Tube preamp inside a guitar

Well, there surely are other possibilities like putting the circuit into a plasticine block.



Nope, I was thinking about using a push/push pot for adding the piezo directly to the magnetic output, but after the volume pot so if I roll down the volume for magnetic pots only the piezo would sound.

Well, I've ordered the tube and the piezos already, so yes, I'll go ahead. Since the preamp will be a module, I'll also order the parts for the transistor preamp, just in case the tube preamp won't fit.

I'll try to model the preamp in the Electronic Workbench. Unfortunately I am a software developer and know almost nothing about electronics (Except of the Ohm's law) so I will really need as much help as I can get.

You are right, plasticine is a bad idea, it melts when heated and makes an ugly mess.

This sounds like an interesting project, Dunk. Onboard tube preamps have been done before, and successfully, so I don't think this is a fool's errand.

For a chunk of money, one can buy a Takamine acoustic with the the Takamine "Cool Tube" Preamp. Of course, size considerations are not as big an issue in an acoustic. But this unit was apparently designed to use with mag/piezo setups similar to what you are contemplating. No schematic that I found, but clearly this is a more complex circuit that what you are planning.

Takamine's description notes that:



It runs on 4 AA batteries, said to give 24 hrs of playing time.

I also found this:

Tung-Sol 5672 submini pentode

I don't know what type of tube the schematic you reference is calling for, or its size. But I saw discussion of using the above as an onboard tube pre.

These have a heater requirement of only 1.25V at 50ma, so would presumable be fine for battery operation. This is a tiny unit, (1.5" X .35"), but I don't know enough about tube amps (or preamps) to know if it would be suitable for your purposes. There are/were apparently similar Soviet-made units as well. Additionally, they do not require vertical mounting, as some tubes do, which is something to consider, since you would presumably lay the guitar down horizontally with the power on, sooner or later.

I have a sneaking suspicion that Some Guy out there has a considered opinion on the utility of one of these for such an application. ;D

In googling about, there were various mentions of a 60's- era Rickenbacker that supposedly had an onboard tube pre- but no one seemed to know anything about such a unit.

In a Strat-style guitar, if one sacrificed the trem for a hardtail, such a unit could probably be mounted in the trem cavity, and the backplate cover could be vented for heat dissipation (and to warm your belly in drafty venues!)

Rather than mounting it in a plasticine block, how about a small aluminum block for heat dissipation? Drill holes to mount it and use RTV to hold it in the block? Sort of like a tiny heat sink.

Just some food for thought, anyway.

Okay, I've modified the circuit somewhat using another one.
Is it correct? Any suggestions? I have got some serious doubts about it.

Don't be sorry. I am glad that you've confirmed my suspicions - that means that I finally understand a bit of electronics.

I've asked a clerk at Conrad (the German equivalent of Radio Shack) today what he would suggest for my situation. He was so happy that someone still works with tubes that he didn't try to sell me the PSU kit but instead sold me only the IC, printed out the circuit of the PSU and eliminated the parts I don't need. What do you think of it? Is it better than a Zener diode? If yes, could it be further simplified?

Thank you for the help, will order the missing parts this evening.

Certainly, be happy to.

Since you didn't link to the datasheet itself, nor the corresponding example schematic, I'll assume you're referring to the part numbers in your own schematic.

The tube itself is amplifying AC (your guitar signal), and that's getting back to the battery and the input of the LM317, via R3. To keep the input of the regulator happy, we need to filter this down as much as possible - that's C3's job.

R1 is a current limiter. Many simple regulator circuits rely only on this part, and have no filter cap (C2). In this case, C2 is used to, once again, filter out the amplified guitar signal from the regulator's Adj terminal. Without it, the DC voltage to the cathode/G3 would vary up and down in time with the signal, and that's not acceptable.

Transient response refers to the ability to follow the signal. In your case, it has to be large enough to handle all the possible frequencies that your circuit will likely see. I should think that 1µf would be sufficient for this task, but your diagram shows 10µf. That's a bit of overkill, but if you already have this value cap in your "junk pile", then go ahead and use it, it won't hurt anything.

