Hello there,
I'm asking for some help in understanding the capacitive divider used in this circuit.
It appears that the "division" occurs at the node between the two 8.2nf capacitors to the power transformers CT.
Why does this node need to be referenced to the middle of the reservoir capacitor stack?
I wish to use this design, but I want balancing resistors for my series reservoir caps to ensure that one does not charge/discharge dispraportionately.
I'm assuming adding these resistors would mess with the DC conditions of the circuit.
Any thoughts?
I'm asking for some help in understanding the capacitive divider used in this circuit.
It appears that the "division" occurs at the node between the two 8.2nf capacitors to the power transformers CT.
Why does this node need to be referenced to the middle of the reservoir capacitor stack?
I wish to use this design, but I want balancing resistors for my series reservoir caps to ensure that one does not charge/discharge dispraportionately.
I'm assuming adding these resistors would mess with the DC conditions of the circuit.
Any thoughts?
That is not a capacitive divider but simply 2 x 8200pF noise suppressing capacitors, each in parallel with one transformer winding.
Zero influence on DC voltages of any kind.
Zero influence on DC voltages of any kind.
You can add a resistor across each 220uF, but they won't affect the operation much, or later the DC voltages.
The circuit is self-balancing, and is used to get the two different positive DC voltages from the same winding.
The circuit is self-balancing, and is used to get the two different positive DC voltages from the same winding.
So nothing with the transformer itself is involved in the splitting? Why is the CT connected to the junction of the two 220uf capacitors?
so the circuit does not necesarily need balancing resistors? Would it still be best practice?
The circuit is actually a variation of the standard bipolar power supply that gives equal and opposite DC voltages.
The only difference is that the system ground is referred to the "negative" supply terminal instead of the secondary center tap.
This creates two positive DC voltages, one twice the other, and is why balancing resistors are not needed.
The topology forces the capacitor DC voltages to be equal.
The only difference is that the system ground is referred to the "negative" supply terminal instead of the secondary center tap.
This creates two positive DC voltages, one twice the other, and is why balancing resistors are not needed.
The topology forces the capacitor DC voltages to be equal.
Ah, I see now. If the circuit ground was made to be the CT, then you would have 232v +/- rails instead of +465v and +232v.
Only the +465v rail is full bridge recitified, while the +232v rail is half wave rectified.
The capacitors across the PT secondary leads threw me for a loop but I see now that they are just for noise suppression and aren't required.
Would such capacitors need to be the X rated type or is that only for the mains side of a power supply?
Oh ok so the diode on the center node is just to keep the 100uF Y node capacitor from discharging into the series capacitors?
In some amps, the lower magnitude driver stage DC voltage supply is fed from a diode and has substantial
energy storage afterward. This is to prevent the LV supply from being pulled down when the HV supply
is heavily loaded by the output stage sourcing high current. Older ARC amps like the D76A and D150 did this, for example.
energy storage afterward. This is to prevent the LV supply from being pulled down when the HV supply
is heavily loaded by the output stage sourcing high current. Older ARC amps like the D76A and D150 did this, for example.
That makes sense. Since the +232v rail is so much lower, you wouldn't want it to sag at all or things might start to get hairy.
Thank you for walking me through this. It's interesting how many design decisions go into such an innocuous looking circuit.
Minimalism is an art in itself.
I LOVE it when an apparently "simple" design does 2-3-4 jobs well.
It takes LOTS more thinking than just throwing parts at a problem
As, in, say, designing a circuit to be naturally stable or balanced instead of requiring a servo, etc.
Or when you need 0.1% precision parts to work at all.
I LOVE it when an apparently "simple" design does 2-3-4 jobs well.
It takes LOTS more thinking than just throwing parts at a problem
As, in, say, designing a circuit to be naturally stable or balanced instead of requiring a servo, etc.
Or when you need 0.1% precision parts to work at all.
