Hum issues on transformer output preamp

[bob orbell]

Andy, I may have said before but, do the cables from the transformers to the output sockets connect to ground, the output should float, IE, "NO" connection to ground, that is as far as my thinking goes. BOB

[bikerhifinut]

yes Bob, completely isolated outputs floating and meter checks that out.
The Fault lies either in the internal circuitry or is being somehow coupled directly into the output transformers themselves.
There's a heck of a lot of gain in the 2 stage amp and any slight amount of hum would be boosted accordingly but would be attenuated in the output transformer along with the rest of the signal.
I havent screened the wiring to and from the Volume control, but again that shouldn't make a blind bit of difference at line levels. I've never really needed screened wiring and the hum disappears as soon as the connection to the PSU and I assume mains earth is unplugged, this happens whether the PSU is switched on or off, i.e the Preamp is switched off but still connected up. In fact the hum increases when the pre is turned off.
Disconnecting the remote control and bypassing the input relays to make a direct hard wired connection to the volume control has no impact either. This because its an after market job off ebay and that set of boards hasn't given me problems in the past anyway.
Which boils it down to earth loops as far as I can see.
I tried a chassis earth lift inside the case, an adapted HD3 case used for one of the early prototypes and bought off Nigel a couple of years ago so I am following the Pre 4 layout. That made no difference either and I didnt expect it to. I tried moving the signal earth from the input sockets from the earthy end of the Volume control to the star earth point on the HT decoupling capacitor, this is a 60 +250uF job and I used it as an extra CRC filter on top of the already ripple free regulated Ht supply. There are no reasons to suspect that.
The PCB does seem to have a lot of earthing points on it but I followed the instructions on this anyway, If there is a loop there it's designed in and surely would have been spotted in the development stage.
The transformer screens are all earthed as per the diagrams, and the transformer frames make a good chassis contact. Primary and secondary windings measure correctly and are isolated from each other and the frames.
It leaves me with only 1 option which is to use some screened wire from input to pot and down to the board. I've got a length of the blue stuff that was supplied with my KT88 amp and loadsa other high quality 2 core screened microphone cable which is what the blue stuff is. If that's the issue then sorted but if thats the case, then those transformers must be Radiating an induced field as if they were picking it up and coupling into the secondary, the truns ratio which must be someweher between 100:1 and 200:1 assuming a gain between 10 and 15 on each ECC82 stage would push any induced noise right down into the background.
Obviously the kits must work and work noiselessly or there'd be a rash of threads on here to that effect.
I bought the the boards and transformers from a member here, I forget who now and they were all brand new unused so we can eliminate damage I would think.
I'm Baffled.
If I don't get to the bottom of it soon, I'm Scrapping the transformer circuit and reverting to the known high performing circuits, basically a decent CF, probably based on the Broskie active load designs as theres room for the extra valve, with a low gain front end based on something like a 6CG7 or 6SN7 which I have in the parts bin.
I think a complete rework is in order with attention to earth runs and if that fails, then there is something awry else where.


Andy.

[bikerhifinut]

I think I sorted it.
However i departed from the WD build instructions.

I knew it had something to do with Earthing and signal ground and there was a hum loop of some sort.
Now you would think that the "floating" output from the transformers would isolate this problem and that somehow the hum was induced elsewhere.
I noticed the hum got really bad when the PSU was connected and even worse when the PSU was plugged in to the mains but not switched on. This had been baffling me for weeks and in truth still does.
Anyway I got curious and thought I would try a short wire link from the outers of the output RCA's i.e. the CT of the output transformer to signal ground at the star earth point. Nothing to lose.
Well all the nasty 50hz hums disappeared and its as quiet in the background as you'd like it to be.
So I wonder why this is and why as a standard build, following the WD build instructions, the preamp was a nasty noisy thing?
It's fine fixing something, but if I don't really know the reasons why then I will probably go out and make the same errors again as for all I know I am merely masking an existing issue?

A.

[bob orbell]

Well this is strange Andy, but as you say in your PM to me, it is possible that the double insulated power amp is responsible. Some time back I looked at a Pre III and Phono III for one of our members, he had hum issues when connected to a KAT6550 power amp and he had made this earth connection that you have just made, on me removing it silence prevailed, any way, good luck with the building work. BOB

[bikerhifinut]

yeah bob,
I think I might need to wire in a small earth lifting switch at the back of the unit.
That will cover all options.

