FACT: The go-to devices for +/-9 volts from a 9 volt supply and +18 volts from a 9 volt supply in stompboxes are MAX1044, TC1044, LT1054 etc...........
I was just re-reading some old threads I'd bookmarked and this caught my attention:
Quote from: maarten on July 30, 2009, 05:43:02 PM
Hello Rick, for making a negative voltage using a 555 timer, have a look at:
http://www.circuitsonline.net/circuits/view/139
Should be good for up to 30 mA, which should be enough for negative biasing, I suppose. 1N4148 should work, according to the text: will give you a slightly higher voltage (-11,8 or so, I guess). You could add two voltage multipliers to this to get to to -36Volts (or even closer to -33 when using 1N4148's). Just observe polarity of diodes and capacitors: should be directed opposite to your positive multiplier.
Maarten
Here's the voltage doubler link: http://www.circuitsonline.net/schakelingen/138/voedingen/spanningsverdubbelaar-met-ne555.html
.....and here's the negative voltage link: http://www.circuitsonline.net/schakelingen/139/voedingen/negatieve-spanning-met-ne555.html
The text is in Dutch, so you'll need to use Goggle's translation feature, but the schematics are in the international language of electronics.
Compared to a dedicated charge pump chip, the additional parts count is negligable, and the 555 is very cheap and possibly to most plentiful chip on the planet - and unlikely to go out of production any time soon.
I was actually re-reading that post to see if there is a way to get a negative voltage from a 555 already set up as an SMPS, I'll have to do some experimenting there.....
Maybe it's also possible to use additional voltage doubling stages to get higher voltages too?
So.........why are we using MAX1044, TC1044, LT1054 etc to do this job, and not NE555? I know there are issues with 555's crowbarring the supply - but this can be worked around pretty simply. Additionally, 555's are used in a few stompboxes already, eg, Crash Sync, Uglyface.
Using a 555 for a negative supply definitely works; I build one 3 or 4 years ago for an 40x2 display LCD that required a negative reference. Not much current was required, of course.
I don't see why a doubler wouldn't work...
Both the linked circuits run at around 20kHz, if whining occurs (like in an unboosted MAX, or an ICL7660) it should be possible to lower the value of the timing cap to get the 555 oscillating at a frequency higher than human hearing.
I'll get around to testing this out with my circuits this week. I tried using a 4049 that output +/- but the current draw was too much for it to handle and it worked very sloppily. I wonder if CMOS 555 would be better to use.
The 555 works fine as a driver for a doubler/inverter/pump. This has been a favorite dodge of the EE in trouble for another voltage at the last minute for decades. I once designed in a 555 to run a capacitive discharge auto ignition. You appreciate that this was before cars all came with electronic ignitions, right? :icon_lol:
The problem is that a stock 555 does not play well with audio. It produces current spikes on the power supply and ground leads when it switches that are quite difficult to keep out of audio circuits. This is why the CMOS 75555 is recommended every time someone tries to use a 555 circuit here and finds out that it ticks.
Getting it over 20kHz is a good start, but even then you can see aliasing back down to audio in some circuits.
The 555 is an amazing chip. It's been the electronic tinkerer's wonder for almost 40 years. But be prepared for those spikes.
I just successfully added a -12 volt output stage to the 555 based SMPS I'm using for my submini tube amp experiments. Actually, just a touch under -12 volts, more like -11.7 volts. Just two 10uF caps and two 1n4148. That's cool! Now I need to add some more stages to it to get to around -33 volts to negative grid bias my 5902. Thanks
maarten - it didn't sink in right away, but I'm on it now!
Quote from: R.G. on November 01, 2009, 09:12:57 AM
electronic tinkerer
For electronic tinkerer read monkey with a machine gun!
Quote from: frequencycentral on November 01, 2009, 09:39:12 AM
Now I need to add some more stages to it to get to around -33 volts to negative grid bias my 5902. Thanks maarten - it didn't sink in right away, but I'm on it
Haha! Cracked it!
