LSK389 - Gfs=20mS, Ciss=25pF, Crss=5.5pF, En=0.9nV/rtHz
LSK489 - Gfs=1.5mS, Ciss=4pF, Crss=3pF, En=1.8nV/rtHz
BF862 - Gfs=45mS, Ciss=10pF, Crss=1.9pF, En=0.8nV/rtHz
2SK3557 - Gfs=35mS, Ciss=10pF, Crss=2.9pF, En=1.3nV/rtHz (based on Dmitri's measurements)
BF862 is the best of all worlds (largest Gfs/Ciss ratio, lowest Crss, lowest noise). 2SK3557 (and the dual version) is the second.
If it were for my own use and I needed the matching, I've got hundreds of 2SK389V duals.
I did use some LSK170 parts in the front end of a power amplifier, primarily because they were available in SM. The design used a d.c. servo, and as well the negative resistance generator I showed earlier to compensate for the stiff loading on the JFET. It all worked extremely well.
The amplifier company suffered from personality conflicts and four of us quit abruptly, and are waiting for the corporation to time out. So it goes.
Probably so, although at least I'm not asking for a new chip, just doing something on a spur off the main production somewhere using equipment already developed for bipolars, and an existing package.
In the meantime, the topologies I've been working on lately can work well with loosely matched parts, although some adjustment can be required.
The accelerometer has to be characterized first.
Here is the B&K article (pdf pages 27-36)
http://www.bksv.com/doc/TechnicalReview1976-2.pdf
It answers many of the questions here (and generates a few more)
This is another valuable article of them (test records and vinyl issues)
http://www.bksv.com/doc/TechnicalReview1962-3.pdf
Wow. Thanks, to you and to Demian. Great stuff.
Much thanks, as always,
Chris
Hi Waly,
The exact schematic and measured THD numbers are in the LA article. THD is less than 0.005% at an input level of 20mV rms.
Your first simulation was obviously flawed. The onus is on you to get the simulation right before you say any more. I suggested how you can get it right in my last post. My models are available on my website.
Cheers,
Bob
Well ours does, and probably the commercial HSPICE from CADENCE does. I might add probably only because RF circuits with HEMT's need them.
Last edited:
Maybe, but then you didn't show any better verifiable results other than "because I say so" and quoting articles I don't have access to (LA). And snubbing answering/addressing the core issues. I'm disappointed, you seem to have no interest in listening to alternative voices.
Good night and good bye.
Okay. HSPICE uses level 3 to distinguish MESFETs from JFETs, LTspice uses a separate circuit element, Z - MESFET transistor, to characterise these. Or is there something more to it ...?
The issue is available. www.linearaudio.net - home Bob's article is clear, well-documented, and highly educational.
Look Waly, we spend a lot of time on our designs, and when criticism is put forth, usually there is often a problem with the 'criticism' perhaps more than the criticism describes a problem. That doesn't mean that Bob or I cannot learn from you, but be careful and be sure of your accuracy, because you can potentially throw doubt into our efforts, and it could potentially confuse others. Also, it is important that you realize that we are doing OUR BEST as well to help others. Confrontation is not really appropriate, all the time. It is best to go 'lightly' and 'diplomatically'.
If not, we will fight 'fire with fire', and it could potentially hurt your feelings, as Bob has apparently done, just now.
If not, we will fight 'fire with fire', and it could potentially hurt your feelings, as Bob has apparently done, just now.
by eyeballing this curve:
Measurements Rate SMT Low-Voltage n-JFETs Under Consistent Conditions | Power content from Electronic Design
about 70 Hz
Hi Waly,
I spent a lot of time patiently trying to address your questions and doubts, in spite of the fact that you attacked my design and carelessly claimed results from a flawed simulation you did. I even went back and re-simmed some circuits to see if you had a point. You made some personal attacks, but I did not try to insult you. We have reached a point where additional discussion is not worthwhile. I try to be generous with my time, but my time does have value.
Going forward, you really will learn from this if you just put in the effort to do the sims I suggested in my last post. Sanity checks are a wonderful way to learn. After you get those working properly, then make the changes you want to make, one step at a time. I urge you to subscribe to LA - you will learn a lot from it. If you haven't already, buy Doug Self's amplifier book and my amplifier book. If you buy mine, it may help you a lot to read the two chapters on SPICE. Use LTspice.
You will learn a lot from others by questioning them - most will respond in a helpful way. You will learn much less by attacking others.
Cheers,
Bob
To which Mr Cordell definitely Not adheres !
FYI, retail of a J73 or K146 a quarter century ago in the EU zone, was in the order of $6 per dual JFET unit.
Back then, those parts were as readily available around every corner here, as the OnSemi/Philips JFETs are now in the US.
By the time they were discontinued, I had barely scraped enough shillings & pences to buy by the hundred at half rate. Rare few could be picked at 0.1% Idss.
A little over $5/pc for 1k orders sounds less boutique to me than the usual and total bizarre rates of obsolete parts and fragrance components.
Despite that the BF862 is a real radio trash can treasure, buy while you still can.
(NSA, Area 96. I didn't get passed thinking that the designer felt process soixante-neuf would be off the tacky scale)
Well I am feeling evil today. So I'll ask a few simple questions:
1 If I charge a 100,000 uF capacitor to 100 volts, how much energy is stored?
2 If I charge a 50,000 uF capacitor to 200 volts, how much energy is stored?
3 If I place a 100,000 uF capacitor in series with a second 100.000 uF capacitor, what is the total series capacitance?
4 If you don't know what this question should be, just wait.
5 Where did the energy come from?
ES
1 If I charge a 100,000 uF capacitor to 100 volts, how much energy is stored?
2 If I charge a 50,000 uF capacitor to 200 volts, how much energy is stored?
3 If I place a 100,000 uF capacitor in series with a second 100.000 uF capacitor, what is the total series capacitance?
4 If you don't know what this question should be, just wait.
5 Where did the energy come from?
ES
I bought a few hundred 862 with that very concern in mind.
Process 69 is a pretty-nice small signal PNP, the 2N3251A et al. It appears from the characteristic curves to have a decent Vbe magnitude match to Process 23, at least from National at that time. The latter chip is in their 2N3904, which I find a lot of folks still consider good enough for about anything.
- Status
- Not open for further replies.