Author Topic: 6X5GT - Tube rectifier input filter capacitance  (Read 393 times)

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,117
  • Total likes: 36
  • The path is the goal. Mahatma Ghandi
    • Mars Tube Audio
6X5GT - Tube rectifier input filter capacitance
« on: May 21, 2019, 12:18:46 PM »
Hi all tube DIY'ers

This common question has come up regarding input filter capacitance in tube rectifier circuits. Have a look at this PSU that supplies about 20mA:



Anything wrong with this design?

Buizen handboek gives this data for the 6X5GT:

OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Family_Dog

  • Administrator
  • Trade Count: (+28)
  • *****
  • Posts: 9,203
  • Total likes: 261
  • STEEL & GLASS - Everlasting Beauty!
    • Repcom 2-way Radios
Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #1 on: May 21, 2019, 01:48:47 PM »
I haven't looked up any specs but I get the feeling that the first cap at 100uF is a tad optimistic for a 6X5. The better-rated GZ34 has a max recommended first cap value of 50uF, which is what I use in my amps.

After the series resistance you would be fine using the larger value cap, but those series resistors are very high-value for the HT line. The current drawn by the attached circuitry must be minimal.

-F_D



-Eric

That Guy in South Africa...
*************************************
http://groups.yahoo.com/group/vintage_tube/
*************************************

Offline handsome

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #2 on: May 21, 2019, 03:50:34 PM »
the first capacitor in a power supply like the above draws current in quite narrow pulses, the bigger the capacitor the higher the pulses will be. Tube rectifiers have a maximum DC current rating and a maximum pulsed/AC current rating: IFRM (current forward repetitive maximum). It is this rating that determines the size of the first capacitor in a capacitor coupled PSU. Its easier to just use the datasheets' maximum capacitance but you can work out the transformer impedance and, depending on the voltage and currents and safely use a higher capacitance.....but this requires much calculation and reference to both manufacturers' and Schade's graphs. A resistor connected between the cathode and the first capacitor (or resistors connected between anodes and transformer winding) is recommended in the datasheets whose value depends on the output voltage/current required - this is to stay within IFRM limits. But it is generally a good idea to always use it anyways as it ensures a longer life for the rectifier.

The 6X5GT is rated for a maximum of 40uF....

High resistor values in the filters are fine provided you do not have too much current variation. In other words preamps. Large current variations will result in large signals on your HT line, quite the opposite of what the HT is intended to be (steady). Filter resistors are usually sized to get the required DC voltage.

Offline El Sid

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #3 on: May 21, 2019, 04:49:29 PM »
High resistor values in the filters are fine provided you do not have too much current variation. In other words preamps. Large current variations will result in large signals on your HT line, quite the opposite of what the HT is intended to be (steady). Filter resistors are usually sized to get the required DC voltage.

The 20 mA current draw quoted and the 6SN7 heaters in the schematic suggest it is for a preamp (or phono?).

Karel couldn't you drop the first cap to the recommended 40 uF and increase the other two? It's too late in the day for me to think about the effect on ripple, regulation, etc.

Offline Hasren

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #4 on: May 21, 2019, 05:45:23 PM »
The 20 mA current draw quoted and the 6SN7 heaters in the schematic suggest it is for a preamp (or phono?).

Karel couldn't you drop the first cap to the recommended 40uF and increase the other two? It's too late in the day for me to think about the effect on ripple, regulation, etc.

It is for a preamp El Sid. I agree with you that it is prudent/safe to drop it to perhaps 40uF, but I have 47uF Panasonic caps which I think it won't be that bad to have them across the network. In terms of the ripples, the CRCRC network should be enough to add enough ripple reduction factor...  :2c: @FD with regards to the values of the resistors I have some 4,7K Vishay resistors lying around, would these be fine across the network? i.e replacing the 10K 

Offline Family_Dog

  • Administrator
  • Trade Count: (+28)
  • *****
  • Posts: 9,203
  • Total likes: 261
  • STEEL & GLASS - Everlasting Beauty!
    • Repcom 2-way Radios
Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #5 on: May 21, 2019, 06:22:34 PM »
Hasren, you would need to calculate the value of the resistance required by the voltage drop required, so whatever voltage leaves the cathode of the 6X5 minus the required voltage (220v?) divided by the current drawn. 20mA sounds quite a lot for the preamp but could possibly be so if it is a stereo preamp with at least 2 valves per channel. FWIW, a 12AX7 typically draws about 1mA per anode if I remember correctly. 


