Still in need of cooler 6L6's....advice please

[Garydean]

I'm still working on a MIJ bass amp from the 60's.... yes I know it's been 2 weeks ( on and off though ). There is a 4ohm tap off the OT that is not being used. The wire is dangling inside the chassis. Currently the 8ohm tap is connected to the speaker jack and functional. I was wondering if disconnect the 8ohm tap and swap it with the 4ohm tap and run a 4ohm speaker would this cool my output tubes or vice-versa or no change?

Lafayette Bass Head (Univox 1236).

My main concern right now is, "slight" orange plating on the two 6L6's when tested. At idle (line voltage at 116V) the plates are reading 660V. I've been through the amp lifting and replacing components that are off spec by 20% or more. My line voltage on the bench is 123V and the schematic measurements were performed at 117V. I know this will bump the 6L6 plate voltages up about 40V+. I tried to adjust the bias at idle to make them cooler, but I've pegged the pot, no room left. Does biasing them hotter at idle make them run cooler when they are pushed (thought I read that someplace)?

I wouldn't usually be "as concerned" with an old amp made in Japan, figuring, it is, what it is and leave well enough alone but.... the amp has a very unique sound which is actually quite pleasant and worthy of using on smaller gigs.

I'm not savvy enough at this point to alter the circuitry enough to cool things down for modern day usage. Soooo.... I would most appreciate any knowledge given forward, from the guru's within this forum. Many thanks, once again.

Regards,

Gary

http://i7.photobucket.com/albums/y275/garydean57/univox.jpg

[Twist]

Changing the speaker load and tap wont change your red plating. You need to increase your bias voltage potential ( for example... -32 volts to -42 volts). According to your schematic the two cathodes of the output tubes share a 10 ohm resistor to ground. For a 660 volt supply each tube should be idling around 20 ma. Adjust your bias till you read about 400 milli volts across the 10 ohm resistor. If the tubes are matched that should be around 20 milli amps per tube. If you can't adjust your bias higher then increase the resistor values in the bias supply (33K to 47K). Hopefully that will do the trick. Remember this is just my advice. I'm sure someone will probably disagree with me. Be careful and good luck.

[Garydean]

Thanks Twist for the advice. Would changing the 10ohm cathode resistor value be another way? The reason why I ask is, when I got the amp in for repair, someone had previously removed the original 10ohm 2w resistor and replaced it with 2 10ohm 2W sand resistors in parallel to yield 5ohms. I removed that mod and went back to what the schemo called for. I hadn't taken any readings with that config.

[Twist]

I would leave the 10 ohm in there and work on the bias voltage levels.
Try a 50KL pot in place of the 33K resistor (in the bias supply) and instead of grounding one leg of the pot, lift it off ground with a 15K resistor. It might not be a bad idea to also replace the 10uf 50 volt cap with a 100uf 100 volt cap.

[Garydean]

nice idea... I already changed the 10/50 to a 10/150, actually, I replaced every cap on the board.

Thanks again!!

[Chuck H]

At 660vp your going to cook those tubes pretty fast, even if you bias them cool. In fact it sometimes happens that the cooler you bias them, the higher the vp will go, making things even worse.

Are you sure the amp is designed to run 6L6s? Where did that schematic come from?

Doesn't matter. I'd change to 6550s or el34s at least.

Chuck

[sportster4eva]

Maybe use the pink tap on the PT secondary instead of the red? Unless it's voltage is a significant amount lower, it might get you back in the ballpark for B+. Man, with 660v and a 10k primary on the OT, the voltage swings on the plates of the 6l6's must be pretty radical when it's really cooking...:eek:

[Bruce / Mission Amps]

At 660vp your going to cook those tubes pretty fast, even if you bias them cool. In fact it sometimes happens that the cooler you bias them, the higher the vp will go, making things even worse.

Are you sure the amp is designed to run 6L6s? Where did that schematic come from?

Doesn't matter. I'd change to 6550s or el34s at least.

