Tube preamp inside a guitar

[sumgai]

dunk,

Your idea has great merit, I like it!

The XLR connector might be a bit big though, be sure to check for enough room inside your axe. Possibly a DIN connector instead?

I think you already know just about everything necessary to make it happen, just from the way you describe each of the parts and the reasons for their use. Check back here for parts values when you actually have the zeners in hand. The calculations aren't quite as easy, because you're actually going for control of the current, and not just how much voltage to drop.

And I'll need to know what your source voltage is, too.

HTH


Oooh, wait.....

No, you're not a PITA, you're a fellow Nutz!! People who want to learn can ask me stuff all day, that's why I'm here. (Trust me, I sure as hell ain't here for the Australian Gay Porn! ;D ;D)



sumgai

[dunkelfalke]

If all else fails I still can solder an XLR cable inside the guitar with the female connector hanging out from the output socket hole. Okay, the guitar will have a tail then but this way I stay compatible to all phantom power units like mixers.

The only thing that might be problematic is that XLR phantom power wouldn't have enough current for all the stuff (AFAIR it can be anywhere from 1 to 10 mA at +48V). Also I have read something about decoupling resistors and stuff so I definitely will need your assistance in designing your circuit - you way overestimate my knowlege about electronics (although I am currently trying to learn it, I have got too much other stuff to learn at my new job and thus my learning capacity is seriously limited at the moment).

[dunkelfalke]

Would it be safe to run the active pickups and the EQ controller with 10 volt instead of 9? If yes it would make the resistor redundant and probably would save some current (and lower the heat).

[sumgai]

dunk,

The tube's plate will certainly handle it, it's good for up to 30 volts in some cases. But you still want to lower/regulate the cathode/G3 voltage down to about 1.5 volts or so. You might get away with as much as 1.65 volts (about 10% overvoltage), but beyond that, I can't be responsible for the fire works.

As for your active pickups and EQ circuits, I can't speak to them, I have no knowledge of any sort regarding these components. But I seem to recall that most EMG pickups and accessories are capable of being run as high as 18 volts - is that true, or do you know that your pieces can't go that high?

HTH


sumgai

[dunkelfalke]

I asked Artec about the EQ controller and they answered that up to 12 volts is safe. Schaller (the OEM maker of my pickups) says that their active pickups shouldn't be used with voltages higher than 9v. Can't tell whether 10v would still be in the safety margin.

[dunkelfalke]

Disregard that, I just found out that Conrad sells 9.1v 1.3 watt zener diodes, thus removing the problem altogether.

Can you modify my circuit diagram by adding everything that I need to use the XLR 48v phantom power with it with the help of a BZX85C9V1 zener?

Thanks a lot :-)

[sumgai]

dunk,

No problem, but I'll need a bit of time, and right now that's the one thing that I'm in short supply of.....

So, if you're tapping into a standard 48v Phantom power supply, then you're intending to run this tube pre-amp into an ordinary mixer? Not that it matters, I'm just curious.


sumgai

[dunkelfalke]

I have ordered a dual PSU for that since I also have got a Behringer Di-Box with a phantom power option and a cabsim. I also own a XLR to USB "converter" so I do intend to do a bit of direct recording, and later use it with an ordinary mixer. This probably will even sound good with piezos.

And since I will have a PSU for that, I'll also be able to use guitar amps with the help of a XLR to 6.3mm jack cable thus having the best of both worlds.

[dunkelfalke]

I just found out that the zener is not a BZX85C9V1 but a ZPY9.1. It should be compatible but has got a somewhat different specs. Here is the data sheet: www.diotec.com/pdf/zpy1.pdf

[limbe]

I don´t know if you have any use for this information,but anyway here it is:If a battery still is one of your options there is an adjustable regulator named ICL 7663 which has an output voltage between 1,3-16 volt ,needs an input voltage of at least 1 volt more than the output voltage,draw less than 4 µA and can deliver maximum 40 mA of current.Good luck with your project which I find very fascinating.

