Can you temporarily clip the AC heater wires in the PS box then hook up your bench power supply set to 6.3V. If the hum remains then you'd know its not AC heater induction via the 30 inch umbilical because the umbilical heater wire is on your bench supply now. If it goes away then at least you'd know its heater related. I don't know if your AC heater is elevated though, but the bench supply would be an isolated heater so shouldn't matter.
Does anyone know of an article or posting that will help me think about the capacitance needs on the amplifier side of the umbilical?
If you use such PSU, where the "last" capacitor (which is has the best parameters) is the energy storage device, the umbilical resistance and inductance degrades "last" capacitor quality (apparent ESR and ESL growing).
In this case is better solution the "capacity sharing", the most better capacitor is in the amplifier box, and the PSU's "last" capacitor is more average.
If the PSU is regulated, stabilized, this layout not so effective.
In this case is better solution the "capacity sharing", the most better capacitor is in the amplifier box, and the PSU's "last" capacitor is more average.
If the PSU is regulated, stabilized, this layout not so effective.
euro21... Would you explain what this means "If the PSU is regulated, stabilized, this layout not so effective.". I am planning a PS with a short 12 inch umbilical and Maida regulator in the PS box, then some polypropylene DC link capacitors in the Amp box branching left and right channels. The PS box is basically transformer, bridge, doubler, RC filter, Maida, umbilical. The amp box adds WIMA DC link capacitance upon entry, then the Amp. I'm hoping you could elaborate on this last point here, thanks.
It's a thermistor Tony, see - https://www.digchip.com/datasheets/parts/datasheet/1950/CL-60.php
Can you quantify the hum? Is it 50/60hz or 100/120hz? Is it on the grids of the voltage amplifier stage or getting in somewhere else? If it were my amp I'd have a sniff about with a scope, check the wires/terminals of the umbilical etc.
On my separate PSU & two 50w PPP 807 monoblocks I used separate heater & DC umbilicals, there's little 50hz hum but I have had issues with noise & other artifacts. I found a lot of that was down to ground/mains earth loop issues. It's a right sod to track down some of these problems.
Andy.
For the capacitance at the amp side of the umbilical, I was initially thinking that I should consider it like another layer in the CRC power-supply filter, consistent with @euro21's reply, so it would be an electrolytic at 18,000 uf. But @Windcrest77's reply seems to indicate that I should think of it as a dc link capacitor, so it would be a film capacitor sized very differently. I'm reading articles online and still not sure how to think about this.
(Apologies to @euro21 and @Windcrest77 if I'm misunderstanding your responses, and I do appreciate that you took the time to respond.)
(Apologies to @euro21 and @Windcrest77 if I'm misunderstanding your responses, and I do appreciate that you took the time to respond.)
Most regulator (designer) define capacitances on output to maintain stability.
For example original Maida (LM317 based) regulator has R-C snubber on output.
If you use (in amp box) additional capacitance, the regulator can work differently at load variations.
BTW I use -not too large- additional capacitance in (pre)amplifier box too, but the HV regulator is shunt type, with 4 wire in umbilical (Kelvin four terminal sensing), and the load almost invariable (CCS loaded tube).
For example original Maida (LM317 based) regulator has R-C snubber on output.
If you use (in amp box) additional capacitance, the regulator can work differently at load variations.
BTW I use -not too large- additional capacitance in (pre)amplifier box too, but the HV regulator is shunt type, with 4 wire in umbilical (Kelvin four terminal sensing), and the load almost invariable (CCS loaded tube).
Should be obvious that in the amplifier box you just add some smaller extra caps for each DC voltage , because of the long cables , not the ones that must be used for power supply ...
Any long cable must be decoupled because it is a small inductor and resistor ... regardless of how "good" the power supply is . Even when you use batteries
10-20uF for high voltage and 100-470uF for low voltage should be enough , of course for something industrial/proffesional could be calculated .
We are talking about a 1-1.5m long cable. It's not transmission line, nor the signal (which is -in good case- DC).
If you build it correctly (individual cold -ground- and hot wires for each supply voltage, each pair gently twisted, possible screening or shielding for each pair -if necessary-, good shielding for umbilical), the disturbances on amp socket side are almost imperceptible/immeasurable.
