More Fun With Magnavox: The 9300 Series

[eickmewg]

And you're about $200 richer! I like the stock opts just fine. I don't think my old ears would hear any difference.
And another shout out to Dave for helping me track down some ground loop/hum issues in my modified 9302 and for all his hard work on the 93XX series.

Bill

The replacement OPTs are markedly better- but, there's nothing fundamentally wrong with leaving the originals.

 

[gadget73]

Basically marry the 6EU7 buffer stage to the early schematic if you want to run stock iron.

For a very rudely hacked together schematic showing just that, I present to you an elegant schematic poorly spliced by an idiot with an image mangler program

 

[dcgillespie]

Gadget, you're a hoot!! But thanks! A couple of points if I may however with the spliced together schematic:

1. The series grid resistance into the AF Amplifier portion of the Driver Stage must remain as 47K as in the original modified design. Its value in part helps to produce the desired HF roll off to achieve the stability level indicated.

2. It could be ambiguous seeing references to both a 320 volt source, and a 340 volt source. In fact, these are one and the same.

For those wishing to keep the original OPTs in place, adding the buffer stage would still be a very worthwhile upgrade.

Thanks again!

Dave

 

[Morrism]

I used my own image mangler to edit Gadget's schematic.
I made the changes Dave suggested.
I also added a volume control. Is it in the correct place?

 

[eickmewg]

I've really been enjoying my 9302. It has been a real learning experience. I've polished the apple with several minor modifications. I did the Dynaco-inspired "yellow sheet" diode modification for the GZ34 rectifier and I've added an NTC CL 80 thermistor for in-rush current suppression. After initially having noticeable hum on the right channel, careful attention to grounding and wire dress, and it is now virtually silent. I think those EL84/6BQ5's are really very musical tubes. Only issue now is that voltages are running a tad on the low side.

 

[dcgillespie]

Thank-you Morris. To show the volume control properly, the .01 uF input cap would go to the wiper of the volume control, the low side (min volume) terminal of the control would be grounded, and the high side (max volume) terminal would go to the center pin of the input jack. The two 470K resistors would go away.

If you look at the input circuit as originally drawn, that is the way the two 470K resistors are connected -- with their combined total representing (basically) a 1M pot placed across the input jack, with the .01 uF connected to their junction, which would represent an equivalent level control with its wiper set to mid position if it were a linear control. With the addition of the buffer stage, now any type of attenuation scheme (fixed, variable, stepped, etc.) can be used at its input, and at any setting, without any degradation of audio quality at mid level settings.

Dave

 

[Morrism]

Thank-you Morris. To show the volume control properly, the .01 uF input cap would go to the wiper of the volume control, the low side (min volume) terminal of the control would be grounded, and the high side (max volume) terminal would go to the center pin of the input jack. The two 470K resistors would go away.
Dave

If I'm following this correctly, there is zero resistance between input and .01uF cap with the volume pot at max?

 

[dcgillespie]

Correct. In that scenario, the input cap would be effectively going directly to the center pin of the input jack, and there would also be (effectively) a 1Meg resistor from the input jack center pin to ground.

Dave

 

[kirk57]

I can either
a) not do these mods because I'm not sure about some stuff or
b) ask really basic questions.

So here goes:

What are the minimum power ratings for the resistors and minimum voltage ratings for the caps in the new circuits (buffer and EFB) where those ratings are not specified? I do see the 330k next to the bias pot is 1W, and the 4.7 uF cap is 35v etc. Is 1/2 watt OK for the other resistors?

Also, the one leg of the voltage regulator is indicated as going 'to other channel' which seems kind of vague, to me anyway. I assume the other two cathodes (pin 9)? Obviously there is another 10 ohm resistor over there somewhere because that's how you set the bias... Would anybody be so kind as to draw that out?

 

[gadget73]

Standard resistor wattage is 1/2 watt on most tube amps. I built mine with that and no smoke came out.
Caps where not specified should be at least 500v. I usually use 630v film caps because its a common value now

The schematic drawn is 1/2 of the amplifier. For the second channel, you duplicate it. The "to other channel" goes to the same pin on the other side. Cathode is pin 3 on an EL84, screen is pin 9. Do not mix those up, ugly stuff will happen.

