Thoughts on dropping B+ in Ampeg V4 (another blasted V4!)

[nsubulysses]

Well I have received a V4 with a shorted power tube that burnt the hum balance pot to a crisp. It was retubed and biased and lasted about 6 practices (not by me!). A plate resistor has exploded and although new 470 ohm screen grid resistors were installed they do seem pretty heat stressed. Also some time long ago the power transformer was replaced. The B+ is about 560V! the high plate voltage is not beacuse of very cold bias. It just is.

I was thinking of dropping the B+ to more regular V4 levels (520-540V) to try to make it easier on the tubes. Never done this before. I know some say put a power resistor in series with the OT center tap. ANy thoughts on this or other methods that are better? I will also replace the 470 ohm screen grid resistor with 1K/5W.

I don't want this guy to have his amp blow up again. ANy thoughts or suggestions on this please chime in!

[olddawg]

An inexpensive way is to put a few 10v 5watt Zener Diodes in series with the B+. They do get warm and you might need to suspend them zigzagged on a perf board. I have been told you can put them back to back in the center tap too, but have never done that.

[nsubulysses]

Thanks olddawg!

Another though I have is if the B+ is so high already, would it just be negligible to drop it 20-40V down from 560V?

[g1]

Was it actually biased cold? If not, that could be a problem right there.
Seeing as it's a replacement PT you don't really know what's what, it could be all the voltages are hot. In that case it is better to drop the AC to the primary with a bucking transformer. A good check is to look at the heater voltage. If it is not high, then yes, just drop B+.
But I think 560V is probably fairly normal for these with modern line voltage approaching 125V.
Could also just be a bad tube, or bias too hot.

[tedmich]

I think if you put it stock and get some decent 6L6GC-STR Tung-Sol's it should be fine. Jim McShane will test them at the exact B+ for not much $
see:
Tube & Information Page

[nsubulysses]

Thank you for mentioning Jim McShane. I just heard of him yesterday in another thread when Leo Gnardo was referencing him. I will definitely need to use his services in the future

The tubes have the JJ label blackened and unfortunately the customer doesn't want to buy a new quad since he just did that a month or so ago. I may end up throwing in a used JJ 6L6 and let him take his chances. My tube tester cannot test 7027s unfortunately, but they've certainly been through the same abuse that made the other one short

I will check filament and other voltages and see if they're all high and take it from there. Thanks for your input y'all

[nsubulysses]

Filament voltages at the power tubes are 6.4 volts and a little further downstream in the preamp they are at 6.1V

Bias checked to be about 36mA per tube at 553V this time around. Basically idling at 19W, so about 66% plate dissipation. And this is just with a quick turn on and check so I would assume it would drift slightly up if I left it on for 20 min. Maybe the hot bias and the 470 ohm screen grid really did it in. If I bias this down to the 50-55% range I assume I will probably exceed 560V

[nashvillebill]

What if you went with 6550's instead? As long as the PT has enough heater current.

[tedmich]

He can't try other tubes as the customer just bought the JJs (that likely killed it) a month ago...its almost always the tubes, fixing everything but the tubes is rearranging deck chairs on the Titanic IMHO, I hope it works...

[g1]

The heater voltage sounds fine, so it's probably just B+ that needs to be dealt with.
Sounds like someone biased it for the "universal" 70% which is not so friendly to tubes running on voltages like that.
Agree with your 50 to 55% range, then see what you need to reduce B+ by.

[nsubulysses]

What if you went with 6550's instead? As long as the PT has enough heater current.

Have you had good experience with this? Unfortunately with my own V4 (72-75 era) I tried running 6550s for a couple months and it broke down every handful of shows/practices. It was less reliable in my experience because the higher current draw of the 6550s compared to 6L6 made everything run even hotter. They were Sovtek 42W tubes. I eventually gave up and went back to JJ 6L6GC and haven't had a problem in about a year.

Mine wasn't a good candidate for 6550s though I think because a section of the PCB was burnt out around the screen grid resistors and with that section gone I think a lot of the heat-sinking from the screen grid resistors to the PCB was lost. I would occasionally have 1K screen grid resistors falling out because the solder would melt. Makes me want to try 6550s again though in my own amp, now that I'm thinking about it. ........biased really cold

I could see how 6550s would resist shorting from the high voltage better but my V4 in particular just couldn't stand the heat.

[nashvillebill]

I haven't tried the 6550's in mine--I was tempted to, but they are so close to the transformer, I was worried about both the heat and the increased heater current. Mine's the early version too. So in the end I just bought the Sovtek 7027's (yes I know they are supposedly just 6L6's).

[Chuck H]

Bias checked to be about 36mA per tube at 553V this time around. Basically idling at 19W, so about 66% plate dissipation.

Not accounting for screens (which I don't usually see when doing the math for big bottles, not sure why) I get 36 x 553 =19.908. If the JJ's are actually 6L6 types in a 7027 bottle that would make their actual max plate more like 25W. So that's about 80% dissipation. At 553V that might explain the brown paint.

