hi,
a friend has asked me to make them a Rusty Box, which i understand is a rip-off honourable tribute to the Traynor TS-50B's preamp section. i've done the first draft of a vero layout (cool tip: forget to label all the resistors, then you'll HAVE to double-check it!).
just wondering: there are a few polarised caps in this schematic which seem backwards (for example, the 4u7 near "Line out"). i understand the voltage system on this isn't the same as your average stompbox, so apologies if i'm asking something dumb here.
(http://i.imgur.com/zVRVdWZl.png)
(full-size here: http://i.imgur.com/zVRVdWZ.png (http://i.imgur.com/zVRVdWZ.png))
i've a long way to go - need to read up more on charge pumps and the whole bipolar thing etc. - but does anyone have any tips on this circuit for a kinda-but-not-quite n00b? would be much appreciated. the Rusty Box seems to only have one input as opposed to the high & low - how would i go about that?
cheers!
I would stick with the original cap orientation, if you look at the 4u7 cap you mentioned there is an 8k2 resistor to -15V on its negative end. Just looking at the schematic, it would be reasonably simple to convert to using a single supply by connect almost all the preamp GNDs to a 1/2 V+. I'll draw up a schematic some time.
Here's a quick rough schematic:
(http://s20.postimg.org/i3z17bbrd/rusty_box.jpg) (http://postimg.org/image/i3z17bbrd/)
If anyone sees any errors, please let me know and I'll fix them. I wonder how this would sound with "normal" guitar?
That looks good to me, you've drawn it as though it's using the LO input if you want HI remove the resistor to ground after the input resistor.
Quote from: psychedelicfish on September 04, 2013, 02:30:28 AM
it would be reasonably simple to convert to using a single supply by connect almost all the preamp GNDs to a 1/2 V+. I'll draw up a schematic some time.
wow, thanks, that makes things a lot simpler! i'm a bit confused - some (well, most) -15V have become VBias, where as others (ie. the bottom of the diodes just before the opamp) have become ground. could you give me a rough explanation of the pattern there? sorry to be annoying! oh - and the FET has disappeared. was that an unnecessary component?
my plan is to go over my vero layout part by part and convert to your schematic. i'm guessing as it's all just a case of reference, a single supply would still have all the same tonal characteristics of the bipolar, so long as it's 15V, right?
cheers!
All points that were grounded and necessary for biasing have been connected to VB.
The FET was used as an electronic switch so I replaced it with a mechanical one to simplify things.
Quote from: psychedelicfish on September 04, 2013, 04:40:28 PM
All points that were grounded and necessary for biasing have been connected to VB.
so what you're saying is that
some -15V points become vbias, while others become ground? my minuscule brain would've just expected one similar rule of change across the board, which is why i think i got confused.
a mechanical switch would make for a lot less hassle, good call.
-15V has become ground, ground has become VB, and +15V has become +9V
It should sound the same, but as we are running it on 1/3 of the supply it will have 3x the distortion
Quote from: psychedelicfish on September 04, 2013, 06:38:38 PM
-15V has become ground, ground has become VB, and +15V has become +9V
ahhh okay, i get you. i could still set up a charge pump (or 15V supply) to get 15V, right? would that even be necessary? (i'm aware 15V isn't the same as to 15 to -15, but still). there's a user on another forum somewhere (can't remember the link right now) who went down that route - not sure if he used a bipolar supply or not. too late for that now, though - i've just redrawn the whole layout lol.
just a few questions, if that's cool: in your schematic, vbias connects to the middle PNP on the furthest-right, whereas that connection was originally on the bottom of that column. doesn't matter either way, right? also, at the very end, where there's a volume pot followed by a cap and resistor to ground, would there be any harm in putting that pot after the cap? and could i just use one 18k resistor from V+ with jumpers branching off of it to the collectors of the PNPs, or would i need to use one for each? my guess is one would be fine
thanks a lot for your help, it's really appreciated!
progress so far, if anyone wants to point out any glaring errors (and i'm sure there are a few)
Updated (but probably still wrong! despite the update, i just noticed 6n8, 1M and jumper just before NPN were a bit out of place, fixed). yellow indicates vbias, black is ground, blue is general jumpers.
(http://i.imgur.com/iZhwihDl.png)
Quote from: slacker on September 04, 2013, 05:49:30 AM
That looks good to me, you've drawn it as though it's using the LO input if you want HI remove the resistor to ground after the input resistor.
I added the resistor to ground after the input resistor to prevent switch popping. I probably should have put it before the other resistor so I don't get a voltage divider happening.
I'll fix up the schematic a bit and add some component values.
Here's an updated schematic:
(http://s20.postimg.org/gw7rt8yfd/rusty_box.jpg) (http://postimg.org/image/gw7rt8yfd/)
R6, R8, C8 and C14 can be connected to either VB or ground depending on what's most convenient for your layout.
thanks a lot Edward, i'll cross-reference that later. i didn't even notice the pulldown resistor in your first schem! there's fairly easy access to ground from there.
in the original schematic, there's the 10k, the 68n, then another 10k (all in series) - is that 2nd 10k before the 68n still meant to be there?
besides that, i'm up to the NPN in terms of double-checking my layout.
Quote from: MrStab on September 05, 2013, 05:21:07 AM
in the original schematic, there's the 10k, the 68n, then another 10k (all in series) - is that 2nd 10k before the 68n still meant to be there?
Good catch! You could probably get away with using one 20k (22k) instead of the two 10k resistors. Before or after the capacitor is your choice, whatever's best for your layout
i have room for two 10ks, guess i'll just go with that because i'm lazy. i only have 22ks, and that extra 2k could drastically affect the cork-like odour! lol
you seem to have reversed the very last (4u7) cap, but not the one right after the NPN, is that right? i'm used to using transistors as nubish switches , forgive the stupidity.
i noticed a mistake in my layout: vbias instead of V+ was going to the NPN collector, so i took care of that. i'm not looking forward to double-checking the jumper hell that is the gain stage (if that's what that is).
why'd you think they might've put 3x18k from V+ to each PNP pair instead of just one? could it have been a layout thing, or would there be some other advantage to that?
cheers
I was just looking at the original ciruit out of interest. what on earth is that stack of 6 transistors doing, top right? band-splitting tone shaping w/ individual buffers?
^ that makes the most sense, now i think about it - maybe that's why the circuit is so sought-after by bassists.
i know it's unrealistic to expect anyone to go through entire vero layouts, but reckon anyone could check the part with the transistors at least? no worries if it's too much hassle tho. still unsure about the orientation of those 4.7uFs...
(http://i.imgur.com/zf0I5rY.png)
WTH is a 33nG cap ? ;)
I asked about this circuit a while back and it was theorized that those 6 transistors were 3 overdriven differential amps and they are eq'd differently to distort differently. I wonder if they will work at 9 volts or will require new biasing to clip correctly. If I were building this I'd use a voltage multiplier to turn the 9V into 30V.
Here's my original thread:
http://www.diystompboxes.com/smfforum/index.php?topic=95021.0
Quote from: Hemmel on September 05, 2013, 01:20:18 PM
WTH is a 33nG cap ? ;)
lol! "ground" - there were just a few 33nF's in the region and i thought i'd differentiate!
Quote from: MrStab on September 05, 2013, 03:03:32 PM
lol! "ground" - there were just a few 33nF's in the region and i thought i'd differentiate!
Oh !!! ;) I thought it was an obscure Austrian way of qualifying caps. Like NanoGustavs or something ;D
Quote from: Hemmel on September 05, 2013, 03:06:26 PM
Oh !!! ;) I thought it was an obscure Austrian way of qualifying caps. Like NanoGustavs or something ;D
hell no! i'm Scottish, we use the McFarad system.
thanks for the link & info, Bill - it's helping shed more light on the subject. i did plan on using a voltage multiplier to get to 15V, but is the reasoning behind your 30V to get the same range as 15 to -15? thought for a second you might've been looking at the power amp section, but that makes sense. i've already ordered two L7660's, one was as a backup but maybe i could link em as per the datasheet (although i suck at numbers). is voltage tripling (almost) practical for battery usage, or would i need to provide a higher-than-9V DC adaptor? i'll have a search. anyone know what input the actual Rusty Box takes?
Quote from: MrStab on September 05, 2013, 03:17:33 PM
hell no! i'm Scottish, we use the McFarad system.
Oh wow.... I read "McFarad" and got hungry.... DAMN YOU RONALD !!!!
Quote from: MrStab on September 05, 2013, 03:17:33 PM
Quote from: Hemmel on September 05, 2013, 03:06:26 PM
Oh !!! ;) I thought it was an obscure Austrian way of qualifying caps. Like NanoGustavs or something ;D
hell no! i'm Scottish, we use the McFarad system.
thanks for the link & info, Bill - it's helping shed more light on the subject. i did plan on using a voltage multiplier to get to 15V, but is the reasoning behind your 30V to get the same range as 15 to -15? thought for a second you might've been looking at the power amp section, but that makes sense. i've already ordered two L7660's, one was as a backup but maybe i could link em as per the datasheet (although i suck at numbers). is voltage tripling (almost) practical for battery usage, or would i need to provide a higher-than-9V DC adaptor? i'll have a search. anyone know what input the actual Rusty Box takes?
My only worry is that opamps scale up and down with voltage quite easily. Transistor do not. Transistors in the differential amp configuration (and also the buffer) may also not like going from a bipolar supply to a single supply. If you aren't going to use the 30 volts and/or a bipolar supply then you might have to adjust the biasing of the differential amps to work off of a lower voltage single supply.
You may want to experiment with it some before committing to a final layout.
The Rusty Box uses a a 40V adapter.
http://www.geofex.com/circuits/+9_to_33.htm (http://www.geofex.com/circuits/+9_to_33.htm)
aha...
Quote from: Hemmel on September 05, 2013, 03:20:06 PM
Oh wow.... I read "McFarad" and got hungry.... DAMN YOU RONALD !!!!
lol their food probably lasts longer than most caps!
damn, the TL072 has a max VCC+ of 18V... in the interests of mimicking the circuit accurately, wouldn't not using a TL07* be an issue? i'm wondering if i'd be better off switching back to the original bipolar setup
Quote from: MrStab on September 05, 2013, 03:24:32 PM
lol their food probably lasts longer than most caps!
Yup, and I got the gut to prove it :(
Quote from: MrStab on September 05, 2013, 03:35:36 PM
damn, the TL072 has a max VCC+ of 18V... in the interests of mimicking the circuit accurately, wouldn't not using a TL07* be an issue? i'm wondering if i'd be better off switching back to the original bipolar setup
I could be wrong but I always thought the 072 could handle a voltage of 36V's. The datasheets just quote the +/- values because that's how they expected them to be used.
