Rare Beauty: Fairchild's 275 Amplifier

[dcgillespie]

Yeah, that's the Sam's schematic. It is making an effort to show both the early and updated versions in typical Sam's style. As drawn, it largely depicts the updated version. Please note however that there are a number of errors present in just this snippet of the schematic:

1. R29 is 15K in the updated version, not 150K as shown.
2. C13&C14 are .12 uF in the updated version, not .012 uF as shown.
3. L1 is not present in the updated version -- although the .33Ω resistor is, connected straight to ground.

Dave

 

[Brice]

I see.
I've made the modification and here is the result.
Again it's just the NBF network response just to give an hint of what it is. The 3 traces are for the damp pot to 0, 1/2 and 1 wiper position.
That's all for now.

 

[dcgillespie]

Thanks Brice!

 

[radioactive]

Hi Dave,thank you for your very in depth analysis and findings on these classic amps.im no tech by any stretch but appreciate what your doing to understand these amplifiers and explaining the different parts of the circuits and why they revised it.

 

[radioactive]

Fwiw Fairchild had a revision for the 260 which used el34 tubes instead of 1614/6l6gc tubes.i could send you the revision schematic and notes plus a original schematic for you to look over when you get a chance as I was thinking of converting mine to el34 or even kt77's if it would make a improvement

 

[dcgillespie]

CIRCUIT ANALYSIS FINALE

1. NFB Loop: Since this thread is about the Fairchild 275, I thought it best to include performance information on the loop of the original design offered as well, since some of those units are surely out there. At the very least, it would satisfy my own curiosity. Therefore, I went ahead and disconnect the networks installed, and tacked in the original NFB/HF stability networks, configured for the max damping (damping off) configuration. If there was any doubt before, there is none now. The changes were clearly made to address significant stability issues, in at least the max damping configuration:

Refer to Post 25 and the first two pic presented: The magazine scope shot of the unit producing a 40 watt 6 kHz square wave, and this unit here with the updated NFB design producing a 40 watt 6 kHz square wave, both with the damping set for maximum. Now look at the following pic: Look familiar? It too is the unit producing a 40 watt 6 kHz square wave, but with the original FB and HF stability networks installed:



The third pic in Post 25 was of a 10 kHz square wave, and for comparison, a 10 kHz waveform made under the same conditions is presented here:



The overcompensation employed is more evident at this frequency, but still, the improvement in transient response is quite evident over that of the updated design. However, as I originally mused in my remarks when the original pics were presented in Post 25, in my experience, square wave presentations of this type and frequency from a feedback amplifier invariably imply HF stability problems, and boy oh boy, the original design had it in spades:

Below is a pic of a 10 kHz square wave, with no load other than a .005 uF cap connected to the 16Ω output:



Here the unit is on the verge of breaking into full blown oscillation. Look closely and you'll see the faint scatter image behind the main image indicating that condition. With just .0075 uF capacitance in place, the unit becomes a 70 watt radio station, disrupting every radio in the lab. So both of the Fairchild NFB and HF stabilization networks had notable issues -- the original design producing good transient performance but poor stability (which btw operates with 16 db of NFB), and the updated version producing good stability but poor transient performance -- and both while set to the max damping position. That means that as originally suspected, simply wiring either network in the damping "off" configuration results in notably less than optimum performance capability.

