Gibson BR6-F buzzing - troubleshooting

[jconover]

Hi - my name is Joel, and I've been a long-time lurker.  I'm wading into the world of vintage tube amps, and finding my way.  There has been an incredible wealth of great topics here (thank you all for your contributions!).  I went to school (in the 80's-90's) for electrical engineering, but sadly at that time, tubes were considered dinosaurs and things they didn't need to teach any more.  I've managed to self-educate myself enough to be dangerous, and also re-acclimate myself to the art of working with high voltage.  All that said, my career and day job took me far from the art of designing or even servicing electronics, so my skills are all 22 years rusty at this point.

I took on a project from my guitar instructor - a gibson BR6-F failed unsurreptitiously at the end of a gig - developing a nasty buzzing noise after being on for about 4 hours straight at a higher than normal volume level.  (He typically does 2 hour gigs at "4", and ended up doing a 4 hour gig at "7+".  No idea if the extended service and louder volume directly contributed to the failure of the amp, but nevertheless, I am now in possession of the amp, and tasked with trying to fix it.  Schematic from the library is attached here for reference - please note that the amp has one additional grid leak resistor (470K) between the volume pot and the grid of the 6SN7 (apparently Gibson changed their amp design several times - the BR-6 schematic had this resistor, but not the BR-6F schematic.  go figure)

Here is a sample of what the amp is doing:   http://youtu.be/4hWoroyZBZE   

Please note this video was shot after recapping and restoring the amp to its original schematic specs - the previous owner had recapped the power supply section and put in substantially more capacitance (50uF vs. 20 uF stock), which was quite a bit more than the  5Y3GT spec sheet recommends.  At this point, i've replaced all the original Grey Tiger coupling caps with Mallory 150 series 0.047s, and the power supply section has to-spec sprague atom caps. I've also tested (by replacement) the 6SN7, 6SJ7, and the 6V6GT and 5Y3GT tubes; this has had no notable change on the issue.  I've tested the field coil on the speaker (983 ohms, 18H), the power transformer (406-0-406 unloaded on the secondary, and 5.7V / 7.2V unloaded on the 5V and 6.3V windings, respectively).  The OT is connected the voice coil and in series with a humbucking coil on the original field coil speaker.  I tested the voice coil at ~3.6 ohms when disconnected from the rest.  The OT itself tests with 166 ohms between yellow and black, and 187 ohms between yellow and green, or 353 across the green and black leads.  Everything seems in order there.

It is possible that the speaker itself is damaged, but if you listen to the video, i get a pretty consistent buzz at every frequency.  It doesn't sound like it is buzzing at specific fundamentals.  Having never heard a damaged guitar speaker (but plenty in my cars growing up as a kid), it just doesnt sound like the speaker is what is damaged. 

I've been working closely with Hal from Mutt Amps in Milwaukee, who has given me tons of invaluable advice, but I thought i'd turn to the community and their experienced ears, and see if I can source some additional insight on where to look next.  I've got a list of things to keep looking after, including testing my grounds, and checking the voltages at the field coil and on the plates ( I haven't done any voltage testing in-circuit yet, only when I had the amp disassembled to recap it).  I also tested every resistor in the amp while I was diagnosing it.  Most were within tolerance, though the 470K's tested high (616K, 690K, 555K), but I dont think that is the source of my problem.

Really appreciate any sagely advice - I took this project on because the circuit seemed simple enough, and it was a good way to get my feet wet.  My next plan is to clone this amp; I have the amp parts, a PT and a chassis, and once I feel confident, I plan to acquire an OT and a speaker to complete my first home-built project; but I'd like to get the original working first.  Again, thanks in advance, and I'll make sure to take plenty of pictures and share my story here.

[shooter]

I'm clueless on field-coil speakers soo...  you can 1/2 split the circuit by pulling the preamp tube.  As long as it powers up and isn't smoking or blowin fuses get a full set of voltage readings.  that will help deciding if it's signal issue or power. 

[eleventeen]

Sounds like a speaker issue to me. If the voice coil has become misaligned in the channel it's supposed to operate in, it will rub on one or more of the items that surround it. That it happens at every volume level and the particular sound it's making----that's my take.


One thing you can do is to leave the internal speaker in place, and run the leads TO THE VOICE COIL out to an external speaker, and see if you get a non-fuzzy signal from the amp. Easy test.


This amp uses what is known as a "field coil" speaker. Instead of a permanent magnet in which the voice coil moves, the magnetism normally supplied by such a magnet is generated by running the power supply (aka "B+") through the "field coil" mounted on the blocky part of the speaker. Another benefit is that this coil acts just like a choke as used in many amps---a two wire part that looks just like a smallish transformer and has the function of helping to smooth the rectified and filtered DC the amplifier's rectifier (usually a 5Y3 tube) makes.


So, the only issue here is that the wires you DO NOT want to cut to make the above test carry HIGH VOLTS! 300+ volts, no doubt. If you want to do the test I suggested, you would either cut the wires going to the VOICE COIL of the internal speaker, temp-splice them to other wires long enough to reach out of the cabinet, and go to another speaker, whether inside a cabinet or not. You could also just solder those long wires to the internal speakers' VC terminals just to make a quicky test. Ideally, you would have long clip-leads you could just clip on to the VC terminals of the internal speaker and run them out to your external speaker. You will know right away if the int speaker is NG.


So, the real question is, can you differentiate between the VC wires and the B+/power supply wires? (we don't know this about your skill level)

If you get a clean signal on the external speaker, the internal speaker is your villain.


If you determine this to be the case, come back for suggestions as to what to do about it.

[sluckey]

I also think the speaker is shot.

[jconover]

eleventeen - thank you - I understand the operation of the field coil speaker and i've had enough time with it to know where the voice coil wires are.  In fact, I unsoldered them to determine that the speaker itself was 4 ohm (I was surprised it was such low impedence).  I had been noodling about how to test the speaker without removing the field coil, because of the obvious increase in voltage in the system, not to mention the loss of filtering from the choke.

I will try this and see what happens tonight.  I don't know if I have another 4 ohm speaker lying about, but I'm guessing I can use something close.

[eleventeen]

You can lash up *any* speaker for a 30 second test, even a 5" speaker out of a junked radio or something. Don't worry about speaker impedance for a 30 second test. All you have to do is to ascertain that you get a clean, non-fuzzy sound at low volume, with the amp turned up to 2 or 3. Half a dozen notes. If you do, you know your internal speaker is cooked. Ideally, you would grab the speaker wires from the speaker in another amp or from a Bassman cabinet or really, just about anything, and run the signal over that way. You could use a stereo speaker, a box from a hi-fi system. Don't blast it, you are just looking for a low level signal from the Gibson amp to play clean.
 

[jconover]

Thanks.  I borrowed the speaker from my Fender Champ, since it had a convenient RCA plug and I happen to have some female plugs lying about. 

Yes, the speaker is cooked.  Thanks for the insight  Let the youtube video stand as a lesson to other people that don't know what a cooked speaker sounds like.

Now the real dilemma:  the owner prizes the amp for its clean sound (Jazz guitar instructor).  My options, I suppose, are to
1) send the speaker to someone to recone and repair the voice coil/spider assembly
2) find another vintage 10" Rola speaker with a 1000 ohm field coil (I am going to search forums here, any suggestions appreciated - ebay seems less than useful here)
3) replace the field coil speaker with a modern alnico speaker, and replace the lost field coil with ... a choke? a choke and a power resistor?  again, I'm going to search forums here for prior art - but any opinions are appreciated

<edit>

Wow - these seems like a pretty sparse topic, and some of the discussions are a bit lively.  I guess I'll need to spend some more time elsewhere to research this.

Wonder if something like this might do the job....
https://www.tubesandmore.com/products/P-T155J

[sluckey]

That choke is a little light weight unless you move the CT of the OT to the first filter cap. You will need a choke that can handle about 100mA if you leave the CT of the OT as is. I would just replace the field coil with a 1000Ω/10W resistor. It will probably sound just fine. If not satisfied you can always add a choke later.

[eleventeen]

There have been a few threads on this. Just have to search. You have the right idea, you need to replace the choke and the speaker. The choke is a profoundly unfussy part as to actual specs, any choke that can carry the current will work; most people just replace it with a resistor. The issue is a little more where and how you will mount it should you wish to keep some kind of choke in the amp---a physical problem. I am assuming your amp owner would prefer no new holes in the chassis or cabinet. There may not be room on the chassis anyway....though it looks like there is.


When it comes to a speaker, keep in mind that *no* brand new speaker will sound like the old one. Speakers have a "break-in" period, and many people "condition" brand new speakers by feeding them a low frequency sine wave that just makes the voice coil move in-out-in-out for a few hours. Once you do that, I see no reason why a typical Weber speaker that you'd replace the internal speaker with would present any sort of problem. No sine wave gen? No problem, http://onlinetonegenerator.com/  ! Get a speaker with a paper cone, not a metal cone if you want mellow.


Now, how you deal with picking a speaker for Mr. Fussy and drill holes in the frame only to find out he doesn't like it...


Ideally....you replace the choke with a Fender Deluxe or Princeton choke, which our host sells, and you figure out a way to mount it to the speaker spider by drilling two holes in the spider. That way, no new holes in anything that is there presently. If it happens that the wires that feed the field coil on the amp come out on a plug-and-socket, that makes things simpler.


Obviously, you have to examine the replacement speaker to see if the metal spider has a flat spot suitable for mounting the choke after drilling holes thru one of the spokes. Also obviously, you have to be INCREDIBLY careful drilling those holes not to poke through the voice coil. The Princeton type choke, sort of on the small side, would work fine.

[PRR]

> 2 hour gigs at "4"
> 4 hour gig at "7+"

That's one way to damage an old speaker. High power for long time.

Before you re-cone, see if you can find centering details. WITH CAUTION, put a few drops of shellac thinner, acetone, or toluene (toxic!!) on the dust cap rim. The dust cap comes off (be very clean!). Look inside the coil while you gently move the cone a wee bit. It should move like stiff cloth, not like scraping your car on the garage wall.

There is a *chance* the coil is intact, and round-enough, but shifted off-center. On the very oldest speakers the spider is inside, has a screw. Pick little bits of paper and shim between coil and pole. Loosen the screw. Re-shim so you have equal thickness shims all around. Tighten screw. Re-glue dust-cap with shellac.

However it is very likely the VC has come apart. High stresses in there. The old glues weakened when hot. Take a vacuum cleaner hose and pull on it. OK. Now slit the plastic between two turns of wire and pull. It comes un-done. I've pulled-out loose chunks of voice coil, hanging by a thread (two wires) so they still carried the current and "made sound", but the sound like after you scrape your car out of that garage and the fender is dragging on the tire or the ground.

> find another vintage 10" Rola speaker with a 1000 ohm field coil

Good Luck. Save your pennies and your $100 bills.

> replace the lost field coil with ... a choke? a choke and a power resistor?

The ugly but effective way is to take the FC off the frame and nail it elsewhere inside, to keep it in the power supply but clear the area for a new PM speaker.

The FC inductance is filtering, true, but not essential (just "free" when there wee no PM speakers). Usually the FC's dumb resistance alone is ample filtering for the power stage. You have additional filtering to the driver and preamp.

_I_ would leave the irreplaceable FC speaker alone (maybe someday we will rebuild it, like those guys I just saw on TV in "WWI" airplanes which were 50%-99.9% new-made). Dang the inductance, 20u 1K 10u is plenty of B+ filtering for a push-pull power amp.

> Wonder if something like this might do the job....

Rated 30mA. I would expect a 2*6V6 amp to be pulling 60mA-90mA.

"...less inductance at slightly higher currents..." -- 90mA is not "slightly more" than 30mA, it is a big bunch more.

Look at the size and heat. 90mA in 1K is 90V; 90V*90mA is 8 Watts. And I bet that FC runs noticeably warm, even hot. Now look at the 155J. It is much smaller, same resistance (even 3% more), and the amplifier is the same, it too will dissipate 8 Watts here. It isn't quite 1.4"*1.4", much smaller than that FC (3"?). The FC is maybe 4 times the surface area of the 155J, it may run 4X hotter.

(Hmmm... four 155Js in series-parallel gets you 60mA at 1K, and is 4X the total surface area of one 155J.)

Certainly to get Teacher's BR6 "working" again, put the FC speaker in a good storage box, throw in a 1K 20 Watt resistor (in a ventilated but finger-safe space) and whatever speaker-du-jour he might like. The amp is probably hardly over 10 Watts, so any *guitar-rated* 10W or 15W model will stand the strain. I would not jump to a 150W because these will have heavy coils/cones, a less lively sound at lower efficiency. Oh, the hot old JBLs will have better efficiency and be Real Fikkin LOUD, but less funk and very collectable (spelled with a $$$).

[PRR]

> my skills are all 22 years rusty at this point.
> I am now ...tasked with trying to fix it.

I hope he knows (perhaps not) that if he needs an EXACT fix, he should go to someone like those WWI aircraft guys, but working with Gibsons. Not a youngster born long after this amp's period who is 22 years out of any electronics. Our mechanic is 1980s-schooled, and we like him for our 1991 Miata, but I would not bring him a 1910 Stanley Steamer. (It happens there is a gal on the other side of the island who IS the SS guru in these parts.)

So your task is to make it play with minimum harm, or disrespect to future restorers.

Big 1K resistor and jazzy new (or stash-room) PM speaker.

Steve said 10W, I suggest 20W. It may not be the 90mA which I used to get the 8 Watt actual dissipation guess. But if you are going to hang with this teacher much longer, I'd be generous with resistor physical size.

10W are easier to find and sometimes cheaper by the handful. Two 470r 10W in series may be handier than finding a 20W.

There are "50 Watt" chassis mount resistors. They don't really take 50 Watts unless on a huge chunk of alloy. However they easily handle 20W mounted tight to a steel chassis, the metal case is reliably insulated, they will not break on impact (like ceramic types may), so you are down to keeping the terminals finger-safe.

[jconover]

Thanks for all your help and suggestions.  Given the speaker was trashed, I decided to go the route of reconing it.  I used voice coils and paper cones from Weber.  In case anyone else ever needs the information, the BR-6F had a 10" ROLA speaker with a 3.2 Ohm 1.005" inner ID voice coil.  I mated that voice coil with the  10" smooth pulp cone, (1.045" VCID) and the Spider was the 1.045 ID, 2.7 OD, .370 ht, medium stiffness unit.  The spider was approximately the right width as the OEM, but it was a little taller (I believe the OEM may have been 0.20 or 0.25 ht).  I used a spare paper gasket to build a shim and raise the paper cone up to accomodate the slightly taller spider.

Here's a quick video of before and after:
http://youtu.be/qAh3zbqbSjQ

[PRR]

Cool!

[HotBluePlates]

Thanks for all your help and suggestions.  Given the speaker was trashed, I decided to go the route of reconing it.  I used voice coils and paper cones from Weber. 

Good job with the recone!

... but I would not bring him a 1910 Stanley Steamer. (It happens there is a gal on the other side of the island who IS the SS guru in these parts.)

First time I read this, I thought you meant "Chevy SS" and said to myself, "I'd dearly love an old Camaro if I could afford one..."

[eleventeen]

You made that work?! Good on ya!!

[jconover]

Thanks!  The amp has a lot more vigor than it did before I started, and now it is producing some interesting microphonics from the 6SJ7 tube.  I narrowed i down by putting my finger on top of each tube while playing a low C - the tube seems to have a tendency to resonate on C.

Here's a very closely mic'd audio sample:
http://youtu.be/zF2HJPGacOM

I noticed these go away if i hold my finger on the tube and push.  I will search the forums, but wondering if this can be fixed by securing the tubes, or if my pressure on the tube is changing the behavior of the glass envelope in some other way?

[HotBluePlates]

It probably is simply a slightly microphonic tube.

Others might be able to recommend tricks for quieting a microphonic tube (RVT sealant, silicone damping rings, etc), as I haven't played much with trying to quiet down a noisy tube. If it is a microphonic tube, the issue could be very sample-dependent.

[jconover]

I have 3 tubes like this lying about; they all produce the same effect; going to look for some more samples