For those who want to slap modern SMD transistors on a curve tracer ---
I just bought this little booger on eBay at a good price. It looks to be just the right tool for characterizing all of those fantastic SMD transistors that won the performance shootouts in Art Of Electronics, 3rd edition.
CAUTION: if you are seduced by the spectacular beta linearity of the 2SA1312BL (as I was), let me warn you: Toshiba shipped them to Mouser (and Mouser shipped them to me) four months late. Arrgh.
I just bought this little booger on eBay at a good price. It looks to be just the right tool for characterizing all of those fantastic SMD transistors that won the performance shootouts in Art Of Electronics, 3rd edition.
CAUTION: if you are seduced by the spectacular beta linearity of the 2SA1312BL (as I was), let me warn you: Toshiba shipped them to Mouser (and Mouser shipped them to me) four months late. Arrgh.
The famous 2 transistor current source oscillating in LTSpice
You need to add a little bit of inductance to get it to oscillate, and a bit of stimulus to reliably kick it off, but you can see the two transistor Vbe referenced current source oscillate quite nicely. 145 nH was about the minimum to get it to oscillate. The frequency was 193 MHz. Node 2 is where the 145 nH meets the collector.
You need to add a little bit of inductance to get it to oscillate, and a bit of stimulus to reliably kick it off, but you can see the two transistor Vbe referenced current source oscillate quite nicely. 145 nH was about the minimum to get it to oscillate. The frequency was 193 MHz. Node 2 is where the 145 nH meets the collector.
Attachments
Last edited:
Dan, I got it to oscillate in real life, both on a PCB and also on a solderless protoboard, with no extra inductance: post #7525 shows the waveforms.
I also got it to oscillate in LTSPICE simulation, by choosing the "right" transistors, by connecting a low impedance load to the current source output, and by adding ten picofarads of stray capacitance to the output transistor's emitter: post #7585
I also got it to oscillate in LTSPICE simulation, by choosing the "right" transistors, by connecting a low impedance load to the current source output, and by adding ten picofarads of stray capacitance to the output transistor's emitter: post #7585
LEDs need about 10-20mA to realize a reasonably low impedance. Attached is a link to a thread where I measured incremental impedances for 2 strings of Green GaP leds and one of NOS Litronix GaAsP leds. The greens stack up somewhere in the vicinity of 14-15 ohms, the red legacy leds had a much lower impedance, depending on bias current. I guess I should try some of the other old GaAsP red LEDs I have floating around to see if there is a big variation between vendors. I also have several flavors of Litronix GaAsP reds. It would be interesting to see if there is a variation in impedance between them. It may be that I lucked out and got one part number (perhaps with a larger die) that had especially low incremental impedance.
I suspect the shorter wavelength GaAsP leds (like amber) would have higher impedance.
http://www.diyaudio.com/forums/tube...w-year-new-amp-argue-about-4.html#post2446543
I was musing about using a diode-connected 2N4401 as part of the cathode bias string for a triode running about 10mA bias current, so I did some measurements with a series resistor and a bench supply to see what I would get. Starting a a few hundred microamps, the transistor is ~50 ohms incremental impedance. Once you get to 10mA, this pares down to a little over 5 ohms.
I suspect the shorter wavelength GaAsP leds (like amber) would have higher impedance.
http://www.diyaudio.com/forums/tube...w-year-new-amp-argue-about-4.html#post2446543
I was musing about using a diode-connected 2N4401 as part of the cathode bias string for a triode running about 10mA bias current, so I did some measurements with a series resistor and a bench supply to see what I would get. Starting a a few hundred microamps, the transistor is ~50 ohms incremental impedance. Once you get to 10mA, this pares down to a little over 5 ohms.
wrenchone, i just had to say how hard I laughed about that one - good enough for a pepsi spray from the nose, which previously had been earned on several occasions by jocko.
mlloyd1
mlloyd1
Put yourself in Bob Cordell's shoes for a moment. He's about to release the 2nd edition of his book. What would you write, if you were Bob, about the 2T current source? I don't think he can recommend avoiding it 100% of the time, since the 1st edition used it in many dozens of example power amps. So perhaps at this point, his only option is to lay out some guidelines and recommendations and rules-of-thumb, that will keep most people away from serious trouble, most of the time. Remember that not all readers are electrical engineers, and many of them employ the Monkey See Monkey Do method of circuit "design" when creating audio circuits.
What would you suggest Bob say?
(Me, I have forsaken the 2T current source. Tossed it into the fire. I'd rather build a current source with a couple more parts, and sidestep the possible-oscillation issue entirely. But I'm not writing a 2nd edition either.)
I seem to have lost the plot here.
Is the 2T current source what's shown in Figure 2.9a in Bob's 1st edition?
I have to admit, I've used the Vbe referenced current source in many designs, and never saw it oscillate...so it was kind of interesting when the subject came up...that's why I fooled around to see if I could see in in LTSpice.
Now, even though 145 nH is about 6" of wire...you probably wouldn't have anything close to that on a well laid out PCB...but it's certainly good to be aware of the possibility!
Now, even though 145 nH is about 6" of wire...you probably wouldn't have anything close to that on a well laid out PCB...but it's certainly good to be aware of the possibility!
Hi David,
I use the DN2540N3 and N5 quite often, and the LND150N3 as well. I haven't had any oscillation problems with them without a damping resistor. Having said that, I should mention that I use a 1206 or 805 size resistor with the N3 case (TO-92) for the source resistance. There is almost zero inductance with that arrangement. The N5 case is a bit of a pain to use. I mount a 1/4 watt resistor across the leads on that one. No oscillation issues there either.