SIDEBAR: This is hardly a slam against Fairchild. The mid 50's was a time of great change in the High Fidelity industry. The power wars were beginning, stereo was on the horizon, new and higher performing tubes were becoming available, and the very definition of what a stable NFB Amplifier looked like had yet to be developed. Up until that time manufacturer's wrestled with amplifier stability, and not just a few of them. Nor was it limited to the also ran names, either. Fisher, Heath, Eico, Acro, Scott and others all had their headaches with amplifier stability issues, resulting in numerous versions (or revisions) or even new model numbers to distance some units from their previous history. Often -- and I do mean often -- in trying to deal with stability issues, the manufacturer would make revisions that took certain aspects of performance from one end of the spectrum completely to the other, causing just as many, but now different performance problems in the process. Fisher did this with their 50A design series, Heath with their early Williamson series, and others did so with some of their models as well. Some may remember threads I've done on these and other amplifiers that were dealing with exactly the same thing, as a proper balance of performance attributes had not been struck at the time. So the performance issues identified are hardly unique to the Fairchild 275, but the result of changing understandings that were happening overnight. As with so many of the affected models however, by the time even the earliest understandings of what good NFB stability represented came to be, previous models were either by then out of production, or new designs were replacing the old, so that what was, was simply, what was.

CONTINUING: With this understanding of both Fairchild networks then, development work on the new NFB loop and HF stability networks is now finished. In every case where this type of modification work has been done, the performance improvements are audible and notable: marginally stable NFB amplifiers tend to sound harsh in the upper registers, while overly compensated amplifiers ten to sound dull or lacking in detail. When a proper emphasis/compromise is placed on all the relevant issues that make up total amplifier performance, then performance is optimized, and the listening experience enhanced.

Every effort was made to use as many elements from both Fairchild versions as possible to minimize the design changes, and simplify the modifications required. The results are quite significant:

1. Frequency response is now ruler flat from 20Hz to 20 kHz (+/- 0.1 db), with a smooth roll off above and below this band, being down 0.5 DB at 40 kHz, and 6 db at 2 Hz (LF unchanged from original design).

2. Square waveform (10 kHz) as previously presented, with 3 µs rise time, less than 5% overshoot, excellent damping, and a very flat top -- all indicative of a highly stable NFB amplifier in the supersonic range.

3. Fast settling with no bounce under pulsed conditions.

4. Absolute stability into any conceivable loading condition (short, open, capacitance only).

5. The level control does not cause any significant change to a 10 kHz square wave at any midpoint setting.

Details of the new NFB loop and HF Stability networks will be given at the end of the circuit analysis section.


OUTPUT TUBE TESTS

EL34 and KT88 tubes were tried in the output stage. EL34s produced a slight power loss (~ 8 watts), but was most notably more challenged at 20 kHz. This no doubt is due to the winding capacitance within the large output transformer, that the higher current capabilities of the 6550 and KT88 tubes are better able to deal with. Somewhat surprisingly, the KT88 and 6550 tubes performed virtually identically, so these tubes can be used interchangeably. For those wishing to use EL34 tubes in 275 amplifiers, they can do so if an 18K resistor is connected in parallel with R19. Otherwise, there is not enough bias control range on the low end of the control to accommodate EL34 tubes.


Just a few mop up details to present next time, but with analysis of nearly all the critical circuit elements finished now, a restoration plan that optimizes the amplifiers can now be made, and restoration work begin in earnest.

Dave

 

[dcgillespie]

ODDS N ENDS

CHASSIS ADJUSTMENTS

The procedure Fairchild published for adjusting the Bias, Current Balance, and Grid Balance adjustments can all be improved upon in simple ways to improve the accuracy of the adjustments. Some improvement is obtained by way of providing greater detail of the adjustment process, some by using more sensitive indicator methods, while others come from a procedural change in performing an adjustment. The information provided here then is all new information, and works (as appropriate) in conjunction with that originally provided by Fairchild, by either augmenting or replacing their procedures:

1. Always perform all three adjustments when the output tubes are changed. Replacement of all other tubes does not require adjustment of these controls, nor do the adjustments affect any of the other tubes.

2. Always start the adjustment process with the Bias control set full clockwise, and the other two controls set to the center of their rotation.

3. Always make the adjustments in this order: Bias, Current Balance, and finally, after those two adjustments settle out (they are interactive in nature), the Grid Balance adjustment. After the Grid Balance adjustment is made, touch up any settings as necessary for the Bias and Current Balance adjustments. No further adjustment of the Grid Balance adjustment is necessary once set for the output tubes installed (unless they have become well worn with use), so that any future quick checks of the output tube bias settings only needs to concentrate on the Bias and Current Balance adjustments.

4. Always make all three adjustments with the 12AV7 AF Amplifier/Phase Inverter tube removed.

5. If possible, make the Current Balance and Grid Balance adjustments using a scope or headphones connected directly across the Com and 16Ω terminals. Otherwise, use the most sensitive speaker available, positioned so that you can place your ear directly up to the speaker board.

6. When making the Bias adjustment, insert the test plug (with connected voltmeter) just far enough into the jack, so that the tip of the plug just touches the tip connection within the jack, but does not engage with it to snap the plug into place. So installed, adjust the Bias Adjustment so the meter reads 0.50 volts. No hum will be heard in the speaker or headphones when making the adjustment this way. This is an improved procedure, that allows for setting the output tube bias under truly quiescent conditions. Fairchild's procedure has the adjustment being made with the phone plug fully inserted, which then applies a test signal to the output tubes (unnecessary for setting the bias), that then necessitates making the adjustment to a higher .8 volt setting to compensate for the increased current draw produced by the introduction of the test signal. The procedure presented here is much more accurate, where as the Fairchild approach by comparison is a general approximation that unnecessarily depends on other factors for accuracy. The modified approach eliminates those factors to produce a more accurate result.

MEASURED POWER OUTPUT AND DISTORTION PERFORMANCE OF ORIGINAL BUILD

In establishing part of the base performance information for the design, both THD and IMD were measured -- THD at a level (customarily) 1 db below maximum power output (in this case 60 watts RMS into 16Ω), and IMD at the onset of clipping (57.45 watts) using a 60 Hz and 7 kHz mixed 4:1 IM test signal:

THD
@ 20 Hz = 1.55% Maximum power output = 60.1 watt RMS.
@ 1kHz = .32% Maximum power output = 60.1 watts RMS.
@ 20 kHz = 2.40% Maximum power output = 54.4 watts RMS.

IMD
1.9% @ 57.45 watts equivalent RMS power output.

THD performance for the 275 was never published by Fairchild, while IMD was, and is specified as <2% at 70 Watts. While measured IMD was within specifications, power output was not. This is with the unit set for 125 vac operation while operating from a 122 vac line, and the unit being tested with early style Russian 6550 tubes installed, with many such tubes historically being shown to be only capable of producing at best 90% of the power output that the original American pieces could produce (although this is getting better). Also of note, power output ratings of the day were most often based on the power output as developed at the plates of the output tubes -- before the output transformer -- not the power actually delivered to the load. So an understanding of the day and times must be had to put the power output rating of the amplifier into perspective. Manufacturers of 6550 tubes indicate that operating with 450 volts of B+ under full power conditions in push-pull UL mode, they are capable of delivering 70 watts of power into the output transformer. This is no doubt the basis that Fairchild used in rating the power output of the 275 amplifier. Their operating conditions uses a slightly lower load impedance than the published value, but their B+ voltage is not regulated. So there are deviations from the published operating conditions, but ultimately tend to cancel each other out. All in all then, with an output transformer of 90% efficiency, a power output of about 60 watts RMS should be considered as typical, and is typical of many other amplifiers of the time period employing UL operation of 6550 tubes.

AC VOLTAGE ADJUSTMENT

With my line voltage here (pretty typical at 122 vac), it was found that direct operation on the 125 volt tap produced the closest correlation with schematic voltages. Most notably, main B+ is specified as 440 vdc for both the original and updated version, and direct operation on the 125 volt tap produces a B+ of 450 vdc. As well, direct operation on the 125 tap produces a heater voltage of 6.36 vac at the output tube socket terminals furthest from the power transformer, which is ideal.

For comparison, operation was also tried using the 117 volt tap, with a CL-80 current limiter installed. With this scenario, B+ rose to 473 vdc, and heater voltage rose to 6.67 vac, measured under the same conditions. In view of minimizing under hood temps, maximizing dependability, and with knowledge of what type of speakers these amplifiers will be used with (horns, so power output is a guaranteed to be in plentiful excess), I plan to leave the units set to operate directly from the AC line and applied to the 125 volt tap.

RESTORATION PLAN

With all the necessary information gathered now, a plan has been devised for the restoration of these amplifiers. Besides the usual practice of cleaning all controls, sockets, and jacks, etc., the following points will guide the restorations. Throughout the process, every effort will be made to match the two amplifiers for maximum stereo effectiveness:

1. Replace noisy 6AB4 tubes in both amplifiers. Use matched quad of Tung Sol 6550 output tubes provided. 5V4GA rectifier and 12AV7 driver tubes are quite good. KT88 and GZ34 tubes can be used if desired.
2. Replace missing fuse and fuse post cap on one amplifier. Replace missing tag board hardware on both amplifiers.
3. Replace input jacks and power cords with new items provided.
4. Replace dual first filter cap with single CE Manufacturing 30-30-30-10 525 volt can configured for single 90 uF or 70 uF as desired, as used in original or updated design respectively. 525 volt rating exceeds normal peak voltage rise at turn on by 30 volts. Exceeds B+ produced with output tubes removed by 20 volts. Offers 75 volt safety factor during normal operation.
5. Replace secondary can with CE Manufacturing 40-20-20 500 volt can configured for 40-40 uF with improved safety factor over original 450 volt can. Use the new grounding scheme developed for lower noise.
6. Remove old Damping/NFB/HF Stability circuits and replace with new NFB and HF Stability networks developed. Install hole cap where damping control was previously located (client preference).
7. Add Screen Stability resistors to output stage.
8. Add EFB™ Control Grid regulator circuit for control of output stage bias voltage. Circuits to be added where mount clip for second cap of dual first filter cap resided (no longer needed).
9. Replace all coupling and bias caps.
10. Replace all current carrying resistors resistors, and any other resistors as drifting would indicate or matching would dictate.

Chassis controls will be set, and new components allowed to burn in for 1 week. Chassis controls then checked and adjusted as necessary, and final performance of finished restorations measured and documented.

So this will form the core of the restoration effort to work from. The plan will be modified as/if needed along the way. As I get into it, I'll post some pics of the progress being made.

Dave

 

[kward]

Add EFB™

Okay, cool, and on a monoblock amp to boot.

On the KT120 mark III amp that I scratch built using Triode Electronics clones of Dynaco A431 OPTs (with 33% distributed load taps), I added EFB and (thankfully) was able to find a distortion null at about 90 mA quiescent cathode current (with 475V quiescent plate voltage). This was "quite a bit higher" than I expected. My amp measured 0.4% THD at 1 KHz at 50 watts (max) output at 60 mA quiescent cathode current per tube, both channels driven, w/o EFB. When EFB was added, I obtained 0.15% THD (or close to it) under same drive conditions, at 90 mA quiescent cathode current per tube. So EFB was definitely doing its handy work.

With the OPTs on the Fairchild at 25% distributed load taps and being monoblocks, I am really curious where you will find the distortion null and what THD you will get at that bias point. Watching with interest....

Also I'm impressed with the detail of your build plan. Most people (me) wouldn't take the time to write it down to that level of detail. Something for us plebs (me again) to aspire to.

 

[radioactive]

i will be following and interested with what you decide to use for coupling caps,resisitors and whatever else parts you replace.also thanks for your indepth anaylisis and your approach to stting the bias etc..

 

[radioactive]

Fwiw Sam never included the bias setup instructions for some reason for the 275 but they did for the 260.it took me forever to find the manual for the 275 with the way they wanted you to do it.your way is very much appreciated and if I ever get another 275 which I foolishly sold I'll use your way.

