Dropping B+ Zener or Resistor - Difference in different architectures discussion

[kagliostro]

A friend is building (since a long time) a new amp (he revised it a lot of times) and, at the moment, he is trying to drop B+ as to obtain a better work point for the Power Tubes (the PT was planned for a different planned B+)

The drop of B+ can be achieved in different manners one is the use of Zener the other the use of resistor (I think the use of a Mosfet circuit will act as a resistor)

So I was thinking, which can be the difference on reaching the drop via a Zener or via a resistor in different topologies of amp ?

In a PP circuit we have a current variation between idle and full power, in an SE amp the current between idle and full power will vary only a bit and also at idle the current consumpion will be near the full power consumption

So, which can be the difference using a Zener to drop B+ in a PP that hase a current fluctuation instead of using a resistor ?

I think that in a PP amp with the Zener there will be no SAG, instead, using a resistor will give as result a SAG effect

otherwise, in an SE amp to use a Zener or a resistor will perform at the same manner

Am I correct ?

I'll like to hear what do you think about this

Many thanks for any contribute

Franco

[shaun]

I experimented with zeners to drop B+ a few years back. I cannot speak to the more technical aspects, such as any math for the amount of sag etc., but I found the following...

My first priority was to keep heat down; of course, both zeners and resistors can get very hot when dropping high voltage (at least in my experience). When I realized I couldn't avoid heat dissipation as a result of voltage drop, my concern became that of tone. Now, I can't remember the details too well, but I greatly preferred the tone when using a dropping resistor rather than using diodes. I don't know whether this was because of sag caused by the resistance, which can sound pleasing to some ears. I suspected there was more to it than that, but I didn't know enough at the time to comprehend it fully (probably still don't :).

I tried the zener experiment a couple of times in different builds, and I always opted for the dropping resistor(s). Partly, this was because I felt comfortable re-designing the overall filtering stage so that, if I needed to drop voltage, I could add an extra RCR or CR component before the first B+ node, so that I got voltage drop and also some filtering benefits. I guess I just felt more in control of the design that way.

[AlNewman]

It's funny, I've been playing with a dropping resistor on my DRRI today, (cause I was bored mainly), but my B+ was running at a fairly high voltage at the bias I wanted to run them, and I wanted to bring voltages down closer to the schematic.

The schematic for the DRRI states 396V b+, 391 at the plates, and -37V at the grid.  The original AB763 schematic shows 420V, 415 at the plate, with -35V at the grid.  I'm not sure what the original transformer specs are, but in the DRRI, mine reads about 200ish ohms DC resistance across the primary.

On my amp, where I was happy setting the bias, (around %60), I was getting around 431V B+, 427V at the plate, and -41V at the grid. 

I ended up using a 150R resistor off the rectifier, and got my B+ to around 406V, and my plates just over 402V, My grid is now sitting at -38V or so.  What I found surprising is there's only a 10V or so drop across the 150R resistor, but it had a fairly large impact on the voltages through the rest of the amp at the bias setting I like.

Does it sag more?  Hard to say, I kind of doubt the 5AR4 would sag much anyways, with 65mAish of current, so the resistor might have added some sag in.  It seems to have more dynamics, and pick attack makes a larger impact, but that could be just because of the lower voltages across the entire amp.  Everything dropped by close to 25V.  Sounds minor, but that's at every node, it seems to make a big difference.

I do have some 16V/5W zeners here, I never even tried them, if I had read your post earlier I probably would try clipping them in to see, but now it's pretty much buttoned up.

[HotBluePlates]

So, which can be the difference using a Zener to drop B+ in a PP that hase a current fluctuation instead of using a resistor ?

I think that in a PP amp with the Zener there will be no SAG, instead, using a resistor will give as result a SAG effect

otherwise, in an SE amp to use a Zener or a resistor will perform at the same manner

Yes, you already knew the answer.  Resistor ---> Sag when there is Changing Current.

A friend is building (since a long time) a new amp (he revised it a lot of times) and, at the moment, he is trying to drop B+ as to obtain a better work point for the Power Tubes (the PT was planned for a different planned B+)

I have gotten to the point where I have no patience anymore for schemes to "drop B+ voltage."  These days I say, "Use the correct power transformer, or don't bother building the amp."

That's not because B+ reduction won't work, but more because people go too far with it.  Or they underestimate how much correction they will need, and end up higher than they wanted anyway.  I almost never see anyone say, "My B+ voltage wound up being a lot lower than I'd like."

[kagliostro]

My friend amp PT story is very long

The amp was planned in a slightly different way about Power Section and also the available currents were a underextimated for a part of the winding, the more the Power Tubes require a low voltage G2 respect to B+, so the way the PS was realized was changed

Now seems he obtained what he was looking for but he is looking for the best way to dissipate the heat produced dropping the voltage

About this aspect I think that to use a Zener, a MOSFET in a linear circuit, a resistor on B+ or in AC before rectification will result in the same, the W dissipated  for a specific Voltage Drop will be the same in any way you reach it (I think so, correct me if I'm wrong)

Franco

[bmccowan]

I think you are right Kagliostro. Power that is wasted is wasted as heat. So unless he wants to replace the PT, the choice boils down to how to handle the heat. With either a big resistor or a MOSFET, establishing a heat sink is the method that occurs to me. Sluckey dropped a lot of voltage in Rocky https://sluckeyamps.com/rocky/rocky.htm using a 30W resistor. I have used a VVR that includes a heat sunk MOSFET on a couple of projects - but in order to lower volume and retain tone for "bedroom" playing, not to get the overall voltage down. The VVR MOSFETs have held up well, but not everyone shares that experience. Some here have had failed VVRs I think.
Best solution is swap out the PT and save it for another project IMO. That is until someone figures out how to unscramble eggs.

[kagliostro]

Ciao bmccowan

The amp is a prototype so my friend will keep this transformer, we are planning the correct one for the next construction

here is a first approach to the new PT solution









Franco

[HotBluePlates]

The amp is a prototype so my friend will keep this transformer, we are planning the correct one for the next construction

here is a first approach to the new PT solution

170v AC x 1.414 = 240v Peak  (minus about 1v for the bridge diodes)

G2 Supply is really going to pull 340mA?!?  That doesn't seem likely...

I don't see how these voltages are going to drop down in the ways hoped, certainly not the G2 supply sagging 20v compared to the winding output (which will likely be higher voltage than planned, due to under-loading).  This feels like forcing bad design choices in an attempt to recover for an earlier design/purchase mistake.  It's not likely the end-result will be satisfactory when it starts out poorly in this manner.

My friend amp PT story is very long

The amp was planned in a slightly different way about Power Section and also the available currents were a underextimated for a part of the winding, the more the Power Tubes require a low voltage G2 respect to B+, so the way the PS was realized was changed.

I think you're trying to be helpful to your friend.

Sometimes the best lesson is how to "cut your losses" or "avoid throwing Good Money after Bad."

[PRR]

G2 Supply is really going to pull 340mA?!?  That doesn't seem likely...

The two 220VDC supplies are stacked to make 440VDC.

If we assume the total cathode current is 340mA and Ig2 is 10% of that, the upper winding can be only 306mA. Which is probably same-as 340mA for practical purpose.

The windings can be re-stacked to save four diodes but the cost is trivial and there may be reason to limit the circulating current.

[kagliostro]

Hello Friends

G2 Supply is really going to pull 340mA?!?  That doesn't seem likely...

No ....  G2 consumption on that circuit is 2.4mA for each tube (tubes are two) 

Plate consumption is ~100mA for tube

So we estimated 205mA DC x 0.6 = 340mA AC

As the two windings are staked (as in Hiwatt architecture) the current abilities of the two windings must be able to supply the whole current consumption of the circuit

The solution of two staked windings is to be able to supply the G2 at a specific lower voltage (experimented via simulations and also in the real world) respect the Voltage on plates

The transformer proposal you can see on my porevious post is the new version that solves the problems encountered with the transformer that is on the amp now, wich has far different architecture



To feed G2 with that PT we used the CT connection (as in some old Philips PA amp)



But also that is a last moment solution, the 190V-0-190V initially was planned for G2 supply purpose (and I donìt really know why the 190V winding was planned for 270mA and the following 300V winding for 220mA)

Franco

[danhei]

Could he use choke input instead of capacitor input?

[kagliostro]

The PS schematic posted is only the PS at the begining of the circuit, follows gyrators in place of chokes

Franco