OK, I just posted a schematic for a SMPS with a -33 volts charge pump section: http://www.diystompboxes.com/smfforum/index.php?topic=74088.msg661075#msg661075
What I want to do now is to develop a 555 charge pump with multiple doubling sections delivering around 80 volts. So far not so good - just fried a 555. :'( Any suggestions?
This??? (http://analogguru.an.ohost.de/001/schematics/Analogguru_SymmVoltMultiplier.gif) :icon_razz:
Quote from: puretube on November 01, 2009, 05:24:15 PM
This??? (http://analogguru.an.ohost.de/001/schematics/Analogguru_SymmVoltMultiplier.gif) :icon_razz:
Yeah - but how to add more +ve stages?
Quote from: frequencycentral on November 01, 2009, 05:47:11 PM
Yeah - but how to add more +ve stages?
Google "%^&*croft-Walton".
Hello Rick,
have a look again at the link Renegadrian provided - http://w1.859.telia.com/~u85920178/power/vconv_01.htm. I am curious to learn what a negative bias will do to the 5902, currently I am still enjoying experimenting with just cathode bias. This works fine for me, even though it is an SE amp, it still puts out more noise than my family cares for.
Let me know if you need some help with the dutch text - I picked up some knowledge of this along the way...
Maarten
Quote from: R.G. on November 01, 2009, 06:15:39 PM
Quote from: frequencycentral on November 01, 2009, 05:47:11 PM
Yeah - but how to add more +ve stages?
Google "%^&*croft-Walton".
Gotta love that autoedit. The C0ckcroft-Walton multiplier is a nifty little circuit. It's an inexpensive way of deriving negative and high voltage supplies without taking up a lot of board real-estate (IC based designs still need those big caps!). More inverters used equals more drive capability too, which is nice to know when you're "HV" line starts sagging under load :).
You could also replace the inductor in your switcher with one side of a transformer. If the other side is center-tapped, you can get symmetric + and - high voltages out. Just be mindful of the current going through those windings (i.e. cut power if you smell burning).
If you're trying to supply any amount of current then that is pretty bad practice. Typically one goes through a whole process of laying out the specifications for their switcher design before you start and part of the process is calculating things like inductance (Volt-seconds per amp), number of turns, wire size, winding area of the core, required magnetic flux density and permability of the material and the temperature rise of the core. This is all for a disconinuous-mode flyback supply.
Now, a discontinuous boost supply is VERY similar design-wise to a discontinuous flyback and the conversion from one to another isn't all that complicated, but there are a lot of factors that one needs to take into account to make a truly safe and reliable supply. Throwing in a random transformer will probably work if the inductances and ratios work out correctly but you're going to have a pretty lossy supply that probably won't be able to deliver its full load current. Also, you want to avoid accidentally making a continuous-mode boost supply (current control) which has a right-hand plane zero that is impossible to completely compensate for. If one ends up accidentally in continous mode with their boost supply (inductor current never goes to zero) then you're going to have a very unstable power supply as there is excessive phase shift in the feedback loop (aka you have an oscillator instead of a regulated supply).
Quote from: frequencycentral on November 01, 2009, 02:31:59 PM
What I want to do now is to develop a 555 charge pump with multiple doubling sections delivering around 80 volts. So far not so good - just fried a 555. :'( Any suggestions?
Quote from: R.G. on November 01, 2009, 06:15:39 PM
Quote from: frequencycentral on November 01, 2009, 05:47:11 PM
Yeah - but how to add more +ve stages?
Google "%^&*croft-Walton".
I think I confused myself yesterday doing the negative charge pump first, all those wrong-way-round caps and diodes. I'll take a fresh look later this week. It should be easy.....
Quote from: frequencycentral on November 02, 2009, 01:28:39 PM
Quote from: frequencycentral on November 01, 2009, 02:31:59 PM
What I want to do now is to develop a 555 charge pump with multiple doubling sections delivering around 80 volts. So far not so good - just fried a 555. :'( Any suggestions?
Quote from: R.G. on November 01, 2009, 06:15:39 PM
Quote from: frequencycentral on November 01, 2009, 05:47:11 PM
Yeah - but how to add more +ve stages?
Google "%^&*croft-Walton".
I think I confused myself yesterday doing the negative charge pump first, all those wrong-way-round caps and diodes. I'll take a fresh look later this week. It should be easy.....
Should this work? I haven't got the opportunity to breadboard it for a few days...............
(http://i210.photobucket.com/albums/bb292/frequencycentral/555ChargePump80.jpg)
It's gonna be a lot of work for the 555 to charge up 60uF (all those 10uF caps in parallel) initially, but after that I don't see why it wouldn't.
If the 555 can't handle it, you might try adding some buffers or converting part of the cap section to series (at the cost of increased ripple).
Quote from: frequencycentral on November 02, 2009, 06:57:51 PM
Quote from: frequencycentral on November 02, 2009, 01:28:39 PM
Quote from: frequencycentral on November 01, 2009, 02:31:59 PM
What I want to do now is to develop a 555 charge pump with multiple doubling sections delivering around 80 volts. So far not so good - just fried a 555. :'( Any suggestions?
Quote from: R.G. on November 01, 2009, 06:15:39 PM
Quote from: frequencycentral on November 01, 2009, 05:47:11 PM
Yeah - but how to add more +ve stages?
Google "%^&*croft-Walton".
I think I confused myself yesterday doing the negative charge pump first, all those wrong-way-round caps and diodes. I'll take a fresh look later this week. It should be easy.....
Should this work? I haven't got the opportunity to breadboard it for a few days...............
(http://i210.photobucket.com/albums/bb292/frequencycentral/555ChargePump80.jpg)
Rick did you get this or the inverter working yet?
I just breadboarded a bipolar 555 timer and its very noisy and unregulated. In a voltage inverter topology, i was able to see -13V to -4V over a 0-70mA load. RG's comments about spikes were definitely true, there was lots of clock bleed through when i powered an opamp. I also measured a lot of spikes on the 9V and generated negative supplies. I'm gonna try the CMOS version tonight.
I was able to add a -33 volt charge pump output to a 555 based SMPS, which I detailed in this thread (halfway down page 6): http://www.diystompboxes.com/smfforum/index.php?topic=74088.100
I had a couple of stabs at using a 555 as a charge pump for positive voltage doubling with multiple doubling stages, but didn't get it completed before (as usual) I got sidetracked by another project.
Quote from: frequencycentral on December 03, 2009, 04:50:50 PM
I was able to add a -33 volt charge pump output to a 555 based SMPS, which I detailed in this thread (halfway down page 6): http://www.diystompboxes.com/smfforum/index.php?topic=74088.100
I had a couple of stabs at using a 555 as a charge pump for positive voltage doubling with multiple doubling stages, but didn't get it completed before (as usual) I got sidetracked by another project.
I was looking for that thread too, thanks!
Did you apply your -33V to anything (tube, transistor, any other amplifying device)? I'm just curious about the noise you might be getting from that unruly BJT-type 555. It was just awful in my findings, noise on the order of +/-100mV with a TI NE555P.
OK, I have succeeded in using a NE555 as a charge pump to generate 80.5 volts. Cool. :icon_cool:
(http://i210.photobucket.com/albums/bb292/frequencycentral/555ChargePump80-1.jpg)
The circuit is running from a 12 volt supply. Not tested under load yet, I'm hoping to run a Murder One from it, but I'm out of output transformers at the moment. I'm also assuming that the 100pF cap is putting the oscillator well above the audio range.
If all works out it will be pretty cool - I need 12 caps and 12 diodes to get 80 volts out of a MAX1044 - this circuit does it with less components, and the cost of an NE555 compared to the cost of a MAX1044 is a joke.
> So why don't we use NE555....
The newer chips are better promoted. And documented for this chore.
The 555 is a big crude beast. It can dump 200mA, much more than the new toys (far-far most than CD400x CMOS gates). OTOH it wastes about a Volt, which hurts when working from 5V or 3V supply. And sucks quite a few mA even when not doing anything; the new-toys are far thriftier on small loads.
