Comments for page: A New Look

<<first - <previous - 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - next> - last>>

Posted by Dave October 12, 2016 - 01:27 pm
Hi Bill -- Of the caps you are considering, the 5 uF @ 35 vdc is the most ideal -- but all of them should work fine.

The capacitor's function is to help maintain stable operation of the regulator -- just as virtually all 3 terminal regulators require in most applications. It needs to be big enough to accomplish that task, without being so big as to limit response time to changes in B+ voltage. It would be very hard to distinguish any performance difference between the specified value, and that of any of your caps.

One other point to consider however is the rated working voltage of the cap you use. The cap normally sees about 15 vdc across it, making a cap rated for 25 volts ideal, and one rated for 35 volts perfectly acceptable. But as you go to caps rated for increasingly higher operating voltages, the capacitor -- over time -- won't hold its form as well as one that is more appropriately rated for the voltage it actually sees in service. Granted, it can take years for issues with form to occur -- but they will occur. Just something to consider.


Posted by Bill October 12, 2016 - 09:59 am
Hi Dave, one more thing: one of the caps in the EFBcircuit calls for a 4.7uF rating. I have 3, one measuring 5.0 uF @ 35 VDC; another that is 5.0 uF @ 50 VDC; the third that measures 5.9 uF @ 70 VDC. I like the build quality of the 5.9 uF cap but wonder if it's too far out of range?

Posted by Bill October 02, 2016 - 10:38 am
Thanks Dave

Posted by Dave September 30, 2016 - 06:42 pm
Hi Bill -- You can increase the value of the resistors if you like, but moving in the direction of increased resistance only serves to reduce power output, and work to reduce the effects of the Ultra Linear connection. Installation of Screen Stability resistors are by all accounts sonicly neutral, and once their value is large enough to damp out both large and small scale parasitic oscillations, then increasing their value beyond that point serves little purpose other than adding some insurance. Typically, tubes of the 6BQ5 class will remain stable with a Screen Stability resistor value of as little as 47 Ohms in most applications, so there is already some insurance built into the 100 Ohm value specified.

If you do increase the value of these resistors, I would not increase them beyond a value of 330 Ohms. Within the 6BQ5 family, EL84 tubes are true pentode tubes, and therefore draw more screen current than their beam power brethren. Raising the value beyond the limit suggested will only serve to limit amplifier performance when these tube types are used, while adding little in the way of insurance for the tubes.

Good luck with your project!


Posted by Bill September 30, 2016 - 03:22 pm
How far can one deviate from the 100R value for the screen stability resistors in this circuit. I have read where it "might" be better to go to a higher value, like 220R or so.

Posted by Dave July 12, 2016 - 07:54 pm
Hi Allan -- We need a lot more information to diagnose your problem, that would include:

1. Is the motorboating in both channels?

2. Is it present on all inputs?

3. Is it present at all volume control settings?

4. Do the tone controls change the nature of the motorboating?

5. Have you always used electrostatic speakers -- or did you just recently change to this type of speaker? If so, did the motorboating start when using electrostatic speakers?

6. Is your amplifier an SCA-35 or ST-35?

7. Are your electrostatic speakers (likely) internally transformer coupled with their own HV power supply?

The answers to these questions can shed significant light as to just what -- or what not -- the problem might be. However, if the output tubes are biasing correctly, this would indicate that the EFB(tm) portion of the amplifier is operating correctly. Any more detailed suggestions at this point however would represent no better than stabbing at the wind. With more information, more precise suggestions can be made.


<<first - <previous - 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - next> - last>>