SA100 Rebuild Breaking in:Red Plate Protection?

Dave,
Am loving the amp "as is" so far but caught a 7189A glowing slightly red today. FWIW, I haven't dismissed your EFB mod, I just want to get to know this amp in it's stock form before I go and change things.

Inspired by your work, I separated the two (Channel A and B) test points into four so that I may check each individual tube's bias as opposed to checking pairs (something you thought Fisher should have done in the first place.) As with the original design, jumpers are placed from each test point to ground during operation thus bypassing and removing the resistors from circuit. (I state the obvious and created a new thread for those who may have experienced a red plate and would like to protect from damage).

I checked the individual tubes and of course found that the tube in question was drawing more current (close to 70ma), while sucking current away from the others. Can't really find a reason why this occurred as the problem seemed to have been corrected as I readjusted the bias and DC balance controls of each channel. Maybe the components are still breaking in, maybe the potentiometers need a good cleaning. For the time being, I am running the amp without jumpers and checking the current and DC balance of each channel often.

I noticed that in your version of separating the test points, no jumper is used to bypass the 10 ohm resistors during operation. I am currently using 1 watt resistors and am wondering if reducing their power rating to 1/2 or 1/4 watt and leaving the jumpers off would afford any protection should a tube begin to draw too much current? Would one resistor "pop" and if so, would there be any consequence to the other tubes? As both channels share the same bias supply, I've noticed that decreasing the bias of one channel increases that of the other. If one channel were to shut down, would your EFB design maintain regulation for the other?

I don't think I have a bad tube as the problem hasn't recurred. The overall bias seems stable enough, drifting only very slightly. I have less than 25 hours on the amp, so I think it's still settling in.

What do you think? I am particularly interested in modes of failure and their consequences.

Rich

Hi Rich -- First of all, congrats on doing a great job on your rebuild. I think it is always most appropriate to first restore any unit to its original condition, and then use it for a good period of time so that you know how it performs. Then, you can determine if any modifications suit your needs. Otherwise, you won't have any reference to compare to if you do any modifications in the future.

Regarding the 10 ohm resistors, I in fact used well matched 1/4 watt devices, because that rating will easily allow the amplifier to operate at sustained full power output without concern for overheating the resistors. Also, as you suggest, the smaller rating will cause the resistors to act as a semi-fuse in the event of severe overload, burning open much sooner than a higher rated device would. To that end, metal film resistors are the best devices to use in this location, since they tend to open rather immediately with a decided "pop" upon overload, versus carbon film devices that tend to take more time, overheating, cooking, and then finally burning open.

Regarding the grounding straps, I'm not sure if my pics showed it or not, but I did reinstall them in the finished amplifier, but not in the manner you might think. Many years ago, Hafler showed (and I have reconfirmed) that standing the cathodes of the output tubes in a fixed bias amplifier on a small COMMON resistance to both tubes can -- and will -- act to significantly lower IM distortion. Significant typically equates to cutting it in half. Therefore, if the pics didn't show it, the original straps were reattached on each output strip (where I made the test points available) -- not to ground the test points, but to connect the test points back together again in each channel after the adjustments are made. Therefore, during the adjustment session, each tube stands on its own resistor, but during operation, the cathodes in each channel are connected together, grounded through a common 5 ohm cathode resistor. This approach allows for maximum performance from the output stage, allowing proper setting of the quiescent current in each tube individually, while also allowing a small amount of current sharing between the tubes during normal operation to minimize distortion products.

One additional benefit of this approach, is that the common cathode resistance that is retained in normal operation still acts as a semi-fuse against high current overloads as well.

Regarding the one tube that red plated, you should check the usual suspects in such events: the tube socket itself for tight coupling to the tube pins (particularly pin #2), and the connections from pin #2 back through the 2200 ohm grid stopper resistors and 220K grid return resistors to the DC balance control that serves the socket where the tube that red plated was located. A dirty DC balance control generally would not cause this condition in this amplifier, since if the wiper lifted from it's track in this control, then both tubes served by the control would red plate. Of course, a crack in the carbon track in the control would surely cause it however. Also, don't rule out the tube itself. If it goes off the ranch again, move the tubes around, and see if the red plating stays with a certain tube, or a certain socket. Finally, I assume you replaced the coupling caps, but of course if one of these goes south, the tube it serves will red plate for certain.

