The AM Forum

Well I received my two 572B's from RF parts today. Physically they looked good. I tried both of them in my test jig. The first one the filament did not appear equally bright along its length. I left it cook for awhile and tried putting current through it.

The test setup is to tie the grid and plate together and apply a regulated supply positive to the plate and negative to the filament transformer center tap. While measuring the applied voltage and current you raise the voltage so it draws 250ma. At that point the voltage on a good tube should be about 44-45 volts.

So back to the bad tube. It took 59 volts to achieve 250ma.  I then checked the second tube. The filament looked good., equally bright along its length. The current test was right on - 45 volts for 250ma.

So here we go. A call to RF parts tomorrow. These tubes were marked with equal numbers indicating they were tested and matched. I wonder how the filament could get that way. If they really checked them it obviously could not have been that way when they did.

Any thoughts on how a filament could get this way and would this be something that could just happen or is it always a manufacturing defect?

Doug,

   I have an 812A tube that is half lit. The filament is in the shape of a capital 'M'. On my tube the bottom ends of the 'M' are bridged together, such that 6.3V is applied between the bridged bottom ends and the bottom center.

  It seems that one side or the other filaments is open. It kind of sucks I know. I bet your tube was good when tested....its just that sh_t happens. Mine is an RCA.

  Maybe some sort of soft start on the filament could reduce these type of failures. Then again, it could have been open already after being shipped. Did they double box, and use soft packing material or bubble wrap?

Jim
WD5JKO

Jim,

  Yes it was well packed. There appears to be no physical damage.

Both filaments are lit but not evenly. Down in the center looks much brighter. It looks like the entire length has some brilliance but the top 1/2 inch of one and a little less of the other are dim. I should try to take a photo of it.

I was able to get some photos by manually adjusting the aperture and speed of my digital camera. I think you can clearly see there is uneven heating in the filament.

Doug.....in my experience.....RFP can be a bit squirrely when it comes to warrantee on these tubes.
They always ask a lot of questions to see of you know what you are talking about etc......I wouldn't tell them that you tested it this way.
Yeah, yeah....I know.....
But....that's my experience.
73

Yes, I was thinking the same thing Dave. I have heard that they stand behind their warranty but I will see.

My real question is could this phenomenon happen in shipping? They claim 100% testing and matching. My understanding was that this was done in CA but maybe not. Maybe they task that to the Chinese also.

If the testing is done in CA and this cannot be blamed on shipping then I don't care much for their supposed testing.

Its the luck of the draw. I have heard of many out of the box failures on these tubes and others have no problem. Often it is filament related.

I will say though that the majority of the appliance operator hams that buy these would just stick them in and hit it with RF and probably never know there was a problem until months later when the hot tube failed. Then they might have a warranty issue. 

 Doug,

  I am curious where you got this information? I compared what you say with the RCA 811 plate curves, and the Svetlana 572B plate curves. I post them here edited.

  The variation in color and brightness is likely due to variation in the tungsten wire thickness resulting in the filament being brightest in the thinnest areas. Less likely is a surface contamination issue obscuring the filament surface making it look less brilliant, cooler, and more orange.

  Possibly the tube you say is "Good" is the one with the thinning partial filament, higher temperature (more yellow), and greater emission. Perhaps this is the tube that will fail first?

Look at these curves...

Jim
WD5JKO

edit: Perhaps picking a match point where the slope of the plate curve lessens. Testing at +50 volts (plate and grid) is where the curves are near vertical, and therefore I'd expect large variations tube to tube.

Typo In Svetlana image...Where I put +45v in red, it should have said +65V.

Hy

is this tube china made,I have experience using 813 tubes ,after using a month the filament goes out,but after checking the filament pins at the tube,I have found out that the solders is bad,after filing this pins and after soldering again
The filament come to light again.

Maybe a bad solder connection at the tube filament pins ?

Gito

No, that wouldn't cause the filament to be half lit the way it is. Take the tube and look at the filament without it being lit and see if there are any physical differences along the length of the filament that you can see. I've never saw one do what they are doing in your pictures, but I have saw where they would be just slightly dimmer around where the filament wire bends around the hanger hook at the top of the tube.

Jim,

  The DC current test I did is not made up by me. It is from PA0FRI and is after testing many 572B tubes of differing types. I have tested at least four tubes here. I had two well used Cetrons that came in around 48 and 52 volts for 250ma. They both had filaments that were visually of constant brilliance.

The bad Taylor is way up at 57 volts for 250ma and that would coincide with poor emission which is what you would expect if the filament was not hot along its entire length.

http://pa0fri.home.xs4all.nl/Lineairs/SB200/sb200eng.htm#TUBES

Actually that link should be -

http://pa0fri.home.xs4all.nl/Lineairs/SB200/sb200eng.htm#REACTIVATION

  Doug,  That is an interesting website for sure. I wonder if you or others have the actual Cetron published plate curves? Also the curves for the RF Parts 572B (Chinese) tubes would be interesting to see as well. Of the two plate curves I posted, neither one would pass the paofri test. It's interesting to me that the old RCA 811 as shown has a much higher pervience cathode than did the Svetlana 572B. I always thought the Cetron 572B was like the RCA 811A with a graphite plate.

Jim
WD5JKO

RF parts actually mentions PA0RFI's site at their 572B page -

http://www.rfparts.com/old_site/572bprices.html

I have been happy with the tubes I bought in the past from K5SVC but he doesn't appear to have 572B tubes anymore. 

