Designing my first tube guitar preamp: a few questions

[jaywalker512]

Hello everyone, as the title implies, I am in the process of designing my own tube guitar preamp. I have never used tubes in a circuit before, so I am sure to have some general questions to ask as I go along in the design process to make sure I dont blow anything up/hurt myself. I have a general understanding of most things I'll be doing, and most of the questions I'll have will just ask for confirmation to my own ideas, with maybe a few asking for an explanation. So, here is my preliminary schematic as of now, with power supply and input gain stage:



The tubes I am using are 6247 Subminiature triodes I picked up on ebay. (got 6 of them for $13, not a bad deal I thought).

First questions are regarding the power supply:

1. My transformer is one I took from an old 25 watt SS amp that I broke (accidentally... of course). It was used as a step-down from 120v to 36v. I want to use it as a step up to give me the 250v plate voltage I plan to use, by dividing the mains voltage in such a way that It multiplies to the exact voltage I need. Seeing as how I am only designing a preamp, and not an entire amp, the total power demands on the transformer should be less. Yes the voltage will be higher, but the overall power demands should be much less. I have zero documented information about this transformer since it is quite old. Do you think it will be safe to use my transformer in this way?

2. In my power supply (before the transformer), do you think 5w rated resistors will be adequate to divide the mains voltage? I dont know how much current comes from the mains, only the voltage, so I chose this value just to be safe (and I hope it will be).

3. About the heater supply, is it alright to supply my heaters in this manner using a voltage divider since I dont have a transformer tap to use instead? The selected values should deliver about 6vdc at 200ma, well within the heater specs for these tubes. Also, it is ok to power heaters with DC voltage correct?


Those are really the only questions I have for now, though I will undoubtedly have more as my design progresses. Im pretty sure my input gain stage is as straightforward as it gets, pretty much center biased (slightly warmer than it could be, but still pretty much center). It allows about a 4vAC signal to go through without clipping, so I think it will work pretty well as an input gainstage (and considering all of my guitars have active pickups that can deliver a pretty large signal, this large headroom should prove useful). The plate resistor and voltage make my load line almost identical to the "normal" specifications as provided by the datasheet for these tubes. Tube datasheet can be found here if someone needs it:

http://frank.yueksel.org/sheets/138/6/6247WA.pdf

Very much thanks ahead of time to anyone who can answer my questions

[tubeswell]

Hi jaywalker512

Check out Merlin's excellent valve wizard site for designing tube amps here.

How to design valve guitar amplifiers

Without going into your circuit too much at this stage, I will note that having 1k2 as a voltage divider for your heater elevation is not a good idea. You don't want nearly all the B+ going to ground at that point (which it most surely will do with that load. So I suggest you try aiming for a voltage divider that is going to give you a total load of at least 220k. (That way it will act as a decent bleeder resistance path for discharging your filter caps safely at power off as well).

Also, having only one 10uF filter cap after the rectifier will probably leave too much ripple on your power rail. I suggest you use a CRC or CLC filter there.

[Robert M. Martinelli]

Hi jaywalker,
no disrespect intended, and I don't want to bust your bubble, but I think the design to have several flaws.

First of all using a voltage divider at the primary is out of question, as it's not safe IMHO, second, you're using a Graetz full wave bridge, so with 252 VAC at the secondary, you' ll have much more than 250 VDC ( some 355 VDC ) after the rectifier/filter cap. Third, using the +B rail for the heater supply makes for some serious problems as well....the 1.2 K dropping resistor should dissipate ( 250-6 )V * 0,2 A = some 49 Watts Apart from the enormous waste of power, you'd need a 100 W resistor to keep its temperature within reasonable limits.

You will make life much easier to yourself, and will spare yourself some serious headaches ( and it will result in a much better design in the end IMHO ) if you will use a couple of very cheap wall-wart ( 200 or 300 mAmps ) transformers by putting them side by side. ( "mirroring" them : P-->S-->S-->P ).

With two 120VAC--->6 VAC you'll have an "intermediate" voltage of 6VAC to use for the heater ( so that you can freely decide to use AC, DC, stabilizers, or whatever you like to supply voltage to the heater ) and 120 VAC out of the second transformer for your + B rail ( some 165 VDC after the rectifier/filter section ). Should you need a higher +B you can add a voltage doubler easily enough, as the triode's current demand is very low.

