Rectifier tube for U70?

simon wagstaff · « **on:** 23 Jul 2010, 02:33 am »

Read a few threads back and it seems there is some opinion that the rectifier tube can have a significant effect on sound. I am looking for suggestions for a replacement since mine just blew up. I am surprised at the price difference between the cheapest ones (16 bucks) and the most expensive (100 bucks plus). Any suggestions for a good replacement (other than the Philips which is one of those 100 buck plus ones) would be appreciated.

Lefty052347 · « **Reply #1 on:** 23 Jul 2010, 05:06 pm »

I would recommend keeping a Sovtek 5AR4 on hand for comparison since that is what AVA supplies for the U70 and Ultravalve amps. To do better would require a NOS 5AR4, IMHO.

Wayner · « **Reply #2 on:** 23 Jul 2010, 05:13 pm »

I'm with Lefty on this one. Oneinthepipe has had some success with other brands, maybe he will chime in.

Wayner

simon wagstaff · « **Reply #3 on:** 23 Jul 2010, 06:11 pm »

The Sovtek is about 15 bucks, the NOS at the Tube Depot is $199.00! Yikes! I am poking around a little bit on e-bay but don't think the best price will be there. a couple of used tubes are there, but I don't think I want a used tube.

avahifi · « **Reply #4 on:** 23 Jul 2010, 06:14 pm »

We have recently made a running improvement in the Ultravalve amp (applicable to the Ultimate 70 too) that should provide better service life for the 5AR4 rectifier tube.

The improvement consists of placing a 1N4007 diode in series with each incoming high voltage AC power transformer lead from the power transformer at the 5AR4 socket. This provides partial rectification of the AC power before the rectifier tube and we have been told this will provide better service life for the tube.

Looking at the rectifier tube socket from the bottom (wiring side), pin 1 is closest to the "keyhole indicator" and just to its left clockwise. The 8 pins are numbered clockwise from this, with pin 8 being just to the right of the keyhole.

Do the following with the amp turned off, unplugged from AC, and after it has sat off for several hours to discharge the supply capacitors. Remove the 5AR4 tube before starting this project.

1. Locate and disconnect the red transformer lead from pin 6 of the 5AR4 socket.

2. Connect a 1N4007 diode from pin 5 to pin 6 of the socket. The banded end goes to pin 6. Solder connection at pin 6 only.

3. Connect the red transformer wire to pin 5 of the socket and solder both the lead and the diode lead here.

4. Located and disconnect the red transformer lead from pin 4 of socket.

5. Connect a 1N4007 diode from pin 3 to pin 4 of the socket. The bended end goes to pin 4. Solder connection at pin 4 only.

6. Connect the red transformer lead to pin 3 of the socket and solder both the lead and the diode lead here.

Check your wiring work carefully. You have now installed a 1N4007 diode in series with each red transformer lead, using the unused pins 3 and 5 as tie points.

Reinstall the rectifier tube and re-set bias after a 30 minute warm-up.

Note that this project will work well for a stock Dyna ST-70 amplifier too.

Regards,

Frank Van Alstine

P.S. There is nothing special about a 5AR4 and there is no good reason that any brand, old or new, should provide better sound than any other, cheap or expensive. However, the rectifier tubes do vary in output voltage from sample to sample, and a low voltage one will affect the sound. As the output voltage goes down, the power goes down, the tubes are not biased on as hard, and we could expect poorer sound quality with a very low output rectifier tube. We have had good service life and consistently good high output from newly produced Sovtec 5AR4 tubes and see no reason to spend any more for one than these cost.

Wayner · « **Reply #5 on:** 23 Jul 2010, 08:14 pm »

Here is a graphic of the change:

Wayner

boycephoto · « **Reply #6 on:** 23 Jul 2010, 09:25 pm »

I have used Mullard 5ar4's or Sylvania Fat Boy 5ar4 with excellent results, never a failure in 30 years. I rebuilt a friends ST 70 and tried a JJ and it blew the fuse at start up. After several more tries with new fuses, I tossed the JJ and put a Sylvania in, problem solved. I have recently tried the Sovtek and found great results. Keep in mind that AVA needs a steady supply of new tubes, that's why he uses the Sovtek, as long as it works. I will do the rectified update this weekend. Thanks guys,

Frank, I am still wondering about the choke modification for the Super 70i, like or simular to the U70.

Dave

avahifi · « **Reply #7 on:** 24 Jul 2010, 12:51 am »

Dave, I will check out the Super 701 supply wiring early next week.

Frank

Charles Calkins · « **Reply #8 on:** 24 Jul 2010, 12:58 am »

God you guys are smart!!!!!!!!
When God passed out brains I was last in line. And I can prove it!!!!
No way I'd ever try to do what you guys do. My way is to buy it. Bring it home. Plug it in and enjoy!!

Cheers
Charlie

rcag_ils · « **Reply #9 on:** 24 Jul 2010, 03:48 am »

Thanks Frank, these are the things the I like, DIY that works, give me the feelings of accomplishment.

simon wagstaff · « **Reply #10 on:** 24 Jul 2010, 12:51 pm »

I know how to work a soldering iron, sort of.

I would say my skill level is average to below average at best. I had no problems with the previous wiring modification do want to say again that it made a truly significant difference in the sound, almost go so far as to say it transformed it.