[dunkelfalke]

Can't find that one in German shops. Thanks for the good luck wishes :-)

[sumgai]

dunk,

OK, got some time, let's use it constructively......

Assumptions:

We have +48vDC coming in on the XLR connector. (This would be sharing the ground connection, not always a good idea - what if you want/need to lift the ground for the audio portion? Can't happen if it's also part of the power supply.)

We want +9vDC for use on the plate (anode) and 1.5vDC for use on the cathode. (Thanks to the internal connection to G3, the cathode's polarity is specific - you must hook the positive voltage to pin 5.)

Ideal solution:

A zener diode regulates the incoming voltage to produce the desired voltage(s). In fact, since the 9.1 volts will be regulated, we can use a simple voltage divider to obtain the 1.5 volts from that, thus eliminating the need for a 1.5 volt regulator of any kind.

Now, I've spent some time both online and off, and I've come up with a "nifty" circuit that will do what you want, but there is a requirement. I've never paid attention the "phantom power" other than the fact that it's available on nearly every mixer made in the last 10 years. However, regardless of what the voltage might be, the current is almost never stated, and that's a sad thing. In this scenario, we're gonna need about 10-12 ma., and possibly a bit more. (I'd be surprised if it took more than 15 ma. though.) Many, if not most, mixers use 'current limiters' in their circuitry do prevent damage to a microphone, should a connection be damaged (or wired backwards by your drunk roadie!). I've seen some units claiming to limit that output to 2 ma., and others offering up to 10 ma.

You mentioned that you've ordered a "dual psu". OK...... what's the expected current availability of this thing? If it's not up to what we need, then I hope you can exchange it for a unit that will deliver the requisite amperage.

Onwards.....

Thankfully, only one or two things are somewhat critical here. In fact, we can use this circuit, with one or at most two changes, on any of the standard phantom power voltages we're likely to find -12, 24 or 48 volts. I'm showing the values for a 48vDC source, per your post above.


OK, the circuit itself:



Note that I've rearranged the part numbers to align them in the order found, reading from left to right. (Usual industry practice.)

R1 - 1MΩ, sets the input impedance
R2 - 18Ω, limits the current on the cathode/filament
R3 - 100KΩ, the load resistor,
R4 - 6.8KΩ, the upper half of the voltage divider for the 1.5vDC
R5 - 1KΩ, the lower half of the voltage divider
R6 - 1500Ω at 1 watt, drops the incoming voltage down to where the zener can work with it

C1 - 10µf electrolytic, bypasses the cathode's AC to ground
C2 - 1 - 10nf, optional, used mainly to suppress RF transients that might become audible
C3 - 10µf electrolytic (can be polarized or non-polarized), keeps all of our DC inside our own rig
C4 - 10µf electrolytic, smooths any small fluctuations found on the incoming phantom power voltage

D1 - zener diode, select to provide the desired voltage

...... and of course, the 6418 tube itself. (You should try to find a socket that fits this thing, for soldering purposes if nothing else.)


Discussion:

I've eliminated the capacitor on the incoming signal - we won't encounter any DC for it to block.

The input impedance is set very high by R1, as as specified by the tube manufacturer.

The cathode/filament current must not exceed 10 ma. I take that to mean that there is a tolerance, and that if it goes up say 20 or 25%, then the tube will still be OK. If the 18Ω resistor is too low in value, like say half (10Ω), the current would be about double the rated value, and the tube will likely imitate a Roman Candle.

The bypass cap on the cathode ensures that no DC will be shorted to ground, but that all AC will be shunted straight to ground. This helps to eliminate fluctuating bias in the grid/cathode circuit. The input impedance resistor and the signal itself are all that's needed to set the bias properly.

The suppressor cap might not be needed, but just you should use it at first, then try it without. If things still sound OK, then leave it out. If the tone suddenly gets shrill, then put it back - the tube is trying to oscillate, and is generating square waves way up there in the frequency range, like really really high! (This phenomenon is also called "ringing".) (And no, the bypass cap can't react fast enough to remove them, that's why the small value cap.)

