Callins capacitor?

You seem to know what you're talking about. :)

Although its ramp and triangle outputs are now restored, the Schmitt trigger's still a work in progress. It's (manually transcribed from a PDF and as typo-free as possible) Design Analysis is shown below. The last paragraph in the analysis leads me to believe the Callins capacitor was plausibly used by design.

DESIGN ANALYSIS

The central feature of the 2720-2A VCO is the self-zeroing, summing voltage to current converter comprising of IC-1, Q1 and associated components. Unlike the more conventional inverting summer, the feed back voltage does not come directly from the output of the amplifier but rather from the emitter of the current source transistor Q1. In operation positive voltages applied to one or more of the summing resistors (R1 - R3) force the output of the amplifier to go to whatever negative voltage is necessary to make the same voltage appear at both the inverting and non-inverting inputs of the op amp. Feedback circuits are always a balancing act and the balanced point in any linear operational amplifier circuit is reached when the two inputs are at equal voltages. Since in this case the non-inverting input is grounded (0v.) the inverting input must also go to zero volts.

Since the base emitter junction of Q1 is inside the feedback loop of the summing amplifier, both the natural non-linearities and constant voltage drop of the junction are eliminated from the response of the current source. Both sets of resistors that serve as emitter resistances for the current source (Range trimmer R7 in series with R6 both paralleled by R4) terminate at ground and a virtual ground so that for zero control volts input the source must generate zero current.

The output of the current source charges capacitor C1 which in conjunction with Unijunction Transistor (UJT) Q2 forms a relaxation oscillator. As increasing voltages are applied to the control inputs, the current supplied by Q1 increases causing C1 to charge more rapidly thereby increasing the frequency of the oscillator.

The ramp waveform that appears across C1 is applied to the input of the darlington emitter follower consisting of Q4 and Q11. The high input impedance of this emitter follower is important in presenting negligible load to the timing capacitor C1. A second emitter follower Q4 in conjunction with zener diode D3 performs a level shift so that the ramp is transposed to slightly above ground potential while a third emitter follower (Q5) provides a low output impedance buffer to couple the signal to the rest of the waveforming circuitry.

The ramp waveform is used three ways. First, it is applied to the voltage divider string consisting of R11, R24, R23 and R22. Between R11 and R24 the string is capacitively coupled through C5 directly to the "RAMP" output jack J1 where it becomes available as a signal source.

Secondly, the ramp is applied to the Schmitt trigger composed of Q9 and Q10. A Schmitt trigger has a low output or a high output depending on whether the input voltage is above or below a pre-set design level. As the ramp input to the trigger begins to rise the output remains low until the voltage exceeds this level and then abruptly changes to the high state. The output of the trigger, then, is a rectangular pulse at exactly the frequency of the ramp input. By varying the amplitude of the ramp you regulate the duration of the pulse by changing the relative point at which the trigger changes state.

Finally, the ramp is applied to the input of the differential pair Q6 and Q7. In the differential configuration the voltage at the collector of Q7 is in phase with the input ramp and the voltage at the collector of Q6 is inverted. The diodes D1 and D2 "select the higher of the two collector voltages and apply it to the base of emitter follower Q8. During the lower half of the input ramp's excursion Q6's collector voltage is higher and that section of the ramp is presented in an inverted form to the base of Q8. There is a slight rounding at the bottom of the wave during the cross over between Q6 and Q7 and a slight pip at the top during the ramp "flyback" but neither of these imperfections are audibly noticeable.

The most voltage sensitive portions of the circuit are powered from the simple series voltage regulator consisting of zener reference diode D4 and pass transistor Q12. Less critical parts of the circuit are powered by the decoupling networks R33/C6 in the positive supply line and R35/C8 in the negative supply.

Danke,

Derate tants about 3:1 on a supply rail. That seems to be reliable.

If you have to derate, there goes the size advantage. If the tant is at the end of a long power supply lead, there could be a spike when the power is attached. See Williams for more info.

With the risks of tantalum, why not use poly instead?

"BULLSHIT". phil allison told me to let you know this.

Whoops, it turns out the (loose) 25 V Callins belongs to another module. Although the VCO's original 6 V Callins is long gone from my PCB, Inet images show it present on other people's VCOs.

It also turns out there's a typo with D4. Its correct value is 5.6 V. By inspection the emitter voltage of Q12 is then determined to be ~ -4.93 V, which jibes with the measured drop across C7 of 4.87 V.

Danke,

Someone elsewhere mentioned how late designer John Simonton tried to use as few parts as possible, presumably to keep the total price down. So my earlier Woz analogy may be more spot on than first realized. Woz's Breakout game design reportedly earned a bonus from Atari because it kept its chip count below 120. Supposedly way down, somewhere in the neighborhood of 50 chips, it's said. Legend has it although Atari paid the bonus (to Steve Jobs) Atari didn't understand how Woz's circuit worked.

