Sorry guy's still wading thru this post: What do you mean reverse cunduction? -Ids?
Seems to me I've seen an N JFET cunducting like a 2sk170 data sheet?
Your, replying to a dog. A smart dog but hey, how many dogs can read and understand most of this transistor theory? let alone type? Spelling well
Yes, reverse conduction is negative drain current. My dog isn't that smart.
The two images are interesting. "R085" shows the sub- to near-threshold characteristics of the depletion-mode channel design. This gives reason behind the notion that a SemiSouth depletion mode JFET can play the role of a tube and thus benefit from the higher voltage rating of the device. My business partner is an accomplished blues guitar artist and is well acquainted with the musicians love of the tube amplifier. We've been talking about the possibilities. SemiSouth put out a 650 V JFET before the end...
The "depletion jfet" image is interesting too, but if it is from the same R085, don't get too excited. This is what a common-drain (source follower) amplifier imposes on the transistor terminal characteristics. If it came from an R085, the polarity is wrong as this can't happen for positive Vds and positive Id unless Vgd is being stepped. The legend seems to show Vgs is being stepped from a negative bias. Maybe there was just a mix up in connections? Otherwise, if from the same device the two images are mutually exclusive.
Attached is an image taken of the screen of a Tektronix 370B curve tracer measuring one of my R085s with correct polarity when stepping Vgs. (Vds is on the horizontal axis at 2 V per division and is negative, Id is on the vertical axis at 0.5 A per division and also negative, Vgs is stepped in 2 V increments starting from -18 V.) I added an eight Ohm load line using the cursor function of the 370B. You can see the potential for linearity, but this is to be expected from 100% source degeneracy. The negative feedback "straightens" out the non-linearity of the transistor by using the full gain of the device.
The image represents the JFET in reverse conduction, a property of the device under intense study by my research team right now (got no choice, I have to present this research at the PCIM-Europe conference in Nuremberg in May). One caution operating this way in forward conduction by stepping Vgd: You could accidentally overstress the reverse bias on the gate-source diode of the JFET and that is a good way to create a dead JFET.
I was saving this aspect of operation for a later article, but I'll add a short section in the next one to tip the hat to the "discovery" by Buzz.
Two graphs are from different sources. Main limitation of the curve tracer is current capability at reasonable Vds.
Sorry buzz, like I said, I didn't understand enough to agree but I'm not gettin into any arguments or insulting stuff without a much better reason. I totally appreciate what you did but, I wanna try to see it myself or verified by another party. Funny, we havent seen N.P. nudge us back on coarse or throw out any crumbs
It's time to try something myself... Gotta go help my brother and brave the storm at the moment...
It's time to try something myself... Gotta go help my brother and brave the storm at the moment...
The one was mine. The other is from the same thread by somebody else. My graphs still look attractive to me, but unfortunately, i dont think thats the whole story. Seems cascode modulation could yield some nice results, but it doesntvreally need that element for any other reason than distortion manipulation. If only ther was another way to play same game. Oh well, glad we have brought in some big guns.
Last edited:
I had 'em too. Could of been from you? They disagree with each other because they are not the same like Mike said. I did play around with the R085 graph and I think I liked what I was seeing at the time. I gotta admit, gotta sit down and grind thru what was goin on today.
I'm just beginning to understand what's under the hood of these things as it goes with audio. I'm from the knuckle dragging species. We try to power through the subtle features revealed by you all as fast as possible. Kinda fun learning about it. Like waking up and discovering there is something more to audio listening pleasure than an MP3 player and ear buds.
Thanks for sharing your stuff on this thread.
I'ld like to quickly comment on this:
Most musicians weather it be harp, solid/hollow body elctric Guitar, or several others, are not interested in linear distortionless amplification!
The acoustic guitar produces harmonics from the almost triangle wave of the strings due to the body shape, the wood, and construction etc. The electric only guitars/instruments depend on the harmonics created by their amp. For instance, some amps have a knob on them called overdrive. What do you think that is? As I said before about tube amplification, The soft clipping is a much "Nicer sound". Tube guitar preamps purposly over drive (to clipping) the signal and at various places in the circuit there are filters and other raz-ma-taz that gives them their signature sound. It is'nt a what goes in comes out transfer curve. I eluded to odd order distortion previously. If you have an amp with a gain(overdrive?) you can crank up to 1000, but it would be clipping at a gain of 10, your guitar will sound like Jimi Hendrix! No Sh$%^t! A square wave is essentially a fundamental sine wave, with a 3rd harmonic of 1/3 the level of the fundamental, and a 5th harmonic of 1/3rd the 3rd, and a 7th(John Curl's favorite) of 1/3rd the 5th, etc., etc, etc, A pure square wave is 33.333% THD! Pretty close to what Jimi used
It's not that their on another planet, but, it's a different goal altogether and Semiocunductors have not been able to satisfy half the guitarists out there for years!
Actually, here are a couple of details which might interest you.
1) What starts off as a triangle (or any other wave) on a guitar string does not remain the same shape! The harmonics of a stringed instrument are always a bit sharp, with higher harmonics being sharper than the lower ones. That's why pianos are tuned "stretch", with low notes lower than would be assumed from an A-440 reference, and conversely upper notes tuned a bit sharp -- it's to keep the harmonics of lower notes aligned with the fundamentals of higher ones. Hence that perfect waveshape falls into a chaotic looking mess in about a second.
2) The spectrum of a square wave is third harmonic 1/3 the fundamental level, fifth harmonic 1/5 the fundamental level, and so on for every odd harmonic. Hence the harmonic contribution is 1/3 + 1/5 + 1/7, etc etc, which amplitudes sum to 100% of the fundamental over the infinite series, so a square wave can be considered to have 100% THD.
BTW, I'm willing to bet a low-feedback solid state amplifier (hi Nelson!) might sound just as good for guitar as a decent tube amp. Tube amps tend to have much lower feedback than most SS amplifiers, and the "hardness" of SS clipping is partially due to lots of feedback linearizing the transfer function until there ain't no more transfer to function.
A tube guitar amp doesn't make a perfect 50/50 square wave. Grid current tends to shift the baseline of the incoming signal. A perfect square wave sounds hollow. Many of the most liked guitar amps will make a 40/60 wave at the output with power supply ripple modulation.
The Amp Camp Amp sound good, especially with the feedback disconnected. A low damping factor and moderate impedance to the rail when clipping helps. Most solid state guitar amps now days use current and voltage feedback to raise the output impedance.
I'm just beginning to understand what's under the hood of these things as it goes with audio. I'm from the knuckle dragging species. We try to power through the subtle features revealed by you all as fast as possible. Kinda fun learning about it. Like waking up and discovering there is something more to audio listening pleasure than an MP3 player and ear buds.
Thanks for sharing your stuff on this thread.
Whatever you do, do not listen to any of these amps. If you do, then you will have a problem
I want to thank all of you folks for weighing in on the question of guitar amps. I'm quite interested in seeing what higher voltage SiC technology might do where tube amps are used now, and so your input on the role of the musicality of the amps and the link to measurable quantities is very helpful for me to learn what you all know.
One comment. In the engineering literature I'm familiar with, THD is defined by the "RMS" method in which the square of the harmonic amplitudes are summed together and then the square root of the sum is divided by the magnitude of the fundamental. I suppose other definitions are used, but that is the one I am familiar with.
One comment. In the engineering literature I'm familiar with, THD is defined by the "RMS" method in which the square of the harmonic amplitudes are summed together and then the square root of the sum is divided by the magnitude of the fundamental. I suppose other definitions are used, but that is the one I am familiar with.
, at least recently in fashion , can be read in and between lines of Pa's writings in last several years 085 as preamp
Thanks for the great info...I'm just getting a preamp together with the 085 biased at only 10ma VDS about 40V with 1000ohm drain resistor 2-200ohm resitors in series with one bypassed for 13db gain . The signal looks great!! IT drives a 15K load wiht no problems. A triangle wave is perfect. I havn't listened to it yet...
Which leads me to this...What is the best topology for these JFETS, voltage amp or current amp?
The two images are interesting. "R085" shows the sub- to near-threshold characteristics of the depletion-mode channel design. This gives reason behind the notion that a SemiSouth depletion mode JFET can play the role of a tube and thus benefit from the higher voltage rating of the device. My business partner is an accomplished blues guitar artist and is well acquainted with the musicians love of the tube amplifier. We've been talking about the possibilities. SemiSouth put out a 650 V JFET before the end...
The "depletion jfet" image is interesting too, but if it is from the same R085, don't get too excited. This is what a common-drain (source follower) amplifier imposes on the transistor terminal characteristics. If it came from an R085, the polarity is wrong as this can't happen for positive Vds and positive Id unless Vgd is being stepped. The legend seems to show Vgs is being stepped from a negative bias. Maybe there was just a mix up in connections? Otherwise, if from the same device the two images are mutually exclusive.
