6V6 20 watt OT

[Kaaskop]

I have a 6200 Pri ohms/8ohms Sec rated for EL84’s at 15 watts
I want to use it for a Push/Pull 6V6 build at 20 watts, will it work?

https://www.hammfg.com/files/parts/pdf/1750Y.pdf

Thanks for any input.

[sluckey]

yes

[Kaaskop]

What I'm trying to figure out is what would give me more 2nd or 3rd order distortion.
The tube data sheet says a push/pull's effective load resistance plate to plate is between 10000 < > 8000. So  which direction would give the desired effect if my pate voyage is around 300V? Is there any chance of damaging the OT is it's miss matched?

[HotBluePlates]

... I want to use it for a Push/Pull ...

What I'm trying to figure out is what would give me more 2nd or 3rd order distortion. ...

Push-pull cancels even (2nd) harmonic distortion created at the output tubes.

Hi-fi designers would skew towards "more 2nd harmonic" knowing that this is then canceled, leaving a cleaner output stage.

I have a 6200 Pri ohms/8ohms Sec rated for ... 15 watts
I want to use it for a Push/Pull 6V6 build at 20 watts, will it work?

As Sluckey says, it will "work" and the amp is guaranteed to make sound.

As for "20 watts" Load Impedance is only 25% of what you must know.  You also need to know about the power transformer, and the output voltage & current it can supply (50% of what must be known).

Then you look at the Output Tube to determine if it can pull enough current through the OT at the available voltage to result in 20w of output.

Regardless, "20 watts" is only 1.25dB SPL more at your speaker than "15 watts."  You won't be able to tell the difference, and a different speaker will likely deliver much more loudness-increase (if that's what you want).

[tubeswell]

What the others said, plus, if you make a 2 x 6V6 amp that is a Deluxe Reverb, it will be capable of pushing out more power than a 2 x 6V6 amp which is a Princeton reverb (which is the corollary of what HPB is saying). And the '15W' rating might be a bit conservative (depending on the OT), and the OT might be good for 18W, or 22W - depending on how much you push it. You could always put a 125mA HT fuse in

[Kaaskop]

 This is good information (thanks), now let me see if I can wrap my head around it. 

You also need to know about the power transformer, and the output voltage & current it can supply

I take this to mean it depends on the transformer supplying plate voltage of the 6v6's and the current  set by the bias, which will effect the plate dissipation for the output (in watts).

if you make a 2 x 6V6 amp that is a Deluxe Reverb, it will be capable of pushing out more power than a 2 x 6V6 amp which is a Princeton reverb

If we take the Princeton Reverb at 15 watts (OT Pri 8500) and the Deluxe Reverb at 20 watts (OT Pri 6600) the difference seems to be the OT Primary and how hard the circuit pushed the tubes, since both use 2x6V6.

The 15 watt OT I was referring to was for a Vox AC 15 / EL84 (Pri 6200) so I'm thinking its closer to the Deluxe Reverb at 20 watts (Pri 6600). The turns ratio is basically the same but the Vox 15 watt has more iron (better?).

All with 8 ohms taps
Deluxe Reverb at 20 watts (OT Pri 6600) 28.72:1
Vox AC 15 / EL84 (Pri 6200) 27.87:1

Princeton Reverb at 15 watts (OT Pri 8500) 32.6:1

[tubeswell]

  the difference seems to be the OT Primary and how hard the circuit pushed the tubes, since both use 2x6V6.

Yes, its the whole circuit, number of gain stages and gain at each etc (as well as that a standard DR PT is 6k6 rated for about 20W and a standard PR PT is 8k rated for 12-15W, so there's also a bit more iron in the DR, which carries more LF without saturation)

[Kaaskop]

This is illuminating. I had not considered the PT relationship to output but it now makes sense since the PT secondary goes straight to the OT after the rectifier.

[tubeswell]

This is illuminating. I had not considered the PT relationship to output but it now makes sense since the PT secondary goes straight to the OT after the rectifier.

Most of the total current that is present at the B+ reservoir cap is being drawn from the PT through the OT primary (via the output tubes). The remainder of the current that supplies the screens and pre-amp tubes is a relatively small proportion of the B+ current.

[Kaaskop]

I have read that Valve Wizard article a few times and in past build have observed and applied that path of the HT, now in this discussion however there was a shift from knowing to understand ... the penny dropped, thankfully.

[HotBluePlates]

...  You also need to know about the power transformer, and the output voltage & current it can supply ...

I take this to mean it depends on the transformer supplying plate voltage of the 6v6's and the current  set by the bias, which will effect the plate dissipation for the output (in watts).

I do not mean anything to do with bias.  An idling tube makes 0 watts of audio output power.  When everything else is done with respect to designing an output stage, the bias is determined in order to keep the output tubes from overheating & melting when making their maximum output.


How much power transformer voltage?  That tells us how much rectified DC Volts is possible.

     A Deluxe Reverb has a roughly 400vdc power supply.  It will not make 20-22 watts with only 200vdc (using the same OT primary impedance).  "400vdc" requires at least 400 / 1.414 = 283vac from the high voltage winding, and more if we drop volts across a tube rectifier.

     The tube needs to have some voltage left at peak output, so something less than full supply voltage is usable as AC Volts output at the 6V6 plate.  We might estimate that we need to reserve 75-100v, leaving 300-325v available for peak output voltage.

     If we're targeting the Deluxe Reverb, we might already know it uses a 6.6kΩ OT.  For ClassAB, we calculate max output power in terms of one side of the push-pull stage:  a single 6V6 drives 1/4 the plate-to-plate primary impedance at the moment of peak output.  6.6kΩ / 4 = 1650Ω

     Let's say we expect 25w RMS:  Volts = √(Power x Impedance) = √(25w x 1650Ω) = 203v RMS
     Volts Peak = Volts RMS x 1.414 ----> 203v RMS x 1.414 = 287v Peak

     A 400vdc supply leaving 80v across the 6V6 is still capable of sustaining a 287v Peak plate output from the 6V6, even while sagging over 30v.

     How much current?  287v Peak / 1650Ω = 174mA Peak.  RMS Current = Peak Current / 1.414 = 174mA Peak / 1.414 = 123 mA RMS.  (Check: 203v RMS / 1650Ω = 123mA RMS). This is just the output tube plate current, so the PT high voltage winding needs to include extra current for the screens and preamp.  Another 15mA?

     Time to check:  (Audio) Power (Output) = Volts RMS x Current RMS ---> 203v RMS x 123mA RMS = 24.97 watts.
     
     We look at Hammond's replacement Deluxe Reverb PT and find it is rated for 330vac (above the required 283vac & allowing for a rectifier tube), and 138mA (providing the required 123mA plus extra for screens & preamp).

     In all this, we haven't considered the tube:  is screen voltage high enough to support our required peak plate current?  Looking at the top of Page 5 of the 6V6GTA data sheet, a screen voltage of 250v supports a peak plate current of ~103mA at about 100v on the plate (400vdc - 287v Peak).  We can see from the other curves that we can expect more than 20mA additional peak plate current if the screen voltage is 300v or higher.

     The above is "considering the PT voltage & current, and the tube's capabilities, along with the OT primary impedance to determine power output capability."  As you see, idle bias wasn't discussed because it doesn't tell us much that directly pertains to power output.


We haven't attempted to figure out "minimum required volts" or "minimum required milliamperes" for the goal of 20 watts.  But does the PT you're planning on using supply 300vac (or more) and 135mA (or more)?

[HotBluePlates]

If we take the Princeton Reverb at 15 watts (OT Pri 8500) and the Deluxe Reverb at 20 watts (OT Pri 6600) the difference seems to be the OT Primary and how hard the circuit pushed the tubes, since both use 2x6V6.

Follow the process above & use your numbers:
     Princeton Reverb requires 178.5v RMS (252v Peak) and also ~84mA RMS.
     Deluxe Reverb requires 181.7v RMS (257v Peak) and also 110mA RMS.

We can assume the tube will need the same voltage left across it, and so the similar peak voltages point towards a similar plate suppy voltage.  Indeed, we find this to be true in the schematics for each amp.  But we need more PT current for the Deluxe Reverb.  Changing OT Impedance alone doesn't yield more-power.

     84mA x 84mA x 2125Ω = 15w
     110mA x 110mA x 1650Ω = 20w

[Kaaskop]

Thanks HBP and all for this info. I’ll have to scratch my head for a bit to fully understand the implications of your reply.

