FET emulation of tubes without trimmer?

[puretube]

anybody got experience in sound
with other methods of biasing a FET
in tube-clone-circuits
than an annoying trimpot RD,
while maintaining the possibility of replacing
FETs with large production spreads?
:?:

[freebird1127]

On this topic, I'd like to mention that I've heard Lateral-MOSFETs (L-MOSFETs) are quite like tubes in their clipping characteristics.  Something to look into I guess.

[RDV]

I think that will be tough considering the variance in JFETs. One can always set the voltage with a pot and replace with the appropos resistor.

RDV

[R.G.]

Yes, I have.

Sorry, I have nondisclosure obligations, too.

[puretube]

I think that will be tough considering the variance in JFETs. One can always set the voltage with a pot and replace with the appropos resistor.

RDV

1. you wouldn`t wanna do that in a production run - that`s the reason why I asked;

2. let`s see if there`s more than one...;
3. I`m aware of the fact that there is constant-voltage, constant-current, and "combo" - biasing;
4. I`m nosy if there`s a less-MOJO-factor in FETs running on one of these biases, comparable to the (well: quite different aspect there) question Peter Snowberg asked in the discrete opamp thhread: "~...whether it makes a difference in sound, with the transistors running on resistors instead of constant current sources.."

[Joe Davisson]

Try one of these:

-A voltage-divider bias (probably 10M/1M) will reduce the variance, but requires an input cap.
-Larger source/drain resistors will reduce it more, but increases the gain.

A combination of the two is probably best. The increase in gain can be counteracted with an input resistor, and the emitter bypass cap should be smaller. The usual common bond between complexity/results Look at this for an example:
http://www.diystompboxes.com/analogalchemy/pedals/jfetvulcan.html

The idea here was to get the gain up to compare with the transistor version, as well as "just work" without biasing JFETs. (Ignore the diodes, works without them.) The outputs will never be exactly on target, but the variance can be measured in tens-of-volts instead of volts. You can better your chances that what works for the first 10 or so JFETs in a batch would work for the run.

-Joe

[puretube]

Thank you Joe!
yep, that`s one of the mentioned combo-biasings described
on page 5 of this:
http://www.colorado.edu/physics/phys3330/phys3330_fa04/pdfdocs/AN102FETbiasing.pdf,
and what I was wondering about, if it is of any disadvantage to the MOJO...

haven`t tried your circuits personally (IIRC), but from your reputation,
I reckon it`ll sound quite well.


BTW: I love coupling/blocking caps for f-response "shaping" reasons...

[Peter Snowberg]

BTW: I love coupling/blocking caps for f-response "shaping" reasons...

Spoken like a tube head.

[puretube]

oh, Peter, I did build me some solidstate all-the-way-through-DC-coupled amp some 30 years ago: all the way through the speakers :evil:

[puretube]

I think that will be tough considering the variance in JFETs. One can always set the voltage with a pot and replace with the appropos resistor.

RDV

yeah, Rick (-y ?) - if I`d have good contacts to a chinese sweatshop production "plant", the "variance-spread" wouldn`t hold me back...

[puretube]

JOE: I solemnly promise, not to put your diodes in there,
should I ever go into the FET-emulates-tube circuitry biz!

(I`m already glad to have mentioned the awareness of that biasing method in time!)

[Joe Davisson]

The voltage divider actually improves things, JFET gates tend to "stick" (at high-gain) without a little positive bias. You'll notice the difference more when playing fast, it's less constricted.

The diode trick works better with bipolars. Try sticking a resistor (100k?) in place of the diode, designed to counteract the gain increase of using large source/drain resistors, and adjust the overall frequency reponse of the stage via bypass capacitor. Just a thought.

-Joe

[Peter Snowberg]

oh, Peter, I did build me some solidstate all-the-way-through-DC-coupled amp some 30 years ago: all the way through the speakers :evil:

DC in the speakers? :shock: Watch those offsets!

I prefer the cap approach myself, however this does not limit to transistors or tubes.

Here's a combination of tubes, DC, and high performance for slow signals. The missile guidance computer I used to work on (for museum restoration) used about 70 of these dual opamps.

[WGTP]

I built Joe's JFet Vulcan, but ended up using 47k resistors instead of the 33k's with the J201's and J202's I had, because the bias was closer.  The cruxt of the problem!   :twisted:

It sounded marvelous.

Side bar.  Joe, how do you think the Jfet Vulcan sounds compared to the BJT?  

[Joe Davisson]

I built Joe's JFet Vulcan, but ended up using 47k resistors instead of the 33k's with the J201's and J202's I had, because the bias was closer. The cruxt of the problem!

JFETs are evil.

Side bar. Joe, how do you think the Jfet Vulcan sounds compared to the BJT?

I do like it, but not quite as much as the bipolar version.

[puretube]

Peter, is that a Philbrick circuit?  

[Peter Snowberg]

Nope, It's the DC amplifier from a surface to air nuclear missile guidance computer.

The system was called the Nike Hercules and there were about 300 bases world-wide IIRC including several in Germany. A sweeping style of RADAR was used for acquisition of targets, then two point-source RADARs measured the azimuth, elevation, and distance (in otherwords the 3D position) of both the missile and the target. The computer sent guidance commands to the missile to fly it to the position of the target (actually about 10 meters above and in front of the target)

For more.... http://www.nps.gov/goga/nike/ & http://www.ed-thelen.org/

I have the Philbrick application notes here if you would like.

[puretube]

that`s what the state variable filter was developed for...

[Peter Snowberg]

That's also where Doppler weather RADAR came from.

Mr. Norton (of opamp fame) was one of the engineers who worked on the Nike guidance project. http://cmc.rice.edu/docs/docs/Joh2002Sep1TheEquival.pdf

The hydrogen thyratron that fired the 1.2 megawatt Nike search radar is still my favorite tube of all time. It's glass and it's HUGE!  :shock:

Much of the engineering in that system was TOP SECRET classified at the time it was built and used. (1945-the mid 1960s)

[puretube]

FETs don`t like me...   :cry: