My "Tube SEPIC II" project

ask your general tube related questions here

My "Tube SEPIC II" project

Postby LW1DSE » Mon Aug 11, 2014 8:25 am

Hi all folks. This is an useless experiment I currently working on. Useless because it has no utility, it is only to self learning, and the simple pleasure of see it working. Yes, a switching power supply using ONLY vacuum tube (valves) in all stages. No SS at all.

WARNING: DON'T TRY TO COPY THE PROJECT WITHOUT HIGH VOLTAGE AND HIGH FREQUENCY KNOWLEDGE‼

Here is the schematic of the current version of my Tube SEPIC II converter for my audio amplifier project.

Although a bit complicated in the drawing, it is a quite simple design.

It has all the normal parts (bloks) as a SS PWM switching PS, but all functions are implemented using vacuum tubes (Valves).

From left to right, first is the voltage reference section. It comprises the current source V4 (6C4) for the neon lamps, the voltage reference itself. Ne1 has a current of about 100 µA, giving set by the 1M resistor R41. This gives about 60V to V4, from which cathode, a reference voltage of +189V for the circuitry. About 400µA flows through R42 giving 500µA for Ne2 and Ne3.

Ne3 gives the +60V reference for the DC amplifier. It is made about V3. Triode section II decouples this reference from pentode (The true DC amp) giving about +63V for stabilizing cathode-earth voltage of the pentode. Section I is a SRPP for the pentode amplifier, biased at 500µA. I(V1I) is about 3 mA. Triode section I of V2 is a screen supply, adjusted via P102 to optimum pentode biasing, about 15-18V screen to cathode's voltage. Note NO capacitor (AC) decoupling at cathode or screen, as it does not work in DC, all decoupling is made around cathode followers. Screen's source impedance is about 1.3K viewed from screen to triode's cathode, and about 300R from pentode's cathode to the triode's one. R43-C42 reduces noise generated at neon lamps (Instability).

From the low impedance of the SRPP's cathode, a voltage error is obtained, comparing reference at cathode, with a sample of the output voltage via R47/R48/P101.

V2 is the "PWM" and driver section. It is an RC coupled multivibrator. One of the grids is tied to the reference 189V and is the OFF time (fixed) of the supply. The other (pentode) grid is wired to the DC amp's output. More voltage at anode is more current to charge C21, then, the ON time is shortened. Pentode's screen acts as the plate for the RC oscillator, around 80-120KHz. From plate of the pentode, sharp pulses are reduced in amplitude, and amplified in current via T1, a flyback CCM transformer. Pulses are coupled to final stage 33GY7's pentode, who acts as the main switching device. The SEPIC inductor is in his plate (L1), and diode section and L2 are the SEPIC rectifier. V2 is the Cuk section, via L5. For more details, see http://www.analog.com/static/imported-files/application_notes/AN-1106.pdf/of the operating style of the output stage.

The supply actually is operating from a single +275V supply, plus a 6.3VAC and 33VAC for the 33GY7 and the 35W4 devices. In the final project, V1 will have its own heater secondary to not violate heater/cathode isolation voltage. It also gives ±350V @20mA in its output, no measurable 50 or 100KHz ripple, and below ±1V tracking between outputs, having in mind the different rectifiers at both outputs (a 35W4 50Hz rectifier vs a 33GY7 damper diode, aldo with different current capabilities).

Actually V2 and V3 are different from the types depicted here, I suppose that in any time I will have this tubes, as our fucking "presidenta" allow me to buy the tubes beyond my country boundary.

Capacitors and inductors showed as NU are still "not used", and reserved for future filtering, if needed. 100K resistors fix a minimum load for the outputs (3.5mA) in each output.

Soon, photos of this Frankenstein tube circuit
Attachments
Tube SEPIC II.GIF
Tube SEPIC II.GIF (52.34 KiB) Viewed 4241 times
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina

Re: My "Tube SEPIC II" project

Postby LW1DSE » Tue Aug 12, 2014 7:37 am

Here, some photos of the monster. They are of the output stage.

The smaller tube is the 35W4, and the bigger one, the 33GY7. Also, clearly can be seen, the 3 inductors for the SEPIC & Cuk, L1, L2 and L5. Also, the capacitors C1, C2, C3, CM1, CM2. Also, screen drop resistor R1, near the end of the board.