C1 and C4 are quite different values, for a reason. C1 allows AC to pass to ground. Note the 47µf value - it effectively shorts all AC to ground, thus there is only DC on the cathode/G3 pin. I think this will affect the tone, but how much I can't predict. I'd say that you should experiment with the thing in the circuit, and again without it.... see which way sounds best to your ears.

(Advanced study: note that older tube pre-amps often used a 'bypass' capacitor in parallel with a resistor on the cathode. This was a "selective" AC grounding, meaning that certain frequencies were bypassed to ground, while others went through the tube just fine. I'd caution you not to get too obsessed here, because you can kill a lot of time, trying out various possible combinations of resistors and capacitors! My advice is to keep it simple at first, get the circuit working and sounding the way you like it, then maybe "hack" at it a little bit. )

Sometimes things go 'bump in the night', meaning that other signals get into your circuit, from outside souces. Things like AM radio stations are good source of this type of interference. C4 is intended to squelch that kind of noise. If you're environment is pristine, then you can eliminate this cap too. If you suffer 'CB' radio coming out of your speakers, this is the first place a technician works on, in order to get rid of that kind of crap.

There, I think that covered all of what you asked, anything else?

HTH




sumgai

I´m glad to see people with a lot of knowledge who are willing to share what they know in a forum on the internet!I agree with sumgai that using the LM317 regulator isn´t the optimal solution.I wonder if you have looked at regulators that are made for battery applications.You might find one that would fit your project.

Most interesting- those tubes are tiny, not much different from an old-time transistor!

Ive got a few questions (BTW Im a newbie at tube circuits - so I dont have any advice to offer, except It seems like fun!)

So what is the 1.5V supply for? - it seems to go to the cathode, or is it the heater coil? Could that be done with the forward voltage of two silicon diodes in series instead? Does the circuit really need C1 and C4 in parallel (is it about better rf suppression to have a small cap in parallel with a large one)? And what sort of gain is expected?

Onward!

cheers

John

We need sumgais to make the call on this stuff, so still speculating:

BTW. here is a datasheet for a Tung-sol version of the tube:
tubedata.tigahost.com/tubedata/sheets/127/6/6418.pdf

The main current in these circuits is the 10mA filiament current, which needs to be supplied at around 1.5V (the datasheet says 1.25V +/- 20%). I like the idea of two batteries. That way, you can supply the filament from a single AA battery, which has much more current/time capacity than 9V batteries, particularly if you use a lithium type like this:

data.energizer.com/PDFs/L91ULT_EU.pdf

These also have a very stable voltage output as they decay, so the regulator is not needed, and you could get at least 200 hours run time or more. Then you can use a 9V battery for the rest of the circuit. Theres much less power wastage overall, its a simpler circuit with fewer parts.

John

dunk,

1) Yes, if you so wish.
2) The values on your current drawing are good, but the 10µf C2 could probably go down to 1µf and still do the job. If you already have the 10µf and want to use it, put it in C1, in place of that 47µf - that thing's way overkill for your needs here.
3) Most anything but the cheap disc ceramic type. Tantalum's a good choice, and I think MKP will probably be your most cost effective choice.


~!~!~!~!~
limbe, dunklefalke's issue is that he's not always able to get parts cheaply. I can't blame him for wanting to use what he can buy locally without a lot of extra cost and a long waiting time (for shipping). If it weren't for that fact itself, I'd stomp on him until he relented and went with the zener diode. After all, the LM 317 is nothing more than a glorified adjustable zener!


~!~!~!~!~
dunk (redux),

Yes, you do need that 18Ω resistor. It does help to drop the voltage a bit, but it's also a current limiter, within reason. Leaving that out would be almost a guarantee that the tube will have a very short lifespan.