As for

We haven't attempted to figure out "minimum required volts" or "minimum required milliamperes" for the goal of 20 watts.  But does the PT you're planning on using supply 300vac (or more) and 135mA (or more)?

I’m using Hammond transformers and I can either choose from the 290 Series which are Designed for drop in replacement of original units or the 270 Series which are generic but have a lot of choice and are generally beefier Transformers. I was looking at one supplying 300 vac and 150 mA.

[PRR]

Most of the total current that is present at the B+ reservoir cap is being drawn from the PT through the OT primary (via the output tubes). The remainder of the current that supplies the screens and pre-amp tubes is a relatively small proportion of the B+ current.

Lovely image! But he didn't do it to scale. This is a 125/19/6 rendering:

[acheld]

I’ll have to scratch my head for a bit to fully understand the implications of your reply.

I'm still scratching my head after years of thinking about it!    I recently purchased Richard Kuehnel's LTSpice Simulation book to (maybe) help me understand better.  The hard part in all of this is to visualize what happens when you make a change in one part of the circuit -- to the rest of the circuit.   HBP and others on this forum really can do this.

The good news is:
     1.  Tube amps have a lot of tolerance.  You can be pretty far off spec in some cases and do just fine.
     2.  There are so many great circuits out there to copy and use as "modules" for your own design.
     3.  Trying stuff out is fun!

[Kaaskop]

Code: [Select]

The good news is:
     1.  Tube amps have a lot of tolerance.  You can be pretty far off spec in some cases and do just fine.
     2.  There are so many great circuits out there to copy and use as "modules" for your own design.
     3.  Trying stuff out is fun!
Well said. I'm not in any hurry to learn but I do like to move ahead with as much certainty as possible. I guess I'm often trying to figure out what is law and what has as you say, a lot of tolerance.

[Kaaskop]

I couldn't resist responding to PRR's revamped illustration. As a visual artist I like visual clarity, here's one showing the body senses to scale ...but thats getting off topic.

[HotBluePlates]

I’m using Hammond transformers and I can either choose from the 290 Series which are Designed for drop in replacement of original units or the 270 Series which are generic but have a lot of choice and are generally beefier Transformers. I was looking at one supplying 300 vac and 150 mA.

Do you wish to mimic an original Deluxe Reverb?

I used to own a 1964 Deluxe Reverb.  I found that having 120vac of wall outlet voltage I got 394vdc at the plates & screens.  I also measured the winding resistance of my amp's PT and found it to be higher than Hammond's specs for their 290BX.  Lower winding resistance means higher B+ voltage, and less sag.

In 21 years on this forum, I still don't think I've seen a single member complain their new build had a too-low B+ voltage.  Almost always, folks complain their voltage is higher than expected/desired.

So I would personally use build-out resistors between the PT and rectifier to mimic the original winding resistance.  That will yield a "softer, vintage" response.  Omit those if you want "bold & solid."  I'd have to search the forum to find my figures, but likely less than 50Ω per side going into the rectifier.

Though note that Fender changed the power transformer several times (at least 3-4) during the production of the 1960s blackface Deluxe Reverb.  My 1964 DR's 125P23B power transformer had different specs than another members 1965 DR with the 125P23C power transformer.

I’ll have to scratch my head for a bit to fully understand the implications of your reply.

I'm still scratching my head after years of thinking about it!

I made it "harder than it is" for about 15 years, then Simplicity got the better of me.

To design a vacuum tube Output Stage you need to know about:

• The difference between Peak and RMS (and maybe "Average") values of Volts and Current
• How to convert between Peak and RMS values of a Sine wave
• Ohm's Law (Volts = Current x Resistance)
• The Equation for Power (Power = Volts x Current)
• Basic algebra to Combine Steps (substitute a small formula in place of a variable)
• A couple simple rules about the effective load in Class A and Class AB
• A couple simple rules about limitations of a tube's operation
• A couple simple rules for working with data sheets
• Rule: Treat the Output Transformer Primary Impedance as though it is a Resistor

Designing the output stage is almost always about, "Can I hit the desired power output," and almost never about Teh Toanz.  "Tone" is a happy accident, and afterwards the designer asks, "What caused that?!?"