As for the adjustment process, it is quite common in stereo amplifiers with a common power supply, to have the quiescent current increase in one channel, if it is reduced in the other. This is due to the use of a unregulated power supply -- and particularly as it relates to the screen grid. The screen grid controls current flow through a pentode tube as surely as the control grid does, so as the screen voltages drop when current flow is increased in one channel, current flow in the other channel will be seen to drop, and visa versa.

When EFB(tm) is installed, this effect is not completely eliminated, but it is greatly reduced. This is because the screen supply regulator portion of the EFB circuits acts to largely maintain the screen voltage in spite in increased current draw from the screen grids. With the screen voltage held to a much higher degree of stabilization then, the variation from channel to channel as adjustments are made is greatly reduced.

Finally, regarding failure modes, the good news is that catastrophic events in output tubes rarely involve the plate, meaning that output transformers are inherently protected by this fact. The typical output transformer that goes south in a high fidelity setting is almost always due to years of storage in damp places, with moisture doing the dirty work. Once insulation breaks down in the primary winding, arcing begins, and windings either then burn open, or can even short to the laminations or secondary winding.

The vast, vast majority of spectacular output tube failures involves the screen grid. If these short to other elements or go open, there will potentially be a light show, and the resulting transient style change in plate current caused by that tube in the portion of the output transformer that serves that tube will certainly cause a notable pop in the speaker -- and even cause small cathode resistors to burn open -- but such events rarely take out an output transformer.

To help prevent such occurrences, a screen stability resistor can be placed in series with each screen grid for added protection. You can see these added to the SA-100 I modified earlier. If you would like more information on screen stability resistors, you can reference an article I wrote on the topic some years ago now, wherein I first suggest their use to eliminate output tube arcing events. You can find that article here:

http://www.tronola.com/html/maximize_tube_life.html

I hope this helps!

Dave

That did help by making things a bit clearer.
Every passive component on this amp is new, except for the tranny's and tube sockets (I read they should be considered a "component" as well). How many hours of use or power on/off cycles do you estimate it takes till an amp is "burned in and stable"?
The red plate only occurred tonight and I've only listened with the straps disconnected for a few hours. In that time I've measured only very slight DC Balance and Bias Drift. Initially last week I set the bias to 0.3 per tube and only noticed the tube was red tonight when I went to flip an album side. I tried to lower the bias but noticed the bias adj control for channel B is fully counterclockwise, while the control for channel A is only slightly advanced (channel A had the red plate) and the lowest I can get both channels is around .290ma. I'm going to have to go through the amp again and check the voltages. I used carbon film resistors in most of the positions as in the original, I wonder if the heat is making them drift. Any suggestions where I might benefit from using a more stable resistor (mf/mo) in the supplies? I used a WW between the first two capacitor sections and I'm pretty sure MF's to the output tube plates.

If you maintained the original terminal designations of the bias controls in your rebuild, then advancing (i.e. turning clockwise) the bias control in either channel acts to increase current flow through the respective output tubes. Therefore, running with the controls set full down (that is, fully counterclockwise), the bias circuitry is providing maximum negative bias voltage to the tubes. If you are only able to reduce current flow to 29 ma in the stock design at this setting, here are some things you might check. Assuming a rather "normal" (for today anyway) line voltage of about 121-122 vac, then:

1. The screen grid voltage (pin #9) at the output tubes will be about 300-305 vdc.

2. The control grid voltage (pin #2) will typically range anywhere from -10 to -13 vdc, depending on the particular tubes you are using.

3. The voltage at the output of the bias diode will be about -28 to -28.5 vdc.

In the amplifier I documented my work on, once I corrected the original problems with the unit, I found no problem to easily reach the low distortion set point in that unit (about 32 ma per tube) in stock form, with plenty of range available on either side of the optimum setting.

