Unusual Bias Method

If you'll check into the LTspice forum, you'll find many discussion on how to run my circuit without it having to be in .asc format... including multiple times, on that forum, that I have demonstrated how easy it is to do.

Also, you probably need to read up on how to include a subcircuit on _your_ schematic.

You, too! ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

It's a simple circuit that could be a single .asc file. You're going to make it as hard as possible to run because you know it's broken.

70 uA * 0.02 duty cycle is 1.4 uA. Explain how that can bias up the Darlingtons. Do that now, instead of working at making your simulations inaccessible.

Nope, not broken... except for you, you're too incompetent to run the MODEL I provided, at your insistence.

Hand waving computations... must be the "new math"... you're wrong. So sad. Must be that cut and paste schooling you get >:-} ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Multiplication is "new math" ? Explain what's wrong with my math.

You had a good idea and botched the implementation. It doesn't work. I'd feel sorry for you, you lame old git, if you weren't so slimy.

On a sunny day (Thu, 28 Nov 2013 13:32:07 -0700) it happened Jim Thompson wrote in :

Just post a LT spice .asc file, then we can all see that it does not work. better stop spouting insults and blowing smoke Jim. You really get a bit too much.

Or else see a brain doctor.

Do you know how to use LT spice?

The amplifier will actually work well enough for a car-noise-blaster. The bias servo doesn't work.

To simulate my circuit in LTspice, open a text editor and type the following...

• Jim Thompson's 1977 Z Amplifier

* ** Analysis setup ** .tran 0 10m 0 100n .OPTIONS ITL1=1500 .OPTIONS ITL2=2000 .OPTIONS ITL4=1000 .OPTIONS STEPGMIN .OP X1 IN OUT VCC 0 My_1977_Z_Amp VCC VCC 0 13.3V VIN IN 0 SIN 0 4 1K 0 0 0 .INC "C:\InsertYourPathToCopyOf\My_1977_Z_Amp.sub"

• .END

Save as whatever name rings your chime, say...

"JimThompson'sMarvelousAmplifier.cir" >:-}

Then open LTspice. On the Tools/Control Panel/Save Defaults section check both Save Subcircuits... check-boxes.

Then Open "JimThompson'sMarvelousAmplifier.cir"

Then Run

View whatever node voltage or device current you like.

Irrespective of Larkin's stone throwing, it doesn't fail for several reasons... one specifically because it was 1977. Can anyone guess what that was?

Interestingly it takes LTspice _much_longer_ to run this circuit than it does PSpice, particularly the bias point calculation is butt slow.

Note that you _do_not_ need to draw a schematic in LTspice (or any other Spice, for that matter) to simulate someone else's circuit. Many of my clients only have LTspice, so I just pass them a PDF schematic and a netlist, and they can verify my work just fine. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

I'm really puzzled why you keep claiming that. Don't you realize other folk have checked out the design. It works. You know it does (maybe ?>:-}, you're just being your usual asshole self.

PROVE it doesn't work, or STFU.

BTW: You're overlooking a very important circuit behavior. PROVING, yes PROVING your ignorance of circuit analysis >:-}

(And I certainly know how to use LTspice better than you or mentally-defective Jan.) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Oooops! If you missed getting the subcircuit file, it's at...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

On a sunny day (Fri, 29 Nov 2013 08:30:20 -0800) it happened John Larkin wrote in :

I have not studied the bias servo, but the thing should RF oscillate, they always do without Boucherot filter. I am not under NDA, but do not want to say too much about it. That whole thing can be entered in LT spice in half an hour, maybe if you are old and need help like Jim 1 hour.

'Partial simulations' are no good, where is the limit? Will he take every inductor and capacitor and show in a partial simulation it actually IS an inductor and capacitor?

Only problem I see for him is if there are no spice models for the opamp and transistor, then he will have to do some web searching. Maybe a local kid can help him.

On a sunny day (Fri, 29 Nov 2013 09:48:17 -0700) it happened Jim Thompson wrote in :

Why do you always call anybody who disagrees with you defective? What sort of argument is that? If you had spend the same time drawing that LT spice diagram as you did painting insults on usenet, we now would all know it does not work, and even you would have noticed. :-)

On a sunny day (Fri, 29 Nov 2013 09:40:31 -0700) it happened Jim Thompson wrote in :

And where is My_1977_Z_Amp.sub ?

On a sunny day (Fri, 29 Nov 2013 09:51:05 -0700) it happened Jim Thompson wrote in :

Vout versus Vin looks like a half wave rectifier to me :-)

I have, several times.

Make it out of 5% resistors, and the Q3 current drops to near zero. Try it.

Let Q2 self-heat, ditto. Ditto.

Your 1 KHz example doesn't work. Using your own 70 uA Q3 current and the duty cycle from your PDF, the average Q3 current is only 1.4 uA, which makes 140 mV across C1. There's not enough bias available to turn on the darlingtons to the target current.

You refuse to address that last point. The numbers are simple and unambiguous. If that simple calculation is wrong, show us why.

The bias servo doesn't work now and it didn't work in 1975 [1]. You are so proud of it that you refuse to look at the math.

As an amp, Q5 and Q6 and the base current into R7 are enough to make it sound OK in a car. The bias servo does work intermittently, and THD goes up for a while after a long, loud stretch. A simple Vbe multiplier would work a lot better than the flakey mess of the servo.

Lots of bad engineering practise!

[1] You could argue that it works for voice and music that has frequent low-level segments that allow the servo to work. That's often but not always true. It would work better if you got lucky in the resistor tolerances such as to crank up the Q3 current. Or maybe, in 1975, you tweaked it to work. R4 sure looks like a tweak on top of lucky tolerances elsewhere.

But really, if anybody else posted anything this flakey, you'd trash them.

On Sun, 24 Nov 2013 08:53:30 -0700, Jim Thompson wrote:

This is enough to do the DC analysis.

You can frob the resistors R14/R1/R10/R5 by 5% and pretty much turn off Q3.

You can ctrl/rightclick Q2 and change its temperature for self-heating, 30C or so, and ditto.

The nominal Q3 current is only 60 uA, which is not enough to make the servo work in your 1 KHz example.

There's only 13 mV across R3, which sets the Q3 current. With discrete transistors and resistor tolerances, that could swing around wildly. It's flakey.

Version 4 SHEET 1 1752 680 WIRE 144 -256 -96 -256 WIRE 352 -256 144 -256 WIRE 544 -256 352 -256 WIRE 784 -256 544 -256 WIRE 1040 -256 784 -256 WIRE 1248 -256 1120 -256 WIRE 1568 -256 1248 -256 WIRE 784 -208 784 -256 WIRE 1568 -144 1568 -256 WIRE -96 -128 -96 -256 WIRE 784 -96 784 -128 WIRE 144 -80 144 -256 WIRE 544 -64 544 -256 WIRE 1248 -16 1248 -256 WIRE 1408 -16 1248 -16 WIRE 784 48 784 -32 WIRE 912 48 784 48 WIRE 1408 48 1408 -16 WIRE -96 80 -96 -48 WIRE 352 96 352 -256 WIRE 544 96 544 16 WIRE 912 96 912 48 WIRE 1568 96 1568 -64 WIRE 1568 96 1472 96 WIRE 144 144 144 0 WIRE 144 144 32 144 WIRE 288 144 144 144 WIRE 704 144 608 144 WIRE 784 144 784 48 WIRE 784 144 704 144 WIRE 848 144 784 144 WIRE 144 176 144 144 WIRE 1248 192 1248 -16 WIRE 32 224 32 144 WIRE 80 224 32 224 WIRE 544 224 544 192 WIRE 704 224 704 144 WIRE 912 240 912 192 WIRE 1408 240 1408 144 WIRE 1408 240 1312 240 WIRE 1408 320 1408 240 WIRE 352 336 352 192 WIRE 544 336 544 304 WIRE 544 336 352 336 WIRE 144 352 144 272 WIRE 352 352 352 336 WIRE 704 368 704 304 WIRE 912 368 912 320 WIRE 1248 368 1248 288 WIRE 1408 464 1408 400 WIRE 144 480 144 432 WIRE 352 480 352 432 FLAG 144 480 0 FLAG 352 480 0 FLAG -96 80 0 FLAG 704 368 0 FLAG 912 368 0 FLAG 1248 368 0 FLAG 1408 464 0 SYMBOL res 128 -96 R0 WINDOW 0 61 33 Left 2 WINDOW 3 64 71 Left 2 SYMATTR InstName R14 SYMATTR Value 620 SYMBOL res 128 336 R0 WINDOW 0 -61 30 Left 2 WINDOW 3 -59 66 Left 2 SYMATTR InstName R1 SYMATTR Value 22 SYMBOL res 768 -224 R0 WINDOW 0 63 42 Left 2 WINDOW 3 58 79 Left 2 SYMATTR InstName R10 SYMATTR Value 3.3K SYMBOL res 528 208 R0 WINDOW 0 59 40 Left 2 WINDOW 3 53 77 Left 2 SYMATTR InstName R3 SYMATTR Value 220 SYMBOL npn 80 176 R0 WINDOW 0 102 32 Left 2 WINDOW 3 83 67 Left 2 SYMATTR InstName Q1 SYMATTR Value 2N3904 SYMBOL npn 288 96 R0 WINDOW 0 97 25 Left 2 SYMATTR InstName Q2 SYMATTR Value 2N3904 SYMATTR SpiceLine temp 27 SYMBOL res 336 336 R0 WINDOW 0 -65 29 Left 2 WINDOW 3 -65 66 Left 2 SYMATTR InstName R2 SYMATTR Value 22 SYMBOL voltage -96 -144 R0 WINDOW 0 62 42 Left 2 WINDOW 3 62 78 Left 2 SYMATTR InstName V1 SYMATTR Value 15 SYMBOL npn 608 96 M0 WINDOW 0 96 25 Left 2 WINDOW 3 68 68 Left 2 SYMATTR InstName Q3 SYMATTR Value 2N3904 SYMBOL res 528 -80 R0 WINDOW 0 70 45 Left 2 WINDOW 3 76 77 Left 2 SYMATTR InstName R9 SYMATTR Value 1 SYMBOL res 688 208 R0 WINDOW 0 59 38 Left 2 WINDOW 3 54 75 Left 2 SYMATTR InstName R4 SYMATTR Value 10K SYMBOL npn 848 96 R0 WINDOW 0 103 13 Left 2 WINDOW 3 88 51 Left 2 SYMATTR InstName Q4 SYMATTR Value 2N3904 SYMBOL res 896 224 R0 WINDOW 0 63 27 Left 2 WINDOW 3 60 65 Left 2 SYMATTR InstName R5 SYMATTR Value 100 SYMBOL diode 768 -96 R0 WINDOW 0 66 33 Left 2 WINDOW 3 55 71 Left 2 SYMATTR InstName D1 SYMATTR Value 1N4148 SYMBOL npn 1312 192 M0 WINDOW 0 131 51 Left 2 WINDOW 3 99 94 Left 2 SYMATTR InstName Q5 SYMATTR Value 2N3904 SYMBOL npn 1472 48 M0 WINDOW 0 -48 101 Left 2 WINDOW 3 -77 133 Left 2 SYMATTR InstName Q6 SYMATTR Value 2N3904 SYMBOL res 1136 -272 R90 WINDOW 0 82 51 VBottom 2 WINDOW 3 99 49 VTop 2 SYMATTR InstName R8 SYMATTR Value 2K SYMBOL res 1552 -160 R0 WINDOW 0 -78 28 Left 2 WINDOW 3 -93 66 Left 2 SYMATTR InstName R7 SYMATTR Value 100K SYMBOL res 1392 304 R0 WINDOW 0 63 42 Left 2 WINDOW 3 58 79 Left 2 SYMATTR InstName R6 SYMATTR Value 6.8K TEXT -96 296 Left 2 !.op

