2nd rectifier tube in 6 months

[burnedfingers]

My Dynaco has a new Triode Electronics choke and a SDS cap board. The first value of cap is 2 (80mfd ) in series for a total of 40mfd. Now that should be within the specifications for a 5AR4 /GZ34 rectifier.

I didn't think too much of it when the used 5AR4 went bad. Now I am somewhat confused as to why the new JJ 5AR4 didn't last very long.
Is JJ not a good choice for a rectifier tube? Do I need to put something more heavy duty in?
Any ideas?

[TerrySmith]

Here's the thing, a 5AR4 is being run at it's limit just when a ST-70 is at idle. When you add newer driver boards, cap boards etc for better performance, the little 5AR4 just cant handle it.

To keep it stock, you will need a Mullard or Amperex GZ34 / 5AR4, these seem to last a long time.

If you like to run it hard like I do, SS diodes may be a choice. I hollowed out an old metal tube and soldered in two UF-4007 diodes. This way I can switch back to tube rectifiers if I want. I also added a CL-80 to the primary.

If you have a Triode or Dynakitparts power transformer a 5U4 will work great, but with a slight voltage drop.

[burnedfingers]

Terry,

Could you back the the claim for the Mullard or Amperex being rated higher than the JJ or Sovtec or any other 5AR4 for that matter?
Please educate me as to how the standard 5AR4 cannot handle it.

The current draw for my amplifier per channel is 2.4 A which is below the factory rating of 3A per side per heater winding. I am running 2 (6BG6GA per channel as output tubes) and 1 (6sj7/5693 and 1 6SL7 per channel) 6BG6GA X 2=1.8 A 1) 6sj7=.3 1) 6SL7= .3 Total is 2.4 A heater per channel.

The rectifier tube draws 1.9A

Now, given that the input capacitor is no larger than 40mfd there shouldn't be a problem of overtaxing the rectifier tube. The additional power supply capacitor size in nest stages will provide no problem to the rectifier.

Monitoring the current draw for the amplifier I find it to be well within factory specification.

There is no attitude meant by this post. I am just trying to logically look at all factors here and until I can come up with concrete information to prove one particular rectifier tube is better than another I am a disbeliever.

[Ty_Bower]

The 5AR4 is only good for 225~250 mA at the supply voltages of the ST70. Each EL34 eats 50 mA at idle, which doesn't leave a whole lot left over for the driver stages.

You can find all the data sheets you need here:
http://tdsl.duncanamps.com/show.php?des=5AR4

[burnedfingers]

The 5AR4 is only good for 225~250 mA at the supply voltages of the ST70. Each EL34 eats 50 mA at idle, which doesn't leave a whole lot left over for the driver stages.

If you calculate the driver stages which I did I am within the specification.

Still haven't uncovered where the Mullard would be superior.

[TerrySmith]

As Ty said, the DC current at IDLE, with tubes biased @ 50ma each totals 200ma. That's just 25ma away from it's max rating.

Mullard and Amperex tubes are rated the same as JJ and Sovtek tubes. The quality was a LOT better. I'm listening to a '70 with an original Dynaco/Amperex GZ34 as I type this reply! A Sovtek and a Sylvania fat bottle met their fate in this amp at an early age.

And don't even think of using a Chinese 5AR4 unless you like fireworks! :parp:

[TomMcNally]

JJ GZ-34's are crap. I've gone through a couple of them
in various amps. I think the Sovtek's are better.

Actually, I haven't been happy with many JJ/Tesla tubes.
It always seems like one of the quad or octet blows.

... tom

[burnedfingers]

As Ty said, the DC current at IDLE, with tubes biased @ 50ma each totals 200ma. That's just 25ma away from it's max rating

When you check the specifications it is 25-50mA from max depending on what you wish to believe.

The quality was a LOT better. I'm listening to a '70 with an original Dynaco/Amperex GZ34 as I type this reply! A Sovtek and a Sylvania fat bottle met their fate in this amp at an early age.