If there are no component or wiring issues, it may simply be the particular set of tubes you are using. I have always stressed that even though a full complement of controls are available to adjust for optimum performance from the amplifier, any amplifier will still benefit from the use of a matched quad of tubes. This helps to achieve the same performance from both channels, by using tubes that all require (very nearly) the same terminal voltages to produce the same current flow to begin with. The controls are then used to simply tweak for optimum performance, rather than to wrench wildly unmatched tubes into submission.

If your voltages are all normal and the tubes you are using just require more bias voltage to "center" the bias controls, then slightly raising the value of R51, and reducing the value of R56 would be the best way to go. Reducing R56 to 1.8K and increasing R51 to 5.6K should provide a notable increase in voltage available from the negative bias supply.

My own personal opinion is that passive components need little if any burn-in time, beyond a few (single digit) hours to reach any stabilization of performance, although there is no shortage of loud cries to the contrary. However, all such cries are virtually always based on subjective audible impressions only I should add, and are typically never subjected to any double blind testing for real validation. Tubes take longer to stabilize, but from a bias setting standpoint, should do so initially within the first few use cycles (totaling usually no more than 5 to 10 hours at most), and then long term within about 50 hours or so.

Metal film resistors are best used where there is notable current flow, and stability is important versus operating temp environment. Therefore, output tube cathode resistors are prime locations for these components, as are power and bias supply, and phase inverter components. For the general audio path, metal film are fine, although many suggest that they sound bright (only opinions once again), and prefer carbon film in those locations. Personally, I have used both with great success in the audio path, but prefer metal film units in the areas mentioned as specific for them.

Dave

It wasn't till around midnight that I got the chance to look at the amp. I flipped it upside down, inspected everything, saw no obvious cold solder joints or bad grounds and flipped the switch on. Again, V3 began to redplate. Switched the amp off, reseated the tubes, no change. Swapped V3 and V2, problem did not follow the tube, so I feel the problem is at the V3 socket.

The next thing I did was to clean the DC Balance controls with some CRC Contact Cleaner. For now, that seems to have solved the problem. The minimum bias I can set when all tubes are balanced is 27.8ma per tube, which is where I am presently operating. I'm going to let the amp cool off till tomorrow night to see if the problem returns and take more measurements.

For now, these are the measurements I took with the line voltage at exactly 120vac

380 source was 410
290 source was 320
270 source was 311
-13 source was -16
-23v after diode was -24
24vac before diode was 45vac (Sams has to be wrong)
66vac before 2200 2W before diode

Socket Pin 2 V2 was -11.45vdc
Socket Pin 2 V3 was -13.10vdc
Socket Pin 2 V5 was -12.17
Socket Pin 2 V6 was -12.37

And I noticed that the current drawn at V3 seems to climb as the amp heats up, although I didn't continue past 5 or 10 minutes of checking tonight.

I'll get back to you tomorrow with my findings. Best case, even though unlikely, I had a dirty DC balance control, hopefully it's not cracked. I want to see how this amp starts up cold before I proceed further.

Rich

Hi Rich -- All good info. For clarity then on your first statement:

1. You say: "Switched the amp off, reseated the tubes, no change." Does that mean that after reseating the tubes that the original V3 tube red plated again in the V3 socket?

2. You say: "Swapped V3 and V2, problem did not follow the tube......" Does that mean that the new tube installed in the V3 socket also (or now) red plated as well?

Regarding these specific questions, if you actually had a different tube red plate in the V3 position, that until now had not red plated in its original position, then that is a rather strong indicator that something is amiss with the socket/wiring/components at the V3 socket position. However, if you have not actually had a different tube red plate in the V3 socket position, and/or the old V3 tube has not actually red plated in a different socket location, then we don't really know anything until either:

1. A different tube red plates in the V3 socket or,

2. The old V3 red plates in a different socket.

Only then can a clear conclusion be drawn.

Regarding your voltage readings, they all seem good except for the voltage directly after the bias diode, which seems low (I did not measure the voltage directly before the diode in the unit I had here, but the bias tap from the transformer was measured at 65 vac). In any event, if this is the case, it would explain why you can't lower the output tube current readings any lower than what you are currently able to achieve -- whereas I could.

The B+ voltage readings are all higher than Sams due to the higher AC line voltage you used, but are all very much in line with the voltages I measured on the stock unit I had here before it was modified with EFB(tm). With that unit, I was testing it at about 121 vac, or very close to the AC line value you were using for measurement as well -- with the resulting B+ voltages then also very close to yours.

