Power transformer for tube preamp/DI

[Keppy]

I'm designing a tube bass preamp/DI because I'm sick of engineers telling me to plug in direct. My particular concern at the moment is a power transformer. The ones I've found are all spec'd for 6.3v on the heater winding under load in a particular amp. Should I worry about the heater voltage being a bit high, since without the power section and inverter I'll have a much lower load? Is there somewhere I should look for a medium-power transformer (more than a stombox, much less than an amp)?

[thelonious]

Is there somewhere I should look for a medium-power transformer (more than a stombox, much less than an amp)?

Do you know how much current you're expecting to draw? This Edcor will do 284V @ 22mA and 12V @ 1.6A for heaters. http://www.edcorusa.com/p/594/xpwr006

If you use a scavenged PT and your heater voltage is high, you can drop it with big diodes as in one of the illustrations here: http://www.valvewizard.co.uk/heater.html ...but that's kind of a band-aid solution. Obvs getting a suitable PT is better.

Also, Doug has a toroidal PT in stock here in the store, but that takes 12VAC primary input and might be too small...?

[Keppy]

Thanks for the reply! I was just reading the Valve Wizard page and I think I'll be okay, as the transformers I was looking at were below 6.9v unloaded, making them within the 10% that he suggests.

As far as the toroidal transformer is concerned, the reason I'm leaning away from it is that the Ampeg amps that I favor and am looking at for ideas have a higher voltage on the preamp tubes, ~400v. For that reason I was looking at some of the Hammond amp transformers. They're probably overkill, though, since I won't need anywhere near the power they can provide. I'll look around that Edcor site though, it looks pretty promising. Thanks again!

[eleventeen]

You can always use a voltage doubler (to get your 400 volts, if needed) for the HV portion of the power supply. Preamp tubes use miniscule amounts of current, 1-2-3 ma (per side) of a 12AX7.  That's almost nothing.

[Keppy]

You can always use a voltage doubler (to get your 400 volts, if needed) for the HV portion of the power supply. Preamp tubes use miniscule amounts of current, 1-2-3 ma (per side) of a 12AX7.  That's almost nothing.

That's an excellent point. I forgot about AC doubling. It would be awfully nice to use a small transformer like that.

[terminalgs]

If you design a circuit that uses interstage transformers, or a 4:1 type transformer to provide XLR balanced out, be sure that the frequency range of the transformer you select meets your low frequency requirements for bass (freq. range response, be it high or low, is where the money come in on those small interchange, mic, and small output transformers).

[Keppy]

I am a bass player who routinely plugs into a 16 channel mixing board.  Therefore, I am interested in what you are doing. 
a few questions,
1.  Are you playing with active or passive pickups?

Either. My bass has an onboard preamp with a selector for active/passive.

2.  Is phantom power available for DI?

Yes, but I can't count on having enough current for the heaters, so I'm planning on AC power with transformer balanced output to the mixer.

3.  Will your preamp/DI, be tube SS or hybrid? 

All tube. The point is to get at least some amount of tube amp sound when plugging into the mixing desk. When I play amped, I use an SVT Classic, so I dig those kind of tube bass sounds.

One comment,  do a search for "Selecting a DI box for bass", Janis provides some discussion that is interesting. 

Thanks, I'll check that out. I'm actually primarily motivated by the preamp portion, but I figured a DI output would suit my purpose.

Jensen Transformer have the schematics and parts you are looking for.  As a service to Doug, cross check the parts.  I would consider using Doug as the source of parts. 

I've seen the Jensen schematics and I plan to order parts from Doug. In fact, his Tube Pedal is very close to what I want to do as far as being a 4-triode preamp, so I know he sells what I need, except perhaps the DI transformer.

If you design a circuit that uses interstage transformers, or a 4:1 type transformer to provide XLR balanced out, be sure that the frequency range of the transformer you select meets your low frequency requirements for bass (freq. range response, be it high or low, is where the money come in on those small interchange, mic, and small output transformers).

Yeah, everything I've read says not to skimp on the transformers. I'll probably spring for a Jensen, and the ones I'm looking at all go down to at least 20Hz and sometimes lower.


Thanks for the help guys! Since there's some interest I'll try to keep posting updates as I go.

[Keppy]

Here's what I've got so far. I'd like some feedback before I order parts, as this is my first tube project.
The first three stages and tone stack are from Ampeg schematics, the cathode follower is from the Valve Wizard site, and the power supply is from Doug's Tube Pedal.


Questions:

1) Is this safe? There is no earth ground connection using a 12v AC transformer, and this thing will have 240v inside. Doug's tube amp pedal uses this power supply, and that project has been on this site for nearly a decade. Apparently it's worked out so far, but can someone tell me why that's okay?

2) I want a cathode follower at the end to help in driving low-impedance loads. The info I read on these stages makes me believe I might have to elevate the heater supply, but I have yet to see a circuit actually do that. Thoughts? Changes? DC vs AC? 12AX7 vs 12AU7 vs...? I'm leaning towards a 12AU7 for the second tube (first is 12AX7).


I'm sure I had more questions, but I can't remember them at the moment. Feel free to point out what I should have asked.

[eleventeen]

To begin, since it *might* be desirable as you say to elevate the heaters, you may wish to isolate the "7812 zone" ground from signal ground during your build so that it can be done easily. And yet those regulators want very short and robust connections to ground and filter caps....at least this has been the case in the past, maybe they are improved now. So that portion of the ckt will want FAT wires, don't wire with AWG 24.

Is your plan to insert a hi-voltage/high power zener in the ground leg of the 7812 to fool the regulator? I have seen this done but not experimented with it myself. Looks to me like the last cathode will in fact be at a fairly high voltage level.

Does not solving this issue for the last tube impose the same problem, in reverse, for the earlier tubes, whose cathodes are much closer to ground volts?

I don't know the answer. I'm just feeding your thought process, hopefully constructively.

 

[sluckey]

It's safe since you are using transformers. The line is isolated from your 'ground' circuit.

I would not worry about elevating the heaters. Lot of amps with cathode followers don't, ie, 5F6A, Marshalls, etc.

[HotBluePlates]

... The first three stages and tone stack are from Ampeg schematics ...

Can you point to the Ampeg schematic where V2A is lifted from? I'm curious to figure out why they split the cathode resistor and bootstrapped the grid resistor on a stage which doesn't seem like it would benefit from that.

Also, I might add an extra dropping resistor and 22uF filter cap between V1A and V1B (between C12 and R3), but that probably isn't necessary. But every little bit of filtering helps, especially if you're gonna feed a console/mixer/recorder.

Nichicon makes some high quality, small-sized axial and radial caps. I would look for caps rated 350vdc or more, given 230vac * 1.414 = 325vdc.

[Willabe]

I've been buying Panasonic radial filter caps. 22uF@450v, high temp, high ripple, long life (Endurance: 105 °C 8000 h to 10000 h), small size > $1.67 for 1 to 9, $1.14 for 10 to 99.

http://www.mouser.com/ProductDetail/Panasonic/EEU-EE2W220S/?qs=sGAEpiMZZMtZ1n0r9vR22dN7zJS18SGVPwHDeJcuj2Q%3d


              Brad     

[Keppy]

Can you point to the Ampeg schematic where V2A is lifted from? I'm curious to figure out why they split the cathode resistor and bootstrapped the grid resistor on a stage which doesn't seem like it would benefit from that.

Some versions of both the SVT and the B15 used such a stage after the tone stack (see attachments). I suspect it was done to keep the input from loading down the tone stack, particularly since the tone stack is driven from a gain stage rather than a follower, but that's really just a guess.

Also, I might add an extra dropping resistor and 22uF filter cap between V1A and V1B (between C12 and R3), but that probably isn't necessary. But every little bit of filtering helps, especially if you're gonna feed a console/mixer/recorder.

True. I yanked the power supply straight from Doug's tube project, since it's known to be good. I'll probably just try it the way he did it, but making sure to leave room in case I need to change that.

Nichicon makes some high quality, small-sized axial and radial caps. I would look for caps rated 350vdc or more, given 230vac * 1.414 = 325vdc.

