Type 50 triode amp

[nicholaskrautner]

I was given a box of tubes the other day. A bunch of type 10 and type CX350 tubes still in their boxes! I want to make a SE with a 50, transformer coupled 26, and a 24A tetrode preamp. I want this schematic as simple as it gets and I have an edcor 10K 1:1 transformer for the interstage. Anyone want to look this over and see if it will work? Still trying to decide if I will use a 5ar4 or 2x 6d22s tubes with a 400-0-400 transformer. around 530v at first filter cap and then 300v to the 24A plate. 180v to the 26.

 

[Ludwig Haus]

Congratulations on the new find.. but I'd be looking at a split rail DC coupled design = much simpler and without compromise WRT the posted schematic.

LH/S

 

[vega65]

15 k under the cathode of cx-350 ?

 

[vega65]

Maybe change the drawing of the cx-350 from indirectly heated to directly?
And please consider Rod Colemans excellent filament supply regulators

 

[kmaier]

First, check your power supply schematic... why the two HV windings in a such a configuration? It's not a proper full-wave rectifier and you'll likely smoke the lower rectifier. Second, a 15K cathode resistor for self bias is far too high. Based on desired operating point it should be closer to 1.5K and a larger bypass cap as well. Third, the 26 seems a bit light as a driver. Note that you will need to swing around 170 volts pk-pk to the grid of the 50. You won't have adequate drive to the IT to get the voltage/bandwidth.

Regards, KM

 

[nicholaskrautner]

Woops. That's a 1.5K with a humdinger to the direct heated cathode.

 

[nicholaskrautner]

Would a 6c5 or 6c6 in triode mode work to drive the IT? That 15k was supposed to be a 1.5k to center tap 100ohm pot.

 

[nicholaskrautner]

What's a split rail?

 

[nicholaskrautner]

revised schematic

I am very new to HIFI. Guitar amps are much easier, but I feel the need to grow. I revised the schematic to reflect what I hope is some type of quality fit for a type 50 with the tubes I have. The gain of the 24A should be around 30.9 with a grid resistor of 500k into the 26 which should have a gain around 8. That should be enough to drive the 84v or roughly 170 ppv needed for the 50 grid from my record player or tv if I choose. Are we getting closer? Thanks for all the awesome feedback so far.

 

[marcelop]

You are going to have A LOT of hum if you don´t filter the output stage B+. Also it would help if you included the plate, cathode and screen dc voltages and current for the preamp stages.

 

[nicholaskrautner]

50 plate supply voltage 525VDC and 84v in the cathode. Not sure what the actual plate volts are because I have not decided on an output transformer yet. Is the 220uF filter cap not enough for the output stage?

 

[kmaier]

As this is a class A single-ended design, any ripple on the B+ will come through the output stage. A 220uF cap is likely too large for the rectifier tube to feed. You should use a classic Pi filter setup (two caps and a filter choke) to minimize the ripple and improve the regulation a bit. You can use a larger cap on the output from the choke and a smaller one from the rectifier.

Depending on the filament structure on the 50 triode, you may need a DC supply to keep the hum level down to acceptable levels, but I would try using AC initially and see if the balance pot can null it down low enough.

Regards, KM

 

[kmtang]

How lucky you are having the CX350's!!!

JOhnny

 

[marcelop]

Having 7.5V filaments they hum pretty bad, expect at least 3 times more than a 2A3 (when the 60/50hz are already balanced).

Also in a single ended amp as kmaier said any ripple in B+ is going to be almost completely reproduced in the output when using triodes. I simulated your power supply in psud2 and it develops 1,5Vrms of ripple at the first cap, which translates to hum being -34db referenced to 1 watt, totally unacceptable.

As for the first input cap, peak diode current is below the datasheet max so no problem using 220uf.

 

[nicholaskrautner]

downloaded PSUD2. Wish I had known about that before🙂 Looks like 47uf into a 4h choke followed by a 220uF cap. The next stage looks fine at 22uF. I have a 2.1A DC powersupply for the type 50 Heater.

 

[kmaier]

A 4H choke is on the small side, you'll get better results with a 10H choke. The goal is keep hum and noise to a minimum. For low power SE amps, acceptable signal-to-noise is around 80dB referenced to 1-watt. With a DC filament supply and a clean B+ you should be able to manage it.

Regards, KM

 

[marcelop]

For low power SE amps, acceptable signal-to-noise is around 80dB referenced to 1-watt.

+1

With a 6k load you should be aiming for 8mVrms or less at the output stage B+

 

[nicholaskrautner]

revised

Here is what I have from the advise I have received. Just need to find a center tap 1.5v transformer for the 26 tube, and 2.5v for the 24A. Thinking an edcor 5.5k for the output. Is 40-18Khz enough for this amp or do I need a wider range?

 

[kmaier]

Have you drawn some load lines using the plate curves? The data sheet shows your chosen bias point with a 4.3K load. You may want to draw this out first and try a few different points before choosing a load value.

Edcor make some decent iron and is well priced. However, 50 triodes are pretty rare and fetch some fairly high prices. It may actually work out that the output tubes alone are worth more than the rest of the parts.

As for bandwidth, I prefer it very wide. My 45 SET amps are flat within 1dB from 25Hz to 50KHz at 1-watt RMS output. Having a wider bandwidth results in less phase shift at the frequency extremes. However, this also costs more for higher quality iron. It's a matter of choice and balancing your design goals.

Regards, KM

 

[nicholaskrautner]

I will look for OT between 4k and 4.5k load with wider freq.