Avery's "First" Maker: The TA-800

[dcgillespie]

Meant to add the 10 kHz (1 watt) wave forms of the finished power amplifier sections, as injected at the top of the volume controls. Again, Channel A is on top, Channel B on the bottom, with the second pic an overlay of both. Rebuilding both of these sections from scratch afforded me the opportunity of matching each corresponding component between the two channels. The results show on the overlay pic: The two waveforms simply disappear into one. The stability of both amplifiers is absolute: No amount of capacitance only loading will cause the amplifiers to even come close to unstable operation.

Dave

 

[nj pheonix]

DAMN! That explanation on POST # 125 had me feeling I was on a circular driveway getting nowhere fast. Gotta read it a few more times to look for an exit.

Thanks Larry,
You're not alone.
I can't concentrate that hard now. Maybe a dozen times through later i'll grasp most of the concept without the nuances. Then again maybe not

 

[dcgillespie]

The driver tube modification is all completed now in both channels, and the TA is buttoned back up for some weekend listening. In the lab, input sensitivity, frequency response, and channel balance all measure as they were before with the stock driver circuit, but now, it is dead quiet as well. Enjoying a few tunes before heading out for a HS reunion..........

Dave

 

[Rob Thomas]

How does she sound?

 

[dcgillespie]

To my ear, absolutely identical to the stock design, which of course was the goal. As for the sound of the TA in general, it has a mid-range presence that is simply wonderful. It also doesn't have the characteristic soft high end that many of the later units tend to display. It's hardly in your face or brash, but appropriately there. The low end is full, particularly in the mid bass, but low bass is no slouch either. It has no trouble delivering the goods down deep when required. Very listenable, not fatiguing, with that I-don't-want-to-turn-it-off quality!

Dave

 

[Rob Thomas]

To my ear, absolutely identical to the stock design, which of course was the goal. As for the sound of the TA in general, it has a mid-range presence that is simply wonderful. It also doesn't have the characteristic soft high end that many of the later units tend to display. It's hardly in your face or brash, but appropriately there. The low end is full, particularly in the mid bass, but low bass is no slouch either. It has no trouble delivering the goods down deep when required. Very listenable, not fatiguing, with that I-don't-want-to-turn-it-off quality!

Dave

Wonderful! It will be interesting to see what EFB adds to the equation.

 

[monkboughtlunch]

It will be interesting to see what EFB adds to the equation.

Does EFB alter the Fisher sonic signature in any way?

 

[larryderouin]

EFB doesn't alter the sonic qualities of the unit. It tightly regulates, as I remember it, the screen voltage to the plate voltage, and allows the tube to bias at idle at a significantly lower value than stock which increases the life of the output tubes.. Distortion is less, but output power is not appreciably increased, maybe a couple watts. Again it doesn't alter the sonic qualities.

 

[Rob Thomas]

EFB doesn't alter the sonic qualities of the unit. It tightly regulates, as I remember it, the screen voltage to the plate voltage, and allows the tube to bias at idle at a significantly lower value than stock which increases the life of the output tubes.. Distortion is less, but output power is not appreciably increased, maybe a couple watts. Again it doesn't alter the sonic qualities.

The biggest reason for pursuing EFB in this project is the lower operating temperatures and less strain on the components.

 

[monkboughtlunch]

Are there any tradeoffs or drawbacks to using EFB -- or does using it result in a 100% net positive? How much longer can output tubes last with EFB? And does it reduce strain on any other parts besides the output tubes? Also, what does the EFB acronym stand for?

 

[gadget73]

Enhanced Fixed Bias. There are actually a couple versions, one is only controlling grid bias relative to the plate supply, and EFB-II which controls both grid bias and screen voltage. How it controls the grid bias depends on the amp. Ones with a negative DC supply like the Fisher use one circuit, stuff that was originally cathode biased with a big resistor use another type. Gets the same thing done, but you have to get there in a slightly different way.

Dave can far more accurately tell you about it than I can, but it seems to be all roses and moonbeams as far as all that goes.

 

[dcgillespie]

The version of EFB(tm) that is controlling just the grid bias voltage relative to the plate supply voltage is specifically for applications that are either Ultralinear (UL) connected, or are pentode output stages where the screen grids operate at virtually the same voltage as that of the plate voltage. In those cases, because the screen grid is (basically) powered from the plate supply, the use of a Screen Grid EFB regulator is not applicable. However, in pentode applications where the screen grids operate at a notably lower voltage than the plate does (> 5%) -- as is the case in the TA-800 -- an EFB Screen Grid regulator will be required. And, because the output stage employed fixed bias to begin with, it will require an EFB Control Grid regulator as well -- as opposed to a Cathode EFB regulator, which was the initial incarnation I used to originally introduce the EFB concept in the Dynaco SCA-35 integrated amplifier.

In applying EFB to the TA-800, the goals include:

1. Reducing Distortion: With single channel operation, this averaged .75% at 1 kHz at 25 watts RMS power output. However, at 20 watts RMS in both channels with both channels driven (maximum power output in this condition), distortion more than doubles in both channels to nearly 1.8%. With EFB, it is expected that distortion at full power output in both channels will remain well under 1%, and this at greater power output in both channels than the stock design could deliver.

2. Maximum Power Output: Fisher used a rather large dropping resistor for the screen grid supply in the TA-800 -- great for power supply filtering, but lousy in terms of operating the tubes most efficiently. As a result, power output is limited much more in the TA-800 than it is in the later receivers -- especially when both channels are driven. By installing EFB, this problem will inherently be eliminated. Power output will still drop from a single channel to a both channels driven condition (due to power supply droop), but overall power output should still show a very worthwhile increase over stock levels. It is estimated that power output in the modified TA will approach 30 watts RMS in both channels, with both channels driven.

3. Operating Temperature: Vacuum tubes get hot when operating anywhere near properly. It is a basic function of how they work. And placing four output tubes in such close proximity of themselves to each other magnifies the temperatures created exponentially. EFB will hardly reduce temperatures to the point that the tubes are only warm to the touch, or that a fan is not mandatory for such a layout if long term dependability is a goal. It will be. But being that everything is relative, the tubes themselves will still operate at significantly lower internal temps with EFB than in that of the stock design -- fan or no fan. This is because less output tube quiescent current is required to a achieve low distortion operation. And, since the output tubes pull less current, the power supply components all operate cooler as well. So it is a cumulative effect.

4. Output Tube Life: In the stock design of the TA-800 each 7591 output tube operates with a quiescent cathode current of 40 mA. It is anticipated that with EFB, this will be able to be reduced to < 30mA for the same effective distortion performance as was delivered in the stock design with single channel operation. And of course, because of EFB action, at that point, distortion would change only insignificantly when both channels are driven. If quiescent current is then increased towards the original operating point from there, then distortion -- with both channels driven -- actually falls below single channel distortion performance of the stock design with both channels driven. Therefore, with EFB, a range of acceptable quiescent current is created, that can be chosen based on the goals sought: maximum performance, or maximum tube life, or somewhere in between. Experience has shown that an excellent compromise between these two qualities is usually found at about 75% of the original stock current draw level. When the reduced current draw level is coupled with the reduced internal tube temps that produces, tube life is then extended by a factor at least equal to the reduction in quiescent current.

So these will be the collective goals sought with the installation of EFB. We'll find out soon enough next week.......

Dave

 

[dcgillespie]

The last leg of the TA project is going slow -- simply because there is no room inside the TA's chassis with which to add the EFB regulators. The physical installation has been quite a challenge. After a number of different trials that I didn't like, I adopted the time honored approach used whenever you need to add on, but are out of real estate: build up. So, once that was decided on, construction of the EFB circuits on a board cut specifically to fit the available space began. With that work finished, I how have to go back to the TA chassis and remove the last of the last tech's work (DC Heater/Bias supply and dropping resistor), as a first step in installing the new board. Since the EFB circuits will now ultimately generate the bias voltage for the output tubes, the original design of the DC Heater/Bias supply will be reconfigured to function as a DC Heater supply only. To help maximize the capability of the new EFB bias supply, the filter components of the DC Heater supply will be electrically moved to appear in the negative side of that supply, allowing the positive terminal of the DC Heater bridge rectifier to be connected directly to ground now. That helps the EFB circuits because the unique voltage doubler circuit developed to supply output tube bias through the EFB control grid regulator actually uses two of the four diodes (the two now grounded) within the DC Heater supply bridge with which to function. So, a few pics to show current progress:

Below: The existing DC Heater/Bias supply components first need to be removed, so the area can be prepped for installation of the EFB board on top of what will be the new DC Heater supply only. I had already removed the old can cap mounting clip in this shot, that was still in the unit when received.

Below: That was some kind of super epoxy glue the previous tech used, which was just nuts. Besides destroying the old components to get them out (of no matter), it left a layer of the stuff that then had to be cut out with a Dremel before proceeding. New components will be installed with silicone adhesive, which makes for a very durable mounting, but the components can easily be removed if need be. The usable area is not as big as it appears!

Below: The new EFB Board will take up every bit of the usable space.......

Below: Meanwhile, the bone pile for this project continues to grow........

