18-watt power reduction method comparisons?

[PharmRock]

The next project is converting my Marshall 1974x clone to an EF86 variant, and was researching power and/or voltage reduction methods.
As usual, Sluckey has some ideas for this on his website (see page 6):
http://sluckeyamps.com/misc/Amp_Scrapbook.pdf

I was looking at the 6-position switch for changing the LTP ("Power dampening switch").  Instead of a switch, could a 50K pot be placed in series with the normal 820R resistor to give a continuously variable power control?

On that same page is a pretty cool method for reducing power via the EL84 bypass caps (Geezer's Bypass Cap Attenuator).  I was able to find the original thread on this forum and also read about 9 pages of it over on 18watt.  Looks pretty cool.

Was wondering if anyone has compared both of the above methods in an 18-watt style of amp, and which one you prefer.  I've built a couple of 18-watt clones and used Dana Hall's VVR for B+ scaling (I still have a couple of these).  While the VVR is pretty good, I've found that the amp loses a lot of its "sparkle" when dialed down at lower B+ settings. 

If there's anything you can share about the above methods, especially if you've compared one to the other, I sure would appreciate it!

[kagliostro]

Instead of a switch, could a 50K pot be placed in series with the normal 820R resistor to give a continuously variable power control?

Give a look to the Egnater Rebel 20 PI circuit (may be also to the PPIMV preceding the Power Tubes)





Franco

[Bieworm]

If you have a HT center tapped power transformer you could use half of the HT winding to reduce your B+ voltage. You'd have to use a FWB rectifier then. And do the switching between 0V and 290V _ 0V and 145V before the rectifier. Or am I taking nonsense?

[PharmRock]

If you have a HT center tapped power transformer you could use half of the HT winding to reduce your B+ voltage. You'd have to use a FWB rectifier then. And do the switching between 0V and 290V _ 0V and 145V before the rectifier. Or am I taking nonsense?

I'm not quite sure I follow this, but I'm still way down in the shallow end of the swimming pool, so to speak.  So would I switch one leg of the PT secondary prior to the FWB? 

In any case, it's probably best I stick with the EZ81 rectifier for this circuit, so I'm not sure if this is way to go with the power reduction.  But I like the idea as it seems it only requires a switch and no other parts/components. 

[PharmRock]

Give a look to the Egnater Rebel 20 PI circuit (may be also to the PPIMV preceding the Power Tubes)

Franco

Thanks Franco,
It looks like that's exactly what is happening with VR105 (25KB pot) labeled as the WATTS control. 

[sluckey]

I was looking at the 6-position switch for changing the LTP ("Power dampening switch").  Instead of a switch, could a 50K pot be placed in series with the normal 820R resistor to give a continuously variable power control?

You can do that. I predict you will not like power dampening.

[Bieworm]

If you have a HT center tapped power transformer you could use half of the HT winding to reduce your B+ voltage. You'd have to use a FWB rectifier then. And do the switching between 0V and 290V _ 0V and 145V before the rectifier. Or am I taking nonsense?

I'm not quite sure I follow this, but I'm still way down in the shallow end of the swimming pool, so to speak.  So would I switch one leg of the PT secondary prior to the FWB? 

In any case, it's probably best I stick with the EZ81 rectifier for this circuit, so I'm not sure if this is way to go with the power reduction.  But I like the idea as it seems it only requires a switch and no other parts/components.

See the schematic is this topic. This is something similar. But you need the top part (power tubes + PI  marked)
I did a 3 way switching,  but you need 2.https://el34world.com/Forum/index.php?topic=28495.0

[Bieworm]

hi

Here's a schematic of what I mean. By the way, an EZ81 is going to do Jack SH*T on dropping voltages. If you need sag you can put a 100R 25W resistor in series after the rectifier...
On 18 watt amps I never add the EZ81. A good SS rectifier with a 100R or 150R 25 watt resistor in series gives a litte sag and is more reliable than te EZ81. If you use JJ you're in for trouble for sure.. I have had a lot of trouble with those, so I stopped using them in general.
The FWB rectifier is a lot cheaper and easiepeezie to install. Make sure to use a good DPDT switch though.. not the flimsy mini switches. A good Carling is what I'd use

This switching method gives you a nice EL84 overdrive at a significant lower volume. I'm surprised I haven't come across this in other threads...