-F_D



-Eric

That Guy in South Africa...
*************************************
http://groups.yahoo.com/group/vintage_tube/
*************************************

Offline fredeb

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #6 on: May 21, 2019, 07:57:33 PM »
Is it worth considering RC network as low pass filter in the PSU ? In order to filter PSU noise ?
...evolution is the gradual development and stratification of progressive series of wholes, stretching from the inorganic beginnings to the highest level of spiritual creation.
Jan Smuts

Offline Hasren

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #7 on: May 21, 2019, 08:56:15 PM »
Great FD, following your logic I come to 7.8K. So considering that they are in series then R1 = 3.9K and R2 = 3.9K. Yes the 6sN7 cathode current draw is 20mA. So I divided the difference voltage by 0.02A. I hope this is correct. Thanks once again FD.

Offline Ampdog

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #8 on: May 22, 2019, 03:23:51 AM »
"Is there something wrong with the first diagram?"

Yes. A very high value input capacitor serves little purpose. That cap. determines the ripple more than anything else, and for the small current drawn there should be negligible difference between using a 40F and 100F there; ripple would be further reduced by the following sections. The highest capacity value is usually at the power supply output, to give a low impedance here, particularly when a complete amplifier is fed.

The effective series resistance in series with the input capacitance is rather difficult to determine. It includes the transformer winding impedance, which is somewhat higher than the winding d.c. resistance. But it depends on the transformer construction - and remember that it is lower than the winding open circuit impedance, because other transformer winding impedances are not open circuit (the primary impedance has the mains impedance in parallel, which is quite low. Heater windings are loaded etc.) Thus transformer-wise one might simply include the secondary winding's d.c. resistance only and be on the safe side.  There are also spreads ... as said I only use the secondary d.c. resistance  when calculating any series plate resistor where necessary.
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline Hasren

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #9 on: May 22, 2019, 09:38:15 AM »
Thanks, Ampdog,

For the capacitor, I intend using the 47uF across the network i.e in principle trying to get as close to the 40uF spec and I think this should be enough to deal with the ripples. On the resistors, I will try the secondary DC resistance method, but if all fail, I recall in another thread FD saying sometimes there are some solutions that might not earn marks in an engineering paper but may work in real life, talking about how he resolved a hum using a single ground loop  :thinking: or something like that... So in my case, I may have to do a trial and error finding, starting off with my calculated resistor values then ramp them up or down until I measure closer to the desired 220V  :giggle:  :giggle:
 

Offline charles

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #10 on: May 22, 2019, 03:34:51 PM »
From Ampdog

"Is there something wrong with the first diagram?"

Yes. A very high value input capacitor serves little purpose. That cap. determines the ripple more than anything else, and for the small current drawn there should be negligible difference between using a 40F and 100F there; ripple would be further reduced by the following sections. The highest capacity value is usually at the power supply output, to give a low impedance here, particularly when a complete amplifier is fed."

Agree with Ampdog. An example is the Leak 12 Plus: 60ufd/200ufd

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,117
  • Total likes: 36
  • The path is the goal. Mahatma Ghandi
    • Mars Tube Audio
Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #11 on: May 23, 2019, 11:12:55 AM »
OK... time for some measurements:

The toroidal Primary (230VAC) measures 55 ohms and the Secondary (265VAC) 134 ohms. The voltage before the 10k resistor measures 377V DC and after 238 VDC. I am focusing on the "after" voltage, since it should have a much lower AC component than the "before" voltage.

If I feed that into Duncan's PSUD2 I get this result:





To get the voltage in the PSU to match up with measurements, I had to use a 7mA current consumption for the stereo preamp.

@Hasren: can you measure the voltage drop across all six cathode resistors (the 2K2's and 2K7)?
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline handsome

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #12 on: May 23, 2019, 12:11:24 PM »
20mA across 10K is 200V........?
 
20mA for 6x 6SN7s implies ~3mA per 6SN7 that's low, 220V is very, very low for a 6SN7 - so you will be in a non-linear region for those tubes. Why not post the schematic?

Also you don't want to connect the 6SN7 heaters that way in a preamp, you might run into hum issues...

Offline Hasren

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #13 on: May 23, 2019, 01:44:19 PM »
Thanks, Handsome

The said preamp has 4x6sN7 which according to your logic would imply 4mA per 6sN7. As a junior member, I can't attach. Perhaps @Mars can assist in this regards or if you don't mind you can check it under this topic 2019 DIY valve amp workshop in Pretoria?.

Also regarding the 6sN7 heaters connection, how would you rather have them connected?

Offline Hasren

Re: 6X5GT - Tube rectifier input filter capacitance
« Reply #14 on: May 23, 2019, 01:57:01 PM »
@Mars, I will provide the voltage drops across the cathode resistors later  :thumbs: