Startup Hum on R-3.

larryderouin

I'm VERTICAL and Breathing...most of the time.
Subscriber
My R-3 (1951 FISHER Coronet WR-11 or Warwick) has a startup hum that starts at about 15 seconds after I turn the switch on and lasts up to about 1:10 then stops. About 30 seconds later the tubes FINALLY warm up enough to conduct. These are dead cold off overnight times. The times below are what i got on the video, as it had been off about 1 hour.

I've checked all the voltages on the tubes and they are within 3-5% of published voltages in the manual. All Electrolytic Caps have been changed, along with all of the wax, pio's and black beauties. Some but not all of the power resistors have been changed.

Tubes are Old Stock. Rectifier is a 5u4GB, and the output's are Sylvania or GE 6v6gta's in a parallel Push pull configuration. Driver tube is a 12au7.
All tubes have been tested and show above minimum's on GM and emissions on 2 different testers.



Schematic;
attachment.php


Video here of the startup. This is a typical start up. Video is 1:30 in length

Turn on @ 15 seconds
Hum starts @ 28 seconds with couple of small pops then steadys out until about 49 seconds when it fades out (for 21 seconds on).
Music starts about 1:12 or 57 seconds after switched on.



Or if you'd rather watch it on Photobucket......
http://vid408.photobucket.com/albums/pp168/torskdoc/100_4644_zpsfw0uuogi.mp4

Any ideas???

Larry
 
Last edited:
Larry -- what was the source that finally started playing? From the Tuner? Aux input? What?

Also the schematic link never made on your first post....

Dave
 
TUNER. Specifically FM. But it does it in ALL modes.

I'll send you the manual. I may have already , but a 2nd time won't hurt. Post 1 fixed.

attachment.php
 

Attachments

  • FISHER R1-R2-R3 Schematic..jpg
    FISHER R1-R2-R3 Schematic..jpg
    116.7 KB · Views: 112
Last edited:
Yeeeup, you've got a two part problem there. Based on the design, and in my best used car salesman's voice, I can most assuredly say, " They all do it!"

It is a problem of light filtering, coupled with Fisher's absolute devotion to operating the heaters of the audio tubes in the most sensitive portion of the circuit on DC voltage. In this case, it's the old power the phono preamp and line amplifier tube heaters from the cathode current from the output stage trick. In this case, it's the 12AT7 phono preamp, and 12AV6 line stage tube heaters that are connected in series, with the series string then powered from the quad of 6V6 output tubes. However, while the series string wants 25.2 volts to properly power both heaters, the cathode voltage a the output tubes is only 21 volts. Therefore, the warmup time would be veeerry long indeed -- and on all functions, too.

But during the extended warmup time, the unit would go through a series of different hum tones, based on what was happening in the unit at the time. At turn on:

1. There should be nothing other than any hum induced from magnetic coupling between the power and output transformers -- but as I recall, they are far enough apart that this should not occur. B+ voltage should appear within about 3 seconds of turn on, due to use of the quick warming 5U4 rectifier tube, but this is not a cause for any noise. Then,

2. The output tubes warm up (at about the 15 second mark), but as they do, the unequal plate currents through each half of the OPT will cause incomplete cancellation of B+ ripple, so some 120 Hz hum will appear, but should fade rather quickly with good tubes as they heat (assuming they are well balanced as well). Then,

3. The 12AU7 phase inverter tube is also heated by now, but two conditions cause it to pick up and amplify the hum signal that endures until sound is produced:

A. The AC heater circuit of this unit utilizes one winding for all heaters. Since one side must be directly grounded to satisfy the proper decoupling needs for the RF/IF circuits, this prevents the use of balancing resistors to minimize AC hum in the audio circuits. Note that Fisher wisely went to dual heater windings in their stereo receivers, so that the tuner winding could be directly grounded, while the audio winding could be balanced to ground.

B. Until the 12AV6 warms up -- which could take a year or two -- it does not draw any current. But the 12AU7 phase inverter tube immediately after this tube is hot, primed, and ready to go. With the minimal filtering in place supplying the 12AV6, and no voltage drop occurring through the B+ dropping resistor or the plate resistor for this stage, it means that the 12AU7 is effectively amplifying the noise on the MAIN B+ line until the 12AV6 warms up. When it does, voltage drop occurs across the B+ dropping and plate resistors, which improves the noise rejection at the plate of the 12AV6, and the hum goes away. But it takes a year and a half for it too!!!

A couple of points for resolve of both issues:

1. Consider dropping in a small 12.6 vac 1.2 amp R/S transformer, with bridge rectifier, appropriate dropping resistor, and a couple of 4700 uF filter caps to operate the 12AT7 phono preamp and 12AV6 line stage amp heaters from (rewiring the heaters to be a 12.6 volt circuit). You would also have to drop in an appropriate cathode bias resistor for the output stage to replace the heater string no longer connected there.

