I modified the "Road Rage" charge pump circuit, thoughts?

[Boner]

I need to make a "bullet proof" charge pump circuit that I could drop in to any circuit I throw at it. I know that's dangerous thinking since power is a finicky old lady. I also need these to be compatible with boss style supplies and daisy chain power supplies. I know I'm asking for the moon, but I want something that 90% of the time, the power circuit is over overkill, but for that other 10% the circuit is robust enough that there won't be any problems. This is what I've come up with so far. I'm still waiting on some LT1054s to arrive but I'm 99% sure this will work. Wanted to of course run it by all of you to take a look or possibly take something of value from it.


This is a mishmash of "road rage" and things I've learned from here, btw R.G. is a national treasure.....

I imagine the whole idea originates from the klon centaur, electrosmash has an amazing writeup on the pedal if you haven't seen it already.
https://www.electrosmash.com/klon-centaur-analysis



Its LT1054 based for higher current demands than with a MAX1044 used previously. I included (I believe) RGs trick on switching/reverse polarity protection) power switching circuit with a 2n2907. From my understanding the 2n2907 is capable of providing 300mA and the LT1054 can't take more than 200mA so that should not be a problem.

1n4742A is for overvoltage protection (12v input max).

This entire thing needs a "true ground" and thats the grounding pin on the DC jack. Everything eventually hooks back up here via star ground.

Effect is powered on with the TS/TRS switching trick on the output jack. Inserting a jack shorts Vss on the ring to "true ground" on the sleeve thereby turning on the BJT and turning on the LT1054. Each Vss from any and all audio circuits (transistor boost stages or OP amps for example) connect to Vss on the audio jack ring forming a "star audio ground" to the ring. A single wire on the sleeve hooks up to true ground, so this "star audio ground" on the ring is shorted to true ground.

Each Vss pertaining to power (power conditioning caps for example) is connected to LT1054s Vss pin first (pin 3), and then THIS pin is connected to true ground via wire or thick 20mil trace.

Any and all LEDs in the audio circuit connect to true ground via star grounding.

A grounding spring (my personal perference) connects the metal enclosure to true ground (DC ground) vi thick 20mil trace or wire.

You may be wondering why I'm spending so much time on grounding. After reading so many of RGs posts on "sewer ground, I've learned proper grounding SEEMS simple enough, but it REALLY requires great attention to detail, maybe the most attention out of everything pertaining to guitar pedals. Google "site:diystompboxes.com "sewer ground" and just read everything you find.


The 15pF cap connecting pin 6 (Vref) to pin 7 (OSC ref) raises the internal clock outside the audio range to help mitigate "whining". All of this is from the datasheet (They recommend 5pF-25pF I think). In my research here, this whining is apparently a huge problem that is difficult to fix and trace/wire placement of ground with signal traces is critical. Cross talk is a problem with audio ins being next to audio outs and I'm guessing form some weird cross-talk loop. So keeping signal ins as far from possible from signal outs. Remember all this nonsense is just loops. Signal in forms a loop, signal out forms a loop and power is another loop.


1n5817 diodes are low forward voltage types thereby lowering the voltage drop at 18V output. Using 1n4004s or whathaveyou will cause 18v output to be lower, than with low forward voltage diodes.


LT1054s datasheet calls for tantalum caps for C29/C28. I normally go with a fat ass aluminum electrolytic cap and a smaller alumina electrolytic cap, "large" MLCC and a "small" MLCC all in parallel for power filtering like everyone else, however with the datasheet calling for low ESR tantalum caps, I'm not sure if you would then leave out or include the MLCC caps.

C8 in parallel with C33 and C6 in parallel with C34 are cost saving measures. Normally (per datasheet) these would just be large valued, low ESR tarantula caps, however they are balls expensive. So you can go with a fat electrolytic (C8 and C6) each in parallel with a low ESR, lower valued tantalum to save some moolah. All this is from the datasheet btw.

All the remaining 100uF aluminum electrolytic caps in parallel with 100nF MLCCs are decoupling caps for each power point, 18v, 4.5v and -9v. No need for these on 9v since the input decoupling already decouple 9v point.

[antonis]

Can't figure out the need for Q9 series pass transistor..

[Boner]

Can't figure out the need for Q9 series pass transistor..

reverse polarity protection, current can only go one direction. I think also theres less voltage drop like if you used a diode.

I think R.G. came up with it.

[kraal]

Hello,

When I read your original post, I couldn't help but think about this () :

https://youtu.be/7YyBtMxZgQs?t=15

If it works as you expects (I'm not able to help), I wonder why you need, so much filtering, and such a drop-in solution (from a cost / effort / efficiency perspective, unless you decide to populate some sections only when you need the corresponding voltage).
But I'm probably missing something and it would be great if you could explain it :-)
Side note: the way you've annotated / oriented your diagram is quite difficult to read (at least for me).
Side question: is 20 mil really "thick"  ? (all my PCBs' tracks use 20 mil as default, with 40 mils for power tracks)

Best regards,

M.

[antonis]

reverse polarity protection, current can only go one direction. I think also theres less voltage drop like if you used a diode.
I think R.G. came up with it.

I also think R.G. will come and tell you that another BJT (n-p-n) is needed for shunting Q9 base current, in conjunction with a voltage set on the Base of the former BJT and a diode for preventing reverse B-E breakdown voltage..

[bluebunny]

btw R.G. is a national treasure.....

Hey!  You can't keep him all to yourself!   

[Rob Strand]

I see it as taking Fig 15 (the standard inverter) but with no regulation, merged with Fig 20 (the positive doubler),
https://www.ti.com/lit/ds/symlink/lt1054.pdf

FYI: As Antonis implied, you should follow RG's BJT version exactly.   The BJT versions will do weird stuff if you just follow the MOSFET pattern.    (You need to think of about what happens with the BE an BC junctions in with all fault cases.  Remember the BE junction won't block like an Si diode it will break down at 7 to 10V like a Zener.)

[niektb]

If you want a low-drop reverse polarity protection then you might also look for a P-Channel MOSFET (with DS body diode) with a low Rds(on) like this:


How this works is that (in normal operation) the body diode will raise the voltage on the source to (f.e.) 8.3V. The gate is 0V so there is a negative Vgs voltage which will open the 'channel' (and which will now have a series resistance of Rds(on)). I've used the BS250. This one has a Rds(on) of 10ohms (which is pretty high actually, you can easily get around 1ohm or even lower), for a low-current overdrive pedal (say 25mA) this will mean a voltage drop of 25*10^-3 * 10 = 0.25V...

In reverse polarity mode, Vgs will be positive and the fet will remain closed

You just need to make sure that the Vgs threshold is low enough so it can actually turn on. Also the maximum ratings for the Vgs should not be exceeded

[duck_arse]

.... low ESR tarantula caps, however they are balls expensive. ....

and, they will bite you if you don't pay attention to their polarity markings.

[Kevin Mitchell]

Wow that really is overkill 

I did this simple layout not too long ago so one could quickly make a drop-in board for any circuit that doesn't have the charge pump on-board already. I usually velcro them near the bypass switch or wherever it'll fit. I've had absolutely no problems with it.
Layouts here

-KM

[POTL]

If you want a low-drop reverse polarity protection then you might also look for a P-Channel MOSFET (with DS body diode) with a low Rds(on) like this:


How this works is that (in normal operation) the body diode will raise the voltage on the source to (f.e.) 8.3V. The gate is 0V so there is a negative Vgs voltage which will open the 'channel' (and which will now have a series resistance of Rds(on)). I've used the BS250. This one has a Rds(on) of 10ohms (which is pretty high actually, you can easily get around 1ohm or even lower), for a low-current overdrive pedal (say 25mA) this will mean a voltage drop of 25*10^-3 * 10 = 0.25V...

