Selenium Bridge Rectifier On Geloso Tube Amp

[Phoenix]

Here's a paper from Evox Rifa CAPACITORS FOR RFI SUPPRESSION OF THE AC LINE: BASIC FACTS

Also note that not all Class Y caps are self-healing, and self-healing does not = Class Y.
Class X and Y caps are designed to survive large transient voltages (many kilovolts) without catching on fire or exploding, and Class Y are designed to fail-safe where there's a risk of electric shock/electrocution (line-to-ground).

CDE (Cornell Dubilier), manufacturer of Orange Drops, have lines of Class X safety-rated capacitors (but do not manufacture Class Y), and Orange Drops are not among them.

[Rob Strand]

The *motivation* for putting caps on mains side of transformer power supplies was often to reduce pops when the equipment was turned off.   *Many* schemes were used.   Some schemes probably shouldn't be used.   If you dig around enough old audio equipment you will come across them.   Guitar amps rarely if even use them.   

When used the across the transformer either X or Y caps could be used (provided they are used with the corresponding X or Y connections).  It's probably safer to use X caps but without an inductor (and virtually none of the old pieces of equipment used inductors) you get an inrush which isn't great for the life of the switch, so some equipment had resistors in series with the X caps.


Other than a turn-off pop you aren't going to see any problems pulling those Y caps.


Actually you can see a de-pop cap in this schematic from one of the current threads,

https://postlmg.cc/5H985ms9

[amptramp]

Some people in the audiophile community use capacitors across the highest secondary voltage winding of a power transformer to avoid the ringing when semiconductor rectifiers go below the voltage where they conduct.  Power transformers tend to ring at about 100 KHz once the load is removed.  Bridge rectifiers offer more time in the cycle for ringing.  Those with tube rectifiers do not have this problem as there is no voltage at which they do not conduct.  This may make more sense than putting capacitors across the primary and it has the advantage that it does not conduct to ground if it fails.

[willienillie]

Some people in the audiophile community use capacitors across the highest secondary voltage winding of a power transformer to avoid the ringing when semiconductor rectifiers go below the voltage where they conduct.

Would that be the same as the two .22µF caps in some earlier Marshall amps?



I've never really known what those were for, but of course assumed some kind of noise reduction.  I know the amps work fine without them.

[amptramp]

Some people in the audiophile community use capacitors across the highest secondary voltage winding of a power transformer to avoid the ringing when semiconductor rectifiers go below the voltage where they conduct.

Would that be the same as the two .22µF caps in some earlier Marshall amps?



I've never really known what those were for, but of course assumed some kind of noise reduction.  I know the amps work fine without them.

Yes.  There is no functional difference as far as the user is concerned except designers do not want high frequency noise floating around where it could radiate into the preamp stages and possibly affect bias even if it does not always have a noticeable effect (and the effect would take place at twice the line frequency so it would be like hum from an inadequate B+ capacitor).  The use of such large caps is intended to swamp the oscillation.  In some cases where there is not enough resistance in the secondary winding, the caps are used in series with a resistor.

[Rob Strand]

There is no functional difference as far as the user is concerned except designers do not want high frequency noise floating around where it could radiate into the preamp stages and possibly affect bias even if it does not always have a noticeable effect.

Added secondary caps can also reduce buzz from acoustically noisy transformers.   The winding current relates to the force on the transformer.  When the diodes snap off it creates a sharp change in current and a sharp change in force.  That pings the transformer.   The secondary caps slow down the change in current which reduces the level of the acoustically produced harmonics.   I've played around tweaking the cap to shut transformers up.

Because of leakage primary caps and secondary caps aren't 100% the same.


     

[sarakisof]

The *motivation* for putting caps on mains side of transformer power supplies was often to reduce pops when the equipment was turned off.   *Many* schemes were used.   Some schemes probably shouldn't be used.   If you dig around enough old audio equipment you will come across them.   Guitar amps rarely if even use them.   