Quote from: Bill Mountain on September 05, 2013, 03:42:12 PM
I could be wrong but I always thought the 072 could handle a voltage of 36V's. The datasheets just quote the +/- values because that's how they expected them to be used.
in theory, all i'd need to do to test either configuration would be to just switch wires as per this, right?
Quote from: psychedelicfish on September 04, 2013, 06:38:38 PM
-15V has become ground, ground has become VB, and +15V has become +9V
from the Rusty Box manual (http://www.tronographic.com/manuals/Tronographic-RustyBoxUserManual.pdf (http://www.tronographic.com/manuals/Tronographic-RustyBoxUserManual.pdf)):
"AC Adapter supplied with your Rusty Box. If using an aftermarket AC Adapter, please be sure that it
delivers between 24VDC and 40VDC with a minimum current rating of 100mA".
granted, that includes an LED and a balanced conversion which presumably uses an opamp, but no way can this be run off a battery. i'm questioning whether or not a single charge pump would provide enough current from your average DC adaptor, or whether using 2 in parallel would, or how exactly i would sync their clocks
Quote from: MrStab on September 05, 2013, 04:36:27 PM
from the Rusty Box manual (http://www.tronographic.com/manuals/Tronographic-RustyBoxUserManual.pdf (http://www.tronographic.com/manuals/Tronographic-RustyBoxUserManual.pdf)):
"AC Adapter supplied with your Rusty Box. If using an aftermarket AC Adapter, please be sure that it
delivers between 24VDC and 40VDC with a minimum current rating of 100mA".
granted, that includes an LED and a balanced conversion which presumably uses an opamp, but no way can this be run off a battery. i'm questioning whether or not a single charge pump would provide enough current from your average DC adaptor, or whether using 2 in parallel would, or how exactly i would sync their clocks
I don't see anything too current heavy in this design. I suggest you experiment at alternate voltages and with using a voltage multiplier. You can always use a single voltage multiplier to get up to 30 volts then use something like this :http://sound.westhost.com/project43.htm to get a bipolar supply.
Edit: added an on-switch as recommended by RG on another thread as an extra precaution against noise from the IC
(http://i.imgur.com/y1QmLuK.png)
based on:
http://www.geofex.com/circuits/+9_to_33.htm (http://www.geofex.com/circuits/+9_to_33.htm)
the L7660SCPA seems to have the boost pin from the MAX1044 - would it be okay to just bridge pins 1 & 8 together like i've done? and to leave pins 4, 5 6 & 7 floating?
alternatively, maybe this could work - it's based on a post by FrequencyCentral: (http://www.diystompboxes.com/smfforum/index.php?topic=69886.msg561107#msg561107)
(http://i.imgur.com/JttFA2T.png)
not sure where i'd connect the ground for the multiplier board itself in that situation, though (where i put the question mark). also another TL072 would just be adding to the current draw.. either way, as this 2nd circuit is just an extension of the first, i could start with +30V & see how that goes
^^in the hours since posting this i've learned loads about the whole bipolar/relativity thing. lol
Quote from: MrStab on September 05, 2013, 03:35:36 PM
damn, the TL072 has a max VCC+ of 18V... in the interests of mimicking the circuit accurately, wouldn't not using a TL07* be an issue? i'm wondering if i'd be better off switching back to the original bipolar setup
According to the datasheet, 18V is relative to the midpoint between pin 8 and pin 4. I'd say you'll be fine running it on 30V
Quote from: psychedelicfish on September 06, 2013, 02:22:20 AM
relative
that's what blew open my understanding at one point last night. all i really need to change is my perspective and i can wire it up as i would any other stompbox. so i'm really glad you showed me how to reassign the vbias/ground, thanks!
just as a result of how this has evolved, i'm prolly gonna have both V+ and Vbias (or virtual ground, however you look at it) situated off the main board (ie. on the charge pump board) - unorthodox it might be, but it'll help with saving real-estate.
to make a unity gain buffer for the 30V output as well, is it safe to link both an opamp's non-inverted input and Vcc+, or would i need to isolate Vcc+ in some way? or is there even any point in doing so - i just figured it'd be good to have the V+ and Vbias both conditioned equally.
on that note, i have no 30 & 15V voltage regulators.
mrstab, you can voltage multiply as much as you like. this link has lots on doing with cmos inverters, very helpful.
http://www.gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf
as for the differential pairs, they will run happily from a single supply, if you bias the bases to somewhere near half supply. of course, if you do that with only 9V, you'll not get much headroom.
now I'll go and have a squint at your transistors' layout .....
....... pnpr3 and pnpl3: one is backwards, you pick. there are no cuts showing under the 2 18k emitter resistors and the 47nf to 2k2. otherwise, it looks sane ......
there's a lot of "shorthand" in my layouts - i was aware of that PNP's orientation and traces under horizontal components, but i forgot to put em on the schematic - thanks, though, it made me check to be sure.
i've populated the board (i'd fixed a few things in the layout before starting), but not tested it, although i have just gotten the charge pump working. i only had 3 schottkys so the rest i subbed with a 1n4001 and 2 germaniums (forgot which) i had lying around, which gives me ~25-27V. i wonder if im missing something or if i've just reached the limitation of the diodes. maybe i could get away with 25? i'll give your link a read, cheers duck!