The voltage divider merely cuts the voltage at the top (9.1vDC) down to what we want. The values are high, compared to R2, in order to avoid over-working the zener itself. The ratio is 9.1/1.25, which works out to about 7 to 1. (I'm being conservative here, in anticipation of resistor value tolerances.) In terms of common resistor values, that gives us 6.8KΩ and 1KΩ, a good compromise.

R3 provides a load for the tube. In brief...... this is necesary (in any amplifier circuit) in order to derive a voltage as the current passes through it. (You can all thank Mr. OHM for that one.) Without this process, we'd have no way to distinguish between what the power supply is putting out and what the tube itself is putting out.

The output blocking capacitor keeps our power supply voltage from getting out of our rig and into any other circuits where it would likely damage someting, or at least make our tone go way out of bounds. These days, if you can find a non-polarized high-value electrolytic cap, then it's likely to be cheaper, and work just as well. Otherwise, use the same thing you used for C1. It'll work just as well, tone-wise, but it's usually more expensive.

The heart of the matter:

The zener diode and R6 go hand-in-hand. R6 is calculated to drop some of the voltage from the incoming supply. It must pass a certain amount of current as well, in order to keep the zener working, but it must not pass so much that the zener self-destructs. The upshot here is that no matter what we choose, if we do it correctly, we'll still be building up some heat within the resistor. This is why I specify that the resistor be able to handle at least 1 watt - it must not be any less than that!! I think you'll find that this is still only a fraction of the size of a battery (even a standard 9v unit), so the net effect is still a much smaller overall device. (But it is somewhat bigger than your average ½ watt resistors, just to let you know ahead of time.)

And finally, the capacitor across the incoming power supply. C6 is meant to keep small amounts of AC from getting in and causing our tone to "warble" or some such. Since we have an impeccable source of power in the first place (at least it should be rock steady), I hesitate to include it. However, it is more-or-less standard engineering practice to include it, so I'll go along with the crowd on this one...... but just this once, mind you! ;D

OK, I think that about covers it. I sincerely hope that your 'dual PSU' unit can supply the required amperage, cause if not.......

HTH



sumgai

[dunkelfalke]

Thanks a lot for the solution. It raises some questions, though:

1) I've ordered the parts before you came up with it so I used the values in the previous circuit. I don't care about resistors, they are easy enough to get, but I've ordered a bunch of very good capacitors (Panasonic stacked metal film) with the tolerance of only 5%. Is it possible to substitute some capacitors in your circuit diagram with them? It's not because of the money spent but because Conrad doesn't carry those and I try to use better parts for this guitar.

2) I also will need to feed my pickups with 9v. Is the junction between R3 and R3 the right place for that? Do I need another load resistor for them?

3) Oh yes, the pickups. How I connect their output to the tube output?

4) Since we talk about the phantom power current: let's assume the current is 10mA. But it is the specs of 48v, wouldn't the current be higher after the convertion to 9v? And if not, would it be possible to identify and to replace the current limiter in the PSU with one more fitting to my purpose? Since it is a dual PSU I could do that for one part of it (so the other part stays compatible with other stuff). I don't mind voiding the warranty.

5) It just came to my mind: there are three poles in the XLR connector: positive, negative and shield. Positive and negative is clear, but where does the output signal and the shield connect to?



Thanks a lot for your time.

[JohnH]

Nice circuit sumgai - I'm following this one with interest!

John

[catweazle]

Been watching this thread for a while now.
Cut my teeth (& burned my finger s) on tubes (uk: 'valves').

Sorry but Gai's latest circuit won't work.
The R4/R5 divider will give the required 1.5 volts only when unloaded. If the heater takes 10mA at 1.5 volts, then it has a resistance of 150 ohms. This is in parallel with R5 giving a divided voltage to the heater of maybe 150mV.
If the zener is 10volts or so, I'd suggest removing R5 & changing R4 to 850 ohms - this would be in series with the 150ohms of the heater to give 1000ohms, which should result in the needed 10mA current. The 850 should also act as a limiter when current is first applied & the heater is cold and therefore will have less than its operating resistance.

