I'm designing a tube pre-amp for bass (still a noob!). The only thing missing is a balanced direct output that's to be connected to e.g. the PA or recording equipment. It has to be connected somewhere at the first stage of the preamp (see pic), but I fail to properly do this.
All my attempts to use a parafeed transformer (10k:1k) screwed with the signal; 1. due to the impedance of the transformer changing the cathode load with different frequencies, or 2. when I tried to 'isolate' the transformers with a 10k series and 1k parallel (to maintain low impedance for the transformer) resisitor, it decreased the level way too much.
I'm probably missing something. Perhaps a 10k:1k parafeed transformer is wrong or there is a trick I'm not aware of... I'm out of ideas.
So, my question in short is, how can I implement a line level balanced DI in the given schematic.
I'd really like to keep the basic design of this stage, because it sounds awesome. I know V1A is missing a bypass cap and V1B's grid is only at -0,5V, but this is all with good reason.
All my attempts to use a parafeed transformer (10k:1k) screwed with the signal; 1. due to the impedance of the transformer changing the cathode load with different frequencies, or 2. when I tried to 'isolate' the transformers with a 10k series and 1k parallel (to maintain low impedance for the transformer) resisitor, it decreased the level way too much.
I'm probably missing something. Perhaps a 10k:1k parafeed transformer is wrong or there is a trick I'm not aware of... I'm out of ideas.
So, my question in short is, how can I implement a line level balanced DI in the given schematic.
An externally hosted image should be here but it was not working when we last tested it.
I'd really like to keep the basic design of this stage, because it sounds awesome. I know V1A is missing a bypass cap and V1B's grid is only at -0,5V, but this is all with good reason.
Last edited:
Just take the signal off of C2 and run it across a high valued pot.
Like a 100k or more this way it won't load up the output changing the sound of the preamp.
And then run the signal into a unity gain opamp buffer followed by a gain of 1 inverting stage to get your balanced signal.
You can omit the 100k pot at after C2 as long as the output signal doesn't swing more than 12V to 13.5V peak.
I can draw up a schematic of what I mean if you like.
jer
Like a 100k or more this way it won't load up the output changing the sound of the preamp.
And then run the signal into a unity gain opamp buffer followed by a gain of 1 inverting stage to get your balanced signal.
You can omit the 100k pot at after C2 as long as the output signal doesn't swing more than 12V to 13.5V peak.
I can draw up a schematic of what I mean if you like.
jer
Last edited:
Man, thanks a lot! That's more than I could ask for.
First I was a bit hesitant to use 'silicon' in my tube pre-amp, but thinking about it, your idea makes a lot of sense. Here's my thought and correct me if I'm wrong; An impedance/parafeed transformer will have WAY more distortion then a jfet opamp with massive feedback. The input impedance is very high, so C1 (C2 in my schematic) can be reduced significantly and I guess output impedance will be very low, but that really is a guess, because I have only little knowledge of solid state electronics.
So thanks and I will probably pursue this suggestion.
First I was a bit hesitant to use 'silicon' in my tube pre-amp, but thinking about it, your idea makes a lot of sense. Here's my thought and correct me if I'm wrong; An impedance/parafeed transformer will have WAY more distortion then a jfet opamp with massive feedback. The input impedance is very high, so C1 (C2 in my schematic) can be reduced significantly and I guess output impedance will be very low, but that really is a guess, because I have only little knowledge of solid state electronics
.So thanks and I will probably pursue this suggestion.
Last edited:
I believe that the value of C2 will just set the lowest frequency that it can output.
I can try some different values and check the effects for you.
I think the biggest factor of using the transformer is the reflected impedance loading of the tube and the fact that if it is high enough the output voltage will be very high and causing the transformer core to saturate very quickly on the lower frequency's.
jer
I can try some different values and check the effects for you.
I think the biggest factor of using the transformer is the reflected impedance loading of the tube and the fact that if it is high enough the output voltage will be very high and causing the transformer core to saturate very quickly on the lower frequency's.
jer
Last edited:
Yep !!
very cool,I can tell you this as you lower the value of the capacitor you get less bass.
This is were you may have to do some plug and play to find the values that you will be happy with.
But if you give me some values that you have in mind or think that you like I can put those in the simulation for you and show you the curve.
With the values shown the low frequency knee is about 5Hz.
jer
P.S With the values shown the high frequency drops off as well starting around 5Khz to 8Khz.
With a value of .1uf for C2 it shows a flat response from about 10Hz on up past 20kKhz.
The frequency analyses part of this program is not that great but at least it gives a good place to start.
very cool,I can tell you this as you lower the value of the capacitor you get less bass.
This is were you may have to do some plug and play to find the values that you will be happy with.
But if you give me some values that you have in mind or think that you like I can put those in the simulation for you and show you the curve.
With the values shown the low frequency knee is about 5Hz.
jer
P.S With the values shown the high frequency drops off as well starting around 5Khz to 8Khz.
With a value of .1uf for C2 it shows a flat response from about 10Hz on up past 20kKhz.
The frequency analyses part of this program is not that great but at least it gives a good place to start.
Last edited:
Hi Enzo, I was actually refering to a bypass cap (cathode to ground). I know it's common practice to use a bypass cap to achieve maximum gain, but I chose to utilize the serial feedback by not using one, to increase V1A's headroom. A bass can have a pretty big signal swing! Sometimes well over 1.5Vp (V1A's bias).
The DC coupling between the stages has several reasons. I like to use as itttle caps as possible in the signal path. That's why there's no input cap either. And with this arrangement (both V1A's anode and V1B's cathode at a tuned high voltage), the bias of V2B causes a little constant grid current, which has a small compressive effect (check Morgan Jones' site, good stuff). Sounds awesome for bass!
The DC coupling between the stages has several reasons. I like to use as itttle caps as possible in the signal path. That's why there's no input cap either. And with this arrangement (both V1A's anode and V1B's cathode at a tuned high voltage), the bias of V2B causes a little constant grid current, which has a small compressive effect (check Morgan Jones' site, good stuff). Sounds awesome for bass!
Last edited:
Worked like a charm! There is no significant tonal impact on the signal and it even allowed me to implement a separate volume control for the DI output. So thanks geraldfryjr!
I was wondering if it's good practice to use build-out resistors of, let's say, 100 ohm, to prevent short circuiting the outputs in case of e.g. cable failure.
I was wondering if it's good practice to use build-out resistors of, let's say, 100 ohm, to prevent short circuiting the outputs in case of e.g. cable failure.
Most opamps are S/C protected anyway, the usual reason for including the resistors is to prevent instability with certain types of opamps with certain loads.
However, adding a couple of resistors won't do any harm, so you may as well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.