 

[gadget73]

which then applies a test signal to the output tubes

Where does the test signal come from? Just a tap off the heater supply or something of that nature so it produces hum?

 

[dcgillespie]

Correct Gadget. In this case however, it is sourced from the 75 vac tap on the HV winding that provides bias for the output tubes.

Dave

 

[Dswankey]

I can barely walk and chew gum at the same time so reading this thread and ones similar makes me really appreciate the engineering/electrical knowledge all of you impart & share.

Respect!!!!!

 

[dcgillespie]

Radio -- Many years ago (25 or so), I participated in all manner (quite a few) of double blind tests regarding the effects that various components and claims had on delivered sound. There were about 15 to 20 of us at any one session, made up of folks of all various stripes -- including those quite advanced in audio technology, those who could only plug and unplug, those who said they could hear the slightest changes produced, to those who maintained there were none. There were even a few psychologists in the bunch who helped to maintain proper test format. We were quite a determined bunch who for a couple of years, were all truly intent on trying to get to the bottom of the audio snake oil pit to see what was down there (anything from the "polarity" in which an AC power cord was inserted into the wall -- some of the stuff was just crazy -- to the sound of various tubes, components, wire, and even solder. We had some really great sessions together with lots of opinions of course, but with each claim put through the same testing procedure: the participants had to achieve at least 50% accuracy as to identifying when or if a change was made, and to which option presented. By the time we had disbanded, there were (we virtually all felt) some rather clear findings from our rag tag bunch:

1. When just a change (only) was made between (say) certain types or brands of tubes, or using mylar caps in place of ceramic caps, then there was actually quite credible correlation among those most sensitive to sonic changes, although when the whole group was considered, the results could only be considered as inconclusive. Later tests showed that those producing credible correlation also typically had the widest hearing response as well.

2. When just a change (only) was made between things like type of connecting wire, solder, gold versus conventional plated jacks, etc., there was virtually zero correlation with anyone in identifying any of these elements in or out of the system.

3. What was really fascinating was that when digging into the technical side of why there was any test accuracy shown in #1 above, we were always (i.e., in virtually every case) able to identify a technical reason for it. For example, when different examples of 12AX7 tubes were put to the test in a phono preamp circuit, we only used examples that had first tested for the same Gm indication in both sections -- which the sensitive ears guys were then able to call out with accuracy. Circuit performance was then measured for gain and EQ differences, only to find that the real issue was the AC gain that the various examples of 12AX7 tubes used in the test displayed -- not only between different tubes, but between the sections within the tubes as well. When tubes were then tested between brands that previous had showed credible difference -- but were then chosen for examples that were equalized for AC gain -- then any test correlation fell completely apart. The same thing happened with cap tests between ceramic and mylar: once values were closely matched, previously shown correlation fell apart. It was noted that ceramic caps rather universally read notably low in value, which then produced the differences noted when a true value cap of any typical makeup was used.

The bottom line we took away was this:

A. Manufacturers producing similarly rated components all have production tolerances unique to their own machinery, which can cause -- of their own application -- certain brands to perform with a distinctive signature, for those able to hear it.

B. When those tolerances are appropriately accounted for (equalized), any sonic differences previously attributed to the manufacturer, component construction, or internal makeup of all the different components used virtually disappear.

C. Most important then, was that it was ultimately determined (for us anyway) that it was the circuit that determines a given sound. That within a given circuit, the character of a given sound attributed to the use of a particular component or components was only due to the degree that those components varied from an established electronic standard. In other words, there was no magic contributed by any particular component brand or type -- only (potentially) a generally accepted effect produced from a unique range of production tolerances attributable to certain manufacturers or component types.