[pdf64]

… Here's a schematic of what I mean. …

A 290-0-290V winding into a full wave bridge rectifier and capacitor input smoothing will allow produce a DC output of 820VDC.
Maybe you meant to show a 145-0-145V (or 0-145-290V) winding?

[Bieworm]

… Here's a schematic of what I mean. …

A 290-0-290V winding into a full wave bridge rectifier and capacitor input smoothing will allow produce a DC output of 820VDC.
Maybe you meant to show a 145-0-145V (or 0-145-290V) winding?

Right! Nice catch Pete!! Will adjust the schematic 👍

[uki]

A 290-0-290V winding into a full wave bridge rectifier and capacitor input smoothing will allow produce a DC output of 820VDC.

How to achieve it ? Please educate me, show some schematics please.  thanks.

Sorryyyy 

[PharmRock]

Thanks for the info guys.  My PT is a Mercury Magnetics P18RS, 316-0-316, so if I understand correctly, if I try to use the approach that Bieworm is suggesting where the center tap functions as a "half-way voltage tap" asa non-CT transformer, then what I would actually have is 0-316-632?  That doesn't sound like it would work. 

Regarding the EZ81...thanks for the info on that.  I've done two 18-watt variants, both with EZ81's.  I've never had issues with them, but I also don't have any experience with the SS approach w/series resistor so I can't really compare.  Might be worth trying out.  I agree re: JJ tubes....I've had 2 bad 6V6s out of the box, and I believe one 12AX7 that was extremely microphonic. Its not like I buy a whole lot of tubes so from a % standpoint, its not looking good.  I usually opt for another manufacturer/brand these days.

I'm thinking the approach that Geezer came up with (and several guys have also confirmed over on 18watt) where the EL84 cathodes are split and the caps either grounded or connected via a 2K pot might be the way to go.  It would only require $10 in parts...one extra cathode resistor, cap, a pot, and a DPDT switch.  And would be a simpler implementation than the power dampening approach on the LTP, which Sluckey mentioned I might not care for too much.

Still in the planning stage and doing the board lay out.  I appreciate everyone's comments...keep them coming!

[sluckey]

A 290-0-290V winding into a full wave bridge rectifier and capacitor input smoothing will allow produce a DC output of 820VDC.

How to achieve it ? Please educate me, show some schematics please.  thanks.

Sorryyyy 

Don't need a schematic. Just math. 290-0-290 is same as 580 w/CT. 580 x 1.414 = 820V unloaded.

[Bieworm]

Your 316-0-316 transformer would act like a 632-316-0 transformer.  It's just a coil with 2 ends and a tap in the middle..
I like SS rectifiers more in 18W amps.. cheaper, easier, and you can get sag with a series resistor...

[HotBluePlates]

... I was looking at the 6-position switch for changing the LTP ("Power dampening switch").  Instead of a switch, could a 50K pot be placed in series with the normal 820R resistor to give a continuously variable power control? ...

You can do that. I predict you will not like power dampening.

On another forum, a group of folks rave about 3rd Power's Hybrid Master. 

The Hybrid Master Patent appeared to show a version of the "power dampening" idea added to a standard pre-phase inverter master volume.

[PharmRock]

Thanks for posting the link to that, HBP.  Pretty cool.   It appears they are using a pot vs fixed resistors on the PI.  As far as I can tell, the pot value is not given though....50KL maybe?
When I get to that point in my build, I could simply wire up a switch with an 820R resistor on one side ("stock") and a higher value on the other just to test out the concept and go from there.

