1
\$\begingroup\$

Let's say I have a 74LS161 based counter, and I want an analog ramp generator which will linearly increase until the ripple-carry-output goes back down (so that the ramp goes up the entire time the counter counts, and at 0 it starts over. Trying to avoid yet another chip and I need all the counter bits to count on the up-slope. And I don't want any other ICs for this, just standard BC548 or so NPN or perhaps BC558 PNP, few of resistors and a capacitor.

I thought starting up by something like this:

schematic

simulate this circuit – Schematic created using CircuitLab

If I do this I see the voltage on the RCO pin break down when it goes high, so I supposed I could add a 220Ω resistor in series to the base of Q2.

I haven't tuned the capacitor yet, at this point I have a 1 nF and it definitely charges too fast. That's why I plan 1 μF next.

But then I still have the problem that even if I can make that work, I activate that discharge transistor Q2 too early since the RCO goes up as soon as the output is all high. I have an inverter stage still from a 74LS04 hex inverter on the same board, so I suppose I could run the RCO through the inverter and then build a positive edge detector with another capacitor which would just pulse long enough to open Q2, and then shut off and stay shut off until the next rising edge of the !RCO signal.

schematic

simulate this circuit

Will it fly with some proper selection of R and C values? Maybe there is even a way to avoid the inverter stage?

UPDATE: I already got a better answer with the PNP transistor, but I have almost made it work with the NPN and just showing the scope tracing for this. I used 4.7 μF and the right kind of slow enough counter stage, to get this:

enter image description here

but indeed, with the PNP and 47 nF I get this on the same counter stage: so that is much better!

enter image description here

What I find difficult is to get it to work on the faster end of the counter. Am already down to C1 = 33 pF and still it's too slow for the 15 kHz that I need it for. I guess I need to reduce R3 to something a lot smaller. Is there a lower limit though? I tried going down 10-fold to R3 = 1 kΩ and while I can get a couple of more counter stages down with that I also have a lot less amplitude.

I need one at 15 kHz and the other at 60 Hz (haha, guess what I need those for :) )

\$\endgroup\$
4
\$\begingroup\$

Replace Q1 with a PNP transistor (collector down) and you'll have a constant(-ish) current source.

schematic

simulate this circuit – Schematic created using CircuitLab

enter image description here

It would be better to use something like a 74HC123 to generate the reset pulse if you can't generate it digitally, but something like what you show would work.

Of course you can make the constant current source temperature compensated and do other things to make it better, but this is the simplest. For best results have a volt or two across my R2, however that limits how high the Vramp can go.

\$\endgroup\$
10
  • \$\begingroup\$ Wow! Indeed, that worked! I updated the original question with more findings about that. Specifically it's a bit tricky to get it to work at around 15 kHz where I need it. \$\endgroup\$ Aug 24 '20 at 1:45
  • \$\begingroup\$ Just adjust the current and capacitance appropriately, this kind of circuit can work at very high frequencies. \$\endgroup\$ Aug 24 '20 at 1:48
  • \$\begingroup\$ I went as low as 22 pF and then I could juts pull out the capacitor altogether since the breadboard had enough residual capacitance. \$\endgroup\$ Aug 24 '20 at 3:03
  • \$\begingroup\$ @GuntherSchadow I'm not sure I understand. I just tested (LTspice -- no, I didn't go build it) a 20 kHz ramp with a simple small signal BJT into 10 nF and I just don't see any problems with it. It's linear enough that you wouldn't see a difference on a scope (which doesn't mean anything to write home about) and quite fast. It has no porch of any kind, though. (Because I didn't want to add one.) So it just resets immediately (about an amp of collector current) and ramps up again. This one is ramping from 0 V to 5 V -- because that's what I see Spehro using. \$\endgroup\$
    – jonk
    Aug 24 '20 at 3:33
  • \$\begingroup\$ @GuntherSchadow And yes, breadboards are notorious for stray capacitance. Is it possible that this may be part of your problem?? (I didn't populate my schematic with 5 pF capacitors from every node to every other node in the circuit.) \$\endgroup\$
    – jonk
    Aug 24 '20 at 3:36
0
\$\begingroup\$

Here is my nearly perfect simple solution. I just need to build it tomorrow.

Spehro Pefhany's PNP correction of course set the stage. However the discharging circuit would be a practical problem driving it from the little poor man's edge detector. But wit the Darlington pair I am doing better than with some exotic N-MOSFET.

enter image description here

R4 will of course be a 10k potentiometer to adjust.

Replacing R1 with the two diodes is supposed to make it more stable.

UPDATE: There is a difference between theory and practice. In practice the little pulse generated by the 100pF / 1k high pass filter was not enough to set the Darlington pair off. I could run it through a buffer (like an amplifier to bring it to logic level) or two inverters, and then it would work. But that's wasting two 3 inverter stages total. I tried to use a negative edge detection and one inverter to make a full upward pulse, but the negative spike doesn't go low enough to trigger the inverter.

For now I am back at just using the RCO as a reset pulse. And when I do that I don't need a Darlington pair either. I am wasting one pixel doing that. Not ideal but fair enough for the time being.

Final versions I built on breadboard with fairly stable picture.

enter image description here

I tried these diodes D1 D2 on the V-ramp, hoping it would make it stable, but it didn't help, so I went back to the resistor R1 at 2.11 kOhm, I had a 1% tolerance left over from my ladder DAC project. And after I put it all back together it's a pretty stable picture now.

enter image description here

Some fuzz in the vertical ramp, while the horizontal is nice.

enter image description here

And the picture is stable, perhaps the vertical fuzz could be improved. Will see the next few days when it will be swapped out with the digital DAC based staircase ramps.

enter image description here

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.