First amp build questions

[szabi43]

Hi!

I bought Merlin Blencowe's "Designing Tube Preamps for Guitar and Bass" book and I want to build his ultra high gain design found on page 259. Since this is my first high voltage build (I've build pedals before) and I made a few modifications to it based on the book itself I'd like to get your opinions on them and the power supply section before a cap blows up in my face or I shock myself to death.

The modifications:
Added a passive line out as the only output because I want to use it as a "pedal", so I can add other effects after it or connect it straight to my USB interface.
Removed the two noise gate diodes, its switch and the boost switch because I won't need them. I want to use it for high gain tones only and I have a nice noise gate pedal.
Every tube uses the DC heater supply since the 6.3V output of the transformer I'm planning to use (Hammond 290XEX) is supposed to be able to handle the current load.

Here is the circuit I've come up with:

ultra high gain preamp.png

[nworbetan]

The output impedance is going to be a little higher than ideal without any kind of buffer. That being said, it sounds like you have a fair bit of pedal experience and might be familiar with how to use a buffer when needed. So depending on what you're plugging into you might be just fine without any kind of built-in output buffer. All my books are still boxed up from moving a few months ago and I can't remember what Merlin's original design for the output was, for the sake of comparison.

I'm curious about whether there's enough filtering on the B+ between the rectifier and the output triode stage. I know the ripple voltage is going to be pretty low with the few mA of current the two tubes draw, but I personally would want to add one more filter cap and dropping resistor. I'd probably lower the filter cap values too, 150uf total is more than a lot of power amps have. Those parts of the design are matters of preference though.

[rangdipkin]

that power transformer is a monster if you are just running a few 12ax7s... id consider going with a smaller one from the 200 series (or 300 series if you need a 220v primary). something like the 369BX (300vct @ 86ma) with a fwb rectifier would get you about 420 volts unloaded... not that the two 12ax7s will drag it down much. Also, an extra filter node should be considered mandatory, or even choke the whole supply with a CLC filter. Tiny current draw, so any old fender spec choke would work here. You can fine tune HT voltage post CLC filter with a voltage divider.

[szabi43]

The output impedance is going to be a little higher than ideal without any kind of buffer. That being said, it sounds like you have a fair bit of pedal experience and might be familiar with how to use a buffer when needed. So depending on what you're plugging into you might be just fine without any kind of built-in output buffer. All my books are still boxed up from moving a few months ago and I can't remember what Merlin's original design for the output was, for the sake of comparison.

I'm curious about whether there's enough filtering on the B+ between the rectifier and the output triode stage. I know the ripple voltage is going to be pretty low with the few mA of current the two tubes draw, but I personally would want to add one more filter cap and dropping resistor. I'd probably lower the filter cap values too, 150uf total is more than a lot of power amps have. Those parts of the design are matters of preference though.

The volume pot is 10k so the output impedance should be around that. The original design used only 2 47uF capacitors, the third one between the rectifier diodes was added by me. If you think its unnecessary I'll remove it.

[szabi43]

that power transformer is a monster if you are just running a few 12ax7s... id consider going with a smaller one from the 200 series (or 300 series if you need a 220v primary). something like the 369BX (300vct @ 86ma) with a fwb rectifier would get you about 420 volts unloaded... not that the two 12ax7s will drag it down much. Also, an extra filter node should be considered mandatory, or even choke the whole supply with a CLC filter. Tiny current draw, so any old fender spec choke would work here. You can fine tune HT voltage post CLC filter with a voltage divider.

You are right, a smaller one like the 369BX or 369AX would be better. Also I just noticed that when I calculated the voltages after rectification I used peak voltage but the output of the transformer is rms, so the voltages are much higher. This means I'll have to put a voltage divider after each rectifier anyways. I haven't made any CLC filters so I'll look into it.