*sigh* debug time. full supply voltage on opamp output, that can't be right. no obvious flow to that strip. at one point i got a faint distorted signal which was affected as expected by the pots, so that's a start i guess - but i can't figure out how to repeat it.
recent clues: i get a really faint signal, all affected accordingly by pots, when im using my middle or neck pickup position, but nothing on the bridge, and the signal totally disappears when i enable distortion on my amp (which you'd expect to make things louder).
i put the 4u7 cap after the NPN the opposite way from the schematic, i just can't see how signal can pass through to the PNPs if the negative is pointing away from em. cant think of anything else. no clue about opamp voltage weirdness either. hmm... wonder how much underpowering the circuit has to do with it.
Quote from: MrStab on September 07, 2013, 09:30:07 PM
i put the 4u7 cap after the NPN the opposite way from the schematic, i just can't see how signal can pass through to the PNPs if the negative is pointing away from em.
Which 4u7? The one I labelled C12? I'm pretty sure I put it the right way round, measure the DC across it just to be sure. The orientation of the cap has nothing to do with whether it passes signal or not, but electrolytic capacitors need to have a DC voltage with the correct polarity across them, or they won't last very long, or, if you reverse DC them, explode.
As for the full V+ on the op amps output, make sure you cut the trace between pin 1 and pin 8 of the op amp. You should also double check you've got the chip the right way round (
everyone makes that mistake at least once)
C12 is the one. i'll reverse it back if needs be, although that's probably not the issue if it'll still pass signal
i've tried looking for obvious leakage such as from pin 8, but no dice - iirc there's a few mV difference between either pin. no explosions of note - though it was very tempting for the sadist in me to try a backwards tantalum right at the end of the charge pump. lol
tried 3 TL072s (all right orientation) just to rule that out
i just ate and took a break, maybe i'll notice something if i look again
the plot thickens:
with opamp removed, the socket reads:
(V+ = 24.6)
1 = 2.2
2 = 4.1
3 = 9.9 (bearing in mind vbias is via. 270k resistor)
4 = 0
5,6,7 = 0
8 = 24.6
the negative side of the wrong-way cap is about 2V higher than the positive side, guess that's evidence enough i need to change it.
i've examined all trace cuts compared to a reversed image - they all check out. no apparent solder bridges. i'll switch that cap & see what happens.
also, V+ sometimes drops to about 15V, but this seems to happen after the charge pump & i haven't figured out the cause yet
Without the op amp pins 1 and 2 should have no DC on them, if you look at the schematic there is no connection to VB or V+ from a DC perspective. There is a rather large resistive path between pins 1 and 2, and because pin 2's voltage is higher than pin 1, i'm willing to bet pin 2 is where your problem (or at least part of it) lies. Check, double check and triple check for solder bridges etc. around pin 2. You should also check if the voltage on pin 2 changes with the gain control.
I'll also print off the schematic and breadboard it, partially to make sure the schematic is good, and partially to see how it sounds.
fixed it! (i think) - i tracked down the signal with my "biological audio probe" - no, i meant my finger - and isolated the problem to the bunch of 47nF caps, which i'd actually used 22nFs in parallel for so there must've been one or two sketchy solder joints. the voltage is loaded way down now, at like 11v, but opamp voltages seem to be in order, signal passes, and everything inc. boost works.
it was from touching the treble pot lugs and hearing the "buzz of hope" that i tracked it down. i do actually have an audio probe, but somewhere down the line it got caught up in a headphone - 1/4" plug conversion.
thanks a lot for helping me narrow it down. maybe you should go ahead with the breadboarding & see if things can be improved upon. i still wanna fix the voltage sag so it's more like the original TS50, but i have to spend the day with my lady tomorrow instead of making cool stuff :( lol
observation: with gain at full, and for the last little bit of its rotation (ie. not just shorted to one end), and the boost wires connected together, there's a high-pitched whine. i wonder if this is related to the charge pump IC, despite "S" models of the 7660 having the "power to pin 1" feature to increase oscillation frequency (i know it's also called "boost", but that would've confused things. lol)
annoyingly, as i'm using DPDTs (i'm broke), i won't be able to both add in an extra resistor AND use an LED in boost mode. damn.
on guitar at least, it sounds nice & chunky but with a kinda "glassy" feel. i don't know if these descriptions hold any meaning outside my head. maybe i could record it in a coupla days, but my sound card sucks and i can't find my USB interface. right, it's almost 6:30am - i can finally go to sleep without that nagging sense of failure!
Quote from: MrStab on September 08, 2013, 01:24:44 AM
observation: with gain at full, and for the last little bit of its rotation (ie. not just shorted to one end), and the boost wires connected together, there's a high-pitched whine.
It sounds to me like your input and output wires are too close together. The capacitance between them makes a positive feedback loop and causes the circuit to oscillate. It only happens when the gain is maxed because that's when the circuit has enough voltage gain to start oscillating. If I'm right, then you might be able to hear the whine kind of mixed in with your guitar when the gain is turned up just below the point of oscillation.
It could be the charge pump though, check the AC voltage on your supply rail. Also, I don't see any filtering on VB in your layout, you might want to add some (10uF to ground should do the trick)
Quote from: psychedelicfish on September 08, 2013, 03:56:11 AM
Quote from: MrStab on September 08, 2013, 01:24:44 AM
observation: with gain at full, and for the last little bit of its rotation (ie. not just shorted to one end), and the boost wires connected together, there's a high-pitched whine.
It sounds to me like your input and output wires are too close together. The capacitance between them makes a positive feedback loop and causes the circuit to oscillate. It only happens when the gain is maxed because that's when the circuit has enough voltage gain to start oscillating. If I'm right, then you might be able to hear the whine kind of mixed in with your guitar when the gain is turned up just below the point of oscillation.
It could be the charge pump though, check the AC voltage on your supply rail. Also, I don't see any filtering on VB in your layout, you might want to add some (10uF to ground should do the trick)
they're not physically near on the board, but the jacks may have been close at the time - i'll need to check that out. RG's mention of whine was my initial thought, too - maybe it is somehow the 7660's (pin 1) boost-mode frequency bleeding through.
there's only a 47u from V+ to ground on the main board itself, but there's a 4u7 to ground on both the Vbias and V+ rails coming out of the opamp on the charge pump board. a bit on the low side, maybe worth increasing? i didn't anticipate this and ran out of +35V-rated 10uFs, hence the 4u7s.
right, so check jack position, check for AC, check filtering. got it. cheers! i'll get on it on monday. then to worry about draining the circuit more with 2 LEDs...
here's the final layout i used (forgot to change the pot labelling - iirc Treble and Bass are backwards):
(http://i.imgur.com/qqZjptll.png)
btw, i could've misinterpreted, but didn't you mention at one point that the lower voltage would result in more gain? maybe that has something to do with it? i feel as though that's a less-likely cause, but just throwing it out there.
Lower supply voltage=same gain, more distortion.
To elaborate on my above post:
If you feed a 1V peak to peak signal into an amplifier with a gain of 20, the output will be 20V peak to peak. If that amplifier is running on a 10V supply, the amplifier will run out of supply and the output will be clipped to 10V peak to peak. If that amplifier is running on 40V, however, the output will be able to reach 20V peak to peak without clipping. While in both scenarios the amplifier has a gain of 20, the one running on 10V has more distortion than the one running on 40V. The reason you see controls on amplifiers etc. that control the amount of distortion labelled gain is that they use a variabe resistance to control the gain of an amplifier (think tube screamer), which is overdriven itself, overdrives the next stage, or has its output clipped using diodes. So always remember: gain≠distortion
thanks for the explanation. i know gain and distortion arent interchangeable, but that helped me visualise whats going on here. so my paranoia over preserving the voltage~headroom~tone is unfounded so long as i make accept the increased gain?
just wondering - how problematic in terms of current do you think 2 dim-as-humanely-possible red LEDs might be in this circuit, given that it's already dragging down the charge pump output considerably? the ammeter function on my DMM is screwed following a wrong-setting blunder the other day. can't beat the classic electrical burning smell.
run the leds from the supply before the 766a-ma-call-it cajigger. then you can rack them up to inhumane brightness, without problems.
Quote from: duck_arse on September 09, 2013, 11:04:17 AM
run the leds from the supply before the 766a-ma-call-it cajigger.
i'd thought of that but my tendency towards illogical assumptions thought "that must somehow be bad!" lol. awesome, this way i don't have to worry about the voltage. just got 60 LEDs through in the post. ill get back to work in a few hours to reverse those pots and look into the gain issue. cheers!
first time hearing without headphones and bzzzzzz.
if i use the highly-professional method of holding my stings and touching the 2k2 resistor on the far-right of my layout, it disappears albeit with some tone loss. i'll get a wire holding a resistor to ground and poke around.
update: um... i usually try not to jump to this conclusion, but what are some less-obvious causes of external interference? i ask because at first i noticed that waving my hands above the board affected the sound, and thought nothing of it, but then tried tilting the board vertically while on and there was a gradual decrease in hum. doesnt seem to be any offboard components touching. i've tried turning off my wifi router, monitor, even computer, fluorescent bulb... putting a metal tin over it did nothing. weird. been over the board a few times, no luck. there was a 220nF cap going to ground instead of vbias, but that was one of the ones you said was optional, Edward.
still buzzing, less so when facing upright. now i'm not getting any clean, but that's probably some other problem i've picked up along the way. i have 40VAC on the V+ rail, which surely can't be good.
things i've tried:
-100uF & 100nF caps to ground on both V+ & Vbias
-attaching ground to part of metal enclosure (which it isn't in yet) and covering the circuit - very minor improvement
-hooking up volume pot and jack ground to bias for a more "true" virtual ground - much the same as the normal configuration
all opamp voltages check out. getting about 21VDC from just the bare charge pump and 15 with the circuit load. whilst i hadn't fully covered it in a grounded box, i'm doubtful that's the problem.
You could have a ground loop. These act as aerials and pick up hum and RF interference which are introduced through the ground of your circuit and end up mixed with your signal, resulting in lots and lots of hum. If you only get it when it's plugged in to an amp, it might be to do with the way you're powering the circuit, if you're using an AC wall adapter then it could be the case that it has a mains earth connection on it, which would be your ground loop (all mains powered amps will have a mains earth connection).
Another thing it could be is that the ground wire from your input jack has fallen off, or some kind of earth connection has come loose. It could also be a wiring mistake somewhere.
As far as your AC voltage on your power supply, are you sure you measured that correctly? You easily could have mistook 40mV for 40V, which would make sense. If you do have 40V, however, which I highly doubt seeing as you haven't reported any burning smells, exploding capacitors, etc, if you can get access to an oscilloscope you can have a look at the wave and see what frequency it is. If it turns out to be 10kHz or so, then your problem lies with the charge pump. If however, your 40VAC is at 50Hz or so, then your problem lies elsewhere.