Danke,

Well in this design he didn't optimise enough. I can't see the need for R14, but if needed for adjustment range R15 could be increased in value to 200R. And R19 and R20 are providing an offset voltage of 1.5mV but to a diff amp made from presumably unmatched and non-isothermal Q6,Q7 and whose other input is derived from a vanilla zener and follower. Even across minor room temperature changes the drift could be horrible. I am baffled by that 1.5mV offset, bet you can just tie the base to ground and save two resistors.

Also the polarity of C3 C4 C5 looks strange?

piglet

At some point there was a small ~ 1" x 4" scrap of paper included with my documentation. It was an addendum, which noted how D4 was now 5.6 V instead of 6.8 V. (A fact duly recorded on my paper schematic in ink, but missing from the PDF, which was uploaded then shared with the group.) IIRC it also said Q6, Q7 were a matched pair. Of course, the scrap of paper with irreplaceable information on it got misplaced. Nonetheless, if you seek to match Q6 and Q7 to make the pair as isothermal as possible, will this circuit do the job?

The D3 kludge is easy enough to grasp. But R19 and R20's eccentricity is a little too far out for me. Simonton's next generation VCO removes some unsavoriness:

Danke,

The rubber seals can be pretty robust, if you can spend more. One example:

But if it must be robust, why would you use an electrolytic capacitor in the first place? This one doesn't have any juice inside and is much cheaper than the MLPs:

In the long run, none. If there is no metal to glass seal, the juice will find its way out.

Best regards, Piotr

Thanks - very good info. The poly numbers are amazing. Why use tantalums, especially when JL says to derate them by a factor of three.

So Larkin is the authority in capacitor use? These devices are through hole and relatively huge! I think if you don't have special requirements on ESR or temperature stability, etc, most capacitor selection comes down to $/uF or ccm/uF.

There has to be a rather special requirement to justify the use of a through hole part.

On 12.12.20 22.14, Don wrote:

Hello Don,

I made a LTspice model of the thing:

