Tube Rectifer Plate Voltage Question

[JoeTele]

Is the rated maximum plate voltage for a tube rectifier based on the unloaded PT voltage or the circuit under load? I'm wanting to use a Deluxe Reverb PT in a vintage hifi build that incorporates an EZ80 rectifier. The rated maximum plate voltage is 350 volts for it. My DR PT is rated for 340-0-340, but I measured it and found it's 365-0-365 without the load and I'm will thinking will come in just shy of 350-0-350 in the circuit I'm building. Would that be safe, or should I consider a Zener or a different PT?

Thanks!

Joe

[David Root]

It will be OK at 350-0-350. Assuming a good tube, should be no problem.

Since that is the highest rating, I would recommend calculating the additional resistance you will likely need to add between the PT secondary and each plate input pin. Required total resistance is 300 ohms on each side, as per the datasheet for Rt.

The 300 ohms is made up of half of the PT secondary winding resistance plus the square of the PT voltage ratio, which in this case is 350/120. Add those two numbers and subtract them from 300, and put a 5W power resistor of that resultant value on each plate pin.

This will drop a few volts but prolong the life of the EZ80. This should be all in the Philips datasheet for the EZ80.

[Phil_S]

The rating will be what you measure on the plates in-circuit, so I'd expect you'd be below 350. However, I don't think I'd be comfortable rubbing up against the limit like that. You may want to consider a different rectifier or a different PT. I think you are going to be hard pressed to find a 6.3V rectifier if that is your bottleneck. Why not just go with a solid state rectifier? I'm not clear what you are building, but for hi-fi, I don't believe the tube rectifier adds anything useful. Save a tube, lighten the load.

[JoeTele]

The build is an old Mullard 3-3, which I've posted about. It's here:

http://www.r-type.org/articles/art-003h.htm

I want to keep the tube rectifier for the originality/vibe. Here, there is a 330 ohm resistor right at the dc output of the rectifier, based on a hypothetical 300-0-300 PT. The plan says that the value of that resistor depends on the PT chosen. Is that acting as the protecting resistor here, as there are no resistors connected to the plates?

Also, how do I determine the PT secondary resistance? If I take the measured, unloaded 365-0-365, the Deluxe Reverb current, and the 25v per side drop to 340-0-340 under that load, can I calculate w/Ohms law?

If I did want to move back further from the maximum 350 rating, would there be any problems using Zeners in the center-tap in this type of circuit?

Thanks again!

Joe

[trobbins]

The 3-3 circuit is class A, so there is no significant sag issue.

Perhaps if you download the PSUD2 program and walk through the help/tutorial then you will see how the design and checking of a power supply is done. PSUD2 is used by a lot of us and is well worth the learning curve to get a better understanding of valve electronics.

It will tell you how to measure the DC resistance of power transformer windings, so you can report back with that for us to help further with the performance.

[David Root]

To measure the PT secondary resistance just use a multimeter across the winding. (You are measuring the winding's DC resistance, not its AC impedance).The secondary winding should be disconnected from anything else.
The 330 ohm resistor on the DC output after the rectifier is just reducing the DC voltage, same as any power supply dropping resistor. That is why the note that its value will depend on the PT ie its secondary winding rating.

Any decent old EZ80 in good condition should be quite safe feeding from a 350-0-350 PT sec. The datasheets were not overly optimistic in the old days.

Current production maybe not; not sure if there even is a current production EZ80.

If you do add the two 5W resistors on the plate pins you will not need 330 ohms after the rectifier, but a smaller value. Adjust its value to get the specified raw B+.

+1 on learning to use PSUD2. That is the first thing I do for a new build, simulate the power supply.

Since the circuit uses cathode biased power tubes you can safely put a zener between the PT CT and ground. I would use at least a 5W rating. My experience with this method of B+ voltage reduction is that it doesn't do much until you crank the amp, so for me it would not be my first choice. If you do it don't forget the big .1uF 1kV ceramic cap across the zener or it will be very noisy.

