I'm building a tube-based DI box, specifically for close-miked harmonica. Generally, players use something like a Shure Green Bullet, although sometimes they'll just use a 57 or 58. Either way, the mike is very close to the harp, and usually the player actually tries to make an air-tight seal with the mike. Therefore, the output is really high compared to most dynamic mike applications - more like a guitar output. Also, players usually like some distortion - again like a guitar - so they'll often just use a little practice amp and mike that to send to the PA.
The preamp/DI I've built is based on the Matsumin Valve-Caster, except using a 6111 submini tube instead of the 12AU7. There's also an FET output buffer to keep the PA input from loading the tube stage and changing the tone, and it serves as a transformerless balanced output as well.
This is my first attempt at something like this, and before I toss it into the wild next week at a blues jam, I'm hoping that the wizards here can point out some of the things I've done that need fixing. I've built it, including laying out and fabricating a circuit board, and it does work, but I'm sure there's room for improvement. A couple of things that come to my mind are:
1) I had to add R6 (2.2M) to positively bias the grid of the second stage to get any kind of symmetry at all; without it, the tube was sitting at cutoff all the time. (The 100k plate resistor is I believe around the recommended value, if there can be such a thing with only a 12V B+ supply.) Like I said, it works, but will this kill the tube, and why did I need it in the first place?
2) I dreamed up the FET output stage after finding that a typical PA input has a low enough impedance that it drastically affected the tone as the final volume was advanced. It was at this time that I decided to add the balanced output and make it a true direct box. The FET acts as a phase inverter with essentially unity gain (half of it is just a source follower; the bias point for the gate is established by creating a virtual ground for the bottom end of the final volume pot).
My question concerns the output coupling caps. As is, they should protect the FET from phantom power, but another circuit I saw (op-amp, not FET) emphasized that these needed to be non-polarized. Since phantom power is always positive with respect to ground, I don't really see why - what am I missing?
3) The circuit seems pretty quiet (the 6111 is dead-black silent!), but I am getting a bit of noise from the FET. Is there a better, but non-exotic, part to use, or something I'm doing to create the noise in the first place? (My low-level audio experience is more with bi-polar and op-amps - I'm also getting the feel of FETs with this project.)
4) And what else have I missed?
Thanks in advance for any feedback/criticisms!
~~
The preamp/DI I've built is based on the Matsumin Valve-Caster, except using a 6111 submini tube instead of the 12AU7. There's also an FET output buffer to keep the PA input from loading the tube stage and changing the tone, and it serves as a transformerless balanced output as well.
This is my first attempt at something like this, and before I toss it into the wild next week at a blues jam, I'm hoping that the wizards here can point out some of the things I've done that need fixing. I've built it, including laying out and fabricating a circuit board, and it does work, but I'm sure there's room for improvement. A couple of things that come to my mind are:
1) I had to add R6 (2.2M) to positively bias the grid of the second stage to get any kind of symmetry at all; without it, the tube was sitting at cutoff all the time. (The 100k plate resistor is I believe around the recommended value, if there can be such a thing with only a 12V B+ supply.) Like I said, it works, but will this kill the tube, and why did I need it in the first place?
2) I dreamed up the FET output stage after finding that a typical PA input has a low enough impedance that it drastically affected the tone as the final volume was advanced. It was at this time that I decided to add the balanced output and make it a true direct box. The FET acts as a phase inverter with essentially unity gain (half of it is just a source follower; the bias point for the gate is established by creating a virtual ground for the bottom end of the final volume pot).
My question concerns the output coupling caps. As is, they should protect the FET from phantom power, but another circuit I saw (op-amp, not FET) emphasized that these needed to be non-polarized. Since phantom power is always positive with respect to ground, I don't really see why - what am I missing?
3) The circuit seems pretty quiet (the 6111 is dead-black silent!), but I am getting a bit of noise from the FET. Is there a better, but non-exotic, part to use, or something I'm doing to create the noise in the first place? (My low-level audio experience is more with bi-polar and op-amps - I'm also getting the feel of FETs with this project.)
4) And what else have I missed?
Thanks in advance for any feedback/criticisms!
~~
Attachments
I'm afraid that your gain pot will become scratchy sooner or later, due to the cathode current it has to handle.
Best regards!
Best regards!
I suppose it's possible. Originally I wondered if the pot acted to change the operating point of the triode as well as the gain, and it does a bit, but not enough to be significant. It causes the plate to swing about 0.6V - that seems completely negligible, except that the B+ is only 12V to start with, and the plate seems to hang out around 8 volts.
Perhaps I'll add a 10 uF cap in series with the wiper. to block the DC. It is only about 5 uA, but like you said, over time... Thanks for noticing that!
~~
Cheers!
Perhaps I'll add a 10 uF cap in series with the wiper. to block the DC. It is only about 5 uA, but like you said, over time... Thanks for noticing that!
~~
Cheers!
Well now I do have a dilemma. Upon actually measuring the circuit, it turns out that the cathode voltage varies from 0 to .5 volts over the range of the pot, meaning there's really 10uA flowing through the pot instead of the 5uA I calculated. The other 5 uA must be coming from the grid, or possibly from the filament?
In any case, changing the operating point makes a significant difference in the gain, beyond that due only to the bypassing of the 50k cathode resistor. I get significantly more drive with a direct connection rather than just bypassing the pot with a cap.
The dilemma is that I can really use that extra gain. How much DC does it take to degrade a pot, and how fast? Is there some electrolysis effect or something, or just increased sensitivity to dirt and other irregularities? At the moment there is no problem (not even a significant low frequency level shift) when the pot is turned, but it is new.
Frankly, I'm sorely tempted to leave it as is, at least until problems start to occur....
~~
In any case, changing the operating point makes a significant difference in the gain, beyond that due only to the bypassing of the 50k cathode resistor. I get significantly more drive with a direct connection rather than just bypassing the pot with a cap.
The dilemma is that I can really use that extra gain. How much DC does it take to degrade a pot, and how fast? Is there some electrolysis effect or something, or just increased sensitivity to dirt and other irregularities? At the moment there is no problem (not even a significant low frequency level shift) when the pot is turned, but it is new.
Frankly, I'm sorely tempted to leave it as is, at least until problems start to occur....
~~
Last edited:
Well, duh - I've just answered my own question on the coupling caps. When there's no phantom power, the positive side has to go towards the FET; when phantom power is on, the positive side is away from the FET. So I replaced the 10k resistor across the balanced line with a 22k, and just used 1uF caps, giving a 3db point of around 7.5 Hz - that should be good enough for a harmonica.
please keep us informed as to how this works out for you.
as a fellow harp player I'm always interested in other options.
just building a new amp at the moment based on the lone wolf blues Co. 6l6se design but with an xtra gain stage and spring reverb. be about 10watts with balanced di out.
as a fellow harp player I'm always interested in other options.
just building a new amp at the moment based on the lone wolf blues Co. 6l6se design but with an xtra gain stage and spring reverb. be about 10watts with balanced di out.
I certainly will - the package seems to be coming together nicely, but I'm traveling for a week or so and won't be able to get back to it for that period.
Assuming this works out the way I hope, I also plan to make a related item. It will be essentially this pre-amp, linked to a ~20 watt solid-state amp with a 6" instrument speaker (the Jensen MOD6-15) in a mini practice amp (also suitable for miking on stage). The idea is that all the "tube sound" would occur in the pre-amp, and the power stage would be beefy enough to stay in its linear region.
~~
Assuming this works out the way I hope, I also plan to make a related item. It will be essentially this pre-amp, linked to a ~20 watt solid-state amp with a 6" instrument speaker (the Jensen MOD6-15) in a mini practice amp (also suitable for miking on stage). The idea is that all the "tube sound" would occur in the pre-amp, and the power stage would be beefy enough to stay in its linear region.
~~
- Status
- Not open for further replies.