To Rallming: There are plenty of ways to do this; just pointing to the one, which utilizes a single Op Amp.
You can Google on the search phrase “Square Wave Generator” and refer to the top link, which will bring you to the article, showing the base circuit and providing you with the conceptual description and computational formulas.
You will need to modify the circuit, replacing the feedback Resistor R (see the original circuit) with two resistor and two diodes. First resistor-diode pair, connected in series, will provide charge current path when Op Amp output goes positive (so, the polarity of the diode connection should allow the positive current flow to the capacitor). Second resistor-pair diode, also connected in series, will provide the discharge path when Op Amp goes negative (correspondingly, the diode polarity should be reversed).
Next, to reduce the voltage level to plus-minus 0.1V you can add the resistive divider to the Op Amp’s output.
In terms of components selection, I would recommend to ask Donald first: there is plenty of high quality Op Amps available from the Maxim Integrated Products with low supply voltage, high input impedance and Rail-to-Rail output, so he probably will give you a good advice. Other potential sources of high-quality Op Amps could be: National Semiconductor, Texas Instruments, Linear Technology, Analog Devices, etc. As a practical design example of component selection for highly-stable application you can refer to my old articles:
Alexander Bell. “Improved frequency modulator uses "negatron”, EDN, 7/11/2002.
It uses Op Amp LTC1124, which has very high input impedance and low capacitance. Please note, that it operates in a single-supply mode, which is not required in your application; just the other way round, you should use the typical bipolar supply voltage topology, as in the base Square Wave Generator circuit referred above.
Hopefully it will help.
Regards, Alexander Bell

Hello, all: I want to build a circuit to generate a postive +0.1V pulse output of 0.4 second, and then produce a negative -0.1V pulse output of 0.2 second. WOuld anybody help me out with this? Thanks.

More online findings. Hello again, folks. As FYI: today I ran again Google™ on the same “magic” phrase as in my original posting, namely: “Flip-flop Push Button” and got an interesting result. Link #2 (Top2, skipping the commercial links) instead of “A Report to ENIAC…” mentioned by Donald (actually, “ENIAC…” was #3 on my result list), it is actually pointing to thematically very close article titled “Single push-button ON/OFF power control, Alfredo H. Saab and Tina Alikahi — Maxim Integrated Products Inc.” published in online edition of Portable Design ™ PennWell.

The Circuit described in the articles contains 3 ICs and performs de-bouncing /power-up initial state control functions. I guess, Donald and other folks from engineering community could find the article interesting.

Best regards/wishes, A. Bell

PS. I am not including the external links to the articles for the same reason I explained before: I am not sure that EDN policy allows external links in comments. Sorry for the possible inconvenience.

Hello Donald, Louis and other folks:

1). My post, which was deleted, contained several exact links (URL). I do not want to include them again as it could possibly be against the EDN’s policy regarding external links, but using the same Google ™ you can easily find one of them: just type the search query string as “Flip-flop cpemma” and the refined result set should have the link of interest on the top. In this online resource you could find cost-effective, simple and time-tested solutions addressing de-bouncing and initial state issues.

2). As Donald has mentioned, components used in the referenced online articles (e.g. 4013) are a bit outdated indeed, but, IMHO, the WHOLE TOPIC is a bit outdated, too. In a real world scenario mostly likely all these issues will be resolved programmatically by using uC or digital ASIC; as an example, in a local “Dollar store” you could probably find some cheap calculators having On/Off button perfectly de-bounced programmatically with (optional) off-timer, priced around 1$ (this is the price of the whole device, which can do MUCH more than just Push-Push/Flip-Flop).

Thanks for the kind attention. Best Regards, A. Bell

Alex, It''s a shame that your comments were removed. Though I had nothing to do with the removal, please accept my apologies. I appreciated your feedback and performed the Google search myself. Google returned 243,000 results, however many of them didn''t apply to the topic presented here. I sincerely doubt that "A Report on the Eniac" (#2 result) is relevant.

I looked at some of the relevant results and there are many possible solutions. Some of them use fewer components (i.e. more cost-effective), but don''t debounce the switch or guarantee an output state at power-up. Some use the same principal I did (i.e. flip-flop + switch debouncer), but don''t have a mechanism to turn off the output after a given amount of time. The solution presented here features a debounced switch, guaranteed output-state at power-up, and an automatic turn-off after a programmable amount of time.

Is this the only solution to the problem? Of course not! I believe one returned page from your proposed Google search features a variety of nice designs by a Bill Bowden. Does a better solution exist that does all of the above? Probably… Is there a solution that is more cost effective and still does all of the above? I haven''t found one, but I''m sure it''s out there somewhere.

Unfortunately, I haven''t heard anyone reply with a specific link to a page that features another design that accomplishes the same function as circuit presented here and is more cost effective. The only feedback that I have received is:

a) One link to a circuit that uses outdated components and doesn''t perform the same function as the circuit presented here. (originally posted by Louis and then went missing)

and

b) Your post (titled "Just Google It".) which essentially states that the answer is ''on the internet somewhere''. If you have found something else, please post the URL. I''m very interested and would like to look at it.

Thank you again for your comments.

Don

P.S. Regarding the original comments made by Louis. His comment pointed to a link that detailed an alternative circuit which solved only a portion of the given problem (not the entire thing). Furthermore, he made reference to a simple program crashing in my circuit and suggested that the watchdog timer wasn’t necessary (the circuit doesn’t use a microcontroller or microprocessor). He also used a rather inflammatory remark when referring to the idea presented (something about using tactical nukes to kill a mosquito and insinuating that I was being ‘brilliant’ about doing so). What he did not do, however, was provide a better or more cost-effective solution to the given problem. I found his comments neither thoughtful, nor constructive, nor professional.