However, your voltage reading at the output of the bias supply diode is significantly lower than that of the unit I had here when stock. Of course, errors on Sam's schematic may in fact be in play here, but the unit I had produced -28.2 vdc at the output of the diode, with 121 vac line voltage, and the unit properly biased at 30 ma per tube.

As a result, I would investigate the values and condition of the components used in the entire bias supply circuit. In particular, the value/condition of R51, R56, C2 and M1 will have the greatest impact on what the available voltage is at the output of M1.

I hope this helps!

Dave

I'll check out the bias supply tonight. The problem stayed with the socket, and did not travel with the swapped tube. The amp switched on this morning (approx 6 hrs after last use) without red plating.
Rich

You know....
There's something I only noticed last night but forgot to mention.

Channel A on the stock SA100 has a 1000 ohm resistor connecting V6P3 to the Center Channel Output jack, then onto the 16ohm taps of both OPT's through 12K resistors. The Stock SA100 also has the cathodes (P3) of V6 and V5 tied together, then onto the single external test point.
Channel B also has the cathodes (P3) of V2 and V3 tied together and onto the appropriate single external test point.

Since I separated the cathode pairs of each channel and ran them to four individual test points, there is now an imbalance between the cathodes of V5 and V6 of Channel B with reference to ground during testing when the jumpers are removed.

Do you remember when I initially asked why there were two different capacitor values between Channel A and B on the Sams schematic? Sam's also has a discrepancy as to where the capacitors are placed in the schematic as well. Channel A has 12pf to ground before the 2200 ohm resistor to V3P2, Channel B has 3pf to ground AFTER the 2200 ohm resistor to V6P2.

I know Sams made mistakes in the schematic, and I rewired the amp exactly as I took it apart, but now I'm starting to wonder if there's another problem going on because of that 1000 ohm resistor to the center channel output?

Hi Rich -- The Sams schematic indicates that the 1000 ohm center channel resistor may be directly grounded, or connected to the cathodes of V5 and V6. In reality, how the resistor was "grounded" was likely based on what worked best on the production line.

Technically, the resistor should be connected to ground, end of subject. However, the cathodes of the two tubes mentioned were normally grounded during normal operation anyway, so if it was more convenient point to use construction wise, the cathodes made for just as good of a "ground" point as well. When the straps were opened, the 10 ohm cathode resistor placed in the circuit still represented such a low impedance to ground (relative to the value of the center channel resistor), that it just didn't matter whether the center channel resistor was actually grounded, or returned to the cathodes with the strap either open, or closed.

The same schematic also indicates that C12 is a 12 pf cap connected before the grid stopper resistor for V3, and is standard on all builds. However, its relative for V6 is shown as optional, and therefore could very well be placed as shown, before the grid stopper resistor for that tube, or simply omitted as well.

Fisher had its ghosts it dealt with regarding HF stability in many of its models, as these types of components were known to change more than just a time or two for a given model, both in value, and location. If you rebuilt your amplifier as it was originally built, then that was the best move to make for your particular unit.

Dave

Dave,

Well, I just finished a nice listening session and was honestly impressed. The cold amp started fine, bias for each tube was around 28ma per tube and balanced after a 5 minute warm-up. The sound was powerful, clear and with depth. Can't ask for much more in an amp.

I started listening at lower volumes (as I was also breaking in a new set of speaker crossover caps) and gradually raised the volume as the new caps lost their "boom and sizzle". The caps were very listenable after an hour and I raised the volume accordingly, final volume was just below my wife's "piss off" level, which is actually pretty loud in the morning. This amp has plenty of bass, tight and powerful. High end is clear, never got the impression anything was distorting.

Before I shut the amp down after a few hours of use, I measured the bias currents, which dropped evenly to around 26ma per tube. The DC balance wasn't too far off, varying only slightly over 1 ma. The transformers were on the warm side, not enough to cause a burn or discomfort.

So I think things are working well. Perhaps the DC balance controls and sockets only needed a good cleaning. I'll continue to monitor the amp for another week or two before I put the cage back on.