We also learned that the bias of anticipation is absolutely huge. When you know you have made a change, know what that change is and when it was made, and are personally involved or invested in the outcome of that change, then the perceived results can be extremely skewed. Lots of stuff is sold that way. The double blind testing removed so much of that bias, and brought in so many more elements of observation (participants) that it really helped to put some sanity to it all.

Long story short then, those sessions all left a strong impression on me. Therefore, while I have no issue using any specific brand of quality components (as long as they can properly fit physically into the allotted space to prevent unintended stray coupling), neither do I have a dog in the sonic hunt, either -- other than to use components whose quality is well established, are appropriate for the application, and that are of close tolerance to the intended value -- be they tubes, caps, resistors, whatever. As a result, I tend to use quality components of little fanfare, and through appropriate attention to proper circuit operation or established norms, have never had any concerns or feedback from using that approach.

Just one man's (or group's) findings, but recognize that all opinions are valid!

Dave

 

[x3workshop]

Great thread, Dave. As always, superbly laid out, well considered, and backed up with copious amounts of data!

 

[nj pheonix]

 

[BillWojo]

I've been watching this as well, they are held in very high regard and I absolutely love Dave's approach to rebuilding something. I'm not going to pretend I know a lot like some of the folks on here but his technical discussion on the stock circuits and how they can be improved upon with newer technology is always interesting.
What I find interesting about this amp is the inherent flaws that he measured compared to the sonic reputation that it has. Somehow I think when it's done it will not only measure much better but will sound better as well.
I think there is probably a lot of magic in those output transformers. At first glance it looks a lot like a repainted Mac monoblock. Both Fairchild and Mac spared no expense when they designed and built a product, they were the best that could be made back than.
I'm seriously hoping that Dave decides to do his magic on a Leak Stereo 50 someday. I have probably the rattiest one in existence and it would be nice to have a master lead the way.

BillWojo

 

[dcgillespie]

RESTORATION UPDATE

First a quick followup on the transient performance of the revised NFB/HF Stability networks. I realized I had never posted those from the finalized design:

A 10 kHz square wave into a 16Ω resistive load:


Same square wave frequency, no load other than a .022 uF cap across the 16Ω terminals:


Same square wave frequency again, no load other than a .25 uF cap (the most reactive value) across the 16Ω terminals.


The scope shots show a superb balance of excellent HF stability (the amplifier is showing itself as extremely stable under this most difficult loading condition) -- which the original version suffered in, and excellent HF transient response -- which the updated version suffered in. Together, the addressing of these two performance characteristics represents a significant improvement for the amplifier.


As for the restoration itself, a lot of "infrastructure" work if you will has been done, that forms the basis of a great build. It will also now let the project break out into some real running room. First off, a pic of where the first unit started from after conducting all the base line testing:


Next, removing the clip holding the bottom cap of the dual series cap that made up the first filter cap. By all indications, this was an afterthought as unlike most all of the other small hardware on the chassis, the clip is mounted with a screw and nut. This also jives with the early schematic that simply shows a single 90 uF cap, with the updated schematic showing dual in series 140 uF caps. My thought is that the original 90 uF cap either had too low of a voltage rating for good long term dependability, or the caps themselves were an issue. Either way, it makes it easy to remove the clip and replace the screw and nut so that topside, nobody's the wiser:


Next up, sorting out the power supply ground mess. There are two ground lugs that are mounted down on a stud of the power transformer by the stand-off for the terminal board. One lug is facing the camera, and the other is on it side and above the front facing lug.

