|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
167.192.191.109
I'm having a "flash over" issue with a couple JJ EZ81 rectifiers in a vintage application. The amp otherwise checks out fine, current draw, etc. is all correct, completely recapped, etc. The flash over occurs 8-10 seconds after power-up, and I am thinking it is the tubes, but wondering if anybody else has seen a similar problem.
Follow Ups:
Thanks for everyone's great suggestions. Working with my brother to get this resolved.
Gary
@BofService
Any news on your Universal?
******
http://rh-amps.blogspot.com/
Nope. I'll email you.
Gary
I've used RFT, GE, and Sylvania 6CA4/EZ81 rectifier tubes with NO problems whatsoever.
Steve
You can buy NOS GE or RCA 6CA4, at good price. A friend has the GE 6CA4 in a headphone amp. Has been reliable.
Likely a capacitor input filter?
How many uFs????????????????
How does that compare to ratings ???
JM
I may have moved up to a 47 from the stock 40 on the first cap.
Thanks.
JJ and others like Philips in 1960's, rate 50 uF as the suggested operating amount of uF, and the maximum. Can you TRY another EZ81 tube in there, maybe your JJ is off spec.
Edits: 04/08/15
JJ rectifiers, along with their Octal signal tubes, have serious reliability issues. Get the Jamona Junk out of there, ASAP. Contact AA sponsor Jim McShane or other reliable vendor and get good EZ81/6CA4 specimens.
To be fair, JJ's Noval signal tubes seem to be much less problematic. Their EL84 is good and people say good things about JJ's 12AT7 and ECC803.
It's possible that the series SS diode tweak linked below could solve the arcing problem, but decent EZ81s are available at relatively reasonable cost. OTOH, the 5AR4/GZ34 situation IS dire enough to make the series SS diode tweak truly important.
Eli D.
Eli, the 5AR4/GZ34 Tung-Sol and Genalex reissue tubes are FAR better. The need to hunt down and pay crazy prices for old stock is no longer needed IMHO. If you've got a 5AR4 equipped preamp or some other moderate voltage/current draw device then the Sovtek or Shuguang is fine. But I've put a bunch of U77 Genalex and 5AR4 Tung-Sols out (some into tough duty!) and they have been flawless.To repeat - the new reissues are a MAJOR step up as compared to the Sovtek/Chinese tubes.
I don't have much experience with the EH 6CA4/EZ81 since I still have good stock of RFT on hand. But JJ?? Look at the track record and you understand why I shy away from the vast majority of their stuff.
Edited to correct poor grammar.
Edits: 04/08/15
I got some RFT EZ81's from Jim awhile back great tube....Dan
nt
but to use a cap input of 50uf, the EZ51 requires between 150 and 310 ohms in series with each plate.
With a critical inductance choke input, there is no series resistance requirement.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 04/08/15
Thanks Tre. I assume I would measure between the plate and the CT of the transformer? (not "end to end"). I don't recall what the resistance of the secondary was, but I bet it is well under 300 end to end. Probably under 200. I don't have the amps in hand but that is something we will have to look at.
Just looked at the Sams for this amp. Each EZ81 has 2 plates connected in parallel, and each tube is connected to opposite ends of the transformer secondary. The resistance from each of those plates to ground is shown as 45 ohms. The AC volts on each side of the secondary is 300vac, and the resultant B+ is around 360vdc.
If I am reading the chart correctly, the design of the amp exceeds the spec on the tube quite a bit as it should have at least 150 ohms per leg of the secondary. Even an 8uf first cap would not have worked. Or am I missing something?
"If I am reading the chart correctly, the design of the amp exceeds the spec on the tube quite a bit as it should have at least 150 ohms per leg of the secondary. Even an 8uf first cap would not have worked. Or am I missing something?"What Steve said and....people have been abusing rectifier tubes for a long time.
It's not that it won't work......but those tubes are been hammered on (so is the input cap).
Critical inductance input choke and the hammering goes away.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 04/09/15
Since the two sections of each rectifier are paralleled, the min series impedance is 75 ohm, not the 150 from the chart. The stated sec resistance on the schematic is ~40 ohms so internal impedance is still less than spec. One detail often overlooked is that the primary resistance as reflected to secondary should also be added in to the meas secondary resistance. In this case it might add 10-20 ohms to to the 40 already present. One possible solution would be to add say 20-50 ohms resistor to each sec winding. Another would be to add an NTC IRL to primary for added reflected impedance.
"add an NTC IRL to primary for added reflected impedance."
Might not be the optimum solution... he did say the flashover occurs 8-10 seconds after power on. The fixed resistors you suggested would probably work, and I think he should try a different brand of rectifier tube as well (per Jim's advice).
--------------------------
Buy Chinese. Bury freedom.
Flashover occurring 8-10s after power on is about when the tube begins to conduct. At this point the IRL will still possess considerable resistance because it's not fuly warmed up. Whether or not this prevents startup flashover using substandard rectifiers is unknown. But there are other benefits even if it doesn't prevent the flashover so it would be one of the first measures I'd consider if the amp were mine.
nada aqui
--------------------------
Buy Chinese. Bury freedom.
You could try putting some 1N4007s in line with the high voltage leads if it is a PIV issue
Hmmm.
Here are links to a schematic and the tube data sheet. I'm not technical enough to understand exactly what is happening when the tube first starts to conduct, but looks like it is well within limits at steady state.
http://vintagevacuumaudio.com/schematics-manuals/eico/eico-hf-30-service-manual-schematics.pdf
http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6CA4.pdf
Thanks!
"Here are links to a schematic and the tube data sheet. I'm not technical enough to understand exactly what is happening when the tube first starts to conduct, but looks like it is well within limits at steady state."With a cap input filter each diode is only on for a small portion of it's half cycle so there is no steady state. (turn on surge is a separate matter)
50uf is fine as long as the "total plate supply resistance per plate" is met.
(see page page 2 of http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6CA4.pdf and my post above)
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 04/08/15 04/08/15
With solid state diodes in parallel, the B+ will be INSTANTLY on, before other tubes warm up their filaments.
JM
He said SERIES. I posted a link in another branch of this thread.
Eli D.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: