DC Heater considerations

[pjd3]

Hello, hope ya all had a great T-day.

I suppose its time to get ready to set up for DC filaments in a 2204 type circuit with JEL mods. I assume that noise ought to be something to consider minimizing and seeing that there is no centertap on the heater coil I should probably think about giving the DC heater pcb a whirl that I purchased from Headfirst amplification, at least for the first few gain stages or so.

I would have firstly tried elevating the heaters but, I believe I need a proper centertap for that. (The PT that came with this donor amp has all of the coil centertaps internally connected, anyways - I'm going to give it a try based on reports from others that odd PT's like this has worked for them).

Of course there is lots of information and threads on the subject but I was respectfully hoping to get some warnings on classic things to avoid in terms of layout, wiring and grounding. I have heard that installing DC heaters can introduce new issues that weren't there before so, that's what I'd like to specifically target.

Thank you everyone.
Best,
PJD3

[martin manning]

A faux center tap consisting of 2x 100R resistors is generally more than adequate. You can also do a "hum dinger" where a pot is employed to adjust the centering (or off-centering as needed) of the ground.

The general rules for DC heaters:
Only one ground reference is allowed
V1 is the most critical, and the improvement is marginal thereafter
The DC doesn't have to be all that clean. See below for Marshall's own DC filament circuit on V1 with no heater winding CT from the 4001 Studio 15.

[pjd3]

Thank you Martin,

This looks good, and very simple to put together.

I think I shall give this a go. I like that it skips having to employ voltage regulators and such as the pcb I have does.

thanks again!

Best,
Phil D.

[martin manning]

You can make the 1000u cap 2-4x larger if you like to reduce the ripple and/or tune the measured DC voltage across the heaters.

[pjd3]

Hi Martin, thanks.

Can you give a brief explaination of how enlarging or adjusting the cap would alter the measured DC voltage across the heaters? The only way I have understood a circuit similar to this one is that the ripple continues to decrease as the cap value is increased, or if maybe smaller caps are piggy backed with the larger cap but, voltage would only ever reach whatever the peak voltage was, or is.

Thank you. i would be interested if anything else occurs beyond the ripple voltage decreasing. (it just popped into my head that the impedance could change with a larger cap creating a voltage division of some kind. Maybe?).

Best,
Phil D.

[martin manning]

Can you give a brief explaination of how enlarging or adjusting the cap would alter the measured DC voltage across the heaters? The only way I have understood a circuit similar to this one is that the ripple continues to decrease as the cap value is increased, or if maybe smaller caps are piggy backed with the larger cap but, voltage would only ever reach whatever the peak voltage was, or is.

Sure. Imagine the reservoir has infinite capacitance. Ripple is zero, and the average DC voltage is equal to the peak AC voltage, less two diode drops. If the capacitance is lowered such that there is a 10% ripple, the peaks are the same but the average DC will be reduced by 5%.

i would be interested if anything else occurs beyond the ripple voltage decreasing. (it just popped into my head that the impedance could change with a larger cap creating a voltage division of some kind. Maybe?).

Larger reservoir capacitance increases the peak current in the charging pulses, putting more stress on the rectifiers and producing more forward voltage drop. Simulating this circuit with estimated transformer winding resistance, iN4007's, and constant heater resistance I see about 0.3V difference in average DC between 1000u and 4700u.

[pjd3]

Ah OK, thank you for that explanation.

It makes alot of sense that average DC is going to decrease as ripple increases, as well as current increasing as larger capacitance is opening up the reservoir for more electrons to freely flow. I think that's it.

So what I think I'm getting is that should you be able to tolerate any noise that arises from a decrease in capacitance, lets say if you wanted to drop the heater voltage a tad, you decrease the cap value to lower the heater voltage. Interesting. I wouldn't have thought to do it that way but, there it is.

I have noticed that cap specs will often include a max filter cap value depending on rectifier type.

Thanks again Martin, its nice to have opportunities to look into something like this more deeply than I have. But now, its time to go start another thread about an OT secondary brain meltdown I'm having tonight!

Best,
PJD3

[JD0x0]

The DC doesn't have to be all that clean.

My first go around with DC heaters I had to increase the filtering because the ripple was causing hum. I was using a setup that Bruce Egnater suggested. I think I started with 16000uF (IIRC), and ended up having to parallel in a second cap to get rid of the hum.

[Guy77]

Another great and simple alternative to eliminate heater hum is elevating the existing AC heaters with an injection of DC volts! I have done it on several amps and was pleasantly surprised at how much quieter they got.

Here is the thread that describes it.

https://ampgarage.com/forum/viewtopic.p ... 5&start=30

Cheers
Guy