One the front facing lug, the black w/tracer lead (which all original ground leads are) in the foreground goes over to ground pin 1 of the output tubes. The one bending down out the back originally came forward (relative to the pic) and was the ground for the bottom cap of the dual first filter cap. The third black/tracer lead out the left side serves as the ground for the output tube current sampling resistor. The brown/yellow lead is the CT of the HV winding in the power transformer. And that big 'ol black lead? Well that's the common output transformer lead -- a simply horrible place to ground that lead. With the input stage grounds and the common output terminal all effectively grounded over at the audio grounds, and the OPT common lead being the return of the FB system, it means that the NFB signal has noise injected into it, and the loudspeaker current is flowing through the chassis. Not a good thing in making for a quiet amplifier. On the side facing lug above the front facing lug, attached is the ground for the secondary can cap, the CT of the 6.3 volt heater winding, and the grounds for the bias filter caps:


In sorting this out, on the main lug closest to the chassis are all the high current ground leads: The HV winding CT lead, ground connection for the output tube current sampling resistor, the ground lead for the first can cap (configured for 90 uF), and the ground for the bias filter caps. With these leads firmly soldered together, there is no way any of their noise can appear in the chassis for pickup elsewhere. On the second lug are the no-current ground connections, being the CT lead of the heater winding, and the ground lead for pin 1 of the output tubes. Notice what is missing is a ground lead for the second can cap, and the output transformer common lead. These two items have no business being grounded at the power supply ground point:


In this pic, you can see the test jack has been removed, the new can caps are mounted in place, and the secondary can cap has tape wrapped around the top of the can, so that when mounted, the can is isolated from ground. This allows it to be grounded over with the audio circuits it serves where it should be. Also, you can see at the bottom left corner of the tag board the black/tracer lead sticking up that will connect to the post in that corner to act as the ground point for the new bias caps, not yet installed. Finally, on the other side of the board is that big 'ol black wire that is the common lead from the OPT. It will be extended and connected directly to the Common output terminal (keeping speaker current out of the chassis), and from that terminal will also run a new lead that connects to the audio ground, to complete the FB circuit without introducing any PS noise in the process. The test jack was temporarily removed to gain access to the Common output terminal connection:


Finally, in this pic, you can see that both of the OPT screen tap leads have been removed from their pin 4 connections, and Screen Stability resistors have now been added between the transformer lead and pin 4, covered in heat shrink tubing. The addition of Screen Stability resistors is mandatory when the power supply caps are replaced with modern caps in this amplifier to prevent circuit conditions from forming that would otherwise encourage output tube arcing:


The underside as it currently stands. The blue wire running along the front of the chassis (bottom of the pic) is now the ground return of the OPT secondary winding/Common output terminal, running over to the audio ground in the bottom left corner of the pic. Because the OPT common lead now ties directly to the Common output terminal, the new ground lead carries no current at all, other than that flowing in the NFB circuit. Use of an original type wire lead helps to minimize any modification look of the work. At that ground point then is the audio input and audio level control ground, the OPT Common Terminal ground, the ground for the tag board small signal tubes, and the ground for the secondary can cap. Without the dual cap first filter cap and its mounting clip now, the underside of the chassis is really cleaned up from before -- from that and also from the removal of the damping control. There is plenty of room now where that filter cap used to be located to mount a small circuit board to contain the EFB Control Grid Regulator circuitry. Note too that appropriate tag board mounting hardware has now been obtained as well. The old input jack still needs to be drilled out, and a hole cap installed where the damping control was, but after those things are attended to, then work on the tag board components and their circuit configuration can begin in earnest:



Dave

 

[dcgillespie]

Finally, I've also been meaning to post my working schematic of the modified 275, before EFB™ gets applied, so here that is.

Dave



Note that for those familiar with the updated Fairchild schematic, the 6AB4 plate load resistor is incorrectly labeled as a 270K piece, when in fact it is a 100K as shown here. Also note that on that same schematic, the output transformer primary lead color code is shown as reversed, so schematics should be corrected for these errors.

 

[GordonW]

Man, Dave, that's a pretty nicely square-looking square wave there.

Apologies if I missed it in a previous message somewhere- but what's the effective bandwidth of the amp and/or the output transformers? I'm betting it's capable of several octaves above the conventionally assumed upper limits of audibility...

Regards,
Gordon.