The 232V supply will tend to follow the 465V supply. When the main B+ sags under load so does the lower supply - by roughly the same %. It’s not a precision circuit - just a convenient way of getting 1/2 B+ without a lot of heat in a dropping resistor or mosfet. CR9 and C40 aren’t strictly required, but do suppress most of the ripple at the center voltage node that will appear due to main B+ current. Depending on what you were using that 232V for you may want it cleaner that the 456V B+. If you wanted to suppress that ripple even more, a choke in series with CR9 can be used. Typical uses would be for screen supply for output tubes or a plate supply for CF driver stages - or perhaps just a preamp stage or two.
That setup is common in guitar Music Man amplifiers.
They use 650-700V plate voltage with 6L6 !
Or 6CA7, an EL34 cousin
Yet they work for years 😲
BUT :
* They work in AB2 class 😲
Designer was GENIUS.
In the old days, Philips offered a 100W PA amplifier using only two EL34.
800V plate, 400V screens, 11k plate load, 500 ohm secondary.
I made a lot of them in the late 60s early 70s, very popular with churches and small town "street radio" setups.
They use 650-700V plate voltage with 6L6 !
Or 6CA7, an EL34 cousin
Yet they work for years 😲
BUT :
- They apply 350V to screens, from a similar setup
- Power tubes are biased cold 🥶
* They work in AB2 class 😲
- Grids are always positive 😲 (+16V or +24V) and cathode swings as much as needed.
- Tubes can always fully saturate which contributes to better efficiency.
Designer was GENIUS.
In the old days, Philips offered a 100W PA amplifier using only two EL34.
800V plate, 400V screens, 11k plate load, 500 ohm secondary.
I made a lot of them in the late 60s early 70s, very popular with churches and small town "street radio" setups.
How does driving the cathode of a tetrode work? Does the grid function as a sort of cathode by being held at a static positive voltage ?
Wouldn't it be biased deep into class A with Avery positive grid voltage?
After viewing the music man 2100-65 schematic, why does it feature an all tube power amp on one page and a transistor/pp amp PI on the next page.
Sorry to go off topic, I've just been learning so much from this thread!
1) what wg-ski said.
2) grid is held at fixed +15 or +24V (quoting from memory) which qualifies as positive 😄 but cathode is current fed, not voltage fed (although this drive mode would also be possible).
Then cathode voltage will become "anything needed to pass that current"
Example: suppose with those plate and screen voltages you need -50V grid-to-cathode to pass 10mA (totally made up example).
Then cathode will rise as much as needed to get 10 mA through 😲
If static biased to 10mA, you will measure 50+15=+65V to ground 😲 so grid to cathode is required -50V
Tube is tube and had no clue nor care about where grid and cathode voltages come from, only their difference
Advantages:
1) tube can be driven to death, if needed it can have +15/24V to ground, while cathode sits at only 3-4V above ground, a BRUTAL drive level.
Very worn tubes still give decent performance in a Music Man amp.
2) high linearity comes from master bipolar transistor driving it; a tube would be very non linear in a bad way trying to achieve that.
A GENIUS designer, not sure why no such amps exist in the HiFi world (or in DIYAudio pages).
Music Man amps are famous for being tear your eyes away LOUD and very clean.
do not be so concerned, this is a FWB supply, the center tap was used to connect to the midpoint of the capacitors so that you need not use a bleeder/balancing equal value resistors of say 330k 2watts, be mindful of the voltages involved, the FWB is Vac sec x 1.4 for unloaded voltage, so your series stacked capacitors must be rated at least Vsec ac/80% of the total rectified voltage. Notice that for tubes the idle or unloaded dc terminal voltage is quite lower at full load and has to do with the dc resistance of the secondary winding referenced to primary dc resistance...
I suspect this power supply is for some amplifier that wants a somewhat regulated screen supply. Usually an amplifier will use a resistor/capacitor combination taking the screen voltage from the plate supply.
Screen current draw is not constant with amplifier output so the screen voltage may vary.
Some may use a pass device to keep the screen voltage constant which requires extra parts.
This supply gets a fairly regulated screen supply without resorting to pass devices or a VR tube.
Screen current draw is not constant with amplifier output so the screen voltage may vary.
Some may use a pass device to keep the screen voltage constant which requires extra parts.
This supply gets a fairly regulated screen supply without resorting to pass devices or a VR tube.