For info : to work efficiently, the tube anode should be about half supply voltage with no signal. So with 100k as a load, the load has to drop 5 volts (for a 5 volt supply) which means that the tube will be running with a minute 0.05mA current. I suspect this will be too small to be useful (the output impedance will be very high).

Also - by tying the grid 2 to the anode, the tube is being run as a triode. If another resistor is added between grid2 and the supply voltage, the tube will be run as a pentode. There is a lot of debate about which mode is 'better' for audio applications. Gain versus distortion etc.

Sorry to be negative - I think it's a wonderful idea to add a tube to an axe - but it is a bit trickier than FETs and amplifier chips.

[dunkelfalke]

Hmm, thanks for the insight. The zener is 9.1v by the way.
Also, please don't apologise for being negative, I value every idea and it looks like I won't receive the parts before the next weekend so there is plenty of time to work out the best solution.

My goal is the best sound quality for piezo pickups - so clean sound is what I need. I also need to be able to add this signal to the signal of my active magnetic pickups (the idea being that magnetic pickups sound a bit fuller with a bit more low end so mixed together with piezos I would get the best acoustic like sound).

[sumgai]

dunk,

Tell me what caps you ordered, and I'll be better able to tell you what can be used, and what might need to be changed.




sumgai

[sumgai]

FelineFerret,

You've been here a year now, and if you've been following the general chatter, you should have noticed by now that no one starts out with "that won't work", but instead we start out with "are you sure that's the way it'll work?" Courtesy still counts, at least in the NutzHouse part of the innerwebs. 'Nuff said.

As a former practicing EE, I don't agree with your assessment, nor with your solution. However, I have learned over the years that arguing a fine point is pointless, when the workbench is right before us, and begging to deliver up the proof. After all, the idea is not to 'engineer' things, it's to make an idea work as envisioned. However the bottom line is derived, no one has the right to say that the chosen path is/was right or wrong. One's opinion of rightness or wrongness is of course welcome, but only so long as it remains an opinion. It becomes a fact only when others have physically verfied it (or when JohnH has pSpiced it and published the resulting graphs!).

I will say this however, and that is that at least you offered up your own version of how to accomplish the mission, credit to you for that. I happen to disagree with both your methodolgy, and what I feel are flawed results depending therefrom, but far be it from me to say anything negative - I've not prototyped your circuit on my own bench, so I'll keep quiet.

In closing, let me say only Thank You for your expressed opinion, and I hope that you freely participate much more often here in the NutzHouse.





sumgai

[dunkelfalke]

Feb 20, 2010 14:39:45 GMT -5 sumgai said:

Tell me what caps you ordered, and I'll be better able to tell you what can be used, and what might need to be changed.

10x Panasonic Stacked Metal Film 0,1uF
2x Panasonic SMF 1,0uF
2x Frolyt EKM 47uF 16V
2x Tantalum 10uF, 16V
3x Tantalum 4,7uF, 10V

[catweazle]

I must apologise if the tone of my previous post was brusque rather than terse.
I did not have much time & didn't want Dunk to start something that wouldn't work.

In my defence, I've worked in electronics over 40 years now, design, build and repair & mod, mostly analog. (My first published article was for a tiny guitar practice amp which I'd also built into guitar body). With retirement less than 2 years away, I plan to then catch up on a bunch of guitar-based projects that I hope I can share with ye nutzers. (e.g. I've already made a cute little board with 3 Tillman pre-amps; I have a 30 year old 3 watt tube amp to be converted into a Herzog; a couple of novel effects that may or may not work as planned; etc.)

But to the project in hand...
Full data on the JAN6418 does not appear to be readily available, so some of this is best guess.
IEC spec for phantom supplies is 10mA max, so taking virtually all that for the heater is getting close to the limit. Also, that really restricts what equipment you can use the guitar with. I know you don't want batteries, but an idea I had when helping someone wanting to add an internal amp to his folding (yes - folding!) guitar, was to alter a strap to take a handful of AAA or AA cells bandolier style. Depending on your type of music, that could be made to look pretty cool. Like an ammo belt. Portable radios of the 50s had batteries of 67 or 90 volts for the tubes and a separate battery (or part of the main battery ) of 1.5 or 7.5 volts for the heaters. Remember the heaters are taking 100 times the current that the actual device is making use of!