I guess you missed the 2 second time constant ?? Ever hear of replenishment?

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Don't you think it's time you showed your work? Rather than hand-waving?

It take a 30°C rise to break it... even the 3904 will rise only 20°, and the 1977 version used TO-5 metal can devices (IIRC 2N2219)... the phenolic devices of that era had proved themselves as flaky in an automotive temperature environment... gross leaky after some temperature exposure.

As for resistor skew, I skewed the R2/R5 ratio by +/- 18% and you can hardly notice an effect.

Run the LT version I posted. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

The bias begins to decay as soon as your 1 KHz signal appears. After the high-level signal level drops, it takes a couple of seconds for the bias to ramp back up. Well, how long depends on the Q3 current, which is very poorly controlled. In a car, you won't notice.

But a Vbe multiplier would work better for much less hassle.

5% tolerance resistors kill the servo loop entirely. Try it.

I don't see why you went to the trouble of reflecting the measured idle current down to ground, and then pumping it back up the the bias setting transistors. Everything that you'd need is riding at the Q5/Q6 level already. The Q3 thing, with millivolts across R3, is broken.

You wouldn't let anyone else get away with something this bad. You'd rip them apart, and not in a nice way.

Oh, now you are changing the transistor type to mitigate the temperature problem. It's still a big error with a TO5.

(Plastic transistors were working fine in 1968. Dip DTL/TTL ICs weren't.)

R2/R5 isn't the issue. The problem is the R14/R1 ratio versus the R10/R5 ratio. Both ratios are close to 30:1, boogered by D1 and by the

10K. (What's the 10K for?) The tiny difference between the ratios sets the voltage, 13 mV typ, across R3. 5% tolerances on those resistors will break it.

Try it.

I don't know why you keep saying that I can't do analog electronics. This circuit, your circuit, has serious problems. You are committed to defending a 1975 design, so you refuse to look at it honestly.

Bad engineering practise, to put emotions before math.