[martin manning]

there is a 330 ohm resistor right at the dc output of the rectifier, based on a hypothetical 300-0-300 PT. The plan says that the value of that resistor depends on the PT chosen. Is that acting as the protecting resistor here, as there are no resistors connected to the plates?

Since that resistor is between the rectifier and the reservoir capacitor it is indeed limiting the current flow through the rectifier.

[JazzGuitarGimp]

Someone correct me if I'm wrong, but I'm thinking the max plate voltage for a rectifier refers to peak voltage, not rms voltage. So a 350V(rms) PT secondary actually translates to a peak plate voltage of 494V. Or am I missing something?

EDIT: And I take back everything I said. I should have looked at the datasheet before I chimed in. Carry on, then....

[JoeTele]

I was wondering about that 330ohm resistor, as I found Merlin's page about this and he does mention a single resistor in series with the cathode as an alternative to the 2 resistors on the plates, but I appreciate all of the info on internal resistance, etc. I'm going to download PSUD2 and mull all of this for awhile, as I want to keep growing my knowledge here. I did measure the DC on my PT (which is not connected to anything right now) and came up with 3.2 ohms on the primary and 70.6 ohms on the secondary. Are these reasonable values?

Thanks again to all!

Joe

[martin manning]

With 350 VAC feeding a FW rectifier, the PIV is 2*1.414*350 = 990V, minus the diode drop on the other side. I suspect the 350VAC max is really about the PIV. The resistor is there to address peak current.

[martin manning]

I was wondering about that 330ohm resistor, as I found Merlin's page about this and he does mention a single resistor in series with the cathode as an alternative to the 2 resistors on the plates, but I appreciate all of the info on internal resistance, etc.

In either case the winding resistance is in play.

[JoeTele]

So I downloaded PSUD2 and decided to plug in the stock values from the 3-3 to the best of my ability. The transformer secondary resistance and offload voltage are estimates for a hypothetical 300-0-300 obtained by splitting the difference between values on the 270-0-270 and 340-0-340 taps on my transformer. The currents are combined plate and screen grid voltages from the EL84 and EF86 datasheets. I also downloaded a couple of Fender models whose simulations did not trigger warning messages just to try to determine whether I had extracted the appropriate pieces in my model. I also adjusted various values up and down, and no matter what kept getting the following 2 messages.

A current sink has pulled the voltage below zero for more than 5 mains cycles, at time 0.083477S.
The rectifier IFRM of 0.24A has been exceeded with a value of 0.25A, at time 0.169937S.

Is there anything flawed with my model? The attached image hopefully contains the necessary info. Let me know if I can provide anything else.

Edit: Also, I see that the C3 voltage id 333.48m. The voltage as published in the article (http://www.r-type.org/articles/art-003h.htm) is 210v (With 310 at C1 and 290 at C2). No idea why C3 is down so low.

Thanks as always!

Joe

[JoeTele]

Never mind. I see that this circuit is running the ef86 at much lower than normal currents so will play around with that, study the data sheets, etc.

Joe

[David Root]

Joe, the winding values you measured are reasonable.

[JoeTele]

Ok, I've run some stated voltages in the 3-3 plan against component values to come up with better current estimates and things are looking much better in PSUD. The Deluxe Reverb PT I want to use has a lower tap of 290v (as measured with no load) and, of course, the 5 volt winding so I modeled the circuit with a GZ34 and found that, in series with a 330ohm resistor, gives readings within a couple percent (or less) those with the EZ80. From what I can tell, the 330ohm series resistance will more than cover any protecting resistance deficit in the PT winding (2x50 at 300v plate voltage is needed and I've got 98ohms in the winding), and the 50uf first filter cap is under the 60uf max for the GZ34. Given that that is a more plentiful tube and I wouldn't be pushing it anywhere near it's limits, I think I might like to use it in place of the EZ80. Any concerns I haven't addressed?


Joe