When used the across the transformer either X or Y caps could be used (provided they are used with the corresponding X or Y connections).  It's probably safer to use X caps but without an inductor (and virtually none of the old pieces of equipment used inductors) you get an inrush which isn't great for the life of the switch, so some equipment had resistors in series with the X caps.


Other than a turn-off pop you aren't going to see any problems pulling those Y caps.


Actually you can see a de-pop cap in this schematic from one of the current threads,

https://postlmg.cc/5H985ms9

So, *after having replaced 2 with 3 prong*, could i use one (instead of two originally used) Y cap from neutral (or hot) to gnd for de-pop purposes?

PS. In all my tube equipment, i always replace 2 prong cable with 3 grounded AND replace death/safety/mains caps with appropriate modern X or Ys.

But now i was told that "Including them will only increase earth leakage current (under 400uA on 240V 50Hz for a 4n7 class Y cap, but something to be avoided unless it is necessary)."

I guess de-pop purposes are important enough to call "necessary"?

[Phoenix]

The *motivation* for putting caps on mains side of transformer power supplies was often to reduce pops when the equipment was turned off.   *Many* schemes were used.   Some schemes probably shouldn't be used.   If you dig around enough old audio equipment you will come across them.   Guitar amps rarely if even use them.   

When used the across the transformer either X or Y caps could be used (provided they are used with the corresponding X or Y connections).  It's probably safer to use X caps but without an inductor (and virtually none of the old pieces of equipment used inductors) you get an inrush which isn't great for the life of the switch, so some equipment had resistors in series with the X caps.


Other than a turn-off pop you aren't going to see any problems pulling those Y caps.


Actually you can see a de-pop cap in this schematic from one of the current threads,

https://postlmg.cc/5H985ms9

So, *after having replaced 2 with 3 prong*, could i use one (instead of two originally used) Y cap from neutral (or hot) to gnd for de-pop purposes?

PS. In all my tube equipment, i always replace 2 prong cable with 3 grounded AND replace death/safety/mains caps with appropriate modern X or Ys.

But now i was told that "Including them will only increase earth leakage current (under 400uA on 240V 50Hz for a 4n7 class Y cap, but something to be avoided unless it is necessary)."

I guess de-pop purposes are important enough to call "necessary"?

You can use a Class X directly across the primary (place it after the switch and fuse, or include a resistor in parallel with it to ensure it discharges and the unplugged cord can't give you a zap), then it doesn't add to earth leakage current, and you don't have restrictions on max capacitor value as a result, so you can choose the best fit.

[sarakisof]

Phoenix are you sure i will have earth leakage current if i use both Y2s and 3 prong and how is this bad, how i "understand" this issue in fact? I mean i always use both Y2s and 3 prongs in all my tube equipment and i have never realise any issue, instant "shock" from chassis or whatever. What am i expect to see?

[Phoenix]

Phoenix are you sure i will have earth leakage current if i use both Y2s and 3 prong and how is this bad, how i "understand" this issue in fact? I mean i always use both Y2s and 3 prongs in all my tube equipment and i have never realise any issue. What am i expect to see?

Yes, any line-to-ground capacitor is an AC path through protective earth, so you get earth leakage current dependent on the reactance Xc=1/(2πfC).
Earth leakage current should always be kept to a minimum (there are maximum allowable Class Y capacitor values/max earth leakage current for different device classes for this reason) or avoided entirely when possible, as it can increase nuisance tripping of earth-fault devices (RCD/RCBO/GFI/GFCI/ELCB), or in some cases decrease their sensitivity (if the leakage is DC and the device is type AC not type A), preventing them from detecting real faults. This might encourage some foolish people to bypass safety devices or remove ground pins.
Also increases likelyhood of experiencing ground-loop noise by causing a voltage drop on the protective earth conductor, again motivation for foolish people to bypass safety devices/remove ground pins.

In this case you can achieve your aim (damping transformer inductance to minimze turn-on/off pops) without the need to introduce earth leakage current, so you should do so.