Quote from: psychedelicfish on September 10, 2013, 02:30:01 AM
You could have a ground loop. These act as aerials and pick up hum and RF interference which are introduced through the ground of your circuit and end up mixed with your signal, resulting in lots and lots of hum. If you only get it when it's plugged in to an amp, it might be to do with the way you're powering the circuit, if you're using an AC wall adapter then it could be the case that it has a mains earth connection on it, which would be your ground loop (all mains powered amps will have a mains earth connection).
Another thing it could be is that the ground wire from your input jack has fallen off, or some kind of earth connection has come loose. It could also be a wiring mistake somewhere.
As far as your AC voltage on your power supply, are you sure you measured that correctly? You easily could have mistook 40mV for 40V, which would make sense. If you do have 40V, however, which I highly doubt seeing as you haven't reported any burning smells, exploding capacitors, etc, if you can get access to an oscilloscope you can have a look at the wave and see what frequency it is. If it turns out to be 10kHz or so, then your problem lies with the charge pump. If however, your 40VAC is at 50Hz or so, then your problem lies elsewhere.
my best guess from my limited experience is that it's not ground-related. at least, not offboard. despite the earth pin being standard here in the UK, i'm using a small 2-pin kettle plug with only live & neutral connected. it's a supply i ripped out of some other AV application and stuck some rudimentary filtering on it - i've used it live through amp/cab & PA a few times with about 5 effects hooked up to it with no issues of note. the ground to the input jack had come loose very early on, but i spotted it right away. i guess you could see my other pedals & rig generally as a control batch. my practise amp can be a bit noisy, but i know this is out of the ordinary. in regular-ground-retrofit mode and true virtual ground mode, vbias and true ground behave identically.
i checked with the 200VAC setting on my multimeter. as it's exactly double DC voltage, could that mean that I'm just getting pure AC through and the DC readings are just showing half the ripple/wave/whatever it's called? in that case - why aren't my opamp and transistors dead?! that's a weird one. in a not-so-serious effort, i tried a basic bridge rectifier made from 1n4001s but it just gave voltage drop - given the uncertain AC or DC nature of what's going on, i wasn't entirely sure if i should connect the negative from the charge pump board like you would a bridge rectifier with guaranteed AC input. so maybe that prevented that idea from working. noise is present with a battery, but admittedly i hadn't thought to check AC reading with a battery hooked up.
yknow, i did read a very vague post somewhere saying that voltage double/triplers are AC output only - i didn't know that was implied, and i haven't found anything else on it, so i'm just not sure what's going on.
there's only straight series connection from each section/component to ground. last i tried, the charge pump ground went straight to DC input negative in parallel with the rest of the circuit, to keep out 7660 noise as per (my interpretation of) RG's notes on the matter. speaking of which i've just bridged the PNP activating power from the input ring, as it works fine, but i figure i'll just keep it out of the equation for now.
boxing it up has helped a lot, although since somehow screwing up my charge pump, i've been testing on 9V/4.5Vbias. still a bit of hum, but much closer to acceptable limits, i'd say. almost no buzz at all with gain on full without boost activated, buzzy with full gain + boost on, but seems to just be the characteristic buzz of distortion pedals pushed to the max.
nasty pop when using the boost switch - maybe a FET/PNP would help with that. or can the typical pulldown resistor be used in this situation (i usually associate those with switches passing signal, not DC)?
That's strange, it shouldn't really pop because the only capacitor that gets disconnected by the switch (C3 in my schematic) has a reasonably low value resistive path from one end to the other. Try a 1M resistor across the switch.
everything almost works now, but since trying with the charge pump back in action, there's a weird old-radio/R2-D2-like whine on the clean channel. it's only present if the Gain pot is turned up a little bit, and you can make cool Artoo sounds by turning it. in clean mode, the gain pot weakens the sound the more it's turned up, but works as expected in Boost mode. weird. so something's up when there's no vbias going to the boost section of the board (the 7660S has pin 1 connected to V+ to theoretically prevent audible whining, plus the problem disappears in Boost mode, so i doubt it's that). so tempted to use NPNs or a 4053 i have lying around to just kill the gain pot in clean mode lol. i'll have a look at the original switch on the schematic.
pop seems to have disappeared as well. fingers crossed it doesnt come back, but i'll bear the resistor idea in mind if it does.
i take it back - it probably is the charge pump. upon swapping out the IC for a known defective one, the sound changed. and it seems there is a whine present for just a small part halfway up the gain pot. i can predict trying all sorts of decoupling/isolation/filtering solutions that don't change anything. doesn't seem to be affected by reducing proximity to other components.
at this point, i'm feeling it would be so much easier to try and get cheap 18V regulated supply, seeing how i'm only getting 15V under load on the charge pump anyway. laptop chargers are usually up there. hmm...
Quote from: MrStab on September 10, 2013, 11:16:58 PM
at this point, i'm feeling it would be so much easier to try and get cheap 18V regulated supply, seeing how i'm only getting 15V under load on the charge pump anyway. laptop chargers are usually up there. hmm...
Or you could change some components around and adjust the amount of gain to give you similar levels of distortion running the circuit on 9V. You said you were making this for a friend, so having it run off a battery or a standard pedal power supply would be far more convenient for them. We're changing the supply voltage from ±15V, or 30V to 9V (let's say 10V to make things easier), so we should reduce the gain of the op amp stage by a factor of 3. This can be done by replacing the 1.5M resistor labelled on my schematic as R7 with a 500k resistor, and changing the gain pot for a 250k one.
