SSE 100Hz Hum headache

[cowneko]

Hi everybody,

I finally came back to finish my SSE (mainly UL/triode, CFB switch) and I'm facing a dreadful hum.

Here's my setup:

Speaker: Full range Beyma 12GA50 (8R, 102db) on plan

Mains: 230V AC 50hz (Europe)

PWR trafo: Hammond 374BX (wired for 230V)
OT trafos: Edcor CXSE25-8-5K
Choke: Hammond 193H (5H, 200mA, 65R)
Supplemental cap: Motor run cap 80Uf 440V
C1: 47uF 500V
C2: 120uF 500V
C11/C21: 0.22Uf 600V PIO Ky04
RK: Selector switch (330,360,430,470,560 and 680R values), grounded to R17 only (to avoid ground loop)

Tubes:
Rectifier: Shuguang 5AR4, Svetlana 5U4G / 5U3C / 5C3S, Svetlana 5U4G / 5U3C / 5C3S (black plates)
Driver: Shuguang 12AT7, Tungsram ECC81, Amperex E180CC/7062/12at7/6414
Output: EH KT88 (560R on selector), Reflector 6P3S-E (680R on selector)

Ground Scheme:

-IEC ground to amp metal top plate
-Input signal ground from one of the rca cold to amp metal top plate
-PCB grounded from the rca
-Speaker binding posts grounded to the pcb through cfb switch
-RK selector grounded to the pcb through R17 Ground side

Note: The PWR trafo is oriented 90° in respect to the OTs to avoid magnetic influences.

The hum begins after tubes warming up (more or less 10 sec from switching on) and disappear immediately at switching off (i can still hear music without hum until the caps are empty).
I tried with different output, rectifier and driver tubes, but the hum still there.
It is not influenced by Volume.
With input shorted, hum still there.
The same with supplemental cap removed.

I tried replacing the choke with a 150R 5W on R1 but the hum is even louder.

The Hum is weaker in triode mode (can be heard at 40cm from speaker) but really loud in UL/CFB mode (can be clearly heard at more than 2 meters from the speaker)


So I made some measurements with my scope, here they are:

Measurements made with shorted input, kt88 output tube (560R on selector), 5AR4 rectifier, Shuguang 12AT7 driver, choke and supplemental cap. The loads are 50W 8R resistors.




Triode mode


Triode/CFB



Ultra Linear



UL/CFB

The frequency is an average 0f 100Hz witch seems to be related to 50Hz mains frequency.

Do you have any idea of where to look for this hum to disappear?

 

[JonSnell Electronic]

Try increasing the bias voltage to reduce the current drawn by the output valve. I presume single ended output, if push pull, check the current balance.
100HZ hum is mains power supply related. Half wave bridge rectification with poor smoothing or the main smoothing ground is not in the correct place. Look up star earthing.

 

[spendorite]

Was it working without hum before you installed the two switches ? That is, in the basic triode config was there any hum ?
If not then maybe the cause could be incorrect wiring of the either of the switches. Try installing the switches one at a time.

 

[cowneko]

Yes there was Hum without any switches in triode mode, as well as in UL/CFB (but louder), in fact adding the switches hasn't changed anything.

The hum in triode I can live with (but I know it's here) but it's simply unusable/unbearable in UL/CFB.

I thought the problem was gone a while ago, but in reality I just had solved a bigger problem with a ground loop from the selector thanks to you. The hum is still there.

 

[spendorite]

Okay, Follow George Anderson's advice in the SSE assembly manual.......

""Start with the basic triode amp, get it all working correctly. Then add the choke, the supplemental power supply cap, the rectifier and standby switches (if desired). Then add the UL / triode switch and (or) the CFB switch. Any or all of these options can be added individually or in combination. I recommend making only one change at a time, and retesting the amp. This will uncover any problems before they get too hard to find.""


Forget the switches for now wire the amp in just Basic Triode config as per the assembly wiring diagram below.........

​

A basic Triode connected amp without feedback. A power supply choke is added. This is the most common configuration for a clean basic amplifier. If a choke is not desired, see the next diagram, and use the OPT wiring from this diagram.


Use the amp in this config and see if there is any hum.

 

[cowneko]

Ok Spendorite, I will go through each steps once again

So I removed the two switches and wired the amp in triode mode (not forgetting to ground the negative speaker binding posts to star ground).