In reverse polarity mode, Vgs will be positive and the fet will remain closed

You just need to make sure that the Vgs threshold is low enough so it can actually turn on. Also the maximum ratings for the Vgs should not be exceeded

BSP250 = rdsOn 0,25 Ohm or less
http://www.farnell.com/datasheets/454177.pdf

[antonis]

Datasheets are good..

Real word items aren't so good..

[Boner]

I Like the idea of the mosfet switch! How would you best incorporate it with the traditional "TS/TRS on switch" trick" to power on/off the effect?

[ElectricDruid]

If you're running at 18V, what's the need for the 4.5V "midpoint" supply, when you've already got a 9V actual midpoint?

[Boner]

If you're running at 18V, what's the need for the 4.5V "midpoint" supply, when you've already got a 9V actual midpoint?

honestly none, put it there as a "If I need that, there it is" but will most likely not have it in any circuit

[bean]

If you're running at 18V, what's the need for the 4.5V "midpoint" supply, when you've already got a 9V actual midpoint?

There are a couple reasons I can think of:
1) You rarely get 18v once a load is on the charge pump circuit. It does work best if you use low ESR caps on the pump circuit but generally you are going to end up with something between 17-18v depending on the current demand.
2) Even in best case scenario (you have a steady 18v out of the pump circuit) even the "good" supplies aren't putting out exactly 9v. My One Spot is a steady 9.42v with or without any current draw on it.

Maybe there's a reason not to have virtual ground the same as the power supply input, as well? That's above my pay grade.

[Rob Strand]

I Like the idea of the mosfet switch! How would you best incorporate it with the traditional "TS/TRS on switch" trick" to power on/off the effect?

Here's some of RG's pages showing the basic ideas.

1) Reverse Protection
http://www.geofex.com/Article_Folders/mosswitch/mosswitch.htm

This shows the correct orientation of a P-channel MOSFET for reverse polarity protection.

However, the switching is normal ie. it doesn't use the MOSFET as a switch.


2) Quiet Switching for Stereo Jack
http://www.radanpro.com/Radan2400/Napajanje/New%20Page%202.htm

Diagram has bug: swap ring and tip

It is important to wire the battery -V and the power supply -V directly to the circuit.  Only the base/gate line goes to the ring.


I misunderstood the purpose of the transistor on your original ckt.

[Boner]

reverse polarity protection, current can only go one direction. I think also theres less voltage drop like if you used a diode.
I think R.G. came up with it.

I also think R.G. will come and tell you that another BJT (n-p-n) is needed for shunting Q9 base current, in conjunction with a voltage set on the Base of the former BJT and a diode for preventing reverse B-E breakdown voltage..

I Like the idea of the mosfet switch! How would you best incorporate it with the traditional "TS/TRS on switch" trick" to power on/off the effect?

2) Quiet Switching for Stereo Jack
http://www.radanpro.com/Radan2400/Napajanje/New%20Page%202.htm

So is a second BJT required in the Quiet Switching for Stereo Jack method?

[Rob Strand]

So is a second BJT required in the Quiet Switching for Stereo Jack method?

Good point.   I was thinking to merge the two but unfortunately that won't work!

In order for the reverse protection to work, the direction of P-channel MOSFET passes current via the internal diode under normal polarity conditions.  So that means the power won't turn off!

(With two MOSFETs you could probably add over voltage protection - perhaps a bit of feature creep.)

[PRR]

....even the "good" supplies aren't putting out exactly 9v. My One Spot is a steady 9.42v......

We say "9V" so we don't get a 1.5V or a 67V battery. No battery is ever what it says on the pack. A "12V" car battery floats at 12.6V, is typically cycled 13.8V down below 12.0V.

The One Spot probably aims for "typical battery voltage", NOT "marketing number".

The fresh "1.5V" zinc cell was close to 1.56V (around 1960). Reliable enough to calibrate meters. (They keep dinking the chemistry and it may not be 1.56V now.) So a six-pack would be 9.36V.

Your Spot showing 9.42V is certainly possible for somebody's "9V batt", is well within 1% of my old-old reference.