Reducing filament voltage

[mresistor]

I've got a Baldwin organ amplifier that I built a preamp into and turned it into a guitar/bass amp. The circuit resembles the Ampeg B12N quite a bit. It has a  6SL7 preamp and PI and two 6L6G finals and was built in 1957. With 120V input all the voltages are running high in it.
The filament voltage is 6.95v and I could bring it down to 6.3v if I add resistance to it. To keep it balanced I've calculated to use two .15 ohm 5w resistors. I found some available. Question is, which would be better to use, the sand block type or some Dale wire wounds? I'm sure they're going to get a little warm, so I'll mount them somewhere where they can breath.
The plate voltage on this animal is 455V and it seems like that is pretty high for 6L6's. I'm running the Reflektor 6n3c's and I understand they are rated for 500v on the plate. I noticed though that one of them started to red plate. I swapped the two valves and it redplated in the adjacent socket. I tried a Sovtek 6L6WXT in it's place and it too red-plated, but the one good 6n3c seems to handle the high filament and plate voltage without a problem (so far). I order a couple more. Does anyone think the high filament voltage may have led to the 6nc3 failure?
I plan on bringing the filament voltage down, before I use the amp again. But I am concerned that the plate voltage is pretty high. Any thoughts?
It was running 56 ma one tube and the bad tube was running at 94 ma. I wish I knew what the current was on the bad tube prior to it failing. There's 35.56v on the cathode through a 250k common resistor.
I was thinking of maybe using some KT66's but they are a little on the pricey side and I don't really want to put much into this amp. If I was to replace the PT them I'd probably rewire the whole thing to Ampeg B12N.

[John]

Can't comment on the plate voltage, but to lower your heater voltage, a pair of 1n4007 in reverse parallel works great, it reduces the voltage by about .7 volts. Connect like this...


------->|---------
-------|<---------


on 1 of your filament wires. Cheap, easy. :-)

[eleventeen]

I'd have to opine that the total heater current of 6L6 + 6L6 + 6SL7 is well past the 1 amp rating on a 1N4007. One could feed >>each tube<< with its own diode pair. 6 diodes total.

You say you have a "250K" resistor...you must mean a 250 OHM resistor, which, if it is dropping 35.5 volts implies 142 mils through your 6L6. That will cook them, no question about it. You probably need to double that value, roughly, to get down to 70-80-90 ma through BOTH OF the output tubes: not EACH. Or, you could use a separate 250 ohm FOR EACH TUBE. That could lead to imbalance problems unless those R's are pretty closely matched, but as usual with tube stuff, such things are generally well tolerated.

IMO you're redplating not because of too-high plate volts per se, but because of too-high tube current. 6L6GB/GC (NOT 6L6G)  can run 455 volts fine but you may need to raise the value of the common cathode resistor, or, switch to fixed bias. ("Fixed" bias as you may or may not know doesn't necessarily mean "fixed"--it means bias derived from a separate supply as on a Fender (with its own tap) or just as easily by stealing one of the HV leads. I would lean this way because of the requirement for pretty large wattage cathode resistors. Also, if you crunch the output tube current back to more nominal levels, say 35-45 mils, the plate voltage should rise a tad because you will be loading it a little less. This kind of makes your too-high B+ issue worse if we are saying it is a problem. A little bias supply with a "backwards" diode, a cap maybe 100u @ 160 volts and a low wattage pot, and a couple of resistors isn't a big pile of parts to throw in that big chassis.


Incidentally, I also have a 50's Baldwin organ amp and the PT is 450-0-450, 900 VCT no load! Plus it has a big choke that's the size of a Bassman output tranny, though it's mounted vertically. I did not dismantle the organ...I bought the amp chassis, and I wonder if the volts are so high because the tranny also originally supplied the 40-or-so tubes in the oscillator section.



I prefer the Dale resistors, they are in my opinion a higher quality item, but either should work if you choose to go that way. Sometimes you can find deals on NOS Dale metal resistors on ebay.

What type of rectifier are you using? The tranny can run a 5U4, if it so happens you are using a GZ34, you could use a 5U4 and lose maybe 25 or more volts off your B+ fairly painlessly and I think that's a direction you would want to go.

