• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Rectifier tube 5AR4 vs. 5U4GB?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've got a Ruby 5AR4C rectifier tube in my Simple SE. I'll assume it's Chinese in origin, but it seems to work fine. I have no complaints about it. With a Hammond 374BX power transformer, I get about 450V B+. Total power draw is probably somewhere around 142 mA (61 mA per output tube plus 10 mA per side for the 12AT7).

I picked up a couple 5U4GB tubes for dirt cheap, and I'm thinking of trying them out. Would anyone care to offer their opinions on the pros and cons of each? I expect the 5U4GB will drop about 35 more volts, but I don't think I'm concerned about that. I think I have volts to spare. I also expect the 5U4GB will draw more current on the 5V winding, but the Hammond is rated for 3A so I figure I'm OK there.

I'm curious about the directly heated cathode of the 5U4GB vs. the indirectly heated cathode of the 5AR4. Will my B+ come up "too fast"? Does it really make any difference at the end of the day?
 
You're correct on both counts... and yes, the B+ will come up before the input and output tubes heat up. This will give you a surge on the filter caps (no load) so be sure the caps can handle the surge. I prefer to use either a 5V4G or 5AR4/GZ34 as they typically have a good 10 second delay before B+ builds up and it builds slowly. Your current draw is high for a 5V4G so I'd stick with the 5AR4.

Regards, KM
 
kmaier said:
...and yes, the B+ will come up before the input and output tubes heat up. This will give you a surge on the filter caps (no load) so be sure the caps can handle the surge.

I dunno... it's kinda iffy. The caps are Panasonic TS-UP series. Their data sheet says 500VDC working, 550vdc surge. It doesn't specify the duration of the surge. It looks like my power supply would peak out around 558 volts if it were completely unloaded. I probably ought to just stick with the 5AR4.
 
You could always put a CL90 thermsistor on one of the transformer primaies. That should help out with the surge.

JD


Ty_Bower said:


I dunno... it's kinda iffy. The caps are Panasonic TS-UP series. Their data sheet says 500VDC working, 550vdc surge. It doesn't specify the duration of the surge. It looks like my power supply would peak out around 558 volts if it were completely unloaded. I probably ought to just stick with the 5AR4.
 
You could always put a CL90 thermsistor on one of the transformer primaies. That should help out with the surge.

A CL150 between the rectifier's filament and the PSU filter is (IMO) better. B+ rise is slowed down a tad and a bias supply (if present) still turns on immediately. If the 5 VAC winding has a CT, that's THE place to take the B+ from. B+ from pin 8, when types with cathode sleeves are employed. B+ from a CT, if available, when directly heated types are employed.

A word to the wise, a 5U4 can't tolerate as large a 1st filter cap. as a 5AR4 can. Be cautious.
 
Eli Duttman said:
B+ from pin 8, when types with cathode sleeves are employed. B+ from a CT, if available, when directly heated types are employed.

A word to the wise, a 5U4 can't tolerate as large a 1st filter cap. as a 5AR4 can. Be cautious.

How curious. I suppose it makes sense, although I never gave it any consideration before. The 374BX does have a center tap on the 5V winding. Of course, right now the power supply is wired to pin 8 and the CT is just taped back. What do I lose if don't take your recommendation?

The GE datasheet says 40uF maximum for the 5U4-GB. I guess I should take that seriously, huh? I've got a 47uF input cap right now...
 
I built Eli's "El Cheapo" push-pull design (which is excellent sounding) and "rolled" a bunch of rectifiers to try a tubed design in place of the the original solid-state setup.

For what it's worth, here are my notes on the 5R4GYA, 5U4GB, and 5AR4/EZ34:

The 5R4GYA dropped the most forward voltage, but offers excellent bass and treble, with plenty of clarity in the mids. Stunning sound. Don't try to jack up the input cap to get more B+. Mine arced above the max given by the spec sheet. Use a higher voltage secondary instead.

The 5U4GB came in second and provided slightly higher B+ (less forward drop). Laid back mids, but nice detail with sax and breathy vocals. The bass was full -- a little too much so, almost "loose" in this application. Overall, very easy to listen to, with very little sibilance on some of my borderline female vocal recordings.

The sound of the 5AR4/EZ34, which I ranked third, was similar to the 5U4GB. Bass was more under control, but the setup was less "involving." I'd call it "okay."

I also tried a parallel set of 6CA4s. Leaner bass than all the above, but tight, clean sound overall.

Here's the schematic I used for testing. (Be sure to click on the image to expand it to view at 100% if your browser shrinks images to fit.)

Now, with all of that, I should add that I decided the original solid-state rectifier sounded the best in this amp, so I'm saving these tubes for future projects.

--Jeff
 
I have been using my simple se for well over a year now. Have tried GZ34 and 5U4GB. Use the 5U4GB when I want to lower the voltage. Have standard recommended parts from the tubelab site and using the same Hammond transformer. Have had no problems with any of these rectifiers since getting rid of the standby switch. I think I might try the thermistor on the rectifier filament as per Eli's recommendation...
 
