Power supply for built in effects with NPN and PNP

[SILVERGUN]

The plan is to use one of the toroid xfmrs that Hoffman stocks to build a power supply for 2 built in effects in my amp.
The power supply will be built into the amp chassis, and tapped off of my primary input.
The first effect will be a built in Rangemaster type boost that needs -9Vdc for PNP and will be directly wired to my input jack and then feed the signal directly to the first grid V1 (will be relay switched eventually)

The second effect will be a delay which will be in a D'lator style loop further down the signal chain and uses a NPN circuit

Is there any problem with this design? or implementation? .....grounding issues? ( I do plan on isolating the effect ground path from the amp signal ground)
Any suggestions?.....this will be a first attempt

Should I just use a NPN boost circuit for simplicity?

[sluckey]

-9Vdc for NPN

NPN transistors typically use a positive supply. The collector will be the most positive terminal on the transistor. And the emitter usually connects to ground. You can surely use a 'negative' supply but if you ground one side of that supply you'll have to reference the emitter to -9vdc and the collector will be referenced to ground. Kinda backward thinking to me, but as long as you get the correct polarity to each element you'll be OK.

Everything I just said about NPNs above is exactly opposite when talking about PNPs.

[SILVERGUN]

Everything I just said about NPNs above is exactly opposite when talking about PNPs.

Sorry about that sluckey....I'm going back and editing my post to reflect my "corrected" thinking 

So at least now someone can have an intelligent conversation,,,without trying to fix my backwards thinking first

[sluckey]

Not really backward thinking. Any time you run NPNs and PNPs from a single supply, you always have to think backwards about one or the other. 

[SILVERGUN]

Does that look corrected?

[kagliostro]

Can't you have the effect PS floating ?

K

[SILVERGUN]

Good question K....thanks

I think that was the plan here....

Specifically....the negative for the effect PS will not be connected to the chassis ground

Is that what you mean?

[kagliostro]

I mean nor the negative nor the positive

K

[SILVERGUN]

Yes...floating it is

[kagliostro]

As Mesa Boogie do here (also if for other purpose)

http://www.el34world.com/charts/Schematics/files/boogie/Boogie_Express_525-550.pdf

K

[PRR]

9VAC is not really ample for a 9V regulator.

9VAC*1.414= 12.7VDC
12.7VDC-1.2V= 11.5V (FWB drop)
11.5V-9V= 2.5V across regulator

Many 3-pin regs get pretty sloppy with less than 3V drop.

You would also allow for 10% wall-volt drop when your 2,000 Watt bassist gets thumping hard.

If the load is much less than the transformer rating, un-sag probably saves you.

12VAC would give more margin.

If they are common ground, a single 10V winding and two diodes gives +/-13.8V, which easily reg-sdown to 9V. This will require discovering Negative regulators.

[SILVERGUN]

Thank you PRR

If the load is much less than the transformer rating, un-sag probably saves you.

The spec sheet says the sec. no-load voltage is 11.6VAC @115VAC with a current handling of 89mA

Should that make me feel better (if I insist on using that xfmr), since I won't be loading it down much?

[rzenc]

This will require discovering Negative regulators.

Here you go SG:

http://www.fairchildsemi.com/ds/LM/LM7912.pdf

LM79xx family gives you a nice variety of regulated negative supplies.

Page 12 gives example on how to use and example of bipolar regulated supply employing LM78(+reg) and LM79(-reg).

Enjoy!

Hope this helps



Best Regards

R.

[kagliostro]

Of course PRR unfortunately for us (I've a lot of small not toroidal PCB 9v AC Transformers) at last 12v AC will be required for the supply

Only one thing about the values of the cap used in the datasheet for the regulator, someone must be tantalium cap, read instruction with attention

Ciao

K

[SILVERGUN]

Thanks R and K,
I appreciate you guys trying to help me....

I'm bound to learn something on this trip cause I'm an effects virgin,,,,,and I'm stubborn,,,and curious

If nothing else,,,I have officially discovered negative voltage regulators 
 
so,,I'm a late bloomer 

[DummyLoad]

http://www.ti.com/product/lm317

http://www.ti.com/product/lm337-n

http://www.ti.com/lit/ds/snvs778c/snvs778c.pdf    <--- see pg. 9. ditch the pots. calculate the V divider and there's your BIpolar power supply. :-) 

--pete

[eleventeen]

Few things in life are more frustrating than trying to make a 3 terminal regulator run off insufficient input volts. It completely sucks, usually forces nontrivial component (like transformer) changes elsewhere if you don't have the magic stuff on the input.

DO take note of the different pinout on negative regulators!

Positive ones are so logical: IGO Input Ground Output as viewed from the face of the device, pins down.
Negative regulators are GIO Ground Input Output viewed from face, pins down.

[SILVERGUN]

Thanks DL and 11,,,
I will certainly heed your advice.....