[rangdipkin]

https://www.hammfg.com/electronics/tran ... /power/229

Check these lil guys out too! If you are alright using a 2nd transformer for your filaments, these dudes still have all the HT current a pile of 12ax7 will ever need. A trim 325dc after full wave rectification and filtering, still plenty of jolts! Have a peek at the old sunn model T or the bigger 50s fender tweeds for examples of CLC filtering, or play around with Duncan's psu software!

[szabi43]

https://www.hammfg.com/electronics/tran ... /power/229

Check these lil guys out too! If you are alright using a 2nd transformer for your filaments, these dudes still have all the HT current a pile of 12ax7 will ever need. A trim 325dc after full wave rectification and filtering, still plenty of jolts! Have a peek at the old sunn model T or the bigger 50s fender tweeds for examples of CLC filtering, or play around with Duncan's psu software!

I added a CLC filter based on a Fender amp and a 24k resistor to drop the 355V down to 260ish because the book says the voltage at DC+ should be between 200 and 300 but around 250 is the most ideal.

Now the circuit looks like this:

ultra high gain preamp.png

The LTspice simulation:

Screenshot 2023-11-02 214510.png

[rangdipkin]

Beauty! That is gonna be one quiet power supply... lets the preamp do the talkin'!

[Raja_Kentut]

Achtung, your heater supply will deliver around 9V, its again the RMS vs. peak thing.
Split the capacitor and place a suitable resistor between the both.
First guess is around 15ohm, 6W

Regarding DC+ I think you can save the Inductor L1 and split R21 : first directly behind the rectifier, second between the caps.

[martin manning]

Achtung, your heater supply will deliver around 9V, its again the RMS vs. peak thing.

6.3VAC is 8.9V peak, but there will be two diode drops and significant ripple voltage that will lower the average DC level. I think it will come out about right, if the heater winding has an appropriate AC current rating, about 2x the required DC heater current.

[szabi43]

Achtung, your heater supply will deliver around 9V, its again the RMS vs. peak thing.

6.3VAC is 8.9V peak, but there will be two diode drops and significant ripple voltage that will lower the average DC level. I think it will come out about right, if the heater winding has an appropriate AC current rating, about 2x the required DC heater current.

My LTspice simulation using the diode model from the manufacturer says it's about 7.2V DC after diode drops and filtering. Will it damage the heaters or should I drop it down to 6.3 with a resistor?

[martin manning]

You can get all kinds of numbers out of a simulation, and sometimes they are useful. Do you have an accurate model of the transformer heater winding? If not, I would try it in hardware to make final adjustments. A regulated supply would be the best way to hit the voltage exactly.

[szabi43]

You can get all kinds of numbers out of a simulation, and sometimes they are useful. Do you have an accurate model of the transformer heater winding? If not, I would try it in hardware to make final adjustments. A regulated supply would be the best way to hit the voltage exactly.

I've found a few forum posts where people built the same circuit and they measured 7.2 volts, so the simulation is probably right. One of these posts mention a hack: if I add a 5th one after the 4 rectifier diodes the output should be around 6.4-6.5 which is supposed to be fine based on Merlin's website: "A 6.3Vac transformer supply, after rectification and accounting for losses and diode drop, will yield about 6.5Vdc. You can of course adjust this to something closer to 6.3V with a dropping resistor, but this is rarely necessary."

[Raja_Kentut]

as said above, try it out.
Don‘t UNderheat, slightly overheat is ok.
In some designs they specified 6,5V as normal heater voltage (e.g. some Dynacord designs), and then assume that the mains voltage is +-10%… so it‘s not that critical.
Someone mentioned that the transformer you want to use is oversized, so it is in my opinion really advisable to check the heater voltage and not use it without a resistor after the rectifier. (it also dampens the inrush current of the cold heater filament)
The 6,3V rms are at rated current of the transformer. If you draw less current, you‘ll get a higher voltage.

[TUBEDUDE]

In a preamp where you might need more margin on the heater current vs. the H.V.
you can use appropriate sized separate H.V. and heater transformers.