Version 4 SHEET 1 3144 724 WIRE 848 -320 -864 -320 WIRE 1008 -320 848 -320 WIRE 1472 -320 1008 -320 WIRE 1728 -320 1472 -320 WIRE 1888 -320 1728 -320 WIRE 2064 -320 1888 -320 WIRE 2496 -320 2064 -320 WIRE 2624 -320 2496 -320 WIRE 2912 -320 2624 -320 WIRE 848 -288 848 -320 WIRE 1472 -272 1472 -320 WIRE 1728 -272 1728 -320 WIRE 1888 -240 1888 -320 WIRE 2912 -224 2912 -320 WIRE -272 -144 -352 -144 WIRE -48 -144 -192 -144 WIRE 16 -144 -48 -144 WIRE 192 -144 96 -144 WIRE 304 -144 192 -144 WIRE 432 -144 304 -144 WIRE 592 -144 432 -144 WIRE 1008 -144 1008 -320 WIRE 1472 -144 1472 -192 WIRE 1520 -144 1472 -144 WIRE 1600 -144 1584 -144 WIRE 1632 -144 1600 -144 WIRE 1728 -144 1728 -192 WIRE 1728 -144 1696 -144 WIRE 2064 -144 2064 -320 WIRE 2496 -144 2496 -320 WIRE 2624 -144 2624 -320 WIRE 304 -96 304 -144 WIRE 848 -96 848 -208 WIRE 944 -96 848 -96 WIRE 1600 -96 1600 -144 WIRE 2000 -96 1600 -96 WIRE 2912 -96 2912 -144 WIRE 3024 -96 2912 -96 WIRE 192 -64 192 -144 WIRE 1472 -64 1472 -144 WIRE 1728 -64 1728 -144 WIRE 2912 -64 2912 -96 WIRE -48 -32 -48 -144 WIRE 1008 -16 1008 -48 WIRE 1184 -16 1008 -16 WIRE 1344 -16 1264 -16 WIRE 1408 -16 1344 -16 WIRE 1888 -16 1888 -160 WIRE 1888 -16 1792 -16 WIRE -544 0 -672 0 WIRE -352 0 -352 -144 WIRE -352 0 -464 0 WIRE -272 0 -352 0 WIRE 2064 0 2064 -48 WIRE 2160 0 2064 0 WIRE -112 16 -208 16 WIRE 1008 16 1008 -16 WIRE 2624 16 2624 -64 WIRE 2672 16 2624 16 WIRE 2768 16 2752 16 WIRE -272 32 -320 32 WIRE 432 32 432 -144 WIRE 848 32 848 -96 WIRE -864 64 -864 -320 WIRE -672 64 -672 0 WIRE 1344 64 1344 -16 WIRE 1472 64 1472 32 WIRE 1728 64 1728 32 WIRE 1888 64 1888 -16 WIRE 2064 64 2064 0 WIRE -320 80 -320 32 WIRE -48 80 -48 64 WIRE 368 80 -48 80 WIRE 592 80 592 -144 WIRE 2624 80 2624 16 WIRE 2912 80 2912 16 WIRE 192 112 192 16 WIRE 528 128 432 128 WIRE 1008 128 1008 96 WIRE 1104 128 1008 128 WIRE 2496 128 2496 -64 WIRE 2560 128 2496 128 WIRE 2768 128 2768 16 WIRE 2848 128 2768 128 WIRE 1008 160 1008 128 WIRE -48 176 -48 80 WIRE 128 176 -48 176 WIRE 1472 176 1472 144 WIRE 1600 176 1472 176 WIRE 1728 176 1728 144 WIRE 1728 176 1600 176 WIRE -864 192 -864 144 WIRE -672 192 -672 144 WIRE -672 192 -864 192 WIRE -480 192 -672 192 WIRE 848 192 848 96 WIRE 2496 192 2496 128 WIRE 1344 208 1344 144 WIRE 1600 208 1600 176 WIRE 1888 208 1888 144 WIRE 2064 208 2064 144 WIRE 2624 208 2624 176 WIRE 2912 208 2912 176 WIRE 2912 208 2624 208 WIRE -480 224 -480 192 WIRE -864 240 -864 192 WIRE -672 240 -672 192 WIRE 592 240 592 176 WIRE 784 240 592 240 WIRE 2768 240 2768 128 WIRE 2912 240 2912 208 WIRE -48 256 -48 176 WIRE 1008 256 1008 240 WIRE 1248 256 1008 256 WIRE 1008 272 1008 256 WIRE 592 288 592 240 WIRE 1248 352 1248 256 WIRE 2496 352 2496 272 WIRE 2496 352 1248 352 WIRE 1008 400 1008 352 WIRE 2768 400 2768 320 WIRE 2912 400 2912 320 WIRE -672 432 -672 320 WIRE -48 432 -48 320 WIRE -48 432 -672 432 WIRE 192 432 192 224 WIRE 192 432 -48 432 WIRE 848 432 848 288 WIRE 848 432 192 432 WIRE 1600 432 1600 288 WIRE 1600 432 848 432 WIRE -864 496 -864 320 WIRE 592 496 592 368 WIRE 592 496 -864 496 FLAG -320 80 0 FLAG 304 -96 0 FLAG -480 224 0 FLAG 1008 400 0 FLAG 1104 128 ramp IOPIN 1104 128 Out FLAG 2768 400 0 FLAG 2912 400 0 FLAG 3024 -96 pulse IOPIN 3024 -96 Out FLAG 1344 208 0 FLAG 1888 208 0 FLAG 2064 208 0 FLAG 2160 0 triangle IOPIN 2160 0 Out SYMBOL Opamps\\opamp -240 -48 R0 SYMATTR InstName U1 SYMBOL pnp -112 64 M180 SYMATTR InstName Q1 SYMATTR Value 2N3906 SYMBOL res -176 -160 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R2 SYMATTR Value 10k SYMBOL res -448 -16 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R3 SYMATTR Value 150k SYMBOL cap -64 256 R0 SYMATTR InstName C1