~!~!~!~
John,

Two forward diode junction drops? As compared to what? (As in, compared to exactly what point in the circuit?) The problem with using just standard diodes (reference one of dunk's earlier diagrams) is that the main voltage has to flow through them, and then through a resistor (in series) for the desired voltage to be obtained. But as the battery weakens, the voltage drop across the two diodes does not change, thus the ratio between the drop across them and the drop across the resistor changes. And of course, so too does the current flow change thusly. That's pretty much a definition of an unstable power supply, when observed over time.

C1 and C4 do different things. RF suppression is only one of the jobs, and some folks get away without that particular cap, others need it in place. In most cases, the two caps are mutually exclusive in their actions. I can see a case where C1 might not be necessary, but C4 is. Fortunately experimenting with this part of the circuit is easy, and it might be rewarding.

Sadly, tube circuits like this are open-loop, meaning that they don't yield to simple analysis for gain calculations. Gain is very much a factor of the tube itself, but it's often cloaked under the guise of "amplification factor". The basic formula works only after one has built the circuit, and then measured the input and observed output voltages. But while in the design stage, the engineer has to resort to guessing according to what the datasheet says will happen, given all the parameters presented. (i.e. If you apply X voltage, at Y amps, then gain factor Z will occur.... that sort of thing.) This particular aspect of designing tube circuits is one of the best things about closed-loop op-amps and the R1/R2 equation!


~!~!~!~
dunk (for the final time),

No room for another battery?? I call BS on that one. As I figure it, you've got ungefähr 7 liters of wood there, and surely you don't need absolutely all of it to be solid, do you? Route out the little bit more it will take to make room for a second battery, and you're all done. No more worries about voltage regulation - what's not to love here? ;D

Lithium rechargable? Now we know that you're planning on accessing the battery fairly often (every couple of months), so why not just install a second battery holder and be done with it? (Unless the nice Menschen at Conrad won't take the LM317 back and give you a refund, then I can see your wanting to stay on this course.)

HTH




sumgai

Actually I do. I won't touch the wood of this particular guitar, the guitar is too special and too rare to do that. Now electronics is a completely different beast altogether.




Sorry, I used the wrong word. What I meant was, the resistors and the capacitors have got maximum voltage and current. What values of both should I take?

Oh, by the way, I don't have that many spare parts in my junk pile (except for metal film resistors, they are plenty). The only two capacitors I've got are 47µF and 220µF ;D

P.S. for sumgai, JohnH and others who help me every time I ask for help: I realise that I probably sometimes sound harsh and rude, but it is unintentional. Germans are quite direct, Russians too, and I am a Russian who has lived the larger part of his life in Germany so I'm more or less the worst of both worlds. I really appreciate your help times and times again and am very thankful for it. I cannot count how many times you all have saved my sorry arse. So if you ever have the feeling that I seem to be ungrateful, well, I am definitely not.

dunk,

The maximum voltage ratings for anything you can buy are far higher than whatever you're going to be using within your guitar. In the past of decades gone by, you might have found capacitors with ratings lower than 15wvDC (working volts DC), but in the last decade or two, I think the lowest rating now being made is either 35 or 50wvDC. (Don't worry if they just say "vDC" or even just "v", the rest of it is assumed nowadays.)

Yes, C2 will be polarized. Connect the negative end to the nominal ground, and the positive end to the tube's cathode/G3.

HTH



sumgai

dunk,

My gut feel is that 100KΩ is too high, but every other circuit I found out there seemed to be using this value, and no one was unhappy.....

560Ω is probably too low, but you never can tell. One thing that this range of values will do is increase or decrease the distortion. By looking at the datasheet closely, I found that the specs allow for up to 10-12% of distortion, at the rated output of 0.22 watts. That's pretty far into "not a clean signal" territory, as far as I'm concerned. However, the overall tone may still be what you want, so if the distortion is too much, then increase the value of R3 even higher. This will reduce some of the overall gain, as well as reducing the distortion, so you may have to compromise on R3's value. Use what sounds good to your ears.

HTH



sumgai