Thanks for the tip on the caps. As for the voltage, Doug's description on the power transformer says it creates 280v DC under load with this power supply (and a similar application), but that's still high enough to warrant 350v caps.

I've been buying Panasonic radial filter caps. 22uF@450v, high temp, high ripple, long life (Endurance: 105 °C 8000 h to 10000 h), small size > $1.67 for 1 to 9, $1.14 for 10 to 99.

http://www.mouser.com/ProductDetail/Panasonic/EEU-EE2W220S/?qs=sGAEpiMZZMtZ1n0r9vR22dN7zJS18SGVPwHDeJcuj2Q%3d

Thanks to you too!

It's safe since you are using transformers. The line is isolated from your 'ground' circuit.

That makes sense, but shouldn't the chassis have a safety ground in case of accidental contact with internal voltages? I'm not trying to argue, I just don't get it yet.

I would not worry about elevating the heaters. Lot of amps with cathode followers don't, ie, 5F6A, Marshalls, etc.

Thanks! Does the follower I lifted from Merlin's site seem like a good choice for this application? I chose it for it's low output impedance compared to followers in the Ampeg schematics that were meant to drive tube stages.

[Keppy]

To begin, since it *might* be desirable as you say to elevate the heaters, you may wish to isolate the "7812 zone" ground from signal ground during your build so that it can be done easily. And yet those regulators want very short and robust connections to ground and filter caps....at least this has been the case in the past, maybe they are improved now. So that portion of the ckt will want FAT wires, don't wire with AWG 24.

Is your plan to insert a hi-voltage/high power zener in the ground leg of the 7812 to fool the regulator? I have seen this done but not experimented with it myself. Looks to me like the last cathode will in fact be at a fairly high voltage level.

Does not solving this issue for the last tube impose the same problem, in reverse, for the earlier tubes, whose cathodes are much closer to ground volts?

I don't know the answer. I'm just feeding your thought process, hopefully constructively.

Sorry, I missed this post before my last reply. I didn't have a concrete plan to elevate the heaters. I thought it would be easier to do with an AC heater supply, but I don't know if the 12vAC supply Doug sells puts out a real 12v or a higher voltage that sinks down to 12v under some specific load. I wasn't thinking too hard about that as I was hoping someone would tell me I don't have to worry about it. As Sluckey says, it seems that most manufacturers don't bother.

However, if I DO have to elevate the heaters, there will be plenty of room to do so. Since the 12AX7 and 12AU7 have a 180v max heater-to-cathode voltage, there's plenty of wiggle room to get the voltage up to suit the last stage without messing up the previous stages. I just don't want the hassle of trying to elevate a DC heater supply (or fiddling with an uncertain 12vAC supply) if it can be avoided.

[sluckey]

That makes sense, but shouldn't the chassis have a safety ground in case of accidental contact with internal voltages?

The only internal voltages you have are 12VDC and the B+ rail. Don't you have the negative side of those supplies connected to chassis? If so, any internal voltage that touches chassis will burn up some internal component. No danger to you though.

But if you must have a dedicated safety ground, mount your wall wart transformer (or another chassis mount 12VAC transformer) to the chassis and use a three prong power cord.

Your CF looks fine to me.

About elevating the heaters... Your schematic shows your 12vdc heater supply referenced to ground (zero volts). That's the same ground that your B+ is connected to. So, before you could 'elevate' your filaments, you first must float the filament supply. IOW, connect the negative side of the bridge, negative side of C1, negative side of IC1, and one side of the filament string together, BUT DON'T CONNECT TO GROUND. Once the entire string is floating you can elevate to some positive voltage. But really, just forget about it.

[kagliostro]

Your solution for the PS is OK

but if ~380v DC can be enough for your further realizations give a look to this toroidal (fits U1 rack)

Primary 230V, secondary 270V @0,08A and 14V @2A

Diameter: 78mm - Height: 32mm

http://www.musikding.de/Toroidal-transformer-230V-270V-14V

from the 14v winding is simple to obtain the 12.6v using a pair of resistor (one for each branch)

K

[HotBluePlates]

Can you point to the Ampeg schematic where V2A is lifted from? I'm curious to figure out why they split the cathode resistor and bootstrapped the grid resistor on a stage which doesn't seem like it would benefit from that.