Quote from: psychedelicfish on September 11, 2013, 12:53:24 AM
replacing the 1.5M resistor labelled on my schematic as R7 with a 500k resistor, and changing the gain pot for a 250k one.
that sounds like a plan. i guess that 250k would still need to be a log pot? all i have in the way of 250k is a linear, although i might have a 500k linear somewhere i could taper down (although that usually doesn't work out well for me). a
bit more distortion then the original wouldn't be such a bad thing, as it's meant to be like the Rusty Box but not necessarily identical.
i just remembered something from a couple pages back:
Quote from: Bill Mountain on September 05, 2013, 03:23:57 PM
My only worry is that opamps scale up and down with voltage quite easily. Transistor do not. Transistors in the differential amp configuration (and also the buffer) may also not like going from a bipolar supply to a single supply. If you aren't going to use the 30 volts and/or a bipolar supply then you might have to adjust the biasing of the differential amps to work off of a lower voltage single supply.
it doesn't seem to sound that different besides more distortion on 9V, but think there might be any ground to Bill's concern?
thanks again!
Quote from: MrStab on September 11, 2013, 01:24:09 AM
Quote from: psychedelicfish on September 11, 2013, 12:53:24 AM
replacing the 1.5M resistor labelled on my schematic as R7 with a 500k resistor, and changing the gain pot for a 250k one.
i guess that 250k would still need to be a log pot?
Linear should be fine, unless your friend likes to rapidly turn the knobs while playing :icon_biggrin:
Even then it wouldn't be too hard to swap it out later.
Quote from: MrStab on September 11, 2013, 01:24:09 AM
i just remembered something from a couple pages back:
Quote from: Bill Mountain on September 05, 2013, 03:23:57 PM
My only worry is that opamps scale up and down with voltage quite easily. Transistor do not. Transistors in the differential amp configuration (and also the buffer) may also not like going from a bipolar supply to a single supply. If you aren't going to use the 30 volts and/or a bipolar supply then you might have to adjust the biasing of the differential amps to work off of a lower voltage single supply.
it doesn't seem to sound that different besides more distortion on 9V, but think there might be any ground to Bill's concern?
Other than socketing the transistors, I think you'll be alright. If you have a few PNP transistors perhaps you could try a few out and find a combination that sounds good.
Since you're building for a friend you might want to let him/her play through the circuit before you box it up to make sure it sounds good.
just as i tested the circuit after changing the 1.5M resistor, the whining is back, without the charge pump, just fainter. happens with a battery, too.
it only happens after the halfway point on the gain pot, so maybe i need to tone that back.
screw this, i'm going to bed.
At this point it seems like it's your layout and/or wiring which is causing the circuit to oscillate. Basically, you have the input of one gain stage (probably the op amp) too close to its output. I've just been looking at your layout, and at this point I'm pretty sure that's your problem. You have a big long jumper from the -input of the op amp to R4, R5, R6, etc. which passes right through the middle of the transistor "ladder", which is a later gain stage. The capacitance between the transistor "ladder" and your jumper makes a positive feedback loop which causes the circuit to oscillate. To fix this, I suggest you replace the jumper in question with a shielded wire, you may be able to get away with using a small length of guitar cable for this. Connect the shield to the trace going to pin 4 (DO NOT connect the other end to anything, or you'll make a ground loop) and use the centre wire as your jumper.
that does make sense, and i'll either get that jumper shielded or moved when i wake up later. but there's something i want to bring into the equation first - there seems to be some connection with the treble pot, as though a 10k fixed resistor creating a minimum resistance at the end of the turn would prevent the noise altogether. then i remembered i'd used a 10k pot for Mid instead of the prescribed 15k. a small difference, but not an entirely random theory - you don't think that could be causing anything?
right, i really am going to bed now. lol
Seeing as the tone controls are all capacitors and resistors in the "oscillator" changing their values would also change the oscillation. You might find that turning the tone controls changes the oscillation frequency slightly.
in that case, one quick question: in terms of functionality, should i add a fixed resistance to the lower or upper end of the Mid pot (to increase its theoretical size)? if a 4.7k on one side of that doesn't work (or 2.2k on either side), i'll try the treble pot. i wonder why the mid pot being wrong make treble seem like the culprit, but not itself.
turning the pots do change the sound's frequency - i've had full conversations out of the droid.
well... a piece of shielded guitar lead in place of the opamp inv. input - C4 jumper, with one end tied to ground and the other open, has made it usable at 9V (or so it seems). charge pump still makes crazy whine. sucks, as the circuit seems to lose some fidelity at 9V. but then, it could be psychosomatic and i could be expecting the feel of a guitar pedal.
clearly i'm the worst driller on the planet, but FWIW:
(http://i.imgur.com/0joummbl.jpg)
problems went away after using a shielded wire, so thanks, Edward, and ofc Mr. Faraday! (the whine comes back with the charge pump, but it sounds fine on 9V). i restored the 1M log pot for more gain but retained the 470k resistor i replaced R7 with. there's a bit of interference right at the end of the gain pot turn on Boost mode, but i explained this trade-off to my friend and i'll see what he says.
it sounds more guitar-ish than bass-ish to me, but it sounds crunchy in a nice, organic way. i may get a clip at some point, but i'm a bit busy lately so dunno if i'll be bothered to set up my USB interface and a camera might not do it justice.
I'll knock together a PCB layout soon. Perhaps one of us should start a dedicated project thread for this?
