setting the B+

[cooldude666]

so I'm wondering that means I really need to have more like 800v coming off the rectifier in order to supply my screens and the plate with their requisite voltages.

[jazbo8]

The screen voltage is typically specified for the idle and not maximum current. Also see this for some basic PP stage calculations.

[cooldude666]

so a 5 watt 36k resistor should get me where I want to go. Will try and see what happens.

[Enzo]

I picked 1k because i asked whether you were discussing screen grid resistors or the B+ chain. I then discussed the screen grid resistor, 1k being a typical value, because you mentioned 38ma screen current.

[cooldude666]

Ok, I think I understand what I am misunderstanding. Enzo pointed at this but I wasn't looking. There are two resistors I need to figure for. The series dropping resistor in the B+, and the individual screen grid resistors. The screen grid resistors have a set of common values, between 220R and 1k. So my question is about choosing the series dropping resistor. Which I suppose was the original question in this thread!

Merlin, why wasn't any of this in either of your books??

[Amp Kat]

Ok, I think I understand what I am misunderstanding. Enzo pointed at this but I wasn't looking. There are two resistors I need to figure for. The series dropping resistor in the B+, and the individual screen grid resistors. The screen grid resistors have a set of common values, between 220R and 1k. So my question is about choosing the series dropping resistor. Which I suppose was the original question in this thread!

Merlin, why wasn't any of this in either of your books??

Take into account the screen resistors pull very little current especially 6L6's so the voltage drop across them will be very little to almost none so that is not going to have a big impact on the series dropping resistor. I would worry more about the total current draw of all the tubes combined and put a meter in between and see what you got current wise then do ohms law.

[tubeswell]

I'm reminded of this:



from here The Valve Wizard

[trobbins]

The screen supply current is seen to vary from idle level by a large amount compared to other dropped voltages on the B+ chain. Hence, as screen current varies with signal conditions, it can significantly affect the voltages being supplied to the other stages on the chain - but that is a dynamic consideration, and used in many amplifiers to give them their dynamic character (sag). Similarly, the regulation of B+ itself with dynamic changes to the current drawn from output stage OT and screens, can similarly affect the B+ voltage level. Some designs separate the screen supply from the B+ chain to other stages.

So it is often easier to do an 'idle' voltage design cut, and another design cut based on the sagged level of B+ and the maximum screen current - but this needs to appreciate how dynamic signals affect filtered voltage levels that have RC time constants, and how far the amp can push the output stage in to class B and grid current conditions.

[cooldude666]

I'm reminded of this:

[ATTACH=CONFIG]24929[/ATTACH]

from here The Valve Wizard

I wish this was in his book. This is excellent. This is exactly what I've been looking for.

[cooldude666]

So let me show ya'll my math.
6550 Anode voltage = 480v
Total current draw: 190.4 mA (2x6550: 80mA, 80mA; 2x12ax7: 1.2mA, 1.2mA, 1.6mA, 1.2mA; 12at7 [reverb driver]: 7mA, 7mA; 12ay7 [Pi], .9mA, .9mA)
After power tubes: 30.4 mA
Desired screen voltage = 310v

Drop from anode to screen = 480v - 310v = 170v
R = 170v / 30.4mA = 5.6kΩ
Power dissipated via 5.6kΩ resistor: P = I^2 * R; 30.4mA * 30.4mA * 5.6kΩ = 5.175 watts
So it looks like a 10 watt resistor will get me where I want to go with dissipation to spare.
My maths look alright? I really appreciate the help. Given this is all correct, I finally understand. I'm already bald at 25. Can't afford to rip any more hair out.

[Enzo]

A Margaret has male pattern baldness?


Um, did you include a few ma for the screen current and forget to write them down here? For me all those 1.2 thru 0.9 add up to about 21ma, so did you go for about 9ma screen current to get 30ma? But using the 30ma, I get about 5.2 watts also.

[trobbins]

Caution is required with choosing screen dropper and screen stopper resistor power ratings, due to likely rapid increase in average screen current when amp is cranked. Although 5W operation of a 10W resistor is a reasonable design choice for a static situation, it would be a concern for a screen type application.

[tubeswell]

6550 Anode voltage = 480v
Total current draw: 190.4 mA (2x6550: 80mA, 80mA; 2x12ax7: 1.2mA, 1.2mA, 1.6mA, 1.2mA; 12at7 [reverb driver]: 7mA, 7mA; 12ay7 [Pi], .9mA, .9mA)
After power tubes: 30.4 mA

Huh? I make the preamp to be 21mA total (adding up your numbers). If the 80mA per 6550 is for each of the plates, then you need to figure something in for the screens. The screens need to be added to the 21mA to get the final drop across the plate-to-screen node supply resistor. If you're figuring only about about 5mA per screen, then you will end up with 30.4mA for that node.

To estimate screen current, get the relative plate current and screen current data off a 6550 data sheet. Using this page 2 from this datasheet http://www.dougstubes.com/6550-tungsol.pdf for example, I get 87mA and 4mA respectively with a plate voltage of 400 for a single 6550/KT88. (This datasheet is handy because it also give max estimates for plate and screen current under signal conditions). So your numbers seem about right I guess. Try it and see (Its going to depend on the bias point)

[cooldude666]

A Margaret has male pattern baldness?


Um, did you include a few ma for the screen current and forget to write them down here? For me all those 1.2 thru 0.9 add up to about 21ma, so did you go for about 9ma screen current to get 30ma? But using the 30ma, I get about 5.2 watts also.

yes, i did.