Below: Coming up: With the wrap up of the TA getting close at hand, Rob was able to acquire a nice MPX-100 to go along with his TA, that was nicely, and nearly 100% recapped by the seller (all except the can cap), but acquired with the understanding that separation between the channels was still minimal. In deed it is! This is a shot of the output of that unit connected to my 202T tuner/preamp receiving a Left Channel only 1 kHz signal transmitted from within my lab. The amplitude of the right Channel is nearly as great as that of the left channel. Adjustment of the rear panel Separation Control affected the amplitude of both channels more than it affected the separation. That unit has already been troubleshot to determine the cause -- a problem that likely affects many such units out there -- and so will be expounded on in a quick mini thread once the TA is wrapped up. Family is over for the 4th next month, so I hope for the whole shootin' match (TA and companion MPX) to get a thorough workout that day.........


Dave

 

[AlTinkster92]

A 202T ya say? Pics or it didn't happen

 

[dcgillespie]

Now now, calm down. It happened, but one thing at a time.......sheesh!

 

[larryderouin]

That's tellin' him, Dave!

 

[AlTinkster92]

Dave knows I'm just messin with him... I know all about it...

 

[larryderouin]

I figured as much, how many times have you visited???

 

[dcgillespie]

So almost 6 weeks to the day from when this unit was dropped off, and I can finally all but put a fork in the TA-800 now. Once a plan for installation was finally hatched, installation of the EFB circuitry came off without a hitch being finished up late last night, so that all that remains now is to burn in the EFB circuits for 24 hours, and then document the performance of the final completed unit. So today, I've been enjoying a day of extended listening to the finished unit since early this morning, and plan to dive into the MPX-100 this evening. Then, assuming all goes well today, performance testing of the TA can start tomorrow. If you look back over the entirety of this thread, there is virtually no area of this unit that has not received some level of attention during the course of this project, be it due to either needing a basic or more complex component repair, being part of the routine restoration process, or being part of a complete section makeover for improved performance and more dependable operation. From the as received condition then, in no particular order, the highlights of the project include:

1. Repair dead AM Tuner (open antenna coil, dead tubes).
2. Repair dead FM Tuner (open 1st IF stage screen resistor, defective limiter IF transformer, align IF strip only).
3. Repair sloppy volume control assembly (result of a previous service attempt). Remove, repair, reinstall.
4. Repair broken wiring in an around the volume, loudness, and balance controls.
5. Repair (existing) broken AC power switch.
6. Repair dead Channel A (shorted C114).
7. Replace all PS can caps.
8. Replace HV rectifiers.
9. Replace Selenium Bridge Rectifier with Silicon and adjust for higher silicon output.
10. Replace all mylar caps.
11. Replace all low voltage electrolytic caps.
12. Replace any out of spec resistors as needed (8 total).
13. Add CL-80 current limiter
14. Add individual precision 10Ω cathode resistors.
15. Add individual 100Ω Screen Stability resistors.
16. Add individual output tube bias controls.
17. Balance on-center response of tone controls between both channels.
18. Balance overall 20 Hz to 20 kHz response between channels.
19. Trim balance control for balanced on-center presentation.
19. Performance test all tubes; replace 9 total: 2 AM, 2 FM, 2 Drivers, 2 Output (ordered to match existing 2 new output), 1 Eye.
20. Replace blue LED dial glass lamps with white daylight LED lamps.
21. Clean dial glass and replace glass seal.
22. Clean all switches and controls as appropriate.
23. Develop and install new 7247 based driver tube circuit to replace problematic 7199 tube of stock design.
24. Modify 1st stage tone control amplifier in each channel for more gain to maintain stock sensitivity specification with new driver circuit.
25. Develop and install EFB circuits to control output stage performance.
26. Tightened knobs (falling off).

This was surely one sick puppy when it arrived, in need of some serious help -- but it's in recovery now, and after tomorrow, should be ready to take on all comers! A few more quick pics, and then its onto the MPX-100 to start putting a wrap on Rob's Fisher gear.

Dave

Below: As received:

Below: As finished:

Below: The EFB Screen Grid Regulator Mosfet pass element is bolted to the chassis in a convenient unused hole residing under the volume control assembly:

Below: Time lapse photography in dim conditions with the output tubes now operating under the control of EFB shows no hint of any plate color. Optimum bias setting will be determined during performance verification tests tomorrow. While there is little doubt that the tubes will ultimately end up operating much cooler than with the stock design (~375F avg in open air 76F environment), such insanely close mounting of these tubes mandates a small fan to augment the natural convection cooling for any kind of extended operation:

Below: All set up and running for about 7 hours at this point into an anticipated 16 hour run, with a small fan set nearby to assist the output tubes in heat removal for the extended operating period. It doesn't take much. You can easily lay your hand on the top of all four tubes. After nearly 7 hours of operation, the power transformer (in no direct airflow) reached (an apparently) stabilized 125F:

 

[larryderouin]

Dave: Outstanding job on fixing a Blivet (a blivet is a 5lb sac carrying 10lbs of crap!)
Looks like you did everything but polish the bottom.. Nice choice on the TUNG-SOL 7591 Re-ISSUES.