You could even operate the 12AU7 phase inverter tube from this supply, which would not only maintain DC on the original heaters that operated from this type of voltage, but also extend DC heater operation to the phase inverter stage, AND allow for all of these heaters to operate in a balanced fashion to ground if necessary, because they would be operating from a separate supply. That would kill two birds with one stone, eliminating the extended warmup time, and therefore the long period where hum -- albeit now at a likely lower level -- would be audible.

Short of that exercise, you might also try this (or in tandem with the previous suggestion):

I can't tell what size the filter cap is that is supporting the 12AV6 stage, but up it as necessary to ensure that it is at least 30 uf. Then, insert a common 1N4007 diode between that cap, and the B+ dropping resistor supplying that cap with B+ voltage (NOT the plate resistor). The banded end of the diode should connect to the bypass cap and plate load resistor, while the other end of the diode should connect to the end of the B+ dropping resistor that has been disconnected from the cap. This little trick just might help with the hum issue during the cold winter warm up period, if the heaters of the tubes in question are still powered from the output stage cathode current.

I hope this helps!

Dave
 
I think I got the transformer, bridge, and pair of caps for a power supply. Then pull the heaters off the 12av6,12at7 and 12au7's and connect them in parallel to the DC supply and adjust a resistor to get the voltage to 12.0vdc 2w resistor?. Then drop a resistor into the old loop to compensate for the loss of the tubes (use a pot to set then get the closest resistor to the pot value in 2w?)


My brain ain't functioning this morning....not enough coffee after 4 cups.

C-60 (between the 1st set of 6v6's)is a 50uf/50V. All the Main filter caps are supposed to be 40uf (book value), and I have 47uf /450V Nichicons in it. Which may not be enough? You lost me with the placement of the 1n4007. Can you put part #'s with your designations(B+dropping resistor, bypass cap(the only bypass cap I'm familiar with is the one you bridge a main filter cap with on a Solid state Monster receiver)). I'm gonna have to find a place local that teaches tube electronics with the rig in front of you along with the schematic.

Thanks

Larry
 
Hiya,

Dave .. Avery could have used you back in the day ..

Fascinating read and well put.

Where do you hide all the time to be this detailed ??

Can I borrow some ?? :)

Frannie
 
Hi Frannie -- Thanks for the kind words. As for time, I don't know. I learned long ago that communication is at the root of most ills in the world, so I always try to make sure I communicate well. It is so easy to misconstrue communication that is presented in printed form only, so I try to get it right the first time and save time in the long run that way........

Larry -- In blowing up your schematic so I could read it, the resistor in question is R48, a 47K resistor. Disconnect the end of this resistor that connects to C54A and R47, which is another 47K resistor. That leaves C54A and R47 still connected together, and the end of R48 that was connected there now lifted.

Now insert the diode between the lifted end of R48, and the C54A/R47 connection, with the banded end of the diode towards the C54A/R47 connection -- and that's it. At 40 uF, C54A appears to be large enough as is.

The diode will allow C54A to charge to the peak value of the voltage source provided by way of R48, but won't let it discharge back through R48 as the ripple waveform falls into a valley before rising towards the next crest again. As a result, it blocks the 120 Hz noise appearing on the main B+ line.

Otherwise, the long warm up time of the 12AV6 basically leaves the input grid of the 12AU7 phase inverter tube "open", such that it can basically pick up anything before the AV6 finally begins to conduct. The hum sources it can pick up are: (1) AC from the unbalanced heater circuit, and (2) B+ ripple supplied to the non conducting AV6 until it warms up.

The diode takes care of #2 above, while using the new heater supply should resolve #1. Together then, both of these modifications should make for a very quiet R-3, that actually warms up before you have another grandchild!

Dave
 
ROFLMAO! DAMN, That's gonna be a LOOOONNNNNGGGGG TIME! No more Grandchildren! 3 is plenty enough.
Thanks Dave for the Giggles. I needed some today.

I look at the unit, and then I look at the schematic (or vice versa), and the differences in how it's laid out electrically (scat) vs. how it's laid out on the chassis (physically) gives me a migraine damn near everytime. I traced it all out. R48 was right up against C54A but R47 was hanging on pin7 of the 12AV6. AARRRGGGGGHHHHH!!!!

