Soft Start Using a PFET

[Roff]

I played around with MOSFETs with no success. Here is a circuit that simulates really well. Your output voltages will be about ±6.1V with 70mA loads, which I am betting will be just fine.
You can change the slew rate by changing R1 (make it a pot) and/or C1. You might also have to change C2, which is there to prevent oscillations.
If you change the slew rate of +out, the slew rate of -out will follow it.
Having slow turn-on power supplies does not gaurantee that you will have no pops on your audio, but I guess it's worth a shot.
The voltage-controlled switches are for simulation only.

I have included the .asc file, which is the simulation file for LTspiceIV.

 

[dedjazzgadgetz]

Wow... speechless.

Just WOW.

That's INCREDIBLE, Ron ! I never expected this !

I thought worst case, I'd get a "go do your homework, young man!" and at best, some electronic advice, which would have been perfectly fine... but THIS ?!

Wow. Thank you so much, Ron ! You're a very generous man !

A surprise-job-offer came up (a coupla days) that will keep me from breadboarding this right now (& I do have to find some LM6132 or -34), but I'll get back to you with my results. I'll still try to tinker with the MOSFET idea, 'though (I like its elegance and simplicity), IDK...

Meanwhile, Ron: Thanks again ! I owe you ! Lots !

Sincerely,
Joël

EDIT: hey, I just got IT ! R on, R off... ROFLMAO, great !

 

[Roff]

EDIT: hey, I just got IT ! R on, R off... ROFLMAO, great !

If you figured that out on your own, I'm impressed!

 

[Roff]

Duplicate post.

 

[dedjazzgadgetz]

YA-HOO ! Guess what ? Nailed it !

Hey, Ron (and everyone) !

I finally had some time-off the big job contract and got time to fool around some on the ol' breadboard; I went at it with the P & NFETs first (just had to give it a go, 'though you mentionned not having any success with them... I have to admit, I'm a stubborn kinda guy: no pain-no game, right ?)

So: out a selected output of a MC14017 decade counter (everybody powered Vdd +7 volts/Vss -7 volts) into a simple 10k/10uF RC network (yielding a ramp-up & a little slower ramp-down when going "low"; no biggie) into the NFET (IRF531) through a 22k; the same "positive-going" ramp is fed into a CD4069 inverter (providing the "opposite" ramp) into the PFET (IRF9531) through a 22k as well. Both gates have a 100k to their respective polarity to keep everybody from "flapping in the wind", as it were.

Simple enough, I sez to myself; well...

Works-a-TREAT !

I selected the highest-gain patch I could find on the VG-99 guitar processor (worst-case scenario, I thought) to make darn sure any little click!/pop! would be real ovious... not a peep ! Soft-start AND soft-stop, with the added bonus: I get the full 14 volts swing, with lotsa power to spare, thanks to the FETs. Just have to triplicate for a 1-out-of-3 GK guitar selector... yeah !

So, I would like to thank Mike the OP, Ron, Chris, crutschow; most everyone who pitched-in and got the ball rolling on my little project (there I go sounding like an oscar winner !). I know y'all do it for free, out of sheer generosity (Ron: that schematic sure is outta-sight; sorry I didn't go that way)... rest assured YOU ALL make a BIG difference. I can only hope to make other contributions in the future, sharing with other enthusiasts of the electron brotherhood !

Peace,
Joël

 

[dedjazzgadgetz]

Well: ALMOST got it...

After the initial hoo-ray!/eureka! and all that jazz, I noticed the IRF531 was a little more fussy/noisy on the ramp-up than the PFET ... I tweaked (in bold) the values to get the best s/n ratio (which I have to admit is done using a worst-case/nightmare scenario: a mega-supa-ultra-distorto-bomb-of-a-guitar-"patch") And I'm within 90mA of the full 14 volts. Not bad for an amateur... inn'it ?

"So: out a selected output of a MC14017 decade counter (everybody powered Vdd +7 volts/Vss -7 volts) into a simple 56k/10uF RC network (yielding a ramp-up & a little slower ramp-down when going "low"; no biggie : I added a reversed bias diode parallel to the 56k to accelerate discharge) into the NFET (IRF531) through a 300k to the gate, being pulled up using a 150k -still tweaking those; the same "positive-going" ramp is fed into a CD4069 inverter (providing the "opposite" ramp) into the PFET (IRF9531) through a 100k to the gate, which is pulled up using a 1 Meg. The PFET is very clean."

-Dam ! Ah' knew dem FETs woulda gave me a hard time; grrrrr.

Bye for now,
Joël

 

[dedjazzgadgetz]

I'm sooo CLOSE !...

O.K.: I'm declaring myself officially stumped. (well, not so much anymore: see edit)

The IRF9531 PFET in my GK guitar-selection circuit is behaving flawlessly: just an ever-so-slight "swoosh" (on the +7 volts ramp-up&down) that is barely discernable amidst the baground noise... perfectly acceptable. On to the IRF531 NFET now.

I can't for-the-life-of-me make it behave.

I read through the datasheet. Made sure the GDS pinout was fine. I'm aware of the S-to-D parasitic diode. It does the -7 volts ramp down&up sort-of O.K., only not so noiselessly. Fine so far.

BUT, stange thing is: when I unplug an unselected GK guitar, the meter shows the negative line (which is "up" to zero) rocketing to -7 volts ... when I plug-in again: a "cRUnCh!" in the audio of the currently selected guitar (logical, of course) and the -7 shoots back to 0 ! The positive side exhibits no such behaviour !

Checked, double-checked, trouble-shooted, verified, analysed, subbed other IRF531...

What are the NFET and I doing wrong ? I'm this close to success... help, pretty-please ?

Joël

EDIT: Now, both FETs have a 1 M from control-ramp up/down (+7 volts Vdd<-> -7 volts Vss) to gate as well as a parallel 1 M & 1U (np) to their respective polarity (values sounding good ? Any tips, advice for best s/n ratio ?); The "sshwoomph!" is still noticeable, 'though; wish it was perfect, of course, but... oh well). Oh... and the strange behaviour is -GONE

"I wish you'd stop trying to pin your crimes on me." -IRF531

 

[dedjazzgadgetz]

Any TO-92 p-channel equivalent to n-ch BS170 ?

Hi y'all !

I'm having great success with my simple little +/-7 volts "slow power-up/down" circuit but realized the n & p-ch power MOSFETS I picked for experimenting (cheap and plentyful IRF531 & IRF9531) are a tad overkill and bulky...

I subbed a small BS170 for the IRF531 (at 500mA/1.2ohm Rds-on, gets the job done just fine; and since the whole thing is housed in a small (1590B) Hammond box, the TO-92 package shoor' is appreciated !) So I'd be looking into a "common-place" p-channel equivalent; any ideas ?

See ya !
Joël

P.S.: could also be an altogether different n & p-channel pair... ? very low Rds-on... ? maybe a dual-matched package... ? IDK ! I'm really just waking up to the world of MOSFETs, so would greatly appreciate any suggestions. Thanks a lot !