Regarding the B+mosfet reducer and power ground

[pjd3]

Hello,

Something I noticed last night and got me really thinking but, not yet able to fall on a conclusion.

This is about what happens to Power ground when a B+mosfet reducer circuit as drawn out by RG and company is installed as should between HT centertap and power ground.

As I was going over my latest mods and updates to a Blackvibe 6L6, I realized that the HT centertap was necessarily lifted from power ground and onto the B+reducer circuit, and then back to power ground after the circuit. What occurred to me was that ground returns from the Bias circuit and the filter caps were still connected directly to the original power ground before the installation of the B+Mosfet reducer. This really bugged me as I wasn't mentally able to resolve what this means to the power system, especially as this would (or could, actually) effect the electrical balance of things. My only reconciliation with what I was looking at was that having earth ground connected to chassis as well as power ground would hold some reliance that the ground returns from bias and filter caps were seeing acceptably close to a zero volt, or acceptable ground condition.

The amp appears to have happily accepted the 20v B+mosfet reducer and i was able to get voltages and bias where i wanted them. Perhaps that's enough.

But, what happens in a worst-case scenario say, if the earth ground in some AC well outlet does not have an earth ground connected? Then what would the bias ground and filter caps be seeing electrically, and how would this effect the power conditions of the amp? While I'm 90% confident that the presence of an earth ground keeps things where I want them to be, is there something more I should have done, as in creating a floating HT ground and connect the bias and filter cap grounds to that along with the CT connection to the B+reducer. While it's entirely possible that I'm creating a problem in my head that isn't really there, I thought it would be worth mentioning in case there is something to look out for.

Again, thanks for your time stopping by.
Best,
Phil D.

[R.G.]

I'm having some trouble visualizing exactly how you have it hooked up. Can you post even a quickie, crude sketch so I don't mis-visualize it?

If I'm visualizing correctly, the MOSFET reducer installed in the ground-to-PT/CT should have no effect on the circuit at all - as long as you tie nothing but the negative end of the MOSFET reducer to the PT/CT. The negative terminal of the first filter cap can then stay as the ground reference for all the rest of the circuit, and for tying to the chassis.

I can see where this would get tricky if your bias is a tap on the high voltage winding. In that case, the MOSFET reducer and bias tap do interact, as the bias tap still thinks the CT is the reference. I can construct some scenarios where things are funny ... but then, you say it's working OK. I really need a sketch of how you wired it to do a respectable job of speculating what's happening.

[pjd3]

Hi RG, thanks for posting on this. I am at work now and will try to see about posting a diagram but, I always seem to fail miserably when attempting to post an image on a forum. I need to get on the stick with that.

My power ground was just a small lug strip with HT centertap, filtercap returns (cathodes), bias ground and earth ground all connected together.

All I did was unsolder the HT centertap from the power ground and connect that to the source on the Mosfet. The Mosfet was bolted to the chassis with thermal grease and the mosfet drain was soldered to the power ground, just as the diagram portrays on the "MOSFET B+ Reducer paper, "the one with the PC board, the 20 diodes pics with v drops, and the pc board trace images at the bottom"

So really my power ground was initially done in I believe a traditional way, and all that was done was to unsolder the HT CT and send that to the MOSFET reducer circuit, then from the MOSFET reducer to power ground.

Hope that clarifies the picture. I struggled with the notion, "Should I have kept the bias ground and filter cap ground on the original earth grounded power ground or, have the bias ground and filter cap ground follow the HT CT to the power Mosfet.

thank you RG, hope I was able to clarify and not waste time confusing!

Best,
Phil Dl

[R.G.]

I ... think... that moving only the PT high voltage center tap was the right thing to do for everything except possibly the bias.

How is your bias voltage generated? Is the output stage cathode biased? Is it fixed bias? If fixed bias, does the bias voltage come from a tap on the high voltage winding, or from a capacitor/resistor dropper from the high voltage ends of the high voltage winding? These are things that could matter to the bias.

Give that your bias voltage seem to be OK, the disaster scenario I was worried about seems not to be happening, but the actual wiring would clear it up.

A MOSFET reducer, just like a zener in the same position, just places the high voltage CT one zener-voltage below the negative end of the first filter cap. As long as nothing else is connected and referenced to the CT, this all works fine, as the high voltage winding is galvanically isolated, and works fine with an offset, just like a 5V rectifier heater winding will do. I was worried that if the bias voltage is generated from a tap on the HV winding, it gets offset negative by the same amount that the zener moves the whole winding, so the bias voltage gets different.

But then, I may just need more coffee.

[martin manning]

A MOSFET reducer, just like a zener in the same position, just places the high voltage CT one zener-voltage below the negative end of the first filter cap. As long as nothing else is connected and referenced to the CT, this all works fine, as the high voltage winding is galvanically isolated, and works fine with an offset, just like a 5V rectifier heater winding will do. I was worried that if the bias voltage is generated from a tap on the HV winding, it gets offset negative by the same amount that the zener moves the whole winding, so the bias voltage gets different.

HV CT gets shifted down below chassis ground, so B+ gets pulled down wrt chassis ground, which is the goal. Most bias windings share the HV CT (only a few are a separate winding), so the bias voltage will get pulled down, i.e. become more negative, by the same amount that the B+ is pulled down. An adjustment to the bias circuit is very likely to be required, since the bias voltage will need to be higher, not lower, than it was when the B+ was higher. The bias circuit ground must remain on the chassis, not connected to the HV CT (MOSFET source terminal), because the CT voltage is bouncing up and down with the HV return current pulses.

[R.G.]

That's a clearer explanation of what I was worried about. If the bias is a tap, not a separate winding then the whole bias generation rectifier/filter/is pulled down by the zener setup. The bias voltage is shifted. I - think - what I wrote is consistent with that, at least from the perspective of too little coffee.

I was mentally going through the bias changing by maybe 40V or more. Yeah, the bias filter cap needs its ground tied to the main circuit ground, but the voltage on it goes more negative by the MOSFET/zener shift. For big changes of B+ reduction, some bias change is needed all right.

On the original question, I can't see any issues with signal ground or AC safety ground or whatever with the MOSFET reducer hooked up as described.

At least until I get my afternoon coffee...

[pjd3]

Thanks guys for this information,

This amp is fixed biases, the bias tap is a 50 volt tap off the HT coil, and the bias has both a bias level "master" and a balance, that I believe part of the Rob Robinette Blackvibe 6L6.

And the bias "master" pot had to be rotated nearly 180 degrees to meet the new bias setting. I had to change the resistor connecting the 2 capacitors in order to even be withing adjustment. So, that scenario played out exactly how you described it above. I'm fairly confident that moving the 50vac bias wire from the HT coil from the power ground to where it connects to the B+Mosfet reducer would make the bias less negative (I think) and put the bias pot rotation in a more middle range, as opposed to being an 1/8th of an inch from the end of its rotation as it is now. That wouldn't be very difficult to do and determine.

Then there is still the status of the filter caps and according to both of your explaination, I want the filtercaps to have a reduced differential voltage (that is the sole goal) but, I don't necessarily want the bias voltage scheme to change too widely. So yes, it would make some sense that moving the bias ground with the HT CT to the B+reducer makes sense but, not the filter cap cathodes, as we want to have less differential voltage for the first filter cap B+ by way of maintaining their cathodes at a chassis ground zero volts status of sorts. I will likely try moving the bias ground with the CT to the B+ reducer just to see how it effects the bias pot range. EVen though it is adjusted properly, it would be nice to get the bias pot nearer the middle of its rotation. Whynot. But, I suspect as long as the emissions of the power tubes are where I wan't them either would be OK.

Thanks so much guys, new things to ponder and consider. Its all good to me.

Best,
Phil D

[martin manning]

...I want the filtercaps to have a reduced differential voltage (that is the sole goal) but, I don't necessarily want the bias voltage scheme to change too widely. So yes, it would make some sense that moving the bias ground with the HT CT to the B+reducer makes sense but, not the filter cap cathodes, as we want to have less differential voltage for the first filter cap B+ by way of maintaining their cathodes at a chassis ground zero volts status of sorts. I will likely try moving the bias ground with the CT to the B+ reducer just to see how it affects the bias pot range. EVen though it is adjusted properly, it would be nice to get the bias pot nearer the middle of its rotation. Whynot. But, I suspect as long as the emissions of the power tubes are where I wan't them either would be OK.

I don't think so. The MOSFET source terminal has a large AC voltage on it (~20V in a simulation), and that will cause a lot of hum. I also looked at a separate bias winding with one end connected to the HV CT vs. a tap on one side of the HV coil. The results are the same for both PT configurations with the bias ground on the chassis, i.e. the bias and the HT voltage both come down by the Zener voltage.

[pjd3]

Thank you Martin,

According to what you discovered in simulation, seeing noise at the source of the mosfet, its looks like I should leave everything right where it is, if I am in fact getting the right picture:

1. HT center tap to Mosfet source
2. Bias ground to earth grounded power ground
3. Filter cap returns to earth grounded power ground.

That's what I have now, and although there was a notable re-biasing necessary, I seem to have a good sounding and well behaved amp. I think the only caveat with this is that the presence of the earth ground to the power ground needs to be there to keep everything at the proper reference.

Best,
Phil D.

[dorrisant]

You've got it right, from what I can tell. It can be quite intimidating to connect the B+ Reducer. I always crank my bias to the most negative voltage possible, leaving it grounded to the chassis, before I insert the Reducer, then adjust the bias pot/resistor for the proper bias with the reduced HT. Also, I have had Reduced amps out there being gigged on a regular basis for years now with now issues. As opposed to VVR, I have had zero failures. It may seem a bit weird to wrap your mind around but works quite well and reliablely.

[pjd3]

Thanks Dorrisant,

I think I can relax now. But there was a level or two I needed to wrap the head around to get a fuller picture of what was going on, electrically. I believe it was Martin that mentioned that what the B+mosfet reducer is doing is lowering the HT centertap by the zener drop, which in my circuit would bring down the HT centertap to -20 below the earthground/power ground. Thats what made everything else fall together in terms of where things should be connected.

And yes, I'm very happy with how well the Reducer is working. It was nice to bring the plate voltage down a bit on those cool stubby little 6L6WGB's. And the amp still has 20 times more clean headroom than I'll ever need. On top of that, it's great to have one more very useful and effective trick in the bag should an amp benefit from a little drop. I'm a B+mosfet reducer fan boy.

Well done guys, thank you.
Best,
Phil D.