Basically the 10 ohm resistor goes between the output of the EFB circuit and pin 3 of the output tubes. On the later one, its one per tube, the earlier schematic shows one per pair. You can do either. One per tube tells you what each is doing. I actually need to do that very shortly, once i rustle up a few more small machine screws. My "matched quad" is showing some 50ma on one side and about 32 ma on the other. Some ting wong.

 

[Morrism]

Here is a schematic I made based on the Sam's photofact. It may help clear some of your questions kirk57, as there are both channels and PS.
I just hope there are not any errors.

AK has compressed the image so much it is almost useless. Is there no way to post a full res picture?

 

[Morrism]

Here is a link to a better quality image.
https://www.flickr.com/photos/96793522@N06/31788330814/in/dateposted-public/

I updated the file with Dave's recommendations below.

 

[dcgillespie]

Thanks so much for your work to update the original Sams Magnavox schematic! Your work is so good, that other than the actual (and obvious) circuit changes, it would be hard to know that the schematic has been altered at all. The original drafting style has been maintained to a very high order, which is just neat! To answer the question mark applied to R24, it does in fact remain a 10K resistor. The only other thing that might be useful would be to identify the OPTs as the Dynaco clone Z-565 (48). The "48" designation if given by the manufacturer, and identifies the transformer as the clone version which has an 8 Ohm secondary tapped at 4 Ohms, versus the original device, which had a 16 Ohm secondary tapped at 8 Ohms.

Thanks again!

Dave

 

[GordonW]

Here is a schematic I made based on the Sam's photofact. It may help clear some of your questions kirk57, as there are both channels and PS.
I just hope there are not any errors.
View attachment 860528
AK has compressed the image so much it is almost useless. Is there no way to post a full res picture?

If you convert it to a PDF file (you can do that through OpenOffice or many other programs), I think you should be able to post it in higher resolution.

Regards,
Gordon.

 

[kirk57]

Here is a link to a better quality image.
https://www.flickr.com/photos/96793522@N06/32627249135/in/dateposted-public/

Thanks for posting this. Question: I assume that the change in the feedback circuit from this schematic and the one you posted earlier (post #411) is due to the different OPTs. But why the additional 1k resistors between pin 2 of the EL84 and the .047 cap? I see gadget also has that 1k resistor when using the Edcor OPTs (different thread)

Do I want that if I'm still using the original Magnavox OPTs (for now anyway)?

 

[Morrism]

If you convert it to a PDF file (you can do that through OpenOffice or many other programs), I think you should be able to post it in higher resolution.

Regards,
Gordon.

Thanks Gordon, I'll look for a .pdf converter.

 

[Morrism]

Thanks for posting this. Question: I assume that the change in the feedback circuit from this schematic and the one you posted earlier (post #411) is due to the different OPTs. But why the additional 1k resistors between pin 2 of the EL84 and the .047 cap? I see gadget also has that 1k resistor when using the Edcor OPTs (different thread)

Do I want that if I'm still using the original Magnavox OPTs (for now anyway)?

Hopefully Dave, Gadget, or someone else with tube smarts can answer this.

 

[GordonW]

Thanks for posting this. Question: I assume that the change in the feedback circuit from this schematic and the one you posted earlier (post #411) is due to the different OPTs. But why the additional 1k resistors between pin 2 of the EL84 and the .047 cap? I see gadget also has that 1k resistor when using the Edcor OPTs (different thread)

Do I want that if I'm still using the original Magnavox OPTs (for now anyway)?

Those are grid stoppers- they help prevent oscillation. Yes, I'd leave those in there, regardless.

Regards,
Gordon.

 

[kirk57]

Why is the value of capacitance in the feedback loop so very much different when using the Edcor or Dynaco-type OPTs as opposed to the Magnavox OPT?

For the former it's 680 pF and for the latter 47uF/47000000 pF?

 

[thorpej]

Why is the value of capacitance in the feedback loop so very much different when using the Edcor or Dynaco-type OPTs as opposed to the Magnavox OPT?

For the former it's 680 pF and for the latter 47uF/47000000 pF?

In Dave's Nov 2015 schematic, it's 47pF, not 47uF.

They're different (and connected differently) because the transformers behave different high-frequency performance characteristics.