Anyway. Just finished it up with just the diode. Powered it up. The hum starts about 30 seconds into the startup and stays a bit longer about to 1:15 and about 10 seconds before the sound ramps up. I'm thinking the separate Heater supply would fill the bill.

attachment.php


And drew up what I think is what you are describing for the 12.6DCV P.S.
Edit as needed. There should be a question mark behind the VR1. Not sure it would be needed.
attachment.php
 
Well, it was a thought anyway. The biggest problem is that without any current draw happening in the 12AV6 for over a minute, the input grid of the 12AU7 phase inverter tube is left hanging in the wind, allowing it to pick up anything and and everything it can. Once the 12AV6 warms up, then that grid is "locked down" so to speak, and the ability for it to pick up extraneous noise disappears.

I'm glad you at least tried the diode idea. Once I knew however that the decoupling cap C54A was 40 uF, I knew it would be a 50/50 chance at best that it would help.

The separate heater supply is the real answer to the problem. Before you install one though, you might install a 10 ohm resistor between ALL of the 6V6 cathodes and then circuit connections currently made to them. That way you will then know what the original output stage current draw was, so that choosing the right cathode bias resistor will then be easy.

Dave
 
Is it ok to put in the 10ohm on V14 and V16. The cathodes look to be tied together and to nothing else. I understand putting them on V13 and V15, but am I missing something on V14/V16?? It doesn't look like they are tied to anything else on the scat. (I already put it back in the cabinet, but no problem getting it back out. )
 
Fisher did a poor job of delineation between lines that do and do not intersect on that schematic in some places. In fact, pin 8 of all four output tubes are connected together, with all four cathodes then connecting to:

1. The cathode bypass cap,

2. The heater string, and

3. A 4.7K resistor (R65) to ground.

On the schematic, the line connecting the cathodes of V14 and V16 should have a dot where that line crosses the horizontal line coming over from the cathode connection of V13 and V15. The fact that the dot is missing is what makes it look like the cathodes of V14 and V16 connect to nothing but themselves. If that were true however, then those two tubes could never conduct any current. It is a simple schematic mistake.

The 10 ohm resistor then should be connected in between the four cathode connections, and the connection of the bypass cap, the heater string, and the 4.7K resistor.

Dave
 
Thanks for the clarification. The things I learn from you on an almost daily basis, are one of the things that makes AK and this hobby fun.

I'll get them in today and post voltages and current draws from the tubes.
Thanks Dave.

Larry
 
Boy the 6v6's are all over the place on this one.

WALL VOLTAGE 118.8

Unit on, Volume down to minimum, set to AM and 1000kc as per manual, Wait 15 minutes to warm it up, tested all tube voltages except V1 and V2 (switched to FM) and measured them.

All RF/If tube voltages are good within a volt or 2. Now comes the fun part.
mandated voltages are in parentheses. Mine are open.

V12 (DRIVER) 12AU7
PIN1 (88) 10.71VDC......P2
Pin 2 (0) 0VDC...............G2
Pin 3 (4) 0.547VDC.........K2
Pin 6 (180) 249VDC........P1
Pin 7 (87) 10.7VDC.........G1
Pin 8 (92) 26.3VDC.........K1
Pin 9 AC 6.94VDC...........H

OUTPUT TUBES....... V13...............V14................V15...............V16
Pin 1 SH GND.................................................................................
Pin 2 H GND................................................................................
Pin 3 P...(300v).........280...............281................280................280
Pin 4 GS..(300v).........285..............285...............285.................285
Pin 5 G...(0v) 0.................................................................................
Pin 6..NC........................................................................................
Pin 7..H..AC..............6.16..........6.17.................6.16...............6.16
Pin 8 K...(21v)..........17.01.......16.87................16.86.............16.89

Tube draw in mv across 10 ohm resistor
.................................332.1mv......304.0mv........278.7mv............267.9mv
Equivalent ma............33.21ma....30.40ma.......27.87ma..........26.79ma

V13 is putting out 9.2988W
V14 .......................8.512w
V15........................7.8036W
V16.......................7.5012W

I've got another 10 or so I can test and then get 4 close ones in.
 
I plugged in 6V6's in and gave them a 5 minute warm up and checked the draw. Draws range from a low of 23.47ma to a high of 37.5ma with the majority between 28ma and 34ma.

The quad that is in there now, draw 34.95, 33.39, 33.21, and 33.06ma (V13 to V16) with a plate voltage of 280 and a screen voltage of 285V. 9.78w, 9.35W, 9.30W, 9.26W
 
Output tube current draw looks great now. However, something is seriously wrong with the section of the 12AU7 inverter tube that comprises pins 1, 2, and 3. First thought (if it's not the tube of course), check the value of the plate load resistor connected to pin #1 and make sure it is in value, as well as any B+ dropping resistor that supplies this tube section.

Let us know!

Dave
 
Dave; You can STOP Polishing the CRYSTAL BALL Now! You hit that nail so friggin hard it's in the Next State!!!

