testing JFETS: J201 higher gain than 2N5457?

[zjokka]

I had taken this for granted that a J201 had more gain than a 2N5457, but now I measured a whole stack. Some not impressionist averages to give you an idea:

J201: hFE 160-240
2N5457: hFE 400 and more

Or am I missing something by testing them in my DMM? My Peak Transistor Analyser can identify and measure everything except a JFET. I built the Fat Boostered with 2n5457s without testing them and it sounded terrific. I was aiming at lots of clean headroom, that's why I went for that transistor. It was clean all the way -- until I passed it on to a friend who tested it with Danelectro + lipstick pickup and he got overdrive at higher settings.

Shouldn't the JFETs be biased properly in this Fat Boostered? Why isn't there a bias pot in these designs?

thanks for reading this,
zj

[Seljer]

hfe is "Forward Current Gain", which I don't think is a quality of JFETs, only for BJTs

go read the Fetzer Valve article on runoffgroove, lots of info on the gain of FETs
but as for in the Fat Boostered: the first FET is just there as a buffer, the second section is based off the muamp circuit thingy in http://www.muzique.com/amz/mini.htm

[Gilles C]

The Fat Boostered sound is very dependant on the FETs used and on the input signal.

I just tried mine a few minutes ago with my guitar, and it was indeed clean sounding, but was "warmering" the sound of guitar . But I feel it wouldn't take a lot to make it hotter. I want to install it in a box with a Fetzer Valve to get more "grit".

Try it with a booster before it just for fun.

Gilles

[Jay Doyle]

You are getting into the main difference between BJTs and FETs, BJTs are current controlled and FETs are voltage controled. BJTs have hFE while FETs have transconductance, therefore your DMM will not measure FETs properly.

The Fat Boostered is built upon the Mini-Booster, muamp or SSRP circuit which is self biasing so there doesn't need to be any bias trims in there.

I don't have any links but The Art of Electronics is a great resource and searching around the web for some basic EE tutorials will help you to get a better understanding of the differences between BJTs and FETs.

[zjokka]

thanks for all the answers, I was also very curious why always the same jfet transistors are suggested and the other way around what makes a jfet useable in a stompbox circuit.

we know about the j201 and the 2n5457 and the MPSA14 (right?), but so many others.
that's why I ordered some various jfets like j101 and j113 i believe. the j101 worked in the fb but at highly reduced volume. should all n channel jfet perform in the same setups?

any direct links for in depth info on fets?

thanks
jj

[Ed G.]

To make things more complicated, some devices with the same part number sound different by different manufacturers.
I tried the J113 and liked the sound. I'm not sure which I had, I think it was National, which I can't get any more of.
I ordered Vishay and found them unusable. The J113 is usually sold as a 'chopper' fet, but in some cases, it handles amplifier duty well. If you get the right one.

[Jay Doyle]

thanks for all the answers, I was also very curious why always the same jfet transistors are suggested and the other way around what makes a jfet useable in a stompbox circuit.

we know about the j201 and the 2n5457 and the MPSA14 (right?), but so many others.
that's why I ordered some various jfets like j101 and j113 i believe. the j101 worked in the fb but at highly reduced volume. should all n channel jfet perform in the same setups?

any direct links for in depth info on fets?

thanks
jj

Here is a link to the best web information on JFETs:

http://www.vishay.com/fets-small-signal/ssfanp/

We tend to use the same JFETs because they have proven to work. The J201 was found by, I believe, Jack Orman, to be a good sounding, higher transconductance JFET. Then Small Bear stocked it, so it just became the staple because it was easy to get. There are fewer JFET part numbers then BJTs. I haven't really tried all that many myself. I tend to use either the J201 or the J202.

[R.G.]

"Tone" is not a specification on a datasheet.

A data sheet is the entire sum total of every single thing a manufacturer will guarantee his devices to do. When they say "Rdson max = 200ohms" that's what they mean, and they employ statisticians to do the math and sampling to prove it. And so for every single thing which is shown as a min or a max. If there is a blank place under a min or max column, that's what it means - they will not guarantee any value whatsoever there. "Typical" means... well, it sounds nice and comforting, doesn't it?

Each manufacturer developed his own semiconductor processes and fabrication steps either by doing the original work or hiring people from other manufacturers to tell them how. As you can imagine, there is a lot of secrecy on exactly how X company makes a part and how that varies from Y company's same part number. That's where the datasheet comes in. Big industrial customers insist that if they buy zillions of parts, they all work. And they insist that they not be highjacked by a single source of a critical part. So the datasheet becomes the place where the buyer and two (or more) semiconductor suppliers meet. In general, both suppliers will swear that their parts will do what's on the datasheet. That's all they'll swear, because a buyer who designs in something that the supplier told them the parts would do will come after their blood if the parts won't do it.

So we have the case of independently produced parts from different processes are held to meeting a limited number of specifications. Is it any wonder that they will do exactly what's on a datasheet and no more?

Or that they are not every going to put "tone" on a datasheet, even if there was a concise definition?

[birt]

has anyone ever tried the 2sk30 jfet for audio?

[Gilles C]

Very often... I love them.

In fact, I even used 2 in the Fat Boostered for more gain.

Gilles

[Paul Perry (Frostwave)]

I've never designed (or built) anything with a fet in it (except as a source following buffer) because of the hassle of having such wide variations within the same type of fet.
Maybe I will one day..... but I'm not looking foward to it!

[zjokka]

any direct links for in depth info on fets?