Still not sure which circuits I'm using ,,,,but it will all have to start with the power supply, so I'm ironing that out now,,,and all of this help is priceless

 

[eleventeen]

You can give those 3-term regs a considerable amount of input volts if you are only pulling tiny current out of them eg; 4-5-6 signal transistors or some op-amps. Check your data sheets, I believe 35 volts is the absolute maximum. Once you start pulling real current through them, that's when you have to pay a little more attention to excess input voltage, because feeding them too much implies the excess is dissipated/shunted to ground by the regulator itself. That means heat. That can force you to heatsink the regs which is less than desirable. If you can avoid it. There are clip-on heatsinks which are easy enough to install but they don't do a heck of a lot. Easier overall to chop down some volts before hitting the device just by using resistor(s) What you would really like to avoid, IMHO, is the requirement to mechanically mount the devices to the chassis or other externally-applied heatsink. That forces you to get involved with insulating washers and heatsink goop because the two tabs are ground (for the pos reg) and INPUT (for the neg reg) and thus cannot be common to each other = cannot be bolted to the same chunk of metal, without insulating shoulder washers. They can handle considerable power just standing up in the air, vastly more than you should need for a few transistors, so if you are looking for +/- 12 volts out, if you can get the input on each to say 16-18 volts, you should be fine, with no H/S req'd. In my experience, they need a solid 3 volts over rated output to work right, and 4-5-6 is better.

Most applications, by the way, like to see a tantalum cap really, really close to the regulator. Close. Most folks solder them directly onto the device. I will assume you are going to have these regs grouped close together in a "power supply" area, then run your clean DC power to your transistors. Do *NOT* just figure you can toss in an electrolytic cap or two over by your PNP/NPN stuff, especially if they are sort of distant from the supply zone, and leave the regs naked (uncapped =unfiltered) in whatever configuration you arrive at. That is a formula for oscillation or other noise generation, and assuming your PNP/NPN stuff is going to amplified a lot, that's exactly what you don't want.

[PRR]

> Positive ones are so logical...

They are both logical.... in the Factory. They are not pinned-out for User convenience.

The BODY of the internal chip is soldered to the LARGE tab which is also the center pin. This gets the heat out. With traditional processing, the body is also the most-negative point of the circuit (all other points are isolated with reverse-diodes cooked into the silicon).

That's for the 1970s regulators. The 1980 generation may be internally different, though tended to follow the conventions of their older brothers.

IAC: Print-Out and Pin-Up the data for EACH of your many-leg parts, and keep looking at the pictures.

[PRR]

> get involved with insulating washers and heatsink goop

The hip 21st century trick is to figure-out the order-code for "fully insulated TO220" package. They cut the mold a little bigger, to leave a schmear of epoxy on the bare-metal tab. Slightly lower absolute-Max dissipation, but a LOT less mess for the usual 2W-5W "needs a little sink" situations.

> like to see a tantalum cap really, really close to the regulator. Close.

Close, yes.

But Tantalum has become old-school. Modern Aluminum is MUCH better than it used to be.

Though even with 1970s Al, I never had trouble just using a generous lump of Al *close* to the reg.

I usually had 1000+uFd a few inches from the IN circuit, and that works. 47uFd Al within an inch of the OUT works.

BTW: this is even needed on a "dumb" application: truck fuel-gauge. Classic "6V" gauges used a thermo-cycle-regulator to turn 8V-5V into a steady 5V. 12V vehicles used the same trick and voltage. These thermo-regs rarely fail; so when one does (it goes weak, 4V-3.5V) it is hard to find a replacement. 7805 does the right thing. You drill the TO220 hole to fit the old-reg stud, and you MUST tack a few dozen uFd on the OUT or it makes crazy voltage.

[eleventeen]

Well, I am just thinking to be extra careful wrt noise on a discreet-transistor circuit for audio inside a tube amp chassis. 3 term regs are wonderful devices and I have used them many times on older style +12/-12/+5/-5 volt "LS" chip & uproc ckts with comparators and such, but some of those can be amazingly tolerant of power supply noise as long as the rail voltages are tightly regulated. Which they will be given adequate filtering. But I'd be wanting to be super careful feeding transistors amplifying low level analog audio in an otherwise noisy environment.

[SILVERGUN]

I will assume you are going to have these regs grouped close together in a "power supply" area, then run your clean DC power to your transistors.

Yes...that's the plan, to have a seperate small PS board, with a layout much like the schematic....inside the chassis I'm planning on keeping it over near the input AC

Thanks for taking the time to type guys....the more you guys throw out there, the more thought is generated on this side

[smackoj]

I saw you were thinking of putting a 'rangemaster' effect in the amp.....I have been searching the web for a couple weeks trying to get a good feel for that particular effect. I guess with some big time heavy hitters all claiming it was there for the British Invasion (Clapton, Brian May et al) it's a pretty nice place to start.  The best two places I found for this booster were the following. Hope this helps.

Fuzz Central and look at the 'Range Blaster'  free advice and scheme

Turretboard.org   he's got several 'improved' versions there and free layouts

good luck,  jacko     

[SILVERGUN]

Thanks jacko,
This is the one i'm pretty sold on,,,but i'll look at the stuff you posted
http://www.guitarpcb.com/apps/webstore/products/show/3436382

Check out the demo of this one......it's the "sweet guitar candy" part @ 1:07 in this demo that put me over the edge for this specific circuit,,,plus the fact that I can buy a pre-printed board,,,etc..

http://soundcloud.com/guitarpcb/purp-les-paul-vox-ac30-demo

[smackoj]

looks good. i have a PCB coming from Generalguitargadgets for the RangeM or Brian May treble boost (you can build either one on the same board).