Hope this helps

Best regards

Bob

edit... I have just seen tw's post ( seems like we had another "simulpost" ), and I agree about the 10 uF filter cap to be not enough for proper filtering.

[BackwardsBoB]

Hi, Jaywalker. Be careful out there.

[Merlinb]

Hi Jaywalker,
Before you get too disheartened, your basic amplifier circuit is fine, it's just your power supply that needs work. You might want to look at some existing designs and rob a few ideas. The Real McTube is a popular place to start. It doesn't sound all that good as drawn, but with a little tweaking it can work well. Most importantly, the power supply works and could be copied! http://www.dogstar.dantimax.dk/tubes...s/mctube2l.gif

Depending on what you have available, you may prefer to start with slightly lower voltages. A 36V transformer could provide about 50Vdc,, which is enough to make some cool sounds with submini tubes without being quite so dangerous, and a separate 6V transformer could supply the heater.

Hope you get something from this!

[mooreamps]

I found an applications note for this tube. I hope this helps.

-g



NJ7P Tube Database Search

[jaywalker512]

Hi jaywalker512

Check out Merlin's excellent valve wizard site for designing tube amps here.

How to design valve guitar amplifiers

Without going into your circuit too much at this stage, I will note that having 1k2 as a voltage divider for your heater elevation is not a good idea. You don't want nearly all the B+ going to ground at that point (which it most surely will do with that load. So I suggest you try aiming for a voltage divider that is going to give you a total load of at least 220k. (That way it will act as a decent bleeder resistance path for discharging your filter caps safely at power off as well).

I have found his website to be quite helpfult so far, I would certainly buy the book if most of my funds werent going towards this preamp

I read around a bit after posting and found what you told me about my divider for the heaters. So I suppose the 200 milliamp current (as noted on the datasheet) on heaters is a maximum, and they can run with much less?

First of all using a voltage divider at the primary is out of question, as it's not safe IMHO

Would dividing the 400v secondary down to the proper voltage be a better/safer idea?

Another thing I was wondering about was my transformer, whether it will be able to handle this high voltage. It was originally used as a stepdown from 120-36, and in this situation would be stepping up from 120-400, much higher voltage than before. The leads coming out of the transformer suggest they can handle it (they say 300v and 600v), but im worried about the internal windings.

you're using a Graetz full wave bridge, so with 252 VAC at the secondary, you' ll have much more than 250 VDC ( some 355 VDC ) after the rectifier/filter cap.

This is what I expected as well, (its supposed to multiply the original AC voltage by about 1.414 correct?) however when I tested it with the transformer turned around delivering 36vac, after the rectifier the voltage went down to about 33vdc. What might cause that?


Finally it might help to note that I am hoping to achieve a high gain, hard rock/heavy metal tone from this preamp, so I will probably be using 4-6 total gain stages. Will this kind of sound still be achievable using a lower plate voltage than the original idea of 250v?

[Steve Conner]

Sorry dude, running transformers backwards like that just plain doesn't work. If you try to feed any winding with substantially more voltage than it's rated for, the core saturates and the transformer goes up in smoke.

The back-to-back arrangement already suggested by others is much more practical. You can salvage the transformers out of wall warts and the like: all you need is two 120v to 6v ones.

You can cascade stages for metal distortion, but it'll need some work to stop it sounding like a fart in a can.

[Albert Kreuzer]

Sorry dude, running transformers backwards like that just plain doesn't work. If you try to feed any winding with substantially more voltage than it's rated for, the core saturates and the transformer goes up in smoke.

Now THAT is something I know (don't ask...)


jaywalker: The 6247 needs 200mA heater current at 6.3V. If you are trying to get that from your B+, you still need 200mA. This is NOT a maximum. With "much less" the tube won't work.
1) Re-read Bob's post. 250V x .2A = 50W.
2) Google Kirchhoff's laws.

Cheers,
Albert

[jaywalker512]

Alright, so going with the suggested idea of using two transformers and stepping down to 6.3, then back up to 120, I have a few questions.

I checked out this transformer on mouser:

41FD200

120-6.2v @ .2 amps. Looks good for the heaters, and it would be alright to supply 4-6 of them in parallel like this?

6.2v * .2a = 1.24w, the max plate dissipation of the tubes is 1.25w, the heaters wont be a problem for this will they?

Ok next, I would still like to use 250v as my plate voltage (so that I can keep my circuit as similar to a usual 12ax7 preamp as possible, and borrow ideas from existing tube preamps), would this arrangement work:

[transformer 120-6.3]->[heaters in parallel]->[transformer 6.3-120]->[rectifier, 170vdc?]->[voltage doubler up to 340]->[voltage divider down to 250]->[tubes etc.].

EDIT: provided what I am reading is correct, I dont really need a divider or a rectifier. Am I correct in saying that a "full wave voltage doubler" doubles an AC current into a DC current, thereby rectifying and doubling? If that is the case, I can certainly be happy with doubling 120vac to 240vdc, plenty for my needs.

Disregard this part if the above is correct:This seems like an awful lot to go through just to supply power, but it seems to be the only option I can think of to get this voltage, short of buying an expensive transformer to give me the exact voltage right away. Will I still have enough current to supply the plates? With my current load line, each will only draw about 4ma.

And a question about the transformers on mouser in general, why are there so many with a primary of 115v (like the one i posted), and so few with 120v as the primary? Isnt the standard wall voltage in the US 120v?


Thank you guys VERY much for your help so far, please forgive me for being a tube-noob Also sorry if some of my questions seem stupid or redundant, I just want to make absolutely sure I am doing things right.

[tubeswell]

I have found his website to be quite helpfult so far, I would certainly buy the book if most of my funds werent going towards this preamp

I read around a bit after posting and found what you told me about my divider for the heaters. So I suppose the 200 milliamp current (as noted on the datasheet) on heaters is a maximum, and they can run with much less?

Wait! - I thought you were intending to use the divider to elevate you heaters from the B+ rail, not to create a supply for your heaters!

What the other have already said - you need a separate 6.3VAC supply for your heaters (if you are running them in parallel. And with that method it is common to use an artificial ground reference (or heater winding CT - if your PT has one on the 6.3V winding) to elevate your heaters by between 6V to 90V or so, in order to eliminate hum, or to keep the heater-to-cathode voltages under the manufacturers rated specs. Hence my confusion at what your schematic showed).

Heaters will typically draw far more current than the tubes' plates will BTW.

[Robert M. Martinelli]

Would dividing the 400v secondary down to the proper voltage be a better/safer idea?

Hi Jaywalker,
I won't go again over my other suggestions, as they've been confirmed in their validity by some of our fellows "solder fume breathers", way more experienced and educated than I will ever be.

I just want to answer the ( seems to me polemical ? ) above question of yours.

If you do a simple search across the internet or even this forum, you'll see that no-one uses your system....do you think generations of designers haven't been able to thoroughly analyze already all the pros and cons of any possible PS arrangement for tube amps/preamps?

If you use a dropper or a divider at the primary, apart from the inherent risks of such practice, you need to keep the TOTAL current into account, including heater(s) current ( which can be as high as 5 or 6 amps in a 100 W amp ).

While it's true the voltage at the +B rail can be higher, the currents at the +B rail are usually pretty low, less than 1 Amp in a 100 W output stage ( where droppers are not required, with the exception of the PI stage ), only a few mAmps in a tube preamp/tube preamp stage, so using droppers there doesn't require any 50 W resistor.

Are you designing a preamp or an electrical stove to cook your food or heat your home during the next winter?

Droppers/dividers are rightfully widespread and simple to design, ( as long as you keep the basics in mind; when you design a divider, you need to worry about its dissipated power too... ) but they're practical to use only up to a point....when you have to throw dozens of watts out of the window to supply 1,2 W for a heater supply like in your case, I would not define that as "good designing practice" and I'd try something else.

You also have to keep in mind that transformers don't perform the sole function of giving you the voltage/current you need for a certain design of yours, they also add protection and insulation from mains. By putting the divider at the primary, you're dealing directly with mains voltage/current with all the possible risks involved in this practice.