So, there is no proof the Mullard or Amperex is any better on paper. This is what I was looking for proof wise. To be honest I will not dish out the $100 going rated for an over rated Mullard. I will install two diodes to rectify the AC at the tube socket base and use the tube (5AR4) for the slow turn on of DC voltage.
This will make ANY 5AR4 last a long life.

[burnedfingers]

Specs for 5U4GA

RectifierAC Plate Supply Voltage Each Plate (RMS) ...... 450 VInput Capacitor ............................... 40 MfdDC Output Current ............................. 250 mADC Output Voltage ............................. 460 V

So this means a 5U4 won't cut it either?




--------------------------------------------------------------------------------

JJ GZ-34's are crap. I've gone through a couple of them
in various amps. I think the Sovtek's are better.

Well, I have gone thru 5 Sovtek 5AR4's in about 6 years in my Quicksilver mono blocks. It uses 2) per amp.

I think they are junk myself Yellow_Light_Colorz_PDT_09 I've had better
luck with the JJ's in the Quicksilver amps :) :)

[SDS-PAGE]

Try a WZ68 by Weber (http://www.webervst.com/ccap.html). Here is the description:

What is a WZ68?
In the case of the model WZ68, it is effectively two GZ34's in parallel to emulate the evolution of a higher current, low forward drop GZ34 had that evolution occurred. The WZ68 will work well in a 4 output tube 100 watt circuit and will yield similar sag and other characteristics as a GZ34 in a 50 watt circuit.

It can handle upto 450 mA.

[Ty_Bower]

So, there is no proof the Mullard or Amperex is any better on paper. This is what I was looking for proof wise. To be honest I will not dish out the $100 going rated for an over rated Mullard. I will install two diodes to rectify the AC at the tube socket base and use the tube (5AR4) for the slow turn on of DC voltage.
This will make ANY 5AR4 last a long life.

On paper, they all have the same specs. Whether they meet those specifications in real world use is a different question.

Your idea of using solid state diodes may have some merit. I don't know if it'll help increase the current handling capability of the 5AR4, but I can't imagine how it will hurt anything. The diodes themselves are certainly inexpensive enough. I saw a post on another forum which describes a convenient method for putting diodes directly on the rectifier socket. Take a look at the photo attached to the end of the post:

http://www.diyaudio.com/forums/showthre ... ost1481237

[burnedfingers]

Your idea of using solid state diodes may have some merit. I don't know if it'll help increase the current handling capability of the 5AR4, but I can't imagine how it will hurt anything. The diodes themselves are certainly inexpensive enough. I saw a post on another forum which describes a convenient method for putting diodes directly on the rectifier socket. Take a look at the photo attached to the end of the post:

The usage of diodes along with the rectifier tube will do several things. The diodes take the current hit on start up. The already rectified voltage is then allowed to use the slow warm up from the tube. The tube filters out the spikes from the diodes. Using the diodes along with the 5AR4's has allowed my usage of the 850 mfd filter caps ahead of the choke in my Quicksilvers. Note* The 5AR4's are speced for a maximum of 100mfd cap as the first cap in the supply. The pair of Quicksilvers with 850's as the first cap were purchased by me that way. They come with 330mfd caps as built.

If I knew how to post a schematic I would post what I have in one pair of my Quicksilver amps for power supply modification.

So the best of both worlds. I would say it has merit.

[burnedfingers]

If you do the addition for heater requirements on a stock stereo 70 you will see the output tubes draw 1.5 A ea X2= 3A and the 7199 draws .45 so in stock form it exceeds the 3A heater output on each leg of the stock transformer. The total per side is 3.45A

[TomMcNally]

This is what you need ....



I used a pair of them in my diytube Poseidon/MK-III clones.

[burnedfingers]

Granted I wouldn't mind having a 1/2 dozen or so of the GZ37's but please do tell ......what will I gain? They are rated at 250mA also Yellow_Light_Colorz_PDT_17