We srated pulling stuff apart today and had first inspection off the council.
The builder spotted something in the existing build now its exposed that the structural engineer needs to advise on, it's potentially a time and money saver as we think the existing structure is beefy enough to not need one of the two steel beams that were originally calculated for, but obviously we need the engineers input on this.
And now I find since we changed our internet hub that the epson wireless printer refuses to work either on wi fi or old fashioned USB wires. I so hate computers, they seem to get more flaky not less as they advance.

[John Caswell]

Just reading through this, Pre3/4 transformer secondary is completely isolated from ground, and would normally have the screen of the cable grounded via the input of the following amplifier. This has the effect of not producing any circulating current "hum loops".

John

[bikerhifinut]

Thanks John.
I understand that, and I know I shouldn't get any hum with the secondary "Floating".
This is why I am baffled as to why ALL my amps show a noticeable hum, some more than others, whether double insulated i.e. no mains ground or not.
I can only think there's a problem inside the preamp that up to now has eluded me.
For whatever reason Grounding the CT of the opt transformer completely eliminates hum on all my power amps.
I shall have a good look once the builder is finished and I get the new workroom set up. Methinks the scope may throw up some pointers.

A.

[bikerhifinut]

Quick update.
I had half an hour today so spent it trying the Pre3/4 clone out with a recently accquired cosmetically very tatty but good Quad 405-2 into the 989s which are currently stashed in the conservatory while the build progresses.
Output Grounded and zero hum, and then I disconnected the ground from the secondaries and still no Hum.
From this I deduce that the preamp is, in fact, working correctly and that any hum issues I have with other amplifiers are entirely related to an aspect of their circuitry, most likely the way their signal earths are running. All are either double insulated, i.e floating, or the signal earths are lifted above chassis/mains earth.
The Quad, as far as I can ascertain has it's signal earth at the same potential as chassis/mains earth and this may be a clue?
Another thing I am wondering is that I suspected the ring main upstairs wasn't quite right and as we are in the process of getting the house wiring moved a bit and all new mains circuits upstairs I am hopeful any potential issues will be eliminated. Stuart would be pleased to note that my Sparky doesn't like the British practice of ring mains to "Share" the load and as I wanted radial circuits anyway for the new upstairs rooms he's a happy bunny. As it happens I'll be getting a dedicated spur off the secondary Consumer unit but thats really only because it's convenient as the CU will be situated in the alcove where the amps etc will be situated. I'm not convinced that special mains circuits make any audible difference to the sound system.
As an aside I'm mightily impressed with the 405-2 and I can see why its well thought of. It's going to get a repaint and I'll check the internals out. But I won't fix what's not broken even though there seem to be various "upgrades" on the market.

Andy.

[Richard]

It could be hum on the signal line Andy which would be there floating output or not.
Quad 405-2 had a curious variety of earthing arrangements due I think to Quad changing ranges in the mid 80's and the way their units were linked together. The din version has inputs and outputs at chassis mains earth following the earlier 33/303 system of 2 core linking mains cable and earth via the din cable. The later RCA input 405 had the outputs still to chassis earth but the inputs isolated as the new range 34/306 had 3 core mains connections between units. If you check an RCA version the inputs show 10R to chassis but this isn't the usual situation of floating the whole circuit inside the earthed case, instead that 10R is R2 (which has always been there in 405) connecting the input and output ground tracks on the pcb, and now showing as 10R between isolated inputs and grounded output via the pcb ground track.

[bikerhifinut]

Update.

I have had all sorts of issues with this thing.

Anyway I had a last go at it today and tidied up a couple of earth runs, not expecting anything much, and also altered the HT capacitor inside, basically paralleled it op to make a single 300uF job which is overkill but I had it anyway.

It's also got far too much gain to be any real use with my power amps, the least sensitiveis the Mitsubishi power amp with a 1V sensitivity so for my needs a x2 gain is plenty.


Anyway its got a 620k with a 470k gridleak on the input of stage2 which theoretically should get the gain right down but the volume control still never gets past 7 to 8 o clock.
So I swapped out the existing alps motorised pot for one that I could shunt modify as the latest versions have strange wiring lugs that include common ground tags. The RK 127 is still available for now but for this exercise my old rk116 would have to do and as a shunt mod its shortcomings are much lessened.
That did e the trick with 120k shunt on a 50k pot.

Seems a waste of gain all the same but its all running nice now.


I haven't really compared it to the Aikido ACF cahtode follower for long enough yet, both have their strong points and if I think the cathode follower has the legs then I will rebuild the lot to include the phono stage too.

Andy