20 Years of experience in designing industrial plants have taught me that a good designer ALWAYS takes the "worst case" into account, because sh!t happens.

Hope this helps

Best regards

Bob

[Robert M. Martinelli]

120-6.2v @ .2 amps. Looks good for the heaters, and it would be alright to supply 4-6 of them in parallel like this?

6.2v * .2a = 1.24w, the max plate dissipation of the tubes is 1.25w, the heaters wont be a problem for this will they?

Thank you guys VERY much for your help so far, please forgive me for being a tube-noob Also sorry if some of my questions seem stupid or redundant, I just want to make absolutely sure I am doing things right.

Hi Again Jaywalker,
1 - If a single heater needs .2 Amps, paralleling them would add their current requirements, this means you'd need .8 Amps for 4 tubes of the same type and 1.2 Amps for 6 of them.

2 - You're confusing the heater's power requirements ( 6.3V - 0.2 Amps = 1.26 W ) with the plate maximum dissipated power. They're not the same thing:

-The heater needs a certain amount of power to indirectly heat the cathode and cause thermionic emission from the latter. ( in older tubes, the heater WAS the cathode and these are referred to as directly heated ).

-The plate, which can be ( conventionally ) thought as the collector of the electrons emitted by the cathode, can dissipate only up to a certain amount of power ( without being damaged ) while the tube is performing its function, and, in a triode, just like in any other circuit, the dissipated power is expressed by the voltage drop across the tube multiplied by the current flowing through it.

Hope this helps

Best regards

Bob

[Steve Conner]

120-6.2v @ .2 amps. Looks good for the heaters, and it would be alright to supply 4-6 of them in parallel like this?

For 6 tubes in parallel, you'll need a transformer that outputs 6.2v (or 6.3 or whatever) at 6x 0.2 amps, that's 1.2A. Add on some more because it also has to power the second transformer that's generating the B+. So we're looking for a transformer that puts out 6.3V at about 2 amps.

The reason why they're specced as "115V" is because that's still the USA's nominal mains voltage. It's 120 or more in practice, but on paper it's still 115. The difference isn't really worth worrying about. We have the same thing here, the nominal mains voltage in all of Europe is 230V, but if you stick a DMM into a wall socket in our lab, it actually measures between 240 and 250.

Yes, if using 120:6V transformers back-to-back, you'll need a voltage doubler so that the final B+ is high enough. From 115V AC, a bridge rectifier will give 160V DC, a doubler 320. However, you can commonly find transformers with dual primaries: two 115V coils that you connect in parallel for 115V line, or in series for 230V line. This allows the same transformer to be sold worldwide.

If you buy a pair of these, you can wire the step-up one for 230V out and get 300-320V B+ from a bridge rectifier. You can then drop the 320 down to 250 using dropper resistors, as commonly done in old tube amps. (It's OK to use droppers sometimes, but not to power your heaters off the B+! :-) If you really must do that, then connect all of the heaters in series.)

Or, if you choose the step-up transformer with something other than 6.3V windings, you can alter the voltage that way. Let's say you chose a 9V:230V one. Then the B+ would be 230*1.4*(6.3/9) = 225V.

Or just run it all off 320, some people think that higher voltages give more aggressive distortion.

[jaywalker512]

Alright, so in following with everyones suggestions and doing some experimentation of my own, Ive come to this idea for the power supply:



First transformer drops to 6.3 (this transformer: 41FD030) to power the heaters. Then second transformer bumps it back up to 120, and should have plenty of current left over to double up to 240 and power the plates. I answered my own question about whether or not a voltage doubler rectifies as well, and it does. I wired up a voltage divider on my 36v transformer to give me 8vac, built a voltage doubler and sure enough got about 16vdc on the other end. So this SHOULD satisfy everythings power needs (I hope!) and hopefully I can move on to the gain stages Though one thing I am still not sure of is the 3a current coming from the first transformer to the heaters. Will the the heaters only take what they need or will all 3a flow through them, killing them. If the latter is the case, a certain value resistor in series with their power, lowering current to whatever they need would correct this?