Sorry if my English is not too good.
Attachments
Setup 04.JPG
Setup 04.JPG (149.61 KiB) Viewed 4212 times
Setup 03.JPG
Setup 03.JPG (145.62 KiB) Viewed 4212 times
Setup 02.JPG
Setup 02.JPG (138.52 KiB) Viewed 4212 times
Setup 01.JPG
Setup 01.JPG (145.62 KiB) Viewed 4212 times
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina

Re: My "Tube SEPIC II" project

Postby LW1DSE » Tue Aug 12, 2014 7:45 am

These are the inductors L1, L2 and L5 during construction. They are hand wired, using line filter core and bobbin's, with separators, to reduce interwinding capacitances.
Attachments
Inductor 05.JPG
Inductor 05.JPG (132.38 KiB) Viewed 4211 times
Inductor 04.JPG
Inductor 04.JPG (138.22 KiB) Viewed 4211 times
Inductor 03.JPG
Inductor 03.JPG (132.4 KiB) Viewed 4211 times
Inductor 02.JPG
Inductor 02.JPG (135.97 KiB) Viewed 4211 times
Inductor 01.JPG
Inductor 01.JPG (143.33 KiB) Viewed 4211 times
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina

Re: My "Tube SEPIC II" project

Postby LW1DSE » Tue Aug 12, 2014 7:56 am

Output voltages. Input voltage (not shown): +280VDC.
Attachments
Voltage 04.JPG
Minimum load, 150K in the positive and negative rails. Output voltage at negative out.
Voltage 04.JPG (134.54 KiB) Viewed 4210 times
Voltage 03.JPG
Minimum load, 150K in the positive and negative rails. Output voltage at positive out.
Voltage 03.JPG (131.89 KiB) Viewed 4210 times
Voltage 02.JPG
Full load, 22K in the positive and negative rails. Output voltage at negative out.
Voltage 02.JPG (125.65 KiB) Viewed 4210 times
Voltage 01.JPG
Full load, 22K in the positive and negative rails. Output voltage at positive out.
Voltage 01.JPG (143.06 KiB) Viewed 4210 times
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina

Re: My "Tube SEPIC II" project

Postby LW1DSE » Tue Aug 12, 2014 8:14 am

Some oscillograms. (Sorry because some photos are out of focus‼)
Attachments
33GY7 Anode 04.JPG
33GY7's anode node voltage, similar to 33GY7's and 35W4's diodes, but at different DC levels.
33GY7 Anode 04.JPG (135.05 KiB) Viewed 4207 times
33GY7 Cathode 08.JPG
33GY7 Cathode current, viewed in a 10R resistor. Note the step on a ramp (typical for boost derived) plus the initial spike when the 33GY7's pentode charges parasitic capacitances in the anode node.
33GY7 Cathode 08.JPG (149.5 KiB) Viewed 4207 times
33GY7 Anode 07.JPG
33GY7 grid voltage waveform.
33GY7 Anode 07.JPG (136.71 KiB) Viewed 4207 times
33GY7 Anode 07.JPG
33GY7 grid voltage waveform.
33GY7 Anode 07.JPG (136.71 KiB) Viewed 4207 times
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina

Re: My "Tube SEPIC II" project

Postby Cubdriver » Wed Sep 17, 2014 8:52 pm

That is pretty impressive!! Nice work.

-Pat
User avatar
Cubdriver
 
Posts: 14
Joined: Thu Aug 30, 2012 10:17 pm
Location: Litchfield Co, CT

Re: My "Tube SEPIC II" project

Postby LW1DSE » Mon Sep 22, 2014 6:27 am

Now I'm working in a linear PWM controller, using a "Phantastron" sawtooth generator. Some photos coming soon.
If anything can go wrong, it will.
If anything just cannot go wrong, it will anyway.
If there is a possibility of several things going wrong, the one that will cause the most damage will be the one to go wrong.
User avatar
LW1DSE
 
Posts: 37
Joined: Wed Aug 06, 2014 9:49 am
Location: Barrio Garay, Almirante Brown, Buenos AIres, Argentina


Return to tube 101

Who is online

Users browsing this forum: No registered users and 12 guests