If you build it correctly (individual cold -ground- and hot wires for each supply voltage, each pair gently twisted, possible screening or shielding for each pair -if necessary-, good shielding for umbilical), the disturbances on amp socket side are almost imperceptible/immeasurable.
You want to use LM317 based regulator.
Why not read LM317 datasheet?
BTW in this site are many debate about LM317 output capacitance.
sample:
https://www.diyaudio.com/community/threads/lm317-load-capacitance.313648/
"Big capacitors can cause instability especially since their ESR is lower. The LM317 has a small inductance on it's output which can form a resonant circuit with the output capacitor if there is not enough ESR to damp the resonance."
"Linear Tech. recommends 1uF solid tantalum as an output bypass cap because of its high frequency esr performance. This is useful b/c the V regulators loop gain tapers off. Really big output caps cannot offer any performance compared to the devices negative feedback, plus they fail to deliver at extended frequencies when its needed most!"
Thanks euro21,
My umbilical is short, only 12 inches, it is made from 20mm gold contact 8 pin circular aircraft connector, and the wires in the umbilical are 20AWG silver plated teflon wire military grade wire. I will think of the umbilical as though it were a small inductor in the PI filter and add smallish 10-20 uf of DC link capacitance in the amp box. I dont want to add additional resistance as that would destroy the low impedance output of the Maida.
My umbilical consists of the following 8 pins in the connector:
1) Maida HT out positive (pin 1)
2) Maida HT out negative (pin 2)
3) Regulated 6.3V DC filament positive (pin 3)
4) Regulated 6.3V DC filament negative (pin 4)
5) 5V regulated positive (for relays, powered volume control, arduino, user controls, etc.) (pin 5)
6) 5V regulated negative (for relays, powered volume control, arduino, user controls, etc.) (pin 6)
7) Unused (pin 7)
8) Chassis safety ground wire to make the amp chassis safe to touch if you're standing in a bathtub 🙂 (center pin 8)
I'm swamping the output of the LM317 with an RC snubber 2.7R and 1uf electro to ground near the LM317 pin on the PCB. Tantalum capacitors dont go to high voltages.
My umbilical is short, only 12 inches, it is made from 20mm gold contact 8 pin circular aircraft connector, and the wires in the umbilical are 20AWG silver plated teflon wire military grade wire. I will think of the umbilical as though it were a small inductor in the PI filter and add smallish 10-20 uf of DC link capacitance in the amp box. I dont want to add additional resistance as that would destroy the low impedance output of the Maida.
My umbilical consists of the following 8 pins in the connector:
1) Maida HT out positive (pin 1)
2) Maida HT out negative (pin 2)
3) Regulated 6.3V DC filament positive (pin 3)
4) Regulated 6.3V DC filament negative (pin 4)
5) 5V regulated positive (for relays, powered volume control, arduino, user controls, etc.) (pin 5)
6) 5V regulated negative (for relays, powered volume control, arduino, user controls, etc.) (pin 6)
7) Unused (pin 7)
8) Chassis safety ground wire to make the amp chassis safe to touch if you're standing in a bathtub 🙂 (center pin 8)
I'm swamping the output of the LM317 with an RC snubber 2.7R and 1uf electro to ground near the LM317 pin on the PCB. Tantalum capacitors dont go to high voltages.
Last edited:
Any chance you could point me to an explanation of that calculation? (Just trying to understand.)
Likelihood of AC wiring next to DC voltages sourced by a low impedance supplies causing hum is very small to non-existent. I have a home-brewed two box preamp with an umbilical cable that sources AC heater and DC H.V. voltages and the preamp is dead silent. It would make sense to use twisted pair wiring for the AC heater voltages, though.
I have (old 211) amplifier, where inside, the power supply wires are longer than this. 🙂
Try the amp first without DC-Link. Use it for few weeks, then insert this capacitor to HV.
If the experience (or measuring) confirms this modification legitimacy, use this configuration.
My motto: what isn't broken, doesn't need to be fixed. 😊
BTW which type amplifier it is?