More expensive than the LED-Transistor-resistor type I use, but much more retro-fit friendly on PC boards. Not as cool looking of course.
I set the current using a proto-board jig I made in Ryerson decades ago (1983 actually) and a mounted 1K0 pot. No issues with a couple inches of ribbon cable running together. I'd say they were pretty safe to use.
-Chris
I use the DN2540N3 and N5 quite often, and the LND150N3 as well. I haven't had any oscillation problems with them without a damping resistor. Having said that, I should mention that I use a 1206 or 805 size resistor with the N3 case (TO-92) for the source resistance. There is almost zero inductance with that arrangement. The N5 case is a bit of a pain to use. I mount a 1/4 watt resistor across the leads on that one. No oscillation issues there either.
More expensive than the LED-Transistor-resistor type I use, but much more retro-fit friendly on PC boards. Not as cool looking of course.
I set the current using a proto-board jig I made in Ryerson decades ago (1983 actually) and a mounted 1K0 pot. No issues with a couple inches of ribbon cable running together. I'd say they were pretty safe to use.
-Chris
Walt Jung's schematic is attached; he is careful to explicitly include (a) the load resistance driven by the current source; (b) base resistor R2 which allows pole-splitting; (c) a warning that the circuit can oscillate.
When I plot open loop gain and phase of the circuit in LTSPICE AC analysis, I find that item (a) above is crucial. Small changes in load resistance produce large changes in stability margin.
_
Attachments
Hi Mark,
As I have said before, I will definitely discuss more deeply the stability issue concerning the feedback current source.
Nobody that writes a 600-page book gets everything covered completely. Since its publication I have encouraged readers to give me feedback, and many have, in a very sincere and helpful way. I continue to encourage such feedback. In five years since its publication this is the first concern that has been raised regarding the feedback current source. Some things just slip through the cracks, however.
I agree completely that less-experienced designers should always be considered when recommending and explaining certain circuits.
As we all know, even one-transistor arrangements will oscillate under the right conditions. If you look at the Hartley and Colpitts oscillator topologies, it is easy to see a few of the many ways in which this can happen.
Bear in mind, however, that the feedback current source provides a much higher output impedance than the shunt-referenced CE current source, as explained in my first edition. Some designers like this and are not so quick to throw the baby out with the bathwater. For very high output impedances, there is, of course, the cascoded current source. However, this will cost you headroom.
I'm guessing you will not be buying the second edition
.Cheers,
Bob
I've suspected that LTSpice might not model the base impedance well above 40MHz or so and that may be why simple circuits like this simulate as stable but oscillate on the bench with no obvious parasitic explanation.
T117's simulation shows ringing but it's still relatively damped, if it oscillates there is no small reason for it. I have found that using low feed current helps with stability for 2T current sources, but also reduces performance. I can get T117's schematic to oscillate by increasing the feed current to 5mA.
T117's simulation shows ringing but it's still relatively damped, if it oscillates there is no small reason for it. I have found that using low feed current helps with stability for 2T current sources, but also reduces performance. I can get T117's schematic to oscillate by increasing the feed current to 5mA.
Last edited:
Bob, I suspect it is possible to write up a few recommendations that will help people avoid trouble with the 2T current source, and I was hoping that other readers of this thread would contribute a few of them for your consideration.
I believe it might help a whole lot to point out that a resistor in series with the base of the control transistor, works in conjunction with the control transistor's Collector-Base capacitance, giving classical "pole splitting". The final location of the LF pole depends on the external resistor, the Ccb capacitance, the stage gain (Ceffective = C*(1+gain) i.e. the classical Miller Effect), and other factors. You could suggest, as I occasionally have done, that people select Rbase such that (Rbase * Ccb) > (20 * Tauf) where of course Tauf = 1/(2*pi*fT). It's overkill but it's also very safe.
I believe it might help a whole lot to point out that a resistor in series with the base of the control transistor, works in conjunction with the control transistor's Collector-Base capacitance, giving classical "pole splitting". The final location of the LF pole depends on the external resistor, the Ccb capacitance, the stage gain (Ceffective = C*(1+gain) i.e. the classical Miller Effect), and other factors. You could suggest, as I occasionally have done, that people select Rbase such that (Rbase * Ccb) > (20 * Tauf) where of course Tauf = 1/(2*pi*fT). It's overkill but it's also very safe.
Compensation Miller is a local NFB loop. Such local loops always worsen the parameters of the amplifiers. I was worried that such a loop will reduce the output impedance of the current source when a large capacitance C1 and/or the resistance R3.
monkey see 2T CCS, monkey do 2T CCS
Excuse my somewhat blunt question perhaps bottoming in lack of knowledge, but among other CCS circuits, what are the advantages of the 2T CCS recently discussed in this thread, and in particular in audio amplifiers, and even preamplifiers, where voltage supply headroom is plenty?
Has it become used in audio circuits where voltage supply headroom is of no problem mostly out of "monkey see, monkey do" reasons, or are there any historical reasons, did it appear first in ordinary circuits built on discrete devices or did it appear first in IC's?
Although, I can see its usefulness in IC's where among various design limits, the supply voltage can be even very limited.
Excuse my somewhat blunt question perhaps bottoming in lack of knowledge, but among other CCS circuits, what are the advantages of the 2T CCS recently discussed in this thread, and in particular in audio amplifiers, and even preamplifiers, where voltage supply headroom is plenty?
Has it become used in audio circuits where voltage supply headroom is of no problem mostly out of "monkey see, monkey do" reasons, or are there any historical reasons, did it appear first in ordinary circuits built on discrete devices or did it appear first in IC's?
Although, I can see its usefulness in IC's where among various design limits, the supply voltage can be even very limited.