Ground input with 4069 in GEOfex relay driver?

[MrStab]

hi guys,

this is probably really simple, but i'm running low on 2n3906's so thought i'd ask!

i've used the Millennium Bypass to shunt signals to ground using an NPN before, but i was wondering how i'd use the 4069 in the GEOfex latching relay driver (http://www.geofex.com/FX_images/ltchrly.gif) to ground my distortion's clipping stage.

presumably i'd put the LED anode connection into the base of a PNP, with the signal going into the emitter (via. a 100uF cap?) and ground the collector? is it as straightforward as that, or are there other considerations? the Millennium shunt-to-ground circuit (http://www.geofex.com/article_folders/mill2extn/mil2shnt.gif) seems to have some kinda biasing with resistors going on, so i dunno how or if i should translate that to this situation.

i could just make a whole M. Bypass coming off the relay, but that'd be overkill and the fewer parts, the better.

cheers!

[MrStab]

this seems to work (they say i'm the next Michelangelo):



i just went with 1k resistors on a whim - there are articles explaining how to pick the right values, which i've found hard to understand, but i'll keep reading. are these values okay? if not ideal, can anyone gimme an overview in a stompbox context? i'm grounding the output from a TS-style clipping opamp, as it was the easiest layout-wise, if that helps. no worries if nobody's bothered, though - as i say it seems to work.

[PRR]

> 1k resistors on a whim
i'm grounding the output from a TS-style clipping opamp

The plan in your excellent sketch puts a 2K load on the thing being "shorted" when not shorting. An op-amp won't mind. Some other popular amp stages may quit working.

The 1K resistor can pass almost 9V/1K = 9mA. The transistor multiplies that to over 900mA. It is quite unlikely you need to clamp 900mA. In guitar-cord work you can usually pad-out to 5k, then only need 2mA of clamping. Divided by a low hFE, you need 0.02mA of resistor current. 9V/0.02mA = 450K. Try 220K.

The emitter-follower topology with two equal resistors will clamp to 1.2V. A grounded emitter plan will clamp to 0.05V.

[MrStab]

thanks a lot for the explanation, Paul - the point to be grounded is directly at the output of an opamp, but for safety's sake i'll try a NPN with 220k on the base and see how that works out. just to be clear - the NPN will ground the signal when the LED is off? the reason i thought to turn to a PNP as opposed to a NPN was to activate as the LED goes off. wasn't sure if you could/how to do that with a NPN only how to activate when it's on.

whilst i'm here, i have another transistor-related question about this circuit: i'm having trouble with a second build of the GEOfex latching relay circuit i mentioned - the first one works fine, and the only difference is that i used 2n3904s in the first but BC549c's in the new one (http://www.farnell.com/datasheets/727135.pdf). i'm using the TQ2-L2-5V (datasheet:http://www.rapidonline.com/pdf/60-1308.pdf). it may be something i haven't noticed debugging-wise, and one wiring error might've damaged the transistors or the CMOS, but i won't have time to check that till tomorrow night so i was wondering if there would be any reason the BC549c isn't capable of activating the coils? all i could find was a post by RG saying "2N3904 work fine. The transistor needs to be able to handle the relay coil current, and have a fairly high Vceo. 2N3904 is 100ma and 40Vceo", and the BC549c seems to have a VCEO of 30, and a max collector current of 100 (though it says VDC instead of A/mA on the datasheet, error?). is 30 too low? are there other factors at play? have i just screwed up somewhere else?

the LED switches on & off, but it haven't have time to get voltages yet. i'd think the IC is fine, though. i have 2N3904s and more CD4069's en-route, but i'll try replacing the BC549c's before the new stuff gets here, unless anyone can see a problem with the specs.

cheers!

[R.G.]

Did you take into account the different pinouts of the BC and 2N series?

[MrStab]

Did you take into account the different pinouts of the BC and 2N series?

i did  - although i always tell myself it's just the opposite way round, i always double-check just in case (2N3904/5088 are "the standard" for me, just found 549c's a cheaper sub from one supplier). thanks though, could well have overlooked it. relay and IC are oriented correctly too. i'll swap the transistors out, then, if there's no clear issue with the specs preventing operation.

cheers!

[MrStab]

swapped out the BC549c's and 4069 for fresh ones, and no improvement, so it's probably some error on my part. the relay latches happily if either side is grounded manually, and as i say the LED comes on, so the problem probably lies somewhere between the last few inverters and the relay.

[MrStab]

works! mis-read my own layout and corrected a trace cut, and also removed the 100R resistor i had between the relay ground and signal ground, which worked in my first build (which used a zener for regulation) but not with this (which uses a 78L05). the ground on the 78L05 was connected to "real" ground, ie. not in the 100R-isolated section where its output eventually wound up, so i wonder if that mismatch between grounds on the regulator has anything to do with it. it's the only real difference now. just trying to avoid pop, i suppose.