SYMBOL 2N2646 144 112 R0 SYMATTR InstName U2 SYMBOL npn 368 32 R0 SYMATTR InstName Q2 SYMATTR Value 2N3904 SYMBOL npn 528 80 R0 SYMATTR InstName Q3 SYMATTR Value 2N3904 SYMBOL pnp 784 288 M180 SYMATTR InstName Q4 SYMATTR Value 2N3906 SYMBOL res 576 272 R0 SYMATTR InstName R4 SYMATTR Value 6.8k SYMBOL res 832 -304 R0 SYMATTR InstName R5 SYMATTR Value 10k SYMBOL zener 864 96 R180 WINDOW 0 24 64 Left 2 WINDOW 3 24 0 Left 2 SYMATTR InstName D1 SYMATTR Value BZX84C6V2L SYMBOL voltage -672 48 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 10 SYMBOL voltage -864 48 R0 SYMATTR InstName V2 SYMATTR Value 9 SYMBOL voltage -672 224 R0 SYMATTR InstName V3 SYMATTR Value 6.2 SYMBOL voltage -864 224 R0 SYMATTR InstName V4 SYMATTR Value 9 SYMBOL res 112 -160 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R1 SYMATTR Value 4.5k SYMBOL res 176 -80 R0 SYMATTR InstName R6 SYMATTR Value 1k SYMBOL npn 944 -144 R0 SYMATTR InstName Q5 SYMATTR Value 2N3904 SYMBOL res 992 0 R0 SYMATTR InstName R7 SYMATTR Value 1k SYMBOL res 992 144 R0 SYMATTR InstName R8 SYMATTR Value 500 SYMBOL res 992 256 R0 SYMATTR InstName R9 SYMATTR Value 1.68k SYMBOL res 2480 176 R0 SYMATTR InstName R10 SYMATTR Value 4.7k SYMBOL res 2480 -160 R0 SYMATTR InstName R11 SYMATTR Value 470k SYMBOL npn 2560 80 R0 SYMATTR InstName Q6 SYMATTR Value 2N3904 SYMBOL npn 2848 80 R0 SYMATTR InstName Q7 SYMATTR Value 2N3904 SYMBOL res 2608 -160 R0 SYMATTR InstName R12 SYMATTR Value 10k SYMBOL res 2896 -80 R0 SYMATTR InstName R13 SYMATTR Value 8.2k SYMBOL res 2896 -240 R0 SYMATTR InstName R14 SYMATTR Value 1k SYMBOL res 2752 224 R0 SYMATTR InstName R15 SYMATTR Value 47k SYMBOL res 2896 224 R0 SYMATTR InstName R16 SYMATTR Value 100 SYMBOL res 2768 0 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R17 SYMATTR Value 100k SYMBOL res 1280 -32 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R18 SYMATTR Value 6.8k SYMBOL npn 1408 -64 R0 SYMATTR InstName Q8 SYMATTR Value 2N3904 SYMBOL npn 1792 -64 M0 SYMATTR InstName Q9 SYMATTR Value 2N3904 SYMBOL res 1456 48 R0 SYMATTR InstName R19 SYMATTR Value 150 SYMBOL res 1712 48 R0 SYMATTR InstName R20 SYMATTR Value 30 SYMBOL res 1584 192 R0 SYMATTR InstName R21 SYMATTR Value 5.6k SYMBOL res 1328 48 R0 SYMATTR InstName R22 SYMATTR Value 330 SYMBOL res 1872 48 R0 SYMATTR InstName R23 SYMATTR Value 680 SYMBOL res 1872 -256 R0 SYMATTR InstName R24 SYMATTR Value 3.9Meg SYMBOL res 1712 -288 R0 SYMATTR InstName R25 SYMATTR Value 6.8k SYMBOL res 1456 -288 R0 SYMATTR InstName R26 SYMATTR Value 6.8k SYMBOL diode 1696 -160 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName D2 SYMATTR Value 1N4148 SYMBOL diode 1520 -128 R270 WINDOW 0 32 32 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D3 SYMATTR Value 1N4148 SYMBOL res 2048 48 R0 SYMATTR InstName R27 SYMATTR Value 4.7k SYMBOL npn 2000 -144 R0 SYMATTR InstName Q10 SYMATTR Value 2N3904 TEXT -288 112 Left 2 !.lib opamp.sub TEXT -360 288 Left 2 !.tran 30m TEXT 1816 408 Left 2 ;2720-2A model, some components guessed.

There was no mode for an UJT, so I made an 2N2646.

Version 4 SymbolType BLOCK LINE Normal 1 76 16 80 LINE Normal 1 85 1 76 LINE Normal 16 80 1 85 LINE Normal 48 88 48 112 LINE Normal 16 88 48 88 LINE Normal 16 96 16 16 LINE Normal 48 24 48 0 LINE Normal 16 24 48 24 LINE Normal 1 80 -16 64 WINDOW 0 93 35 Bottom 2 WINDOW 3 110 85 Top 2 SYMATTR Value 2N2646 SYMATTR Prefix X SYMATTR ModelFile 2N2646.SUB PIN -16 64 BOTTOM 8 PINATTR SpiceOrder 1 PIN 48 112 RIGHT 8 PINATTR SpiceOrder 2 PIN 48 0 RIGHT 8 PINATTR SpiceOrder 3

Model from the Net, probably by the late Jim Thompson:

.SUBCKT 2N2646 1 2 3 DE 1 4 EMITTER VE 4 5 DC 0 HVE 6 0 VE 1K RVE 0 6 1MEG BBB 5 7 I=0.00028*V(5,7)+0.00575*V(5,7)*V(6) CBB 5 7 35P

Fantastic! Thank you for all of your hard work. It will take me a few days to digest it all. Meanwhile, my pertinent notes on UJT models are available online, at the link shown below. Among other things, it contains a 2013 usenet thread where the late Jim Thompson compares the UJT model you used with a couple of others.

Danke,