What bias do you recommend for the GE 7189A tubes that I'm using? I assume it's best to set the bias and balance after a good workout.

Rich

Hi Rich -- It is always good to set the final (i.e. semi-permanent) bias and balance adjustments after a unit has thoroughly warmed up.

As for the bias setting, the amplifier will perform its best with each output tube drawing 30-32 ma of quiescent current. That's not to say that it won't "sound" good at a lower setting -- it's just that the optimum setting for maximum performance is in the range I mentioned.

In any event, I would not allow the amplifier to operate with any less than 25 ma of quiescent current per tube, as below that point, crossover (or "notch") distortion can start to form quite rapidly, which is a particularly unpleasant form of distortion. Granted, it's worst effects are only noted at higher power levels, but distortion is ultimately increasing as you move either side of the optimum setting.

Finally, within a give tube "family", which in this case is the EL84/6BQ5/7189 (and others) group of tubes, the optimum bias setting typically remains the same within the family, with the only variable being that of manufacturing tolerances -- which exists from tube to tube even within the same tube type. Of course, that's why all the various bias and balance controls exist on better designs, to adjust for these tolerances.

The primary difference that exists between various examples within a tube class are typically associated with maximum voltage, current, and dissipation levels that are permissible. Such is the case with 7189 tubes, which is rated for higher voltage operation over its other siblings. As such, the amplifier won't really perform any better with these tubes over other tubes in the class, but they will operate with a greater safety margin because of their higher ratings.

Dave

Thanks for the bias range recommendation. I dropped the bias below 30ma after the red plate incident because I was afraid it might run away. When I measured that tube, it must have been drawing close to 70ma and all of the other tubes were very under-biased. That (7189A) tube was tough, I've had other EL84 tubes go red on another amp and the sound just faded out on that channel. The 7189A continued to play as the plate glowed, I wouldn't have noticed it had I not gotten up to change the record. I was playing the amp rather loud yesterday and did notice the distortion as you described. I'll raise the bias level later today and keep an eye on things for a while. I was using Amperex 6bq5's, but the 7189A's seemed to sound better at louder volumes.
I also noticed that the bias levels bounce around during play. I know this was one of the things your mod addressed, but if the bias on one tube gets too high or low, can it or it's counterpart become damaged? Is there a threshold or point of no return that breaks the camel's back? That was my thinking when I biased the tubes low while the amp idled. I was afraid that during loud play the amp might damage itself as the voltages varied.

Hi Rich -- You're confusing two different things with what the EFB(tm) modification does, versus what you've witnessed at the bias test points.

The EFB modification does act to modify the bias SUPPLY voltage -- but only as a reaction to changes in the B+ supply voltage. What you are witnessing with the bouncing cathode voltages during program play is in fact a vacuum tube in action, doing exactly what it is supposed to do.

When you are setting the voltage at the TP points, you are in fact adjusting the negative grid bias voltage applied to the control grid of the individual tubes, so that they will ultimately draw a predetermined amount of current under quiescent conditions, as measured at the cathode test points. Since the cathode test point has a 10 ohm resistor connected between it and ground, then by Ohms Law, .30 vdc across a 10.0 ohm resistor equals 30 ma of quiescent current draw. Quiescent conditions are defined as the amplifier operating, but with no signal present.

However, when a signal is present, the drive voltage that is also presented to the control grid of the tube you are monitoring is acting on the bias voltage that appears there -- whose value is based on the adjustments you made to have the correct voltage appear at the cathode TP point. As this signal drive voltage acts on the bias voltage at the control grid, it adds to and subtracts from the bias voltage, varying the bias voltage around the set point you adjusted it to during quiescent conditions. How much it varies the bias voltage depends on how much power you are asking the amplifier to produce. Since the bias voltage at the grid is therefore varying because of the presence of a signal through the amplifier, so will the voltage at the cathode then vary, because of the varying current the tube is passing between the cathode and the plate. Since the plate is connected to the primary winding of the output transformer, this varying current creates an every changing magnetic force in the transformer, which then induces a signal in the secondary winding the of the transformer, which is ultimately connect to your speaker. Therefore, what you are seeing with the bouncing levels at the cathode with the application of program material is in fact a visual indication of what you are hearing in your speaker.