I should also mention that you could reduce B+ a little more by converting the power supply filter to choke-input....if you are using the choke that the amp came with. That's easy, just move a wire or two.

[HotBluePlates]

I've got a Baldwin organ amplifier that I built a preamp into and turned it into a guitar/bass amp. ... With 120V input all the voltages are running high in it. The filament voltage is 6.95v ...

It's only ~10% high which is still within manufacturer's tolerances...

[jjasilli]

Agreed that the filament voltage is close enough to spec & not too high.  But if you want to drop all voltages, a bucking tranny is an option.

[mresistor]

Thank you for the replies. Yes eleventeen, I meant 250 ohms, don't know how that k got there. Well, guess I was burning up the tubes. I have a lot of respect for these 6n3c's now. 
I left the power supply basically untouched and added in capacitors. Three 22uf and one 100uf caps. The 8mH choke is still in circuit. Here's a pic of the circuit. eleventeen, do you have a schematic/diagram of the Baldwin ps?

and here is a pic of what it looks like physically


The rectifier is a 5U4. I also thought of swapping rectifiers but saw that I already have the best one for dropping volts. I also think I will replace the two tropical .1uf coupling caps.

[mresistor]

I found a 600 ohm ceramic wirewound in my junk parts box, think I'll swap out the 250 ohm and check current level.

[mresistor]

Ok, that's more like it, I installed a 600 ohm cathode resistor and am now getting -48.2 volts.
48.2/600=.0803

Only problem the current isn't split evenly on the tubes, one is running about double the current of the other one. With the current checker I measured .055 on one and .029 on the other. Not good. Not very balanced.....

Another problem is the plates are now running at 506vdc. Time to look at a bucking transformer. jjasilli where would I start?

[eleventeen]

No, I don't have a schem of the unit but it looks identical to yours other than the obvious newer mods = on/off sw & pilot lamp. Got those fat green caps. Narrow, 3" deep chassis, your choke is mounted like mine, Kind of a biggish chassis for (only) a single octal dual triode driving a pair of 6L6 but there you go.

[mresistor]

Yeah it is big and HEAVY - mines got two octals on it now; on the other end, I installed the preamp octal in one of the holes left over after yanking out all the remote cabling.

[mresistor]

OK, jjsilli, if I used this http://www.newark.com/hammond/166n12/bobbins-transformer/dp/92F1318 as a bucking transformer I should be able to get the primary voltage down to 110v ac. My wall voltage here is 122vac.
and wire it up like this
I'm really not sure exactly how this is physically spliced into the amp though.. like...what do you do with the primary wires of the bucking transformer?

.
.
would something like this work  http://www.ebay.com/itm/New-laminated-core-transformer-EI-transformer-120V-AC-to-12VAC-8W-/330981979544?pt=BI_Circuit_Breakers_Transformers&hash=item4d10103598

[jjasilli]

If you're not going to be moving the beast, you could use a "small" variac, rated for the draw of your amp - which might be posted on a UL certificate or the like.


The bottom end for a "global voltage drop" is for filament voltage not to drop too low -- which is 5V (loaded) for a 6V tube.  BTW:  I'm assuming the 6.95V, let's say 7V, is the loaded voltage reading (with the tubes IN their sockets.)


For leeway, let's say you drop to 5.5 filament volts.  That's a drop of 1.5 volts.  1.5 / 7.0 =  a drop of about 20%.  That's the max global drop for this amp, to protect the filaments.  All the voltages will drop accordingly.  This will tell you the range that you can use with a variac or bucking transformer (which functions like a "fixed variac"). 


For another bucking tranny circuit:  http://geofex.com/ > 8/10/99    Vintage AC adapter (on the Home page)

[mresistor]

Oh Ok, I get it. You mount this bucker on the chassis somewhere, bring the ac input in and apply to primary of bucker, then take the bucker secondary and wire it to input of PT. Cool, I think I could use one the steps down 120v to 6v and then have around 116vac on the rectifier. Should bring down the filaments to just about right too.

Something like this may work..  http://www.ebay.com/itm/Step-down-120V-6-12V-30W-Transformer-30VA-TPL-9096301-dual-secondary-/291121258561?pt=BI_Circuit_Breakers_Transformers&hash=item43c82de041

[mresistor]

I'd be better off dropping 10v to 112vac that would be a 10% drop in output voltage, the plates would them be running back at 455v and the filaments would be running at 6.3vac. It appears this organ was made to run on 110-115vac.

[Willabe]

Oh Ok, I get it. You mount this bucker on the chassis somewhere, bring the ac input in and apply to primary of bucker, then take the bucker secondary and wire it to input of PT.

I think most guys build it in a separate box with a AC power cord and a female AC outlet. Then you can use it with anything you want/need to.

Here's a link to Sluckys web site, look on the 2nd page;

http://home.comcast.net/~seluckey/amps/misc/Amp_Scrapbook.pdf



                Brad     

[jjasilli]

I'd be better off dropping 10v to 112vac that would be a 10% drop in output voltage
That's a fine choice. 


I think most guys build it in a separate box with a AC power cord and a female AC outlet.
My need was amp specific so I mounted a 12V Radio Shack tranny to the chassis, wired as a bucking transformer.

[1rebmem]

Can't comment on the plate voltage, but to lower your heater voltage, a pair of 1n4007 in reverse parallel works great, it reduces the voltage by about .7 volts. Connect like this...


------->|---------
-------|<---------


on 1 of your filament wires. Cheap, easy. :-)

Diodes work good to reduce filament voltages by .7 volt. Just use something higher current like a 1N5408 which is a 3A diode.
Zener diodes work good to reduce the B+ if your PT has a center tap, so you can put them in series with it.

I have had the same issues with excessive voltages when converting old amplifiers and I prefer installing diodes rather than the bulky bucking transformer.

[John]

Just use something higher current like a 1N5408 which is a 3A diode.

Good catch, thanks! I still haven't gotten around to swapping mine out with the heavier diode. No issues to date, but my luck doesn't like being pushed, either.

[tubeswell]

Can't comment on the plate voltage, but to lower your heater voltage, a pair of 1n4007 in reverse parallel works great, it reduces the voltage by about .7 volts. Connect like this...


------->|---------
-------|<---------


on 1 of your filament wires. Cheap, easy. :-)

No. You'd need 6A diodes; not 1N4007s!

[mresistor]

Diodes work good to reduce filament voltages by .7 volt. Just use something higher current like a 1N5408 which is a 3A diode.
Zener diodes work good to reduce the B+ if your PT has a center tap, so you can put them in series with it.

1rebmem I think it would be great to get this Plate V down to around 420vdc. I see I could get a 100v 50W 1N3340B for around $13. What kind of heat would this thing generate and how to dissipate it? Mount it on chassis and use chassis as heat sink? This would reduce my plate v to around 416 which would be great.
Looking at the schematic diagram I posted, how would I insert the zener? Would between the choke and ground be OK?
Also starting to like your idea of diodes to reduce filament V.  I like the way this old amp sounds an really would like to stay on the cheap side with it.

[1rebmem]

Well.......I am no expert on anything but I have used zener diodes to reduce B+ on several amps. I have always used a string of zeners something like these:  http://www.ebay.com/itm/4-pcs-1N5353B-16V-5W-ZENER-DIODE-/151268327175?pt=LH_DefaultDomain_0&hash=item23384b7f07

You said you want your B+ at 420v, down from 455v so 3 of these in series will about do it. You won't get a 16 volt reduction with a 16v zener. It will be less like about 12v. Solder them in series to a solder strip, leaving the legs long for heat dissipation, cathode bands toward ground.
As far as the choke in your CT circuit.......I've never seen that before. Seems to me you would put the zeners between the choke and ground like you suggested. I would try it unless someone chimes in and suggests otherwise.

Here is a calculator to figure zener diode wattage requirements: https://taweber.powweb.com/store/vdump.htm
I plugged in dropping 40 volts for a 40 watt amp which computes to 6.4 watts so three 5 watt zeners in series gives plenty of margin.

Edit: I'm confused what your existing plate voltage is. Your first post says 455v and the post above this one implies 500v???
Edit again: I see how you got over 500v on the plates. I'd try putting the 3 zeners in and also the 250 cathode resistor. Getting the B+ down will get the power tube wattage down where the 250 resistor might be ok..