Have you had good experience with this? Unfortunately with my own V4 (72-75 era) I tried running 6550s for a couple months and it broke down every handful of shows/practices. It was less reliable in my experience because the higher current draw of the 6550s compared to 6L6 made everything run even hotter. They were Sovtek 42W tubes. I eventually gave up and went back to JJ 6L6GC and haven't had a problem in about a year.

The 6550's won't draw more current by virtue of being bigger tubes. They only do that if you tell them to (bias, drive voltage and load). All things being equal (-ized, as in the bias MUST be adjusted) the 6550's will draw the same current as the 7027's or 6L6's. So as a bigger tube it should be more reliable. On the other hand, if you just plug them in without any adjustment they could well draw too much current and cook things in the amp. Creating the illusion of lower reliability.

A contributor here could be some assumption that if you buy 7027 tubes from a respected supplier they will be up to the job. Not so IMHE (with el34's) I absolutely could not get any Rusky tubes to handle 465Vp. Three different pairs and three different failures. I just don't know that any of the tubes coming out of that region are at all what they're supposed to be by their nomenclature. I've seen some vague data sheets, but not much comprehensive info really. Prior to about ten years ago there were some more complete data sheets that were clearly copied from old euro or US data sheets and were certainly inaccurate. I just don't at all trust these tubes to be up to the specs of their predecessors of the same number.

I ended up ordering Ruby tubes (Ruby tested Shuguang tubes I think). One is a little microphonic but at least they didn't flash and smoke. I can only suggest Ruby 6550 with a corresponding bias adjustment. Putting old tubes in the amp to get by and hoping for the best may not lead to a happy ending. OTOH if those 6L6's have already proven to work in one of these amps they may drop in, draw safe current and be fine for awhile. But short of trial and error, which is too often the case with modern euro and Russian tubes, IMHE the Ruby product will be much more likely to live up to the specs of the moniker printed on the bottle. Not that other tubes sound bad... We just don't know what the real specs are and it sort of sucks when a tube carrying a designation that should be able to handle 800Vp fails with every effort at only 465Vp.

Sorry, but I think the real repair here is a proper set of tubes with a proper bias adjustment.

[Leo_Gnardo]

I haven't tried the 6550's in mine--I was tempted to, but they are so close to the transformer, I was worried about both the heat and the increased heater current. Mine's the early version too. So in the end I just bought the Sovtek 7027's (yes I know they are supposedly just 6L6's).

Just for a lark I did try 6550's once in a V4. They worked OK, delivered good power but no bonus watts. And the glass was slap up to the transformer iron. I didn't worry about heat so much as the glass getting smashed when the amp gets carried from here to there. Out went the big tubes. 25 years ago when I tried this, NOS 7027 were big bucks, and there weren't any supposed 7027 copies from JJ or Sovtek. I wound up using EL34. At the time you could get good tough EL34 labeled Siemens but I think they were really RFT from East Germany. Back then it was rare to see any of the V-series Ampegs. Now they seem to be popping up often, and my fave is the V2, you can get a lot of the Stones' sounds from the 70's, Exile thru Some Girls .

[nsubulysses]

Not accounting for screens (which I don't usually see when doing the math for big bottles, not sure why) I get 36 x 553 =19.908. If the JJ's are actually 6L6 types in a 7027 bottle that would make their actual max plate more like 25W. So that's about 80% dissipation. At 553V that might explain the brown paint.

The 6550's won't draw more current by virtue of being bigger tubes. They only do that if you tell them to (bias, drive voltage and load). All things being equal (-ized, as in the bias MUST be adjusted) the 6550's will draw the same current as the 7027's or 6L6's. So as a bigger tube it should be more reliable. On the other hand, if you just plug them in without any adjustment they could well draw too much current and cook things in the amp. Creating the illusion of lower reliability.

Chuck, I don't understand what you mean. JJ 6L6GC are 30W tubes. I thought old 7027s are listed as 35W, but the JJ 7027 is just a 6L6GC re-labeled, so shouldn't it also be 30W?

Also if a 6550 is either 35W or 42W plate dissipation wouldn't a 6550 always draw more current than 6L6GC or current 7027, BECAUSE you need to adjust bias. You could bias a 6L6GC to 60% and throw in some 6550s and probably leave it idling around 30-40% and call it the same current draw but no one would do that because the tube is way underbiased. To be at the same operating point as far as idle dissipation you would have to increase the current draw, and since the tube is higher plate dissipation won't the current draw always be higher, since change in plate voltage when biasing is minimal.

Seems that at 550Va a 6L6GC should idle between about 27-38mA and a 42W 6550 should idle between 38-53mA

Also, are you saying that JJ 6L6GC is not really a GC and the 30W rating is generous?
Or are you suggesting to use 6550s and bias them extremely cold (40%?) for reliability?
Not saying your're wrong obviously. I am just confused by your post.