Finally, you will find in fact that the voltage at both cathodes of the tubes in each channel is varying by exactly the same level or amount -- but in exactly the opposite direction. This is the essence of push-pull operation, where one tube conducts more current, while the other conducts less, and visa-versa, with the application of signal. Hence, the classification of "push-pull" is given to an amplifier of this type design.

I hope this helps!

Dave

Dave,
Sorry it took so long for me to get back to you, I've been enjoying the amp so much and reading up. Plus, anytime I do a project, my work area becomes a mess. This was a big project and it took me nearly a day to clean up my mess.

Thanks again for the explanation, I understand the difference between the quiescent bias setting and what happens during play as well as push pull theory. I've monitored the idle levels at startup, after warmup and after play and observed what happens during play as you noted. I happened to be pushing the amp when the red plating incident occurred and was wondering if there was a level not to surpass and whether your EFB circuit acted as a "voltage regulator" or more properly a "voltage limiter". As I mentioned before, my understanding of your solid state circuit is very limited (I know what it's purpose is, I just don't understand "mathematically" how it does it.). I never really spent much time with solid state circuits, at least not enough to understand them comfortably.

I had read many articles about musicians biasing their amps "hot" or "cold", which is why I biased my amp "cool" at 28ma per tube as I observed it during the break in period, especially after the red plate incidents.

Anyway, I think that's all cleared up and I'm on to my next project, a couple of Fisher 481A and 460A amplifiers pulled from consoles for a couple of friends. I would like to install resistors to monitor currents but these amps are cathode or "self" biasing designs unlike the SA100 which is a "Fixed Bias" design. Any suggestions or should I simply set the cathode voltages and let the tubes self bias?

Finally, I attached to this message what appears to be an actual Fisher schematic for the SA100. The voltages are very different and this appears to be a better reference than the more common Sam's Photofacts. If this is the correct schematic, I will rescan it at a higher resolution, dig up the missing parts list and attempt to add it to the AudioKarma Files as a better reference than the Sam's Photofacts.

Hope this helps.
Rich

Hi Rich -- Everything I see on the schematic you provided indicates that it is in fact that of the SA-100. It may be a schematic from a larger group of schematics for a console where the SA-100 was frequently used, but it still appears to be the schematic for that unit. Regarding the voltage indications, one thing that seems to be consistent with Fisher and Sams schematics is that Fisher schematics always show the lowest operating voltages, Sams a higher operating voltage, and then of course there is actually what is measured, which is always higher than what either schematic shows!

If an amplifier is operating properly -- and importantly -- loaded properly, then there is no specific level of power output you should limit the amplifier to. In fact, the classic sound that electric guitarists seek is the result of the power output tubes being driven into the soft clipping produced at the onset of maximum power output.

This type of operation is harder on the tubes -- but only in the same way that causing an engine to produce maximum power output is harder on it as well. As long as both are properly designed, and a proper load is in place -- in both cases(!) -- then the life of the tubes and the engine is simply being used up more quickly than if operated at lower levels -- but they're not be abused. More specifically, with proper design and installation, then continuous operation at full power output for either the amplifier or the engine should not produce any cause for concern.

To punctuate this point, in all of the various designs and modifications I have produced over the years, I rather routinely operate the final design at a sustained maximum power output in both channels, with both channels driven, for a period of at least 1/2 hour to one hour, fully buttoned up with all panels in place, and installed in any cabinets provided for the unit -- this to prove the durability of the components specified. Most recently, this procedure was part of the testing process for my SA-100 clone amplifier, the installation of EFB(tm) into a Fisher SA-100, and the installation of EFB into a Fisher 400 receiver, all as documented here on this forum.

Of course, if an amplifier is not operating properly, and not loaded properly, then all bets are off -- as would be the case for an engine as well. Almost always, when an amplifier is presented with extreme damage from high power operation (melted tubes (literally), overheating, or arc damaged components and transformers), it is because it was extremely abused with regard to maintaining a proper load under high output conditions. Sometimes its simply an unfortunate event, as in a speaker voice coil burning out -- but the results are the same, and so are the facts: If you're going to operate a powerful device at full power, then a durable load must be provided that can adequately and routinely handle that power, so that unfortunate events don't occur.