R53 (100K 1/2W) was out of spec. about 1.45MEGOHM!!!!! I backtracked, found R54 (10K) spec'ed out to 1.01K. R55 (470K), R56,R57(both 47K) were Ok between 47.1K and 47.8K .

I further backtracked to C64B. 273vdc.
273VDC and 237VDC either side of R54 (10K 1/2w).

Sounds a lot better now. Bass isn't as flabby. Looks like I might as well replace the rest of the resistors I didn't replace when I recapped it. Only about 20 or so. This unit is addictive. It's sounding so good now, I don't listen to anything else. It may be that 15" JENSEN CO-AX but I don't think so. Something to be said for a GOOD MONO Hi-FI!

Voltages for the 12AU7 NOW
Open voltages are mine. Spec Voltages in Parentheses.

PIN1 (88)....79VDC
Pin 2 (0)......0.0VDC
Pin 3 (4)......3.5VDC
Pin 6 (180)....145VDC
Pin 7 (87)...79.2VDC
Pin 8 (92)...78.2VDC

Vs the OLD PRE change Voltages
PIN1 (88) 10.71VDC......P2
Pin 2 (0) 0VDC...............G2
Pin 3 (4) 0.547VDC.........K2
Pin 6 (180) 249VDC........P1
Pin 7 (87) 10.7VDC.........G1
Pin 8 (92) 26.3VDC.........K1
Pin 9 AC 6.94VDC...........H
Disregard the circled C54A. I mismarked it.
attachment.php


Thank You Dave, for being there and holding a bunch of hands.

Larry
 

Attachments

  • ScreenHunter_85 Oct. 22 20.47.jpg
    ScreenHunter_85 Oct. 22 20.47.jpg
    96.7 KB · Views: 48
Last edited:
Glad that got it! It's great to know that these older Fishers have such great homes. You've done so much great work on it -- and on such a complicated piece at that. Heck, I'm just happy I can help out in some way. When you've got it back together, I'd love to see a pic of the whole unit --

Dave
 
Will Do. Going to pick up tubes tomorrow for the 1949 Model 24 Anniversary unit. Waiting on a 5K 5W resistor, then install that and bring that whole thing up on the variac/DBT! Gonna be Fun! See if all the work will bring it back from the near dead.
 
Put it back together, and fired it up. Now it has a background pop, pop,pop, almost like a 1950's 5 horse Johnson outboard about 240 times a minute, and you can slow it down or speed it up slightly with Volume, and it gets louder when the TREBLE is turned down all the way. Gets quieter when treble turned all the way up. Most noticable in PHONO and TV. Not noticable when FM and AM turned up slightly. But if you're listening for it you can hear it in the background.

I'm thinking the DIODE is the cause as it's the last thing I did that wasn't a straight replacement and it's straight downstream of the treble pot. I'll pull it in the morning and check operation then.

Awwwww, What the hell. We gave it a shot. Definately gonna get the secondary transformer and do he heater circuit.
 
Normally, the diode should not cause any problems, as the current draw from V11 will keep it biased into forward conduction. Therefore, it becomes "invisible" in normal operation. However, since it didn't help with the hum issue, I'd remove it anyway. On the other hand, if C54A became disconnected from the circuit in the process of installing the diode, then that would definitely cause problems.

From a theory standpoint, you are describing an oscillation. If you can manipulate the speed of the popping sound then clearly oscillation is taking place. No doubt this is now happening since replacement of the plate resistor in the first section of the 12AU7 inverter tube -- which is further evidence that the noise is due to oscillation.

Before, with the (basically) 1.5 meg plate resistor, there wasn't enough gain in the system to support oscillation. With the resistor replaced, the gain has returned, and now oscillation can be supported.

As long as C54A is good, and connected properly into the circuit, it effectively grounds any AC signals that might try to feedback through the B+ distribution system from a later stage, back into the input of the phase inverter stage via R47 and R48. Such feedback would definitely sound like what you are hearing, and would certainly be affected by the setting of the treble control. C54A then not only helps to filter the B+, but also prevents any potential feedback from occurring through the B+ system. As such then it "decouples" V11 from the B+ distribution system, which is why some old timers call such caps a decoupling cap instead of a filter cap. In truth however, it is performing both functions.

Similarly, C57A performs the same function for the phase inverter stage, helping to filter the B+, and preventing any feedback to or from this stage to other stages -- or, within itself. Check also that this cap is good and connected properly into the circuit.

Either of these caps could be the issue. You've been working around both of them with the diode test and resistor replacement, so I'd center my search in those locations.

It sounds like you are sooo close!!

Dave
 
Back
Top Bottom