Quote from: MrStab on September 11, 2013, 05:57:54 PM
i restored the 1M log pot for more gain but retained the 470k resistor i replaced R7 with. there's a bit of interference right at the end of the gain pot turn on Boost mode, but i explained this trade-off to my friend and i'll see what he says.
If you want to get rid of that interference, I would try replacing C2 with a 220pF (or the closest value you have on hand), which would give you a high frequency roll-off of ~30kHz as opposed to ~150kHz (a normal guitar amp will roll of the high end somewhere around 5kHz, so no treble will be lost).
mrstab, seeing your picture, and being blinded by your leds and you saying charge-pump whine has me thinking. not necessarily a good thing. are you running your leds from before or after the charge pump, or have we covered this already? what value series resistor are you using for your leds?
or did you say that whine was constant? I have read this thread as it happened, but can't remember where we are up to ......
i took the charge pump out of the equation altogether, i think my camera's just a bit misleading with the LED brightness. i'm using 1K resistors (one via. Millennium Bypass) - i did some testing with the circuit load hooked up and it seemed like the right balance for this one. i went with red for lowest current consumption.
weirdly, since boxing, interference seems to have gone. i say "weirdly" because it's just unusual in my experience with RF/external interference. i straightened out the knobs a bit:
(http://i.imgur.com/z41odvZl.jpg)
it's for my singer's bro-in-law, so worst-case scenario it's not an outright witch hunt if he doesn't like it. lol
(i was only recently informed that imgur is a hive of adolescent tomfoolery, but i'm too deep into the habit now)
Looks nice! Any chance of a sound sample?
why wouldn't he like it? it's got knobs, switches, lights, shiny bright metal .....
here's a coupla clips on a shit cam, with the pedal going into the front of the amp. i didn't think to record when actually trying it as a preamp. well, going straight through my amp's FX return at least - is a bit more defined that way, but the MG's preamp doesn't adulterate it too much.
please excuse the shitty improv, quality, the fact i'm using a guitar/amp and not a bass, and (what i hope was) hot glue residue which was since cleaned off. lol
the dude just came over and tried it into the Aux input on a Cube 30 Bass amp - much more beef. success!
I wanted to say that I built this using the most recent strip board layout, mostly with spare parts I had laying around. I did need to use shielded cable for the input and jumper and change the value of C2 to get the noise down to a useable level. I'm using a 12V DC wall wart with a 6v regulator for vbias. So far, sounds awesome. We'll see how she teams up with my 1400 watt power amp tonight.
(https://scontent-b-sjc.xx.fbcdn.net/hphotos-ash3/t1/1654336_10100295161700691_1580905746_n.jpg)(https://fbcdn-sphotos-b-a.akamaihd.net/hphotos-ak-ash3/t1/1779930_10100294255486751_2068337598_n.jpg)
I've built this using schematic from this thread and my own PCB layout. It works fine, but it's a little bit too noisy. I use 350kHz charge pump to provide ±15V from standard 9VDC, but noise is audible and this charge pump is
almost noiseless, so I presume it's not the source of the noise. [edit: I tried to decouple V+ & V- to ground with 10u capacitors and noise level does change, so it seems that my charge pump is also pumping the noise. I will investigate this further.]
However, I noticed that another version of this preamp - found in Traynor TS120B, have two differences compared to this on (Traynor TS50B, actually):
- Gain pot is A100k instead of A1M;
- There's a 47p capacitor parallel to gain pot.
So I tried to add that cap and the huge amount of noise is gone. What I want to ask is: is it normal for such a circuit to be noisy, as it has really big amplification factor, or there's something else I should try to fix?
I used TL071, BC550C (don't have MPSA18) and BC640 instead of MPS8599 (similar characteristic). Maybe those BC640 are wrong choice? Some good info could also be found here: https://www.diystompboxes.com/smfforum/index.php?topic=110526.msg1014149#msg1014149
So I tested this with normal ±15V PS (regulated with linear regulators), which is quite quiet :)
Guess what? The noise level of Traynor preamp is the same as with SMPS.
It could be nice to test this with some other kind of transistors, maybe those BC640 I picked are too noisy? The trouble is they use ECB pinout, instead of usual CBE/EBC :(
(https://i.postimg.cc/m1k6y5wc/RUSTY-BOX-LAYOUT.jpg) (https://postimg.cc/m1k6y5wc)
Hi guys, I know this post is old but I have not found other more recent posts so I ask here.
I made a PCB based on the Traynor head scheme, I made a double check of everything and it seems to me ok, I would like to help me understand if everything is in the right place.
I implemented a power section that allows you to feed the pedal with 9V, the potentiometers and the gain switch are directly mounted on the PCB.
If everything should be ok I will post the PDF with the PCB, thanks!
ok, i did it and i can tell you that it is VERIFIED, perfectly working.
I advise those who use ICL7660S on the power supply to jumper pins 1 and 8 also of the -15v chip.
The potentiometers and the input gain switch are mounted on the trace side.
If anyone is interested in the pcb file to etch can contact me, thanks.
some pics of my clone
(https://i.postimg.cc/HJhsGzjQ/IMG-1761.jpg) (https://postimg.cc/HJhsGzjQ)
(https://i.postimg.cc/ygDxW52Q/IMG-1762.jpg) (https://postimg.cc/ygDxW52Q)
(https://i.postimg.cc/9w3m0fd0/IMG-1763.jpg) (https://postimg.cc/9w3m0fd0)