Some versions of both the SVT and the B15 used such a stage after the tone stack (see attachments). I suspect it was done to keep the input from loading down the tone stack, particularly since the tone stack is driven from a gain stage rather than a follower, but that's really just a guess.

The B15 used that particular stage as a self-biased cathode follower, and the lower resistor (connected to ground) is 47kΩ.

The SVT Classic uses the stage to power what looks like a Midrange control, which is in a feedback loop from the plate output of that tube to the cathode output of the same tube. The resistor connected to ground could be 2.2k (like yours), but it would have almost no output from the cathode, and so is very likely a 22kΩ resistor.

... I suspect it was done to keep the input from loading down the tone stack, particularly since the tone stack is driven from a gain stage rather than a follower, but that's really just a guess. ...

The grid resistor of the stage looks like 470kΩ + 2.2kΩ = 472.2kΩ, so a lower than normal input resistance (which would be more of a load on the tone circuit).

Unless that resistor to ground is a large value, no significant a.c. voltage is developed across it, and so the bootstrapping effect never really happens.

If you make your R13 large enough for a bootstrapping effect, gain in the stage will be at/near unity, and you might as well move the output to the cathode instead. With a plate output under those conditions, you get no gain and high output impedance, so the stage becomes not worth having.

[Keppy]

The B15 used that particular stage as a self-biased cathode follower, and the lower resistor (connected to ground) is 47kΩ.

You're right, I copied the stage from the SVT. I missed the difference when I looked at the B15 schematic.

The SVT Classic uses the stage to power what looks like a Midrange control, which is in a feedback loop from the plate output of that tube to the cathode output of the same tube. The resistor connected to ground could be 2.2k (like yours), but it would have almost no output from the cathode, and so is very likely a 22kΩ resistor.

The output of that stage is from the plate, so I'm not sure what problem you're addressing.

The grid resistor of the stage looks like 470kΩ + 2.2kΩ = 472.2kΩ, so a lower than normal input resistance (which would be more of a load on the tone circuit).
Unless that resistor to ground is a large value, no significant a.c. voltage is developed across it, and so the bootstrapping effect never really happens.

Now this I think I understand. It sounds like I should convert that stage to a more normal gain stage, since the bootstrapping isn't providing what I thought it was. Should I just duplicate the second stage, but using a 12AU7? Thanks for the help, you pointed me to a few things I missed in the schematics I was referencing.

[Keppy]

If ~380v DC can be enough for your further realizations give a look to this toroidal (fits U1 rack)

Primary 230V, secondary 270V @0,08A and 14V @2A

I'm in the US, so I'm plugging into 120v. Sorry, I should have said that. Thanks for the suggestion though!

I also should have said before now that I decided to favor a small transformer over my earlier preference for 400v B+, hence my adoption of Doug's power scheme.

[HotBluePlates]

The SVT Classic uses the stage to power what looks like a Midrange control, which is in a feedback loop from the plate output of that tube to the cathode output of the same tube. The resistor connected to ground could be 2.2k (like yours), but it would have almost no output from the cathode, and so is very likely a 22kΩ resistor.

The output of that stage is from the plate, so I'm not sure what problem you're addressing.

The plate and the cathode. See C12 (10uF) going to the Mid pot (P3), which itself connects to an inductor (L1)? That is an active midrange control, which works because the 2 outputs are essentially a feedback loop.

So the stage does something different than what you're trying to do in your DI circuit.

... It sounds like I should convert that stage to a more normal gain stage, since the bootstrapping isn't providing what I thought it was. Should I just duplicate the second stage, but using a 12AU7? ...

I'd still look to a known bass preamp, rather than (me) make random guesses as to what to include.

I notice that the old Alembic preamps were pretty much just 2 gain stages, and looked exactly like the dry preamp of a Fender blackface guitar amp.

The B15N is also only 2 stages and pretty straightforward. The B15S only adds a 3rd stage (between the original 2) because there are extra tone networks to allow selection of guitar, flat or bass voicings (the gain stage makes up for some of the loss imposed by these extra networks).