The OTs outputs are linked this way: Blue wire (8R) is on the positive speaker binding post (the red one) black wire (common) on negative speaker binding post (black)

I get exactly the same result as with the switches (same hum and same Scope measure)

 

[spendorite]

Check your grounding continuity. From the IEC ground pin (not the star ground point) to all the points that should be grounded.
The pcb, both RCA jacks, the speaker ground jacks, all the transformers and choke and aux cap if you have that connected.
This is to ensure that you have a secure ground connection from all these points to the IEC ground pin.

If all of these are okay then maybe you should try some shielded input wire. When I built my first SSE a few years ago I had a
similar hum and after I replaced my unshielded wire with shielded the hum was eliminated.

 

[spendorite]

Another thing you should try if all your wiring checks out okay is recheck your board
to see if all soldered points and corresponding connections are okay.

 

[cowneko]

I checked grounding continuity (from IEC pin to everything that should be grounded), everything is fine, except the supplemental cap housing, so i made a connection to ground for it, but no differences.

I already tried substituting the non isolated input wire for isolated ones, but it has not shown any effect on the hum, anyway, I can try by just shorting the inputs directly on pcb terminal posts, I'll know right away if the input wires are picking something.

When you ask for checking the board, do you mean searching for cold solder joints?

 

[spendorite]

Actually for anything that might be wrong whether cold joints,
solder bridges, solder connections to the ground plane on the board, etc.
Also verify again that the board is only grounded via one of the RCA grounds.
I assume that your RCA and speaker jacks are isolated from the chassis plate.

 

[spendorite]

If you have the cathode resistor selector engaged try disconnecting it leaving only the on board resistor and see if this helps.

 

[cowneko]

I already thought about that, and the selector is removed, I'm in the process of resoldering all solder points.
my rca and binding posts are indeed isolated from chassis and only one rca cold is linked to star point.

 

[spendorite]

The OTs outputs are linked this way: Blue wire (8R) is on the positive speaker binding post (the red one) black wire (common) on negative speaker binding post (black)

So on your OPT's the secondaries are Blue for 8 ohms and Black for common and this is what it says on the label.
I ask because I have the same OPT's, Edcor CXSE25-8-5K where the secondaries are Yellow for 8 ohms and White for common.

 

[cowneko]

Yes I know, but I contacted Edcor concerning the OTs secondary colors, they said they changed the color code on new production and that now blue is 8R and black common.
However, their label still shows white and yellow secondaries... go figure

 

[PDL]

Are your 5v and 6.3v heater taps from the power transformer twisted? I had some mains hum that was drastically reduced by twisting these feeds going into the board.
A picture of your amplifier's internal wiring would help others help you.

 

[cowneko]

PDL:

All the PWr trafo secondaries are twisted (maybe not tightly enough), and you are right about the picture of my amp, so here it is:

An externally hosted image should be here but it was not working when we last tested it.

Note the big cable mess between the binding posts are the triode/ul and CFB switches witch are disconnected for the time being. The same goes to the RK selector.

Spendorite:

I reflowed every solder pads on the board (with flux and a bit more solder) and disconnected the selector. As my board was removed from the case, I took the opportunity to install terminal posts for R17/27 witch will make res rolling easier.
I cleaned the board with Isopropylic alcohol.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

I tested the amp....... no progress, hum still there

 

[cowneko]

hmmm strange thing, r17 and r27 are running hot to the point of smelling and fuming, seems like the tubes are sucking too much current.....don't know what's going on

I'm done for today, going to sleep

 

[cowneko]

measurement taken for B+ and VK:

note: the supplemental cap is disconnected.

B+ (R4 to ground) = 439V DC
VK through R17 = ??? the value is in the mV domain and fluctuate quite a lot, but always lower than 1V dc........big trouble here

all tubes seems to glow normally, I ain't wait for a long time since r17/r27 became quite burning hot after 20 to 30 seconds.
I measured r17 and r27, they are indeed 560R res (rated at 5w)

 

[cowneko]

Rahhhhhh!!!! I'm the dumbest idiot ever!

By mistake I put a 12ax7 in place of the 12at7 hence the problem....

I put back a 12at7 and all is normal again:

B+ = 440V dc
RK = 560R
VK = 40.6Vdc

Dissipation = 28.95W, perfect!

 

[spendorite]

I put back a 12at7 and all is normal again:

So the hum is gone then ?