You should hear less differences between rectifiers in the SE amp since the current draw is more constant.

I have stuffed a bunch of different rectifiers in my Simple SE. Yes some 5R4's will spark out with a 47uF cap. I even had one of those "potato mashers" from the B52 bomber spark out (non fatally). All worked with a 22uF cap. Only one 5U4 sparked, and it still sparked with the 22uF cap so I assume that it was a bad tube. A 5V4 worked without issue.

I got the most voltage from the 5AR4 and all of them came within 5 volts of each other (Chinese, Russian, RCA, Sylvania, and National). The 5U4GB's cost about 20 to 30 volts. Most of them are used tubes of unknown origin. The 5V4 dropped 30 volts compared to the 5AR4, but I only have one of them. It is NOS though. The 5R4's cost 25 to 40 volts, and all are used, mostly pulls from military surplus.

I can hear some differences with my preference going toward the 5AR4, but I suspect these differences are being caused more by the voltage differences than the actual rectifier tube.

In theory the faster voltage rise of the 5U4 can cause cathode stripping in the output tubes. I have not seen this happen in practice and plenty of amplifiers (including Simple SE's) have been made using solid state rectifiers. The CL90 is a good idea, and I have been using them in some of my amps.

A CL150 between the rectifier's filament and the PSU filter is (IMO) better. B+ rise is slowed down a tad and a bias supply (if present) still turns on immediately.

You can wire it in series with the power transformer HV winding's center tap for the same effect. I have put them in both places on some amps. The Simple SE is cathode biased, so there is no bias supply.

If the 5 VAC winding has a CT, that's THE place to take the B+ from.

This requires cutting a trace on the PC board and is probably not worth the effort. It can lower the ripple voltage at the input cap by 2.5 volts if the rectifier tube is perfectly balanced (not likely).
 
A while back I thought I remember reading that the CL80 was recommended... which I believe is the part installed in my Simple SE. How does that compare to the others recommended here? I have one in the B+ CT and one in the primary of the power transformer.
 
I use the CL-80 and CL-70 mostly

So do I. I have the CL-80 in one of my Simple SE's on the primary of the transformer. The 90 gets pretty hot in a Simple SE with fat tubes (EH KT88's) when biased hot (near 100 mA).

I think that these things work best when the average current through them is about half their rating. If operated too cold (not enough current compared to their rating) they never "warm up". Their resistance may remain too high leading to poor power supply regulation. If operated too hot, they may fail. The failure is often ugly involving smoke. It usually takes a lot to blow one, but I did have one explode. The exploding CL80 involved a shorted filter cap and a 20 amp fuse, don't do that!
 
I stuck in an old GE 5U4GA. In another thread I've been noting some of the other tubes I've been playing in my Simple SE. Some of them seem to be on the edge of their dissipation limit, notably the Reflektor 6p3s-e and some brown base Shuguang 6L6GC.

I put in the 5U4 to try to drop the voltage a little, and hopefully ease the burden on the output tubes. I think I succeeded here. Measured supply voltage just upstream of the PSU choke is just under 420 volts DC.

I had the voltmeter attached while I powered it up. As expected, the voltage comes up within seconds of hitting the power switch. I wish I had paid closer attention to the peak voltage, but it definitely went way past 420 and came back down again after the output tubes heated up. (edit: saw a peak of 532 volts on the digital meter)

I suppose the initial surge voltage puts a little more abuse on the power supply caps, but overall it reduces the punishment on the output tubes. I'll probably experiment a bit more and see if I feel it affects the sound.
 
Stupid question... I can add a switch on the 5v filament winding going to the rectifier? Leave it open when I first power up the amp. Give the output tubes twenty seconds or so to warm up, then close the switch. Then the rectifier warms up quickly and starts to conduct.

Or another idea would be to add an IRCL in series with the 5v winding. Might be problematic, since the IRCL should be sized according to the current draw, and that varies from one type of rectifier tube to another...
 
Heating the Outputs before the rectifier filament may make things worse for a directly heated rectifier, since it will face a big current demand during its warmup. 5R4GYs will arc-over very easily in that case.

Check out the RCA 5R4GY data sheet - it suggests warming the rectifier filament for 10 seconds before applying the HT, when higher loads are to be served.

How about a Standby switch in the HT trafo secondary's centre-tap? Power-ON, warm up the filaments & heaters in all the valves, then switch in the HT. If you also use the CL-80, that will be the kindest arrangement for all the components.
 
I am interested in Jeff's observations on the sonic effect of different rectifier tubes. Can anybody enlighten me as to what is causing sonic differenceces? All I can think is the degree of voltage "sag" that each tube has - or have I missed something?

And I also wonder where SS diodes (e.g. the FREDs as optionaly used in the SimpleSE) fit in terms of sonics.

This is of pactical interest to be as I embark on a SimpleSE build.

Cheers
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.