Attached is an image taken of the screen of a Tektronix 370B curve tracer measuring one of my R085s with correct polarity when stepping Vgs. (Vds is on the horizontal axis at 2 V per division and is negative, Id is on the vertical axis at 0.5 A per division and also negative, Vgs is stepped in 2 V increments starting from -18 V.) I added an eight Ohm load line using the cursor function of the 370B. You can see the potential for linearity, but this is to be expected from 100% source degeneracy. The negative feedback "straightens" out the non-linearity of the transistor by using the full gain of the device.
The image represents the JFET in reverse conduction, a property of the device under intense study by my research team right now (got no choice, I have to present this research at the PCIM-Europe conference in Nuremberg in May). One caution operating this way in forward conduction by stepping Vgd: You could accidentally overstress the reverse bias on the gate-source diode of the JFET and that is a good way to create a dead JFET.
I was saving this aspect of operation for a later article, but I'll add a short section in the next one to tip the hat to the "discovery" by Buzz.
Thanks for the great info...I'm just getting a preamp together with the 085 biased at only 10ma VDS about 40V with 1000ohm drain resistor 2-200ohm resitors in series with one bypassed for 13db gain . The signal looks great!! IT drives a 15K load wiht no problems. A triangle wave is perfect. I havn't listened to it yet...
Which leads me to this...What is the best topology for these JFETS, voltage amp or current amp?
Well, I concede I may have some of that wrong. It's been a little while since I even thought of those things. So, I just went to my Pspice and pulled up one of my single ended preamp circuits. I hung some 4.3V zeners on the output. It was already set to output about 20V w .01% THD predominatly about 2mV of H3, 600uV of H2, 100uV oh H4 etc... But, clamping the output like that was a quick and dirty way of producing the idea I was suggesting. I wasn't saying every stringed instrument or even most guitars are making square waves out of triangles. I was exagerating the point of clippling and it's importance. You really don't know how much clipping you actually tolerate with the extreme dynamic nature of the music. And, different gain devices and topologies can greatly effect that.
Back to the "Jimi Hendrix" approach. When I ran the sim with the 4V clamping diodes on the output, it was well on it's way to being a nice square wave in the sims graphical output. The spectra naturally was different. Now we have a fundamental of 1kHz @ of 5.22V, H2 of .13V, H3 of 1.56V, H4 of .12, H5 of .8V, H6 of .1V, H7 of .46V. So, we can see that clipping the Sh$%^t out of your signal can add substantial odd harmonic material. Ohh, Yea, the sim said 35% THD
Tube circuits clip softer, visibly rounder on the scope, and relieve our ears of most of those odd harmonics, whatever their proportions. Solid state circuits as in BJT and MOSFET, tend slam right up to the rail and flatten out the waveform when clipping. Apparently SITs do not do this so bad. I personally feel this is a very important quality due to the dynamic nature of music and limited headroom in some of our electronics. Just my 2C
Imagine a bright future for R100A with DIYers.
Quiet thread! R100A will find a great application in the Pass amps ACA and PLH; in addition to his recent F6. DIYers who have already built the first two, [and the myriad of past others] with power MOSFETs, have raved about their performance. One can only imagine and expect a potential improved performance of these renowned Pass amps with R100A or equivalent NJFET instead.
Best regards
Quiet thread! R100A will find a great application in the Pass amps ACA and PLH; in addition to his recent F6. DIYers who have already built the first two, [and the myriad of past others] with power MOSFETs, have raved about their performance. One can only imagine and expect a potential improved performance of these renowned Pass amps with R100A or equivalent NJFET instead.
Best regards
Based on what's been said already: I think the R100 style amps would be better in almost all cases. By that I mean, anytime you need or are stuck with N-FETs, the R100 will likely be worth tunning into your circuit. Zen V4? what have you Even the R085 can be grafted into your favorite circuit! Even only the bottom FET of a PLH would probably benifit?
I'm collecting the components to try some Ids vs. Vds kinda graphs by the weekend We will see. I guess I'm most interested to see the low voltage R085 Curves?
Even the R085 can be grafted into your favorite circuit! Even only the bottom FET of a PLH would probably benifit?I'm collecting the components to try some Ids vs. Vds kinda graphs by the weekend
We will see. I guess I'm most interested to see the low voltage R085 Curves? Thanks flg. I hope that we continue to give Dr. Mazzola a realistic feel for the potential broad useage of R100A among DIYers. The use of R100A may even expand to producers of power amps tauting discrete components in their circuits; e.g. KRELL etc...diyAudio has a 1/4 million members. Let 10% of them be solely Solid Stater's, and each member will buy ten R100A/year. This is a 1/4 millions pieces multiplied by an affordable price [tbd] to generate a lot of dough.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.