, they are suprisingly quick and easy to solder if you use flux paste to stick them to the boards first. The only extra tool you need is a half decent magnifier. I recently bought a load of 5457 and J201 from Bitsbox here in the UK, and a card of 100 PCB boards from Chromosphere in Australia.

I've only used them on a couple of builds so far, but already I would not go back - I'm a convert.

With the amount of dodgy/fake through-hole JFETs around, they give reliable, quality results. With the price of genuine parts from reliable vendors like Bitsbox, Dr. Tweek, Langrex etc. they work out much cheaper.
And only for the cost of a few extra minutes per build.

[Pavlos]

Thanks Beaker, I think at some point I'm just going to have to bite the bullet and go SMD, in the mean time though I'm still interested in alternatives to what seem to be the main 2 / 3 (J201, 2N5457, MPF102) that seem to be suggested in most builds, realistically how far out of the spec can you go and expect the circuit to work, and what effects do different parameters have on behaviour? is there a particular parameter that implies loose/tight feel or eq, higher/lower gain range etc, I'm all in favour of socketing and experimenting, just rather do so with at least a little bit of an idea of what and why rather than randomly finding out that I'm trying to do the impossible with an unsuitable part. And ultimately at the point I do go SMD then I'll also have so many more choices :o)

And of course anything I find out from my own builds I'll gladly post as well, have been reading all sorts of stuff people have posted about their own builds and problems and have learnt so much already, and though I would only describe myself as having hardly any idea what I'm doing rather than absolutely no idea a few weeks ago, I'm having fun learning and messing about with the breadboard to decide what to build first......and what little tweaks might suit me best :o)

Oh and just had a delivery of a few things from Bitsbox arrive today, pretty good considering it was only ordered at the weekend, good communication and prices are decent too :o) also tried Tayda on multiple reccomendations from the site, took just over two weeks to arrive, and again decent prices though the tracking was a bit patchy, was well packaged though so patience my son, patience ;)

[Beaker]

In reply to your question whether JFETs behave in any way analogous to different valves, the answer is, (unfortunately) no they don't.

Use the wrong JFETs, or not bias them correctly (using the trimpots) and the pedal won't sound like you have used a 12AT7 instead of a 12AX7, it will sound spitty, farty, maybe gated or just dull and flat.

Also bear in mind that an alternative to the stated JFET that works great in one pedal, may not work well at all, in another.

[Pavlos]

Thanks Beaker, was wondering if there were any analogies like that which might help, but at least finding that the answer is no is worth knowing :o)

Regarding biasing, I've noticed a few people under bias and some over bias the FETs according to taste in some circuits (Within what I assume are certain limits of the particulat type of JFET actually working at all) and that this can bring a pedal to life, or tame it, so another question for which there may be no rational or known answer. Other than the Idss (idle drain current) and Vgs(off) (the "pinch off" voltage) parameters for JFETs which from what I can gather are the most important for compatability, are there any other important or critical parameters to take notice of or can I, as long as these are in the same ballpark just swap and experiment with insane abandon safe in the knowledge that things might just get a bit wierd sounding?

[Beaker]

Regards biasing, the "standard" accepted method is to either adjust the trimmers or fixed resistor values to achieve half the suppy voltage at the JFET, i.e 4.5V for a 9V supply, or 9V for an 18V supply etc.

Some pedals (or even batches of JFETs) may sound better at a different voltage (often listed on the layout page), but it's always better to let your ears be the judge, rather than just accept what your multimeter is telling you. You can always bias by ear, then measure and note down the readings for the next time you build one.

You won't do any harm using sockets and substituting different JFETs - try others and let your ears guide you. Most of these pedals use J201 or 2N5457, simply because they are (or were) cheap and easy to find. It is entirely possible to use something else, and get exactly the same sound, for a minimal effort.

BUT... it might entail some major sugery first. You might need to do some extensive resistor and capacitor swapping to get there. That's where a depth of theoretical knowledge and build experience comes in handy, (or a thorough knowlege of circuit simulation programs like LT Spice that will take out the guesswork).

None of which I possess by the way!

I guess it depends on your approach really. Either spend a small fortune on loads of JFETs, and spend hours experimenting, stick to known combinations, but be prepared for a few bad batches of JFETs along the way, or just go SMD.

[Pavlos]

right, looks like you've largely confirmed what I thought from reading through comments on builds and a bit of other light reading. The half supply voltage for biasing is basically a 'vanilla' starting point to start from and tweak to get things just how you want them, and I'd pretty much gathered that there can be some quite big differences in what are supposed to be identical JFET's, though from what others have mentioned the SMD versions are supposed to be far more consistant?

I figured that the widespread use of J201 / 2N5457 was likely to be based on price/availability as you suggested, along with them also giving better results than most others available at the time, I doubt they would have gained such prominance if they'd been rubbish! Good to know that other JFET's will be ok to safely try, and I expected that get either the same or similar behaviour a fair bit of tweaking of the surrounding circuit might be required. I think getting into the whole circuit simulation software and deep theoretical knowledge side of things is something I'll leave to others though :os I'll be more than happy if I can make educated guesses about any experiments and find they do roughly what I thought they would :o)

I think my course has been pretty much laid in where it comes to JFET's by the looks of things. I've already got a few to start with, which should, with a bit of luck (And if I get my act in order at the weekend and knock up the JFET tester) be enough to hopefully get them matched into useable groups and start to have a play, then think about the SMD route for the future :o)

Do you think Therunoffgroove Eighteen and Professor Tweed would make a good starting point for trying amp in a box type circuits? they look simple enough to make reasonable sense to a relative novice and have some useful comments about modifications and sounding good.

And many thanks for all your help, very much appreciated :o)

Though I may get tempted if I can get something that looks promising at a sensible price and see what diabolical creations ensue 0:o)

[Ciaran Haslett]

Beaker's being modest here.  He's done a lot of work with JFETs and was the guy who made the switch to SMD and documented here for all of us.  I'm actually a convert based solely on his experiences and I'm not looking back!  So thanks again Mark

The ROG 18 is a great circuit to begin the journey on as it has trimmers for tuning the FETs and a single tone pot so less wiring.  But with all high gain amp sims, be prepared to use shielded wiring for Input and Output.  Some higher gain settings may introduce noise.

And Mark....have you seen this?  A bit expensive but sooooo convenient.  

[Pavlos]

Thanks Ciaran, I did think he seemed to know what he was on about ;o) one of the things I liked about the site before I took the plunge and joined was that there don't seem to be any prima donnas, just people (Some who are obviousley very knowledgable) having fun, learning stuff and helping others along the way :o)

Ah yes trimmers, maybe wouldn't have them on a final build, but they do look to be an absolute godsend for experimenting and learning, and yep, the less wiring is I think a good way to start, take little steps before going mad on the bigger more complicated circuits later on. Have breadboarded some of the fuzz circuits and found that even on full gain they haven't been too bad for noise even outside a box, and I do tend to prefer to start with lower - mid gained tones and then dirty them up with various flavours of boost/fuzz/drive, so that might help keep the noise a bit more manageable as well, but I'll make sure to have some shielded cable on hand just in case :o)

[Beaker]

ROG circuits are pretty much all good in my opinion, so your choices are great ones. Try the Ginger as well. Fairly simple, and although it is an Ampeg bass amp emulator, it is just as good on guitar as it is on bass. I've discovered that it works even better at 18V than at 9V, so putting the voltage doubler on a switch that Zach (Rocket88) put up on the main page a few days ago, is a worthwhile addition. Just remember to use 25V or over caps in the circuit.

Regarding trimmers, yes you can always use trimmers to tune the bias correctly, measure the resistance across them, then de-solder them from the board and replace with the same value resistors if you want to.

This is the shielded cable I use, and I get it from here as well - cheaper than ebay even when you include postage.

http://www.vdctrading.com/shop/van-damme-cable/audio/Single-Channel-Analogue-Audio-Cable/van-damme-pro-grade-classic-xke-1-pair-install/

[Pavlos]

haha, I think somone must have pulished my wish list ;o) had the Formula 5 has already been mentioned by someone else, and the ROG Ginger is one I'd like to have a go at as well, no idea why it's called Ginger though? and who is/are ROG? have had a good look at their site and there is a lot of very interesting info on there. Nothing wrong with using bass amps for guitar, after all the Fender Bassman is a classic example of a bss amp turning out to be rather good for guitars, think some of the Marshall bass amps have been used to good effect for guitar as well, and if I remember correctly Paul Kossoff used Marshall bass cabs, (But I think he was still using the guitar amp head with them) to great effect.

18V is something else that looks very interesting to try with a few of the pedals I'm thinking of building, just wondering how it affects the rating on power supplies? if for example a circuit draws 100ma at 9V, how much would the same circuit draw at 18V from a doubler?

I think I have some of that cable knocking about in box somewhere actually :o)

[Beaker]

I have no idea why they called it "Ginger" either, but it makes an awesome guitar pedal.

A lot of the Runoffgroove pedals (and many others too) are made to run at 18V, so the layouts invariably include the charge pump on the board, so they can run on a standard 9V supply.

For those circuits that don't include it, use one of these:

http://tagboardeffects.blogspot.co.uk/2016/01/switchable-onboard-power-918v.html

I really don't know how much less current you will get out of these though. Maybe someone else knows?

[Pavlos]

you really do have my wish list lol, this is on there as well!!

[Pavlos]

Right, I managed my first build today http://2.bp.blogspot.com/-zLpU6E6w8xs/UBRUu_pzPhI/AAAAAAAACCo/3XLcB0oReSc/s400/Greatly+Improved+JFET+Matcher+II.png, worked perfectly first time :o) so I have spent some time this afternoon checking my JFET's to see what readings I got, PSU was meauring 9.1V on the output but the question is, how normal (Useable) are these readings? and how closely matched are they?


        2N5457
        VP        IDSS
1 0.388 1.57
2 0.398 1.7
3 0.399 1.7
4 0.409 1.81
5 0.416 1.9
6 0.424 2.02
7 0.425 2.03
8 0.43        2.1
9 0.431 2.12
10 0.435 2.16
11 0.437 2.24
12 0.444 2.28
13 0.45        2.42
14 0.456 2.46
15 0.459 2.53
16 0.462 2.59
17 0.47        2.67
18 0.472 2.74
19 0.476 2.83
20 0.487 2.92

        J201
        VP        IDSS
1 0.531 0.24
2 0.609 0.32
3 0.664 0.38
4 0.791 0.54
5 0.796 0.55
6 0.84        0.59
7 0.865 0.62
8 0.902 0.64
9 0.888 0.66
10 0.939 0.68
11 0.943 0.69
12 0.948 0.76
13 1.034 0.77
14 1.015 0.78
15 1.047 0.79
16 1.029 0.8
17 1.022 0.84
18 1.012 0.86
19 1.084 0.89
20 1.156 1.67


        MPF102
        VP IDSS
1 0.99 0.73
2 1.15 0.9
3 1.13 0.91
4 1.17 0.93
5 1.19 0.96
6 1.32 1.17
7 1.32 1.18
8 1.39 1.21
9 1.52 1.41
10 1.63 1.53