Relating all of this to EFB operation then, the EFB regulators are just that: they are regulators, not limiters. But they are very specific regulators designed to keep all of the voltages to the various tube elements in lockstep with each other, in spite of any variations in voltage from the main B+ supply. This provides a tremendous performance advantage over traditional designs, and is economically possible to implement today because of the availability of components that designers back in the day simply had no access to.

As for your upcoming projects, be careful with those amplifiers if you are trying to separate them from their preamp units for standalone operation. They typically require the connection of the preamp unit for proper biasing of the output tubes unless modifications are performed. If you are attempting this, you might search out a thread where in Derekva performed this operation to a 460A (I believe), and went through the necessary steps for stand alone operation of the amp.

Good luck with your amp, and your projects!

Dave

I wasn't planning on pushing the SA100 or any amp the way a musician might to the point of distortion. I saw a "formula" of sorts used to determine the quiescent bias current for a given tube, something like 25% of it's full rated output, and was wondering what you determined empirically during testing.

Regarding the 460A, 480A and 481A, prior to the SA100 I had already modified a couple of 481A's for "Stand Alone" use. The big trick there was to add a couple of resistors to make up for the missing tuner/preamp chassis (three 12ax7s in the tuner/preamp stage have their filaments wired in series with the bias supply coming from the amp). That was done, all voltages were fine, and I consider my 481A to be the best $125 amp I have ever heard.

But since the bias design is different than that used by the SA100, I was wondering if there was a way to monitor current drawn by each tube instead of measuring global voltages. Presently, I use the chassis mounted molex connectors as my test points.

Rich

The proper quiescent current for an output stage is based on many factors. The only rating of the tube that comes into play regarding establishment of the correct level is to make sure that the proper level doesn't cause the dissipation rating of the tube to be exceeded under quiescent conditions.

In the thread where I was installing EFB in the SA-100, I first corroborated the factory recommended setting for the stock design. In fact, the low distortion operating point came in at ~ 31.5 ma per tube, which is very close to the factory recommended level. After the installation of EFB, I again tested for the lowest distortion operating point, which then came in at 24 ma of quiescent current per tube.

On your 481A, you can basically perform the same modification on that amplifier as you did on your SA-100 to measure individual output tube current draw.

The process is the same EXCEPT:

1. In the 481A, the cathodes are all connected together, and then connected to the bias circuitry. The first step is to disconnect everything from the cathode terminals of the four output tubes.

2. Tie EVERYTHING that is part of the bias circuitry back together again -- as surely as they were all originally connected together before -- but this time to a new isolated tie point. This new isolated tie point will also be the point where the common lead of your meter connects to when making your measurements.

3. With each cathode terminal now completely free of any connections, connect your precision 10 ohm resistors between each cathode terminal, and the new isolated test point.

And there you have it. The new isolated test point is NOT at ground level, but is the proper point to connect your meter to when checking the current draw of each output tube.

Dave

I went back and read over your original EFB thread with a bit more understanding this time. The difference in quiescent currents between the stock and your EFBII mod is what piqued my interest. Nice approach to base the operating conditions around the lowest distortion points. Less heat is nice too.

Thanks again for the 481 test points. I wasn't sure if the different bias methods made a difference.

Rich

dcgillespie said:

As for your upcoming projects, be careful with those amplifiers if you are trying to separate them from their preamp units for standalone operation. They typically require the connection of the preamp unit for proper biasing of the output tubes unless modifications are performed. If you are attempting this, you might search out a thread where in Derekva performed this operation to a 460A (I believe), and went through the necessary steps for stand alone operation of the amp.

Good luck with your amp, and your projects!

Dave

Click to expand...

Here's the thread, Dave: http://www.audiokarma.org/forums/showthread.php?t=461654.

Cheers!

-D

Thanks for the link. I had already worked this out with some help last year over at the DIY Audio Forum.
Great amps, the 481-A. They're what got me turned on to EL84's.

http://www.diyaudio.com/forums/tubes-valves/218569-console-amplifier-pull-voltages-over-spec.html