If you really must avoid batteries, consider a DC-to-DC convertor chip. These can convert high voltage/ low current to low voltage/high current with around 80% efficiency whereas simply dropping 9 volts to 1.5 just burns away 7.5 volts. Not a huge deal here except that you could then use the phantom supply approach well away from the 10 mA limit. Drawback is extra complication - though they are very easy to use - and a possible electrical noise issue because they do switch hard & fast all the time.

I gather the 6418 is spec'd to run from 10 to 30 volts anode voltage, so again 9 volts is not the best starting point (unless that's part of the design to increase non-linearity?) 20 volts is an obvious starting point - or 18 from 2x9v sources.
Back-of-envelope estimate of tube current (from the heater power) is about 0.1 mA.
At a supply voltage of 10 volts (I know you're thinking 9.1, but 10 keeps the calcs a lot simpler & this is just ball-park stuff) the load resistor should drop 5 volts making it 50k ohms. That would be 100k for a supply of 20/18 volts.
What detrmines the actual tube current is the voltage between the grid and the cathode.
A complication with these mini tubes is that the cathode is tied to the heater - so you have to get your head around 1.5volts ( or apparently more correctly 1.25 volts) across the heater but with the cathode end of the heater being a volt or two above ground. I'm guessing the 18 ohm resistor in the heater chain is to help make up this cathode bias voltage.

As I alluded to earlier, the tube is configured as a triode although it is made as a pentode device (tri=3 pent=5 meaning how many electrodes are in the bottle). By taking grid 2 to the supply via a resistor (starting point try the same value as the anode load resistor if the full spec of the tube can't be found) and you should get the extra gain that pentodes were originally designed for. There is debate in the audio world as to which 'sounds' better. I believe some tube amps actually had a triode/pentode switch on them.

For my sins, some of my work involves piezo devices - although in ultrasonics rather than pickups. The starting point is usually as high an input impedance as possible. The circuit in question here has the grid resistor as its input impedance and is therefore 1 megohm. OK for coil pickups & general audio work (which is where the design came from) but a bit poor for piezo. Lower frequencies will be the first to go. A simple improvement would be to try upping the resistor to 10 megohms or more. This may prove a problem with regular tubes (grid current bias effect) but could well work with these mini beasties as everything seems to be 10 times smaller.

So once again sorry to have been so blunt yesterday, I really was rushed.
Today I've taken some time off from fixing the Harley to try to be a bit more constructive.
I really enjoy lurking on this site at lunchtimes and genuinely admire what you guys (& gais) are doing.
I look forward to joining in more once I've paid my dues to the hand that feeds me.

Good luck with tube idea, Dunk. I'd give some serious thought to the 'battery belt' though.

[dunkelfalke]

This whole idea turns out to be much more complicated than anticipated. And way too expensive, too, making me kind of a sad panda.

[sumgai]

dunk,

I think that the real problem with your still-good idea is the heat that will be generated, both by the tube itself and by the "wasted" loss of voltage through that 1 watt resistor, not to mention what will be happening through the zener itself. I hadn't really thought about this aspect, and I should have, way back when you said that space was limited. Well, if the space available is that small, then heat really becomes a consideration of merit, and not just an afterthought.

The components may live OK in that kind enclosed environment, but the heat has to go somewhere, and I now foresee that it might cause problems, at least with your cavity cover, though probably not the wood itself, nor the outside finish.

Yes, this is turning out to be more complicated, but that's because we've hashed through several "second thoughts", a not-uncommon experience when designing things electrical or electronic. (Come to think of it, that would include pretty much any other discipline too.) At first blush, this will work, but the circuit being copied was intended for a box that had lots of air flow, thus heat was not a consideration. Your guitar doesn't have that luxury, hence.......