There are times when the new-toys just won't pull the load. Your "80V" supply has 8X leverage. (Already we see you only get 10.1V action from your 12V supply: that's about half 555 losses.) Whatever current you suck from your 80V node is reflected back at least 8 times bigger (and I suspect 16 times bigger?). So even with the brutal 200mA available, you may only get a few mA at 80V before smokage.
And classic 555 is not protected against all overload.
Note also that small diodes like your 1N4148 may be rated 100mA. The 555 may be able to kill the first one before it quits. If they survive, the first diode has a few tenths more voltage drop (multipled by 8 to the final voltage) than a larger diode. However jellybean 1N4007 won't turn-off quickly at 20KHz rates, which spoils the action.
> Not tested under load yet
That may be more "fun". Wigging the short end of a stick with 1:8 leverage is tough, much tougher when something is on the other end.
Ratios like you propose are more often done with transformer than diode-string. One lossy part instead of eight. But as mentioned, magnetic design is alien to many of us. You can get started by using wall-power iron. 12VDC totem-pole into 6VAC winding, or 12VDC push-pull into 24VCT winding, 100Hz-500Hz, can induce near 120V squarish-waves on the 120VAC winding. Yes, 400Hz vibrator is audible, but common power iron gets lossy long before you get out of the audio band, and you just can't beat the availablity of small power iron. Many-many car radios ran 400Hz vibrators and the whine was not too bad.
Don't want to hijack, just need some clarification.
I also need -9v for an effect I'm building. I know about the MAX1044 but don't have any.
I do have A LOT of 555's.
Anybody have a schematic for a 555 based circuit that will give out +9v and -9v above 20khz that R.G. was talking about?
Thanks
Quote from: alparent on January 10, 2010, 05:49:44 PM
Anybody have a schematic for a 555 based circuit that will give out +9v and -9v above 20khz that R.G. was talking about?
http://www.circuitsonline.net/schakelingen/139/voedingen/negatieve-spanning-met-ne555.html
Not sure of the clock speed but you should be able to tweak it.
Thanks Rick.....................but the clock speed was the part I was wondering about? :icon_redface:
Quote from: alparent on January 10, 2010, 06:46:13 PM
Thanks Rick.....................but the clock speed was the part I was wondering about? :icon_redface:
Google translation says:
........delivers a square wave with a frequency of about 20 kHz. It should work with 1n4148's in place of the BAT85's
What do I do to change the frequency like R.G. suggests?
Do I just change the value of C2?
What frequency should I aim for?
I think C1 is the timing cap. Probably best to google for a quick 555 tutorial though.
Quote from: frequencycentral on January 02, 2010, 01:10:58 PM
OK, I have succeeded in using a NE555 as a charge pump to generate 80.5 volts. Cool. :icon_cool:
(http://i210.photobucket.com/albums/bb292/frequencycentral/555ChargePump80-1.jpg)
The circuit is running from a 12 volt supply. Not tested under load yet, I'm hoping to run a Murder One from it, but I'm out of output transformers at the moment. I'm also assuming that the 100pF cap is putting the oscillator well above the audio range.
If all works out it will be pretty cool - I need 12 caps and 12 diodes to get 80 volts out of a MAX1044 - this circuit does it with less components, and the cost of an NE555 compared to the cost of a MAX1044 is a joke.
Hmmm, some very weird results. I breadboarded and posted this a couple of weeks ago. I just got some output transformers delivered, so I thought I'd try the this charge pump under load. I plugged it in without load and found that I'm getting 52 volts. Whereas previously I was getting 80 volts. It's the exact same breadboarded circuit. :-\
So I breadboarded it again, using new components. Same result, 52 volts. I just can't work out why I was getting 80 volts 2 weeks ago but only 52 volts now. Anyway, under load (a Murder One amp) it drops to 45 volts. It powers the amp well, without any whining or noise. Good result in one way I guess, but I now need to add a few more doubling stages to get up to the desired voltage.
The total energy output of a charge pump comes from the capacitors, through the diodes. They are useful devices for generating high voltages (sometimes millions of volts) at low currents. Ask more current from them and they will sag a LOT. Remember that those 80 volts you measured at no load were coming from the output cap, and not the 555 (at least not directly).