RF Parts really seems like a crap shoot in terms of getting properly working tubes.

Has anyone purchased tubes from MFJ/Ameritron?  I haven't heard complaints about infant mortality failures in their amplifiers and I have been happy with the price and service for RF components from them.

I checked a couple of old 811's that I had sitting around. One was General Electronics name brand. It reached 250ma at 42 volts. The other one was an RCA and it clearly had one filament out. Again I am not sure how that happens unless something falls across it. The voltages and current seemed right though. Are the two filaments in series or parallel inside? Anyhow as expected this tube maxed out my low voltage supply. It reached 236ma at 63 volts. Obviously a show piece only.

I am not sure if the curves are that accurate at the low voltages. They are pretty step there. Remember I am tying the grid and plate together so they are at exactly the same potential.

It would be interesting if others with junk box 811's/572's ran these tests. I would say that anything over about 48 volts at 250ma is probably a weak tube, although if you happened to have two that were match at that level they would probably work OK.

Just got off the phone with RF parts. There was absolutely no problem in getting a replacement pair which they are shipping today.

The way the gentleman was talking they must sell a large quantity of these tubes. They are already in a different lot since last week. I asked about their testing procedure. While they would not tell me how they measure emission they did say that they do 100% testing at San Marcos. The first test is hipot and if the tube passes that they go on to emission and filament testing and then they sort by both physical size and emission. The physical size was surprising. He told me that these tubes are made by hand and that each one is somewhat unique so they must have a sizing issue also. I'd love to see the plant where these are made. Probably some dump with kids working.

I just hope the new set is good. The lot is different but if the emission is the same I could at least make a match pair out of the four I will have before I ship two back.

I still find it hard to believe that this condition could happen in shipment.

Chinese industrial talent seems to be in the newest, high-tech area now.

Phones, HDTV's, eTc.

The retro aspects of technology are not done too well, especially with tubes.  So far, everything I see has filament failure (low emission) in a short while.  I don't believe they have the correct formulas and/or processing for good cathode making.

The tubes from the old Iron Curtain really hold up well, 'cause they didn't copy our designs, they STOLE them and did the same thing we did.  Maybe try and source something Russian from Sovtek or Reflector.

73DG

These aren't tubes from the 1970's RCA plant.  Get over it.  ;)

From discussions over the last couple years on the Zed a picture has emerged...

-most tubes are Chinese these days. They have dirty shops with poor QA, short vacuum draw downs.

-RFP buys tubes by the thousands and at least 50% fail basic checks.  They further test, grade and label them with different brands by quality/price point.

-Even after grading Chinese tubes have a significant failure rate.  RFP compensates with a liberal warranty.

*Don't worry, be happy.  Get a warranty replacement and move on with life!

"Well."
. "(Uttered in George Will's most stentorian voice)
I can't imagine the audio  aficianados up putting with this lack of QC very long. ;D

If you have no other choice than to deal with RF Parts, then thats the way it is I suppose.

I have always been able to find NOS power tubes from 6146s to 4-1000s at the hamfests here in Ontario over the years. We have 13 of them between Feb. and November. Having a good network of radio friends who know what you are looking for, also helps.

Al VE3AJM

 

+1!

73DG

Those audio guys may be crazy, but they aren't stupid.  

They will easily spend $500 for an NOS RCA 845 vs a Chinese copy for $50.

They know the difference, and only use the crap tubes when they can't afford the good stuff or are selling off some gear.

Going out on a limb, I personally believe guys that seriously buy Chinese tubes are ignorant or just lazy. 

They don't know how to find good tubes, maybe the "life is too short to really look" syndrome is their way, or they still think a 'warranty' is all they need to enjoy good tubes.

PFTTT!

73DG

How about $976.50 (on sale) for a matched pair: http://tubedepot.com/kr.html

Then, there's the Shuguang 845 for $300 a pair: http://www.sophiaelectric.com/shuguang/shuguang.htm

I don't remember who it was, but one of the 'usual suspects' selling tubes on the net recently had China 845 copies @ $50 per.

You get what you pay for...and it ain't about pretty...

73DG

Ameritron uses "tested" Taylors and has a lot of problems with low emission and Ive had several failures at turn on. Some customers demand Taylor.

I understand that one of the Chinese companies that made an excellent 572B but quit to concentrate on audiophools is starting a 572B line again. I have no other details but for now Ive been buying Shunguan direct from China 4-8 at a time with no problems....yet...over several years.

Im wondering if RFP is buying as cheap as possible and knowing they are getting a lot of crap and then culling in CA and playing the odds? Meanwhile buying direct gets the best quality as the Chinese are fussy about saving face

It sure seems like a market ripe for an American boutique or niche operation.

What is needed is for someone to make Richardson an offer they cant refuse and walk out with all the Cetron tooling.....paid for or not. I hate that outfit.

Carl

That and know how to make better tubes than the Chinese 8).

It ain't easy, running a tube line.  Charlie Whitener did a good job buying the name and re-creating the WECo 300B, than went belly up before the payback came.  I guess all the machinery is still down South, languishing.

73DG

Every once in awhile I have these dreams about what I would do if I were a multi-billionair. I was actually thinking about this today. Maybe take 10 or 20 million and grab the few remaining tube guru's and buy a shuttered US tube plant and make tubes that people are still using and need. If it broke even or lost money no worry. I could pump a million or two in every year. 

Let me know when you become that billionaire you dream about.  I can then apply for a job at that tube plant you're going to buy.

But, don't wake me up till if happens.

Fred