[HotBluePlates]

... As far as the choke in your CT circuit.......I've never seen that before. Seems to me you would put the zeners between the choke and ground like you suggested. I would try it unless someone chimes in and suggests otherwise. ...

Look at it as "zeners between CT and ground" (before the choke).


The choke is placed where it is to ease insulation requirements: now it has maybe couple-volts from choke to case and ground, instead of 450v+. It has the same effect being in the ground circuit.

[1rebmem]

Thanks HBP

[mresistor]

I installed 4ea 1N5353B diodes today from ground to the one side of the choke.  I also installed a 5R4GYB rectifier. Plate voltage is now 436v. Current on the left 6N3C is .032ma and current on the other one is .039ma. The cathode voltage is 39v and this is with the 600 ohm resistor. I calculate that is 12W + 15W or 27W total. It's a little cool for my taste and it sounds it, I could easily get 13w more out of this.  I could now put the 5U4GB back in and see what that does to plate volts, but I like it at 436v, seems more normal to me. I ordered up a range of some wirewound 8w Clarostat resistors and I think I'll try to bias each tube individually with it's own cathode resistor. Since I don't have a single higher wattage 500 Ohm or 550 ohm resistor. I do have two each of 200, 225, 250, 300, 350 etc.
I like that the power tubes are now running closer together. Filaments are still at 6.95vac though so I'll install the paralleled diodes.

The zeners were running a little warm but not hot, which is great. 

[mresistor]

There are two great big 4 uf can capacitors on this unit. One thing I wonder is why did they use such large capacitors? Was it all they had in 1957 that was rated for 600v, or is there another reason? I would like to know opinions on whether or not they could be replaced with modern caps of the same rating, albeit much smaller? Like a 4uf 630v metalized polypropylene cap? Do you think these capacitors may still be good? I do hear an overriding fizzle type distortion on some of the notes when playing this amp, and I feel it may be capacitor related.
Or could I use 600v electrolytic?

[HotBluePlates]

There are two great big 4 uf can capacitors on this unit. One thing I wonder is why did they use such large capacitors? Was it all they had in 1957 that was rated for 600v, or is there another reason? ...

"Big" in physical size, but not in uF's.

There was a time when iron was cheap and cap uF's were expensive. This amp uses a 4uF cap, then a big choke, then another 4uF cap to feed the entire amp. That's good filtering, because the relatively high-value hefty choke offsets the small capacitance rating.

The other thing you need in first filter caps especially is a high ripple current rating. So what Baldwin did was use (probably) oil caps... the "big cans." In general, if the can isn't leaking the oil, those caps last forever.

The other positive to small filter cap uF's and using a choke is that ripple current drawn through the rectifier is reduced. We look at the rectifier, see a.c. go in and d.c. go out and think of little else. But the rectifier passes current to recharge the filter caps when you play and drain off power to the tubes. This usually results in brief large spikes of current through the rectifier. These spikes are bigger when you use high-uF caps (which is the real reason for the notion of some kind of uF-limit for different rectifiers).

That current spike can be minimized or mitigated in various ways. If you use 100uF caps and a tube rectifier, the power transformer can have high-resistance winding to limit charging current, or resistors can be added in series with rectifier plates. Or the amp can be designed to minimize current spikes by drawing a relatively steady current from the power supply. Yours is the latter.

Someone else has already gone in and installed 22uF caps, which look to be more than indicated on the schematic. I have a feeling that was done without an awareness of what the designer was doing... "I'll update this and make it better!"

I'd check that there wasn't abnormally high a.c. ripple on the d.c. supply and put it back like it was originally. I doubt the oil filter caps would be causing "grainy sound" since there are hoards of audio-nuts that are eager to use oil caps like these in new builds.

[mresistor]

Thank you HPB for the reply.  Makes sense and it's nice to hear someone comment on the oil filled caps. I'll keep them in there.
The three 22uf electrolytics are what I built in there. They are substituting for the three 20uf caps as shown on the schematic in the subsequent PS nodes after the 4 uf cap. I didn't have 20s on hand at the time.  The 100uf electrolytic in the power tube cathode circuit and three 22uf caps were originally all contained in an FP cap can. The type you would typically see in a Fender Champ. I think the it was mounted at the other end of the chassis, opposite the PT. 
Which is where I have the Western Electric style preamp mounted now. With a 6SL7. 
Ripple. I have been thinking of picking up a small handheld O scope, but don't have one yet so no way to look at the waveforms. 

[HotBluePlates]

Thanks for clarifying on the caps installed.

You could use an o'scope to view ripple (which is not a bad idea), but you can also simply measure it with your meter set to a.c. volts.

[mresistor]

The AC component on the DC PS had a largest reading of 0.05vac.

[HotBluePlates]

50mV on 400v+ d.c. is insignificant to the output tube plates.

[PRR]

> two great big 4 uf can capacitors on this unit. One thing I wonder is why did they use such large capacitors?

Electrolytic caps go bad with age.

Oil caps can last decades just fine.

Church Organ is a "critical" application. Your Fender goes bad, you borrow one from your buddy. When the church organ goes bad.... what, can all the Episcopalians go over to the Catholic church and carry their organ back up the street? What if everybody starts service at the same time on Sunday? What about *important* weddings?? The bride won't come in until she hears her "Here comes..." tune. Middle of December everybody's running their Musical Christmas. Middle of June *every* church is booked solid with weddings.

And NOBODY in a typical congregation (present company excepted) can change an e-Cap on Sunday morning.

And a good organ is a many-year Fund Raiser in a small church. If the organ goes bad anytime soon, the organ maker will get unkindly words from the Music Director, the Soloist, the Organist, and even the Pastor. And while there's no soldering Sunday morning, every congregation knows a Lawyer who can file suit Monday morning.

Right until the mid-1960s, most church-class organs used oil caps.

Radio stations did have guys who could swap a cap on Saturday night, and could patch-around most single problems, but into the 1960s there were broadcast gear makers who boasted "no electrolytic caps!"

The caps are *probably* good. Check the solder joints, they may go bad. Temporarily tack a 450V 40uFd electrolytic across-- if the "fizz" goes-away, *maybe* the oil caps finally died, and you should carefully measure the peak cap voltage (often 2 seconds after turn-on) and sub-in an appropriate e-cap.

The e-cap will be fine for self-DIY-ed gear. You probably have other amps. If you *needed* reliability, you'd buy 2 or 3 (always carry spares) Yamaha or Roland bass-amps. But you have a soldering iron. You will probably be re-re-remodding long before the e-cap dies. You don't have to blow-off the bride and call someone in Ohio on Monday.

[mresistor]

PRR, thanks for the help. When I turn the unit on the peak voltage goes up to 550v or so on the first 4uf cap then after a few seconds lowers down to a stable point around 455v. That is a lot of in-rush current/voltage initially. What is it telling me? The big oil cap is charging.  Do I need to put some sort of limiting resistance in the PS? I don't want o burn up good NOS 5R4 rectifiers.

[HotBluePlates]

... When I turn the unit on the peak voltage goes up to 550v or so on the first 4uf cap ...

You said the 4uF caps are rated for 600v. So everything is happy.

When I turn the unit on the peak voltage goes up to 550v or so on the first 4uf cap then after a few seconds lowers down to a stable point around 455v. That is a lot of in-rush current/voltage initially. What is it telling me? The big oil cap is charging.

Here is how a cap-input power supply works:
A.C. volts are applied to the rectifier, and when the rectifier plate is positive of its cathode, it conducts. That passes a rectified a.c. voltage to the first filter cap. 388v RMS of a.c. is not 388v all the time: it has a peak voltage of 388v * 1.414 = ~548v (call it 550v). If no current is drawn from the cap, then the filter cap charges up to the value of that peak incoming voltage, which you measured.

Once the output tubes heat up, they start sucking idle current from the power supply. So the filter cap is a water tank, with the rectifier being a pipe filling the top of the tank, and the tubes acting like a hole at the bottom of the tank draining out some water. The system reaches a steady-state with a total voltage less than the peak you measured when you first turned on the amp.

So the drop from 550vdc to 455vdc is telling you the output tubes started conducting and drawing current away from the power supply.

I don't want o burn up good NOS 5R4 rectifiers.

The (5U4?) rectifier is just fine with all the above.

[mresistor]

Thanks for the explanation HPB.  I have a 5R4 in the unit to lower plate voltage. I like 5R4's, and If they fit I use them in Champs sometimes to do the same thing. I don't know of anyone making new ones and that's a shame.

The 4uf oil filled electrolytics are rated at 600v but the Illinois ps filter caps are rated at 500v.
Both 5u4 and 5R4 are directly heated. They start conducting rather rapidly, no slow ramp-up of B+. And the power tubes aren't heated up at this time. Merlin says "All valve rectifiers must have a resistance in series with each anode" http://www.valvewizard.co.uk/fullwave.html What are your thoughts on this HPB? (This is why I was asking about putting some sort of limiting in this amp.)

Also since this amps pt has a center tap on the filament winding couldn't I add the voltage dropping parallel diodes at the center tap and then the string would be balanced?

[HotBluePlates]

Merlin says "All valve rectifiers must have a resistance in series with each anode" http://www.valvewizard.co.uk/fullwave.html What are your thoughts on this HPB? (This is why I was asking about putting some sort of limiting in this amp.)

Short Answer: Your amp is fine as-is, as long as you don't swap the 4uF caps for more uF's. You don't need added limiting resistance.

Long Answer:

• A cap charged to 0vdc will try to draw an infinite current the first instant rectified voltage is applied to it. The rectifier cannot pass infinite current and would pop; series resistance in the charging path reduces the current drawn to something "below infinite" and preserves the rectifier.
• The circuit already contains "hidden series resistance" in the form of rectifier voltage drop (which represents a resistance) and the resistance of the PT secondary and reflected primary resistance. With reasonable choice of cap values, this admittedly smallish existing resistance is enough to preserve the rectifier.
• In general, the rectifier is only in danger when a tech replaces filter caps with larger-than-stock uF's without a knowledge or appreciation for how that changes current drawn through the rectifier. There is a complex multi-step design process to account for all the factors involved, and generally involves reference to charts to help (the math involved is non-trivial).
• I've not been playing with amps as long as some folks on this forum; I've only been playing through tube guitar amps 20-25 years. I have at some time owned many classic amps. I've never had a rectifier fail in any amp. That tells me the original designers knew what they were doing, and the rectifier doesn't need any extra series resistance as long as we don't get too clever when we tinker. To be fair, Merlin's statement is still accurate, but he's trying to compress a book chapter of info into a paragraph or a sentence. So you may be over-cautious with no other explanation (which is the safe direction to steer you).
• There is no "maximum capacitance" rating for any rectifier; however, there is a maximum peak current rating. More uF's = more peak current. Casual observers note that when the uF's reach a certain point and no other precautions are taken that the rectifier pops. Hence, we get the false "maximum uF" rating. Had extra series limiting resistance been added, the rectifier wouldn't pop. The problem is the design process involved accounts for ~21 different quantities and uses 6 different charts in the Radiotron Designer's Handbook (RDH4). The average modder is unaware of the design process, and gets in trouble.

If you're interested, go to the Library of Information on this site (link at the bottom of this page) and download RDH4. The process of design is laid out in Chapter 30.2.

[mresistor]

HPB, Chapter 30.2 is good reading..  thanks for pointing out that the handbooks are here for download.
My next adventure with this amp is fiddling with the preamp to get the treble and bass controls to work well. I remember PRR saying to
fiddle around with cathode resistor / bypass cap values and things. I've read that the typical Fender 12AX7 preamp circuitry also works
well with the 6SL7, but I note in the Ampeg for instance that the component values are somewhat different.
But this might be a subject for a new thread.

Thanks again to all that contributed.

[PRR]

> "All valve rectifiers must have a resistance in series with each anode"

But they do. The PT winding resistance.

[birt]

so now it has a 5u4 right? if you replace that with a 5R4 it will have less stress from the high voltage and it will drop the B+ a bit more (around 15V lower)




edit: i just read you already did. sorry

[John]

Well then, I removed mine, so there! 

[mresistor]

Birt- it's got a 5R4GYB in it. 

[1rebmem]

mresistor........... food for thought:

Now that you have that string of zener diodes to drop B+ you could jumper one out to bump up B+ if going to a particular rectifier tube drops the B+ too much.