The ways to increase the current are:
Increase the frequency
Increase the capacitance
Increase the current output from the oscillator
Damn! gtudoran tried the 555 charge pump circuit today and got 82.5 volts. I'm soooooooo confused. Need to understand this. I want my 80 volts back!
I could be dead wrong here, but I believe that raising that electros would help with more current?! say 220µF?!
Can I combine caps to get higher voltage ratings like resistors for higher power? I've got a bunch of 50v 47uF but not 100v, can I use 4 50v to rig a 100v section?
Quote from: panterafanatic on January 17, 2010, 09:04:07 PM
Can I combine caps to get higher voltage ratings like resistors for higher power? I've got a bunch of 50v 47uF but not 100v, can I use 4 50v to rig a 100v section?
No. You'd still be putting more than 50 volts across the caps. You can put smaller caps in parallel to make one larger cap. Like 2 100uf caps to make a total of 200uf. That works well.
Identical caps in series, the voltage ratings almost add. Two 50V in series will stand this 84V; altho maybe not 100V.
In this plan the voltage increases 10V-11V per stage. You can go several stages with 50V caps. Start building, measure each stage. When you get to 40V, make your next stage a series-pair.
Quote from: PRR on January 18, 2010, 04:50:21 PM
Identical caps in series, the voltage ratings almost add. Two 50V in series will stand this 84V; altho maybe not 100V.
In this plan the voltage increases 10V-11V per stage. You can go several stages with 50V caps. Start building, measure each stage. When you get to 40V, make your next stage a series-pair.
I don't mean to be argumentative, but I've seen way too many electrolytic caps expload to be comfortable with doing something like that. I'm by no means an expert on the subject, and I'm sure it can work, but isn't it easier to just get a 150v cap and be done with it?
> isn't it easier to just get a 150v cap and be done with it?
YES.
But....
When we get past 450V, we can't use inexpensive electrolytic caps. Paper/plastic caps are easy to get but costly in large values. Many-many-many large tube amps "stack" e-caps to run at 400V to 700V.
panterafanatic has no 100V caps but an abundance of 50V caps. Me, I'd rather get some 100+V caps "and be done with it", the cost is not large (though the shipping-charges are worse than the parts-cost). But yeah I'd bread-board with whatever. (I've made some awful ugly breadboards.)
Does anyone have a vero layout of a voltage inverter for -9V with the 555 chip?
Thanx
Quote from: kupervaser on April 03, 2010, 12:24:37 PM
Does anyone have a vero layout of a voltage inverter for -9V with the 555 chip?
Thanx
Better: you do your best at making a vero and I promise we'll all double check it for you.
The only way to learn is to try it yourself!
Hi,
i know it's an old post... but... i tried this NE555 80V, and i'm getting only 58.5V (without a load)... When i connect a load (and in this case it was a valvecaster), i get only 35V.....
Though, i have few differences in schematic:
- the capacitor from pin4 of 555 to ground is 47uF instead of 100uF;
- the last capacitor on HV out is 22uF instead of 10uF (always at 100V);
- diodes are 1N4007;
Do these differences justify my voltages (with and without load) ?
Thanks
I doubt the cap changes will have a large effect, although keeping the cap across the rails (pin 4) large helps.
The 555 contributes a lot to the sag. Another issue is running the 555 at too high a frequency for the filter caps. Normal electos are the best choice in this circuit. Try increasing the 100pF Cap, say to the 680p to 3.3nF region. You can tune the cap for least amount of drop under load.
Thanks very much, i'll try tonite or tomorrow.
Cheers
The output resistance is a function of both the oscillator frequency and the capacitor size. Frequency up or cap size up translates to higher output current. I would say the issue is most likely related to the smaller caps.
1n4007 might be too slow for this application. You'll have to look at the datasheets to see.
There was an article in either Practical Electronics or Everyday Electronics (perish the thought :icon_twisted:) in the 1980s I should think, called something like Super Tone Control, which contained a photo of Pete Townsend - this used a 555 to generate a negative voltage. I may have posted this at some point in time.
QuoteNormal electos are the best choice in this circuit.
I made a typo here, normal electro's are *NOT* the best choice for this circuit.