Keep at it if you wish, though, it's still a good idea. Hmmm, just had a thought - find or build another cavity cover so as to not destroy your original. Drill ventilation holes in it, and that should take care of the heat issue.

HTH



sumgai

[JohnH]

Why not just do a breadboard build of this, out of a guitar, first? This looks like a circuit where the right design will only be known after testing and tweaking. Just use batterys at least for the test, going back to the idea of seperate 1.5V battery for heater and 9V for main supply - and you can try two of them for 18V too. Try differnt anode resistors etc. Hence the test circuit is very simple and much is learnt.

Im intessted in these also, to see what could be achieved if you overdrive them with a JFET gain stage before.

John

[dunkelfalke]

I don't mind drilling holes in the cavity cover. It is the wooden parts I am not really willing to modify. And yes, I will build it on a bredboard anyway since I don't have any tools for etching circuit boards. I am thinking of a separate board for the preamp and a separate for the PSU so that I could still go with the battery possibility if all else fails (and right now it looks that way). Could I use a button cell for the filament heating? They are pretty small.

[sumgai]

Well, that was weird......

I was sitting in my chair with the laptop positioned appropriately, and suddenly, I went out like a light. Whoosh, Ozone-ville. Couldn't do a thing that needed concentration for several days.

Turned out I had a bad cold, not the flu, and I had taken some OTC cough syrup that reacted badly with my heart meds. Remind me not to do that again! All that happened, really, was that i was drunk, meaning mentally impaired (shut up, you jokers!*), but without the pleasant side effects of loopy giddyness. Was very comfortable, no after-morning headaches and such, but it sure retarded my output, I'll tell you that much.

Normally it takes me only a day or two to get over these things, but this year I see to have caught a bad break. Ah well, life goes on.

OK, back to business.......

dunk, what's the scoop - are you still up for this project, shall we carry on or what?


sumgai


* Especially you, D2o, I know your love for twisting a word or two to evil intent!

;D

[D2o]

I was just about to make a smartass reply to show my concern when I saw the asterisk and the notation. Shtink!

Oh well ... I guess I'll share a similar zonked out story or otherwise then.

1) 10 years ago ; combination of the flu and lyrin laren laerin lyrni larva ... &%! ... I lost my voice! ; several tylenol flu capsules and something else ... not booze, but I can't remember what ; phone rings and I'm thinking I can barely talk but it might be my better half ; remember that I said I am thinking ... my brain was fine, but when I tried my damndest to get the phone I could not move to save my life

2) 20 or 25 years ago ; exam time ; tired ; I heard about "wake-ups" caffeine pills ; they don't hit you all at once ... it was after about an hour that I suddenly realized that I hadn't blinked in about an hour

I suppose if you combined the I'd have been an "awake drunk".

Glad to hear you're feeling better.

D2o

[lpf3]

originally posted by sumgai

Whoosh, Ozone-ville. Couldn't do a thing that needed concentration for several days.

Pffffft..... I make a living in that state. ;D

Seriously, I had a little bout of something this weekend too, but much better now. Glad to hear you are too.

-lpf3

[sumgai]

Feb 28, 2010 22:17:33 GMT -5 lpf3 said:

I make a living in that state. ;D

Yes, and I wish I was back there - the taxes were practically low enough to actually keep some of my money for times of G.A.S.!

;D

[sumgai]

Feb 28, 2010 21:44:31 GMT -5 D2o said:

.......
20 or 25 years ago ; exam time ; tired ; I heard about "wake-ups" caffeine pills ; they don't hit you all at once ... it was after about an hour that I suddenly realized that I hadn't blinked in about an hour

First time your kid comes home and tells you "Daddy, I'm getting married", you're gonna do a repeat performance of that stunt!


(Trust me, I know about this one, four times over!)

[dunkelfalke]

Still waiting for the parts to arrive. Banzai sent them with DPD and that particular parcel service is one of the worst. What makes them even worse is the fact that it is nearly impossible to collect the parcel at their warehouse when you don't own a car because their warehouses are in the middle of nowhere.

Oh, by the way. If you need to take heart medicine, be careful with grapefruits. And get well soon!