Here is a link to the best web information on JFETs:

http://www.vishay.com/fets-small-signal/ssfanp/

Thanks that's what I'm looking for. Will start by digging into that.

"Tone" is not a specification on a datasheet. (...) Or that they are not every going to put "tone" on a datasheet, even if there was a concise definition?

I never used to the word "tone" in my thread. I was asking about what will work. There must be something about the J201 that makes it work in most applications. I really want to quit looking for specific transistors if not necesarry. If the schem says use a 2SB??? should i look for it?

I've never designed (or built) anything with a fet in it (except as a source following buffer) because of the hassle of having such wide variations within the same type of fet.
Maybe I will one day..... but I'm not looking foward to it!

I'm  baffled. Any fet I plugged in just worked immediately. I think a lot of people make great sounding stompboxes with these everyday around here. Don't know what should keep anyone from breadboarding a minibooser.

zj

[Doug_H]

I've never designed (or built) anything with a fet in it (except as a source following buffer) because of the hassle of having such wide variations within the same type of fet.
Maybe I will one day..... but I'm not looking foward to it!

And you can't make me look forward to it either!

[Paul Perry (Frostwave)]

I'm  baffled. Any fet I plugged in just worked immediately. I think a lot of people make great sounding stompboxes with these everyday around here. Don't know what should keep anyone from breadboarding a minibooser.zj

Fair enough, any fet will amplify, and if you want less output, turn down the gain.
I was realy referring to the hassle of matching fets when you want to voltage control a bunch for a voltage controlled filter or phaser.
This is what is wonderful abou DIY stompboxing, and why people should do it: you can spend days - or weeks - tweaking a box till it is doing exactly what you want, with the components that you happened to have on hand. But if you are making a batch commercially, they have to behave exactly the same way...with parts you can order from stock... and it isn't always easy to do that!

[Ge_Whiz]

Incidentally, the MPSA14 is a "Darlington pair", and not a FET.

[zjokka]

Maybe I'm beginning to see the point: I'm building my second tube amp simulator from ROG. Have done the Flipster, now completed the Prof Tweed. With both build I trusted my soldering skills well enough to omit using transistor sockets. Just soldered the J201/2n5457s in without a problem.

Now I never got either to bias right. I changed a trimpot for lower value, Flipster 2/4 biased about right in ProfTweed 1/3. I'm beginning to see that I should have used sockets anyway. It's not the the trimpots causing the problem, but the dfference in the fets. I think that's what you guys are saying right.

Well I'm just a music playin' electroincs ignoramus, but if there are so many problems with these designs it should be noted somewhere. Nowhere on ROG I read about swapping fets till it biasses right.

learning something still everyday...
zj

[Jay Doyle]

Nowhere on ROG I read about swapping fets till it biasses right.

Did you use the correct FETs that are specified in the schem? If not, then you can't really blame ROG can you? 

But the 'Trim on the drain' biasing scheme tends to exascerbate the problem with the variance in specs. Because even if you can get the trim to bias the FET to the right range, chances are that the resistance will be different than the one used by ROG, or anyone else who did sound samples you like, so the gain (and a number of other things) of each stage will be different.

A MiniBooster/MuAmp/SSRP setup self-biases so that cancels out a lot of the problems...

[zjokka]

Nowhere on ROG I read about swapping fets till it biasses right.

Did you use the correct FETs that are specified in the schem? If not, then you can't really blame ROG can you? 

What I meant to says is: I used the exact components listed, but given the fact that fets differ amongst themselves so much, I'm beginning to conclude that I should have swapped the correct types in until I found devices that biased correctly at 4,5V with the trimpots given?

Now with the Prof Tweed one biases more or less correctly at 4,6V a second jumps from 3,3V to 6,8V and the last one only comes down to 9,00V even with the 100k from the trimpot (9,08V on the battery. )  So I took the intuitive route and started looking for a resistor to bring it down more. Tried several and got down to 6,5V with a 3M4. Haven't tried 5M yet, but should listen to the circuit first maybe.

Never got the Flipster to bias correctly but even with half-off bias sounded remarkable, but didn't want to box it up know it could be improved.

maybe I better read the link first.;.
zj

[Jay Doyle]

If you used the exact components then you shouldn't have to swap out FETs. Something is wrong somewhere else. JFETs of the same type are different, but not THAT different. You should be able to get it working if you used the right parts.

What did you mean by 'jumps' from 3.3V to 6.8V?

I think that something somewhere else is wrong. Make sure the pinouts are right, all solder joints are good (reheat them all and dab a little more solder on each) and that you have the right value resistors everywhere.

After this: Post the voltages on each pin of each FET in the two circuits and we can help you more.

[zjokka]

If you used the exact components then you shouldn't have to swap out FETs. Something is wrong somewhere else. JFETs of the same type are different, but not THAT different. You should be able to get it working if you used the right parts.

What did you mean by 'jumps' from 3.3V to 6.8V?

The voltage jumps from 3.3V to 6.8V by a microscopic turn of the trimpot. I had this in the Flipster as well, then some board members suggested using a smaller value trimpot, some adding a resistor in parallel to bring down the resistance.

With the third Pro Tweed trimpot, that was something different, I couldn't get the 9V down with any credible value (100k-800k), but did go down to 6.5V with a 3M4, so I'm sure it's not a solderbridge at hat location, and am sure connetions is ok. I etched the layout from ROG.

Will measure the voltages later today.
thanks still Jay!