[SILVERGUN]

I read through the patent pretty quickly and it sounds like they are using a dual gang 50KL pot to simultaneously raise the value of the PI cathode resistor and act as a pre-PI Master volume.
If the drawing is accurate then the 3.3K in parallel with the 50K pot results in a max setting of 3K in series with the existing 1.5K resistor

[HotBluePlates]

...   It appears they are using a pot vs fixed resistors on the PI.  ...

The picture in Sluckey's scrapbook is a diagram someone else drew up IIRC here a number of years ago.  They drew a switch & discrete resistors to change the value of the bias resistor in the long-tail inverter.

3rd Power's patent shows two resistors in series making up the "bias resistor" of the long tail.  A pot is placed in parallel with one to change its effective resistance in-circuit; cheaper & easier than using a switch.  The other, fixed, resistor is to set a limit on how low the "bias resistor" value can go.

I read through the patent pretty quickly and it sounds like they are using a dual gang 50KL pot to simultaneously raise the value of the PI cathode resistor and act as a pre-PI Master volume.
If the drawing is accurate then the 3.3K in parallel with the 50K pot results in a max setting of 3K in series with the existing 1.5K resistor

When I first saw the patent I thought, "Oh shoot!  They ripped off the idea developed on the Hoffman Forum."

Then I realized maybe they tried it and found it worked for causing early distortion, but only sounded good within some limits.  And within those limits, it wasn't reducing power enough.  Enter the idea of mixing a regular master volume control before, and a limited-range variable bias resistor for the long-tail.

[PharmRock]

I was looking at the 6-position switch for changing the LTP ("Power dampening switch").  Instead of a switch, could a 50K pot be placed in series with the normal 820R resistor to give a continuously variable power control?

So my suggestion of putting a pot in series with the "bias" resistor off the LTP isn't such a novel idea after all. 

Based on that patented approach of scaling the PI power as well as using a pre-PIMV presumably via a ganged pot...since there is a 3.3K in parallel with the pot (and in series with the 1.5K), then the most "range" we could get (say we used a ganged 1M pot vs 50K) would be 1.5K up to just shy of 3.3K.  Compared the 6-position version in which the highest "bias" resistor is 56K, it seems like using a paralleled 3.3K with any value pot would not provide a decent range.

Anyways....I may give this a shot. Ideally, I just need to get my hands on a ganged pot and wire it up as a pre-PIMV along with the LTP dampening, and perhaps manipulate the value of the paralleled (e.g. 3.3K) resistor to see where the sweet spot is.   Am I on the right track?  If I can't find a ganged pot I could just wire 2 pots up of equal value and try that.

[HotBluePlates]

...  Compared the 6-position version in which the highest "bias" resistor is 56K, it seems like using a paralleled 3.3K with any value pot would not provide a decent range. ...

Except users report good function with the range of 1.5kΩ to 3.3kΩ.  And re-read what Sluckey said before:

... I predict you will not like power dampening.

[PharmRock]

I've done a little more digging...it seems like Mojave and Budda amps used this approach in some of their amps but I couldn't find the schematics.  As you mentioned, this went way back on this forum, but I couldn't find that either.  I did find some discussions on 18watt and trinity amp forums going back to 2007 on this (they also reference the Hoffman forum):
https://www.trinityamps.com/phpbb/viewtopic.php?t=827
They mentioned high resistance values "starved" the PI a little too much.

This may be an easy experiment to try out when I get to that point. 
Appreciate all the comments/advice on this.  Thanks!

[sluckey]

Only takes 5 minutes. What are you waiting on? I only did a switch because member Plexi50 wanted a switch. I suggest using a 50K pot in series with an 820Ω resistor. Just use gator clip leads to test. No need to be fancy.

[PharmRock]

Will do.  I started the build tonight.  That approach sounds easy enough.