Then they repeat The Happy Accident.


The Output Section's Power Output is how much heat it can cause The Resistor (output transformer primary impedance) to dissipate.  Then we assume (as a first-guess) all the power "dissipated" in the primary impedance gets transferred to the secondary load.


But if you treat the output section design as an Ohm's Law problem about a Resistor (and the tube pulls current through this Resistor), all the difficulties fall away.

[tubeswell]

To design a vacuum tube Output Stage you need to know about:

• The difference between Peak and RMS (and maybe "Average") values of Volts and Current
• How to convert between Peak and RMS values of a Sine wave
• Ohm's Law (Volts = Current x Resistance)
• The Equation for Power (Power = Volts x Current)
• Basic algebra to Combine Steps (substitute a small formula in place of a variable)
• A couple simple rules about the effective load in Class A and Class AB
• A couple simple rules about limitations of a tube's operation
• A couple simple rules for working with data sheets
• Rule: Treat the Output Transformer Primary Impedance as though it is a Resistor

...
The Output Section's Power Output is how much heat it can cause The Resistor (output transformer primary impedance) to dissipate.  Then we assume (as a first-guess) all the power "dissipated" in the primary impedance gets transferred to the secondary load.

But if you treat the output section design as an Ohm's Law problem about a Resistor (and the tube pulls current through this Resistor), all the difficulties fall away.

Good summary!


I'm ashamed to admit that one of the big lessons for me was the way current changes Pri:Sec in PP Class A vs Class B and the trick about thinking that 'negative current' flowing in a counterclockwise direction in one half of the primary winding  is the same as 'positive current' flowing in a clockwise direction in the other half of the primary winding from the P.o.V of the Pr:Sec phase and total Pri:Sec current. Either way, its all 'current' and its all going in the same direction in the primary at any instant in Class A. I got really confused about this at one point.

[HotBluePlates]

... One of the big lessons for me was ... the trick about thinking that 'negative current' flowing in a counterclockwise direction in one half of the primary winding  is the same as 'positive current' flowing in a clockwise direction in the other half of the primary winding from the P.o.V of the Pr:Sec phase and total Pri:Sec current. ...

There are some cases where it is helpful to consider the entire primary.

But most of the time it is very much more useful to think only of a half-primary.  Indeed, this is the approach recommended in RDH4.

[Kaaskop]

In 21 years on this forum, I still don't think I've seen a single member complain their new build had a too-low B+ voltage.  Almost always, folks complain their voltage is higher than expected/desired.

In my other builds the main tweak I tried to make was lowering the B+ hoping it would soften the amps sound. "bold and solid" felt punchy and shrill.

So I would personally use build-out resistors between the PT and rectifier to mimic the original winding resistance.  That will yield a "softer, vintage" response.  Omit those if you want "bold & solid."  I'd have to search the forum to find my figures, but likely less than 50Ω per side going into the rectifier.

I tried other things to bring the B+ down on intuition. Adding resistance after the first B+ Filter cap which now is clearly not the solution. I Also increased the cathode resistor which helped a bit, but "use build-out resistors between the PT and rectifier" is what I will do/try instead. Get it at the source.

I made it "harder than it is" for about 15 years, then Simplicity got the better of me.

To design a vacuum tube Output Stage you need to know about:
The difference between Peak and RMS (and maybe "Average") values of Volts and Current
How to convert between Peak and RMS values of a Sine wave
Ohm's Law (Volts = Current x Resistance)
The Equation for Power (Power = Volts x Current)
Basic algebra to Combine Steps (substitute a small formula in place of a variable)
A couple simple rules about the effective load in Class A and Class AB
A couple simple rules about limitations of a tube's operation
A couple simple rules for working with data sheets
Rule: Treat the Output Transformer Primary Impedance as though it is a Resistor

I generally learn the most when I come up against a problem because you have to solve it, so that directs my learning some what. Having the above list puts a focus on the learning that you only see after 15 yr. Thats the capacity of knowledge. It get store up and is concentrated. It focuses things.

the way current changes Pri:Sec in PP Class A vs Class B and the trick

I need to be able to visualize something literally have a picture of it in my mind be fore I feel like I understand it. Other wise it feels like guessing.

All these shared understanding are extreme helpful. It's much appreciated.