Whether or not you can get a shock from an unplugged power cord depends on the exact circuit details, ie before or after switch/fuse, and of course exactly what the AC phase is at the time the device is turned off/unplugged, needs to be near peak to get much of a zap.
All inline RFI filters have a discharge resistor built-in for this purpose, check the schematics in this datasheet (randomly selected, first that came to hand).

EDIT: Oh, and your anti-pop cap will be more controllable/effective cross-line rather than line-to-ground anyway as you're not adding the variable of earth-bond impedance and its huge current loop to the mix.

[sarakisof]

Great info once again Phoenix. So, a 4.7nF X2/Y2 (old school safety caps with long leads tend to get disappeared on Mouser, only modern type short leads, so this was the only one found) like this https://gr.mouser.com/ProductDetail/Panasonic/ECQ-U2A472ML?qs=vggogeubx2N5FI9VrUv4ig%3D%3D
across mains after the fuse will be ok? No need for discharging resistor right?

[Phoenix]

Great info once again Phoenix. So, a 4.7nF X2/Y2 (old school safety caps with long leads tend to get disappeared on Mouser, only modern type short leads, so this was the only one found) like this https://gr.mouser.com/ProductDetail/Panasonic/ECQ-U2A472ML?qs=vggogeubx2N5FI9VrUv4ig%3D%3D
across mains after the fuse will be ok? No need for discharging resistor right?

Yep, would actually be preferable to wire it directly to the primary, after the voltage selector switch. Then its value can be chosen for best performance regardless of voltage selector setting, it can be optimised for that winding, and transformer action keeps it working if the selector is set to another voltage. Some experimenting with capacitor value may be required to find the most effective one.
And yes, no discharge resistor needed if it's placed after switch and fuse as the transformer primary provides a discharge path.

[anotherjim]

There was a fashion for IEC sockets with a built-in filter. It was very easy to get an "aftershock" from the socket pins from these and of course, it isn't isolated by the on/off switch.

[amptramp]

There are EMI filters designed specifically for medical equipment that have no capacitor to ground connection.  This is to avoid having your EKG machine feeding line spikes directly into your heart.  They cost a lot because part of the cost has to be allocated to a legal liability fund but also because most other users do not have a problem with connections to chassis ground so the market is limited.

[sarakisof]

It turned to a great discussion thread guys. 

Yeah, upgrade your rectifier, the 280Vrms rating is too low even before safety margin. Should be rated for the 290Vrms secondary (410Vpk) + 10% for mains variation (450Vpk) + an allowance for transformer regulation (I'd allow +20% - so 540Vpk). 1N4007 or UF4007 are rated 1000V 1A, and are only a few cents each, or you can use a bridge rectifier package, 600V rating is common.

About selenium rectifier replacement -thread's first question- i will finally go with this full wave bridge package rectifier  and will screw mount it in one of the two the already existing original selenium's holes.

https://gr.mouser.com/ProductDetail/750-KBPC1510-G

[sarakisof]

Components arrived. Just soldered the X2 4.7nF across mains. Is it normal that i get continuity in  multimeter beep test across mains 🙄

[Phoenix]

Is it normal that i get continuity in beep multimeter test across mains??? 🙄

If the power switch is on and the fuse is installed, then yes, you'll be getting the DC resistance of the power transformer primary.

[sarakisof]

Ok but i get about 16 ohms in there (across primary winding 220/240 and -10/0/10) is this normal? 🙄

[sarakisof]

Ok it's like i forgot basics https://youtu.be/7xvTGd4ysRE

[Phoenix]

Ok but i get about 16 ohms in there (across primary winding 220/240 and -10/0/10) is this normal? 🙄

Quite normal. Here's a datasheet for a vaguely similar power transformer from Hammond. Note the DCR of the 240V primary = 5.059ohms.
You'll also be reading the series resistance of the switch(es) which is non-zero, and the fuse which is typically a few ohms, so very normal.
The DC resistance of a transformer winding tells us very little except its short-circuit current if another winding is shorted, even then it's not the whole story as you need to bring leakage inductance into the equation.