So it seems you could keep a 3rd gain stage, or not.

[Keppy]

The only internal voltages you have are 12VDC and the B+ rail. Don't you have the negative side of those supplies connected to chassis?

Yes I do, but the chassis itself is not earthed. The power supply and output are both transformer isolated.

But if you must have a dedicated safety ground, mount your wall wart transformer (or another chassis mount 12VAC transformer) to the chassis and use a three prong power cord.

That's a good idea, although at that point I wonder if it's easier just to get a transformer that runs directly off of 120v AC, rather than dropping to 12v and then raising it back up.

Thanks for helping me think through this. I've read a whole lot of high voltage warnings at this point, and they've made me want to really understand this before I get started.

[Keppy]

... It sounds like I should convert that stage to a more normal gain stage, since the bootstrapping isn't providing what I thought it was. Should I just duplicate the second stage, but using a 12AU7? ...

I'd still look to a known bass preamp, rather than (me) make random guesses as to what to include.

I notice that the old Alembic preamps were pretty much just 2 gain stages, and looked exactly like the dry preamp of a Fender blackface guitar amp.

The B15N is also only 2 stages and pretty straightforward. The B15S only adds a 3rd stage (between the original 2) because there are extra tone networks to allow selection of guitar, flat or bass voicings (the gain stage makes up for some of the loss imposed by these extra networks).

So it seems you could keep a 3rd gain stage, or not.

Yes, I'm unsure if I need the 3rd gain stage, but I put it in for the following reasons:

1) The designs I'm drawing from don't have gain stages immediately following the tonestack, but they do lead to an inverter which to my understanding has some additional gain.

2) The third stage allows for a master volume that is isolated from the tonestack and the output, though it does not have to be a gain stage to accomplish this.

3) Since I won't get any color from an inverter, power amp or speaker cabinet, some extra preamp drive might be advantageous. It's there if I need it, and I can always turn it down.

4) Since I'm using twin triode tubes, I have four stages to use up. I don't know what besides a gain stage would be beneficial. I did consider a DC-coupled cathode follower in place of the second gain stage as in the attached schem, but even that design had two gain stages before the tonestack.


Basically, since I have that fourth triode, that gain stage stays unless I find a better way to use the stage. The question is what type to use, since I don't have a clear example to draw from at this point. I'll see if I can find some good schems to consider. Thanks again!

[HotBluePlates]

Yes, I'm unsure if I need the 3rd gain stage, but I put it in for the following reasons:

1) The designs I'm drawing from don't have gain stages immediately following the tonestack, but they do lead to an inverter which to my understanding has some additional gain.

True.

But an inverter takes a 1-2v input signal and smacks the output tubes with a 25-55v signal. You won't have output tubes to drive, and line-level is typically 1v or less.

3) Since I won't get any color from an inverter, power amp or speaker cabinet, some extra preamp drive might be advantageous. It's there if I need it, and I can always turn it down.

So this is really your best reason to think about it, as you're right that you can always turn down.

If you want the stage to have gain, replace the 470kΩ with a 1MΩ to ground. Use one resistor in place of R12, R13 and pick a value between 1-3kΩ.

[Keppy]

Yes, I'm unsure if I need the 3rd gain stage, but I put it in for the following reasons:

1) The designs I'm drawing from don't have gain stages immediately following the tonestack, but they do lead to an inverter which to my understanding has some additional gain.

True.

But an inverter takes a 1-2v input signal and smacks the output tubes with a 25-55v signal. You won't have output tubes to drive, and line-level is typically 1v or less.

Yeah, I'm worried less about output volume than I am about any distortion contributed by the combination of inverter/power tubes/speakers (which doesn't sound the same as a single triode stage, I know).

3) Since I won't get any color from an inverter, power amp or speaker cabinet, some extra preamp drive might be advantageous. It's there if I need it, and I can always turn it down.

So this is really your best reason to think about it, as you're right that you can always turn down.

If you want the stage to have gain, replace the 470kΩ with a 1MΩ to ground. Use one resistor in place of R12, R13 and pick a value between 1-3kΩ.

Awesome! Thank you again for your help.