I am happy to hear my experience with fuse blowing was not unique. I had a spare tube and once I popped it in it was all good again.

Thanks so much for the diagram. I could follow it via text but the drawings give me enough confidence to proceed. Doesn't look like a big deal and will order in some of the diodes right now.

thanks for the input on the tubes. I had read one or two posts where they seemed to think that the rectifier tube was the most important one. I will take Frank's word as the final one.

thanks again!

avahifi · « **Reply #11 on:** 24 Jul 2010, 01:27 pm »

I should note that we tried a small batch of JJ 5AR4 tubes a couple of years ago and had an unacceptable failure rate with them. I have not tried them again since then.

Does anyone have current experience with them?

Regards,

Frank Van Alstine

rcag_ils · « **Reply #12 on:** 24 Jul 2010, 02:35 pm »

I tried one several years ago, the pins were too small to make good contact, I coated all the pins with solder to make it work, but I haven't used the amp long enough to test for tube life.

Dynakitguy · « **Reply #13 on:** 24 Jul 2010, 03:00 pm »

Hi Frank & all,
I have been using this diode/rectifier mod now for several months on a stock ST-70, MK III mono's and the AVA U70 most recently. The MK III's have the JJ branded GZ34/5AR4 rectifier tubes. No rectifier tube failures to date. This is a simple modification which can be done by most anyone with good soldering skills. Be sure to install the diodes correctly as shown in Wayner's pictorial and follow Frank's instructions with respect to discharging the capacitors before proceeding.

Kevin @ Dynakit

Wayner · « **Reply #14 on:** 24 Jul 2010, 06:09 pm »

Hi Kevin! I see that www.partsexpress.com has these diodes for .09 cents each. So the cost burden is nothing to save some skin on the 5AR4 tube. I checked Radio Shack, but they did not list it. That kind of sucks to order 18 cents worth of parts from some e-supplier, but maybe you needed something else anyway, like a nice VOM to read your bias correctly.

Wayner

rlee8394 · « **Reply #15 on:** 24 Jul 2010, 08:41 pm »

Frank,

I ordered a pair of the Shuguang Chinese 5AR4 rectifiers from Jim McShane.

http://pages.prodigy.net/jimmcshane/tubes.htm

They measure identical to my vintage Mullard 5AR4s, and as long as you treat them nice, no rapid fire turn-on turn off cycles, they seem very reliable. They do have a brighter filament glow to them, but I think it looks nice. With the 1N4007 wired in series with the secondary AC power leads, they should be just fine. Give Jim a call, he'll probably be able to set you up with production quantity pricing. Of course, if you like them, you could probably deal with Shuguang themselves. Just a thought.

Thanks,
Ron

Opus Flatus · « **Reply #16 on:** 25 Jul 2010, 08:48 am »

I've also had good results with Shaguang (Valve Art) tubes, both rectifier and 6CA7 tubes. They have proved to be reliable, no issues. The 6CA7's were also cheap, $ 47 matched quad from AES. In the U70 I have not heard any discernable difference in sound (tube rolling) between brands -Svetlana winged C (poor reliability), JJ, and SOVTEC (poor reliability).

Wayner · « **Reply #17 on:** 25 Jul 2010, 05:35 pm »

I have finished the rectifier tube mod for both of my Ultravalve amplifiers. If I can do it, most of you can. The most trouble you will have is removing the red transformer wire, as they are mechanically wrapped (and soldered) on the lug. When soldering the diode, use a hemostat to wick off the soldering heat from the diode body, you know, the usual stuff.

The bias is different, as Frank mentioned, so give it some time to warm up after the mod and re-bias. I have it on for 35 minutes and is nice and steady at 1.56 volts.

Wayner

rlee8394 · « **Reply #18 on:** 25 Jul 2010, 06:48 pm »

Quote from: avahifi on 24 Jul 2010, 12:51 am

Dave, I will check out the Super 701 supply wiring early next week.

Frank

I just checked an old Super 70i board that I have, and there is no need to swap the B+ and transformer leads.

Ron

Wayner · « **Reply #19 on:** 27 Jul 2010, 12:11 pm »

Has anyone done the mod, then experienced an audio improvement? It sure sounded wonderful last night, listening to vinyl. Tough to A-B it, tho.

Here is a note on setting the bias. As some of you may know, I built a custom dual meter box to go with my Ultravalve. It is very interesting to observe the behavior of the bias during amplifier operation. I have come to learn several interesting behavior traits. First, at turn on, the bias will be relatively high, like 1.68 volts. During the operation, it slowly comes down to the more normal limits of 1.56 volts, or lower. It certainly reflects AC line voltage, changing .02 volts per every 1/2 volt change in line voltage. This brings up a good point. Bias should be set when the line voltage is at it's design voltage of 120.

In an ideal world, you would use 3 volt meters. One to measure incoming line voltage and the other two (one per channel) to set bias. When you adjust bias, you will notice with the two meters that as you increase the bias on one channel, the other one will go down and vice-versa. If the line voltage is at a true 120 volts, then the bias setting of 1.56 volts on both channels will be in true reference to the line voltage. If you leave your meters on, you will notice as the line voltage goes up or down, so does the bias voltage. The important thing to me is to have the bias setting for the two channels balanced. This would almost be impossible with only one meter. I suspect, (but maybe wrong) that this may have a very small role in tube degradation?

Wayner