ChuckWell I wouldn't panic yet... the 6L6's screen voltage is half the plate voltage in that one.
I have found that those old Univox chassis amps used some inferior coupling caps and many times they would be leaking DC to the grids of the next stage... sometimes that would mean the power tube grids.
If you have those odd shaped, pukey greenish colored caps, I'd just replace them all with any generic plastic cap like a Mallory 150s, Xicon MPPs or even WeberVST caps.
Regardless, I would rebuild the bias supply so I had at least -45vdc to -55vdc available.
Mess around with the bias supply voltage divider.
Remove that 6k2 and 33K resistor.
Try something smaller for the 6K2, like 1k5 or 2k7
Sub that 33K for a 25K trim pot in series with a 15K to 20K 1/2 resistor.
Drop the small value 10uF filter cap and use at least a 100v, 47uF to 100uF cap instead.
**************

Sorry.. I see you've already done most of that.

[Garydean]

Thanks everyone!

To answer a few questions.

6L6 is stamped on the chassis.
Radical?..... well, only played it for a few minutes to see where the voltage was going to end up.
All the coupling caps were replaced with 150's.

I'm liking the idea to first get the bias supply rebuilt so I have some room.

thanks again!!

[Enzo]

What is the actual bias voltage AT THE SOCKETS? Is it the same as at the supply?

[Chuck H]

O.K. perhaps you still need to get the bias voltage dialed in. So when that's figured out...

What could possibly be the harm in NOT running 6L6s at 660Vp. I've never tried it. Maybe it can be done, and I'm always up for an education. But even with the screens low, isn't that about 200 volts too many on the plates? Maybe the amp can be made to work ultimately. But I wonder if it can ever be ideal with that kind of scenario. I'd change to 6550s anyway.

Chuck

[Enzo]

They arc or they don't. Does that model or series have a history of blowing up a lot?

If Leo Fender can run little 6V6s 100 volts over specs, well...

[Garydean]

http://i7.photobucket.com/albums/y275/garydean57/DSCF0041.jpg



Here are my readings. Thanks everyone.

6L6

pin 3 652V
pin 4 329V
Pin 5 -25.4

-36.2 coming out of the diode SD-1Y and -32.5 on the other side of the resistor following it. (see notes below about this resistor).

there is no arcing evident on the sockets.

This amp has been dormant for 17 years so any past history is unknown. I've researched the type and model and not a mention of performance or faults could be found.

the schematic I have is for a Univox not a Lafayette. Research has told me that this is the closest schematic available as the two companies were virtually the same (sorry if I slightly off on this). The schematic I have runs on the money with how this amp is designed with one exception, which I found out last night. The resistors in the bias section are not the same. Where you see a 6.2k resistor on the schem, the actual resistor on this amp is a 3.3K. the other 33k resistor is actually a 26K resistor (I checked it 3 times) attached to one leg of the trim pot and to ground. The violet leads from the transformer to the bias section, as shown in the schem above are not violet in this amp nor can I find -50V on the transformer side of the SD-1Y diode, also shown on the schem above. Those are the only differences.

[52 Bill]

First off, Lafayette was a huge electronics retailer on the East coast back in the 60-70's, not unlike what Allied Radio once was. They had any number of companies build and house brand amps and guitars with the Lafayette label. This amp was probably built for them by Unicord and that's why the circuits are nearly the same.

"The resistors in the bias section are not the same. Where you see a 6.2k resistor on the schem, the actual resistor on this amp is a 3.3K. the other 33k resistor is actually a 26K resistor (I checked it 3 times) attached to one leg of the trim pot and to ground."
The resistor values that your amp has, would allow for a higher bias voltage than shown in the schematic, as long as the input from the transformer was the same. Is the trim pot wired in series to ground with the 25K resistor? If so, when you adjust the trim pot how high can you get the voltage at pin 5 of the output tubes? -33 volts? Is that enough to get the current draw down to a reasonable level?

"The violet leads from the transformer to the bias section, as shown in the schem above are not violet in this amp nor can I find -50V on the transformer side of the SD-1Y diode, also shown on the schem above."
You'll never see -50 volts coming out of any transformer, transformers are AC devices that's just a misprint.

If you can't get enough bias voltage out of the circuit, you could wire up a voltage doubler to increase the bias voltage.

[Garydean]

thanks Bill!!

"Is the trim pot wired in series to ground with the 25K resistor? If so, when you adjust the trim pot how high can you get the voltage at pin 5 of the output tubes? -33 volts? Is that enough to get the current draw down to a reasonable level?"

yes it is wired in series. I never checked the max -volts I was able to get. When I dialed counter clockwise the plate voltages would rise close to 700 so I backed off. I was able to get -33volts with the trim pot half way

[Chuck H]

"When I dialed counter clockwise the plate voltages would rise close to 700 so I backed off. I was able to get -33volts with the trim pot half way"

So it's not so much that you can't get enough bias voltage. You just can't bias it that cool without the plate voltage rising too high?

Chuck

[Garydean]

True. Here is what I did last night.

I removed the 5K 2W dropping resistor on B2, and replaced it with a 2.2K in series with a 4.7K 2W resistor, which helped drop the voltages on the preamp tubes. I subbed in for the 33K resistor, a 15K resistor on the bias adjust trim pot in series to ground.

Below are my findings

116 line voltage

6L6

pin 3.....633V
pin 4.....320V
pin 6.....-20V
.....................................................
adjusted bias via trim pot. Line voltage still at 116
20% orange glow on plates
pin 3.....635
pin 4.....322
pin 6.....-22
.....................................................
line voltage raised to 120V
20% orange glow on plates
pin 3.....652
pin 4.....331
pin 6.....-23
.....................................................
adjusted bias via trim pot. Line voltage still at 120
60% orange glow on plates
pin 3.....628
pin 4.....321
pin 6.....-17.7
.....................................................
adjusted bias via trim pot. Line voltage still at 120
no orange glow
pin 3.....686
pin 4.....346
pin 6.....-28.9

B+ (B1) has now elevated to 696

I'm sure I left out pertinent information. Please advise.

[Chuck H]

Well, I've already been poo poo'ed by Enzo and Bruce, and their perception should probably always be taken over mine, but I'd say you have a very specific problem that needs a different kind of solution from what the original manufacturer did.

Let's say, for example, that with 115v from the wall AND a different 6L6 type tube that is not available today you could get that amp stable. But you don't have the same circumstances to work with as the original manufacturer so you can't get the amp stable.

Change the circumstances. Modify the bias supply to provide up to -60 volts and use 6550s.

This isn't the only amp to have a problem like this. The Fender Deluxe Reverb is fine at 110v, OK at 115v, your pushing it at 120v. Now some people have 126v or more at the wall and it's just too much. You can find discussions one after another, people trying to find the best brand of 6V6 that can take it. The Traynor Guitarmate has the same issue with it's EL84 tubes.

I will take my beating now.

Chuck

[Garydean]

Thanks for sticking your neck out Chuck....LOL. Suppose, for discussion purposes, I do switch to 6550's. I then would need to beef up the filter caps from 100/350's to 100/450's, or would the 350V be OK?. Any other changes required?

Wouldn't my B+ be excessive with the bias beyond -35V?

thanks,

Gary

[Enzo]

Oh dear, I am poo pooing. I hate that shit... All friends here.

I myself worry about dissipation. Note that while the B+ climbed, at the higher bias setting your red plates went away. Now if the resulting elevated B+ is too much for the tubes, then they will arc or otherwise self destruct. But voltage alone does not cause red plating, it is the power dissipated.

There are various techniques for dropping the B+.

YOu can pull the output tubes and measure the B+. That will be as high as it can go.

[Chuck H]

"Wouldn't my B+ be excessive with the bias beyond -35V?"

No, 6550s require more bias and drive voltage for the same current draw as a 6l6. Also, 6550s have a max vp of 600 (IIRC). So your not going to be near as close to max as you would be with a 6l6 (max 375v) or even a 6l6gc (mav 500v).

Enzo is right. It takes both voltage and current to create watts. If you keep raising the bias voltage your plate voltage will go up (to a point) but your current draw will go down by a greater degree. I suppose in that case that you may end up needing beefier filters even if the 6l6s can be made to work. But, since your already rigged for it, try Enzo's idea. Just keep raising the bias voltage until the plates stop glowing red. If nothing arcs or otherwise fails, your done. If that makes you too nervous then order a big diode and drop the B+ by 80volts or so. You may need to change some other dropping resistors in the rail to bring the preamp voltages up. But then you would surely be O.K. as long as you use 6l6GCs and not just plain ol' 6L6s.

Chuck

[Enzo]

Looking back at your schematic, what sort of voltage is on that pink wire from the PT? COuld we swap the pink wire for the red one and reduce the voltage that way?

Aside from the points I was making, I do agree that 660-700v is way too high, just not the cause of the red plates.

[J Martin]

I have used 6550 tubes in those amps without a problem. Just got another one into the shop with a blown output transformer. The original was long gone and the recent replacement was way undersized. Here is my chance to optimise.

[Dave Curtis, dB AudioTech]

At 660vp your going to cook those tubes pretty fast, even if you bias them cool. In fact it sometimes happens that the cooler you bias them, the higher the vp will go, making things even worse.

Are you sure the amp is designed to run 6L6s? Where did that schematic come from?

Doesn't matter. I'd change to 6550s or el34s at least.

Chuck

I doubt the tranny can supply the extra filament current.

-DC

[Dave Curtis, dB AudioTech]

I've fixed up more than a few Univoxes very similar to this one; get the dissipation down to less than 70%, R&R the baked OP tubes, crank it up and stand back. ;o)

Be careful of the board traces when swapping components...

-DC

[Chuck H]

Well, I've never seen one of these so I cannot say. But the 6550s will only draw an extra 1.4 amps. I hould hope the filament winding is overdesigned by at least that much.

Also, Didn't another poster already indicate that he swaps those amps to 6550s regularly.

I can say this. If the amp is designed to run that close to specs then the designer had no business putting 660Vp on a pair of 6l6s.

Chuck

[Dave Curtis, dB AudioTech]

Well, I've never seen one of these so I cannot say. But the 6550s will only draw an extra 1.4 amps. I hould hope the filament winding is overdesigned by at least that much.

Also, Didn't another poster already indicate that he swaps those amps to 6550s regularly.

I can say this. If the amp is designed to run that close to specs then the designer had no business putting 660Vp on a pair of 6l6s.

Chuck

Although the trannies are fairly good sized, I'd want to see how hot the PT gets during a 4 hour gig with 6550s in it (could always spring for a filament XFMR). They don't have to be biased very hot anyway (nor do the 6LGCs). I'm just saying I have happy customers running good 6L6GCs in these amps; as they were designed. It won't sound the same with 6550s or EL34s (I do like the sound of 6550s in my bass amps, though). The Univox U-1226 and U-1061 I worked on early last year had 640V and 625V B+ respectively, and they're still going strong. Just make sure the circuit is good (bad caps & resistors galore; watch those PCB pads), and don't bias it hot. Oh, and definitely no upward impedance mismatches.

[Garydean]

Thank you everyone for all the information. I've tweaked the bias at -28.9v with no orange plating. The voltage #'s are a bit higher that normal (6L6's ..684V) but the line voltage is at 121v rather than 117v as suggested by the MFG.

Enzo, to answer your question about the pink wire.... despite what the schematic shows, this amp has no pink wire. The schemo shows 11 wires coming from the PT, this one has 9.

if I'm crazy.... replies are most welcome.

Gary

[Amp Kat]

That schematic also supports 6L6GC's which are 30 watt tubes and should take the voltage. What I'm confused about is why is changing the bias voltage effects large plate voltage swings. Somethings not right there. -33 is way to hot for those tubes. With that much voltage you should be up in the -50 to-60 range and you may have to create a voltage doubler to get it there but you really need to get up in that range. 6l6's will take it if you get the bias right but I'd make sure you use a GC type in the 25 to 30 watt range or the 5881 Sovtek that will take up to 800V or so even if it's just for a test.

[Garydean]

I have Sovtek GB's in there now. I have a set of EH GC's I'm gonna throw in...... so I'm not done?

thanks Amp Kat

[Garydean]

I threw 6L6GC's in last night and played it for a bit. Plates got slightly orange. Prior to plugging it in I did fire it up on the bench and took some readings with the GC's in. I tried to bias the tubes beyond -30 and the B+ got to around 711+. My line voltage last night was 122+.

Seems I can't bias the tubes without the B+ going through the roof.

But the amp sounds great......

Gary

[52 Bill]

Wow 711 volts! Without doing mods to the amp, you're really going to be testing the limits with this one.

As you increase the bias voltage, the output tubes draw less current and therefore the supply voltage drop is lessened and the voltage goes up. This is the way it always works, more current-less voltage, less current-more voltage.

Have you taken a voltage reading of the power supply with the tubes pulled? This will let you know what the maximum voltage level can be.

Edit: I just looked at the supply schematic. When you recapped this, what voltage rating caps did you use in the first filter/doubler set? The original 2 added up to a total of 700v, already below the voltages you've measured.

[Garydean]

Hi Bill, Unfortunately I did not put in higher voltage filter caps. I never imagined seeing over 700V.

When I remove the tubes where do I measure in the power supply? At the filter cap?

This ones scary....I can't let it go like this.

thanks,
Gary

[52 Bill]

Measure at the top of the filter cap totem pole, where the center tap of the output transformer attaches. Or at the tube sockets, you just need to know what the maximum available voltage is. Be careful and take a quick reading.

[Garydean]

Gottcha, will do.

[Garydean]

I popped out the output tubes to measure 809VDC on the plate socket. My line voltage is now up to 124.6VAC in this building. Threw the tubes back in and still measured 809VDC on the plates!!!!

[sportster4eva]

I popped out the output tubes to measure 809VDC on the plate socket. My line voltage is now up to 124VAC in this building. Threw the tubes back in and still measured 809VDC on the plates!!!!

That's 100 volts over the cap ratings.:eek: Be careful man.

[Garydean]

Most careful. I had it on for no longer than 5 seconds. I also measured
6.8VAC across pins 2 and 7.

[52 Bill]

That's one heck of a lot of B+.

Does anyone know if any new manufacture 6L6's are rated anywhere near 800v max plate voltage?

It seems like if you are going to get the output tube current draw down to a reasonable level, you're going to be somewhere in the area of 725 volts on the plates.

You will have to upgrade the filter cap voltages as well.

[Garydean]

I'm just guessing, or stabbing, what would it take to redesigning the rectifier, and would that be an angle so I don't mess with 800V anywhere in the amp.

[Chuck H]

I was just starting to think the same thing about this thread. I don't know if you could keep the same character to the sound...entirely. But at least it will be usable.

If you change from the voltage doubler to a four diode bridge you should get about 400 volts on the plates. A much more typical and sane voltage. You'll need to put the screens in series on the B+ rail. I think adding a choke instead of a resistor would be best to help firm up the tone after the plate voltage drop. reduce the screen grid Rs to 470 ohms. The preamp voltages will tank unless you change the dropping resistors in the B+ rail so they match the schem. Other than mounting a choke, I don't really think it would be that bad. And I don't think the tone or volume would suffer much if at all.

Chuck

[Enzo]

Just a thought, I mentioned the pink wire before, and there is none on yours, but is there any chance the amp DOES have the 100v tap on the PT primary? That would be what Japan runs it on, and if 120VAC at the primary was connected to the 100v tap, all your voltages would be MUCH higher than they should be.

Just having a hard time imagining how this amp makes 800v.

[Chuck H]

I thought about that too. But a couple of posts ago it was mentioned that there was 6.8 volts on the filaments. If the PT primary was on the 100v tap that should be much higher. If it is on the 100v tap now and it is changed to a
115v (or 120v) tap that would place the filaments below 6v.

Chuck

[Enzo]

Yeah but it is worth a shot.

[Chuck H]

The PT could have a 110v tap.

That would help. That would knock down the B+ about 60 volts!!!

Chuck

[Chuck H]

Just looked at the schem again. It shows a 100v and a 115v primary option.

But hey, if the pink wire isn't there...

Chuck

[Garydean]

Absolutly no pink wire on this PT.

http://i7.photobucket.com/albums/y275/garydean57/DSCF0045.jpg

2 greens
2 blues
1 white
1 red
1 black
1 orange
1 brown

Gary

[Garydean]

As I read through the threads, I'm pondering with the idea of somehow lowering the B+. I don't have experience with altering a SS rectifier or adding a choke. I do need a plan and I'm willing to do whatever it takes to get the B+ down to about 500V. I do understand it will change the sound.

[sportster4eva]

How about a bucking transformer setup before the power tranny? If you can locate a 120:12v unit with enough current capacity you could use it to drop the 12volts giving you somewhere around 112v at the input of your PT. Based on the readings you gave earlier in the thread your PT has a ratio somewhere around 6.5 to 1. That should get you back down to 700v unloaded, with the filaments being somewhere around 6.1v A 24v trans would get you down to 100v in and 650v out, but the filament voltage would be a little too low, unless you switched to the 100v tap. Do you have a variac you could test with?

[Garydean]

I do have a variac.

[Chuck H]

I can only imagine how frustrated Garydean feels. So now that it's come this far, let's burn it down:

Unless a bucking transformer of greater than 12v is used the filter caps will still be under considerable stress all the time. If a bigger bucking transformer is used then the filaments will be too low. And by the wiring given I'd say that the PT doesn't have any primary options either. Nor does it have a CTR tap on the B+ secondary so a big zener or a bucking transformer there aren't options. And obviously the amp can't be made reliable the way it is. I don't know of a rectifier arrangement that will provide 500 volts.

So, the way I see it, there are only three options.

One is to change to a bridge rectifier and series the B+ rail with a choke and different screen and dropping resistors. That should be good for about 400 volts on the plates.

Two is to change to 6550s and beef up the filters. Provided the filament winding can handle the extra current.

Three is to change the PT.

I don't think I missed anything in the details. And I wish it could be easier than that. But if I'm way off someone will be around to correct me soon enough.

Chuck

[Garydean]

Chuck, thank you!

RE: frustration.... :eek: ... lol

I like option "one" as it seems to be the least costly and gets me where I would like to be as far as high voltage.

[Enzo]

Gary. we were past the pink wire, I was looking for that yellow wire on the primary side in your drawing. The orange is the 115v and the yellow is the 100v primary tap. if it isn't there it isn't there.

[Garydean]

Thanks Enzo. I did realize that a bit later as I was re-reading everything..

[Garydean]

FINAL RESULTS

Here's what I did.... I replaced the 100/350 caps with 100/450 caps, threw in new 6L6GC tubes, and sent the amp on it's way. The report from my customer was great. He rehearsed for 2 hours with the amp on 10 and didn't have to call the fire department. I did stop by his rehearsal room last night to have a look at the plates and no signs of orange plating...phew.
If you're in Philly on the 31st of Jan, stop by and listen to the amp.....bring a fire extinguisher..

The chilling Details
January, 31 2008 at The Fire
Philadelphia, Pennsylvania

Look at the name of the venue....haha, they have no idea.

Thanks everyone for your advice and education. Anyone wanna buy a soldering iron.....just kidding.

gary

[J Martin]

I got my 1236 with the blown output transformer fixed also. I put in a Hammond 1650R and a pair of 6550's so its got plenty of power now. The amp looked like it had been left out in somebody's back yard and had dirt caked inside. Customer emailed and said he is very happy with it.

I am only 15 mins from the Fire in S Phila.