silly ground question

[77seriesIII]

Calling it a night and I just finished my first amp in a metal box...usually use wood.

I fired it up and I got a horrendous hum noise - no volume pot as it will be used w/ a pre-amp but am currently controlling volume w/ an ipod which is not plugged in. Hum took a while to kick in everything had to pass the rectifier, when I pulled the rectifier, no hum. System is a DIY Tube ST 35 rev D w/ a rectifier.

I am going to start w/ a ground plane, after I check connections just to make sure.

On a metal chassis do you isolate the speaker outputs so they do not touch the chassis, this goes for the rca inputs as well - correct?

as long as everything is connected to the chassis I do not need to ground it, correct? Well, other than the one ground. the chassis is a Hammond powder coated thinking I may have to sand away some of the pwdr to get a good connection on some of the poitns - if all other aspects dont work.

too tired to play any more w/ electricity...heading to that flat thing w/ pillows.

./e

[TomMcNally]

Erick -

The "hum"may not be hum at all ... but feedback.
As a test, disconnect the feedback input that comes
back from the speaker jacks. If the noise goes away,
your feedback may be backwards. All you have to do
is reverse the output transformer primary leads from
end to end (including screen lead) common problem !

As far as grounding, the speaker negative doesn't have
to be grounded to the chassis, BUT it does need a path
back to the board ground for the feedback to work right.

Typically, most people ground the third wire of the power cord
to the chassis, then ground at the RCA jacks ONLY.

... tom

[77seriesIII]

Tom,

Thanks just removed the negative feedback but it didnt work, ugh.

Need to recheck connections, ground and see what happens.

Thanks for the quick reply.

./e

[TomMcNally]

Hmmmm .... or is it Hummmm .... a couple of things:

if you have no input connected the amp's input is wide open,
it's better to have an input connected (even if off) or shorting
plugs.

Which power transformer did you use? I am assuming you followed
the info under "docs and mods" on tube rectifier in the ST-35 section
of the forum ? Make sure you counted the pins right on the rectifier
tube socket, make sure the high voltage center tap is grounded, and
the 5 volt filament winding tap is taped off.

Yep ... check everything and you'll find it.

... yom

[azazello]

Hummmmm is from, IMO, +B - not good filtering. You can connect
a chock 2 - 5 Hn with parallel non polar cap 1-2 uF, in PSU, instead some low resistor between 2 electrolitic caps.

[77seriesIII]

Tom,

I'm using a trafo toroid power supply. I may pull the board out and put one in that I have bench tested. I thought I had bench tested the 3 of them but maybe only hit two...age and fatigue.

Did a quick test removed the input RCA jacks from the chassis an just wired them in w/o chassis connection. I loaded the inputs w/ an ipod...no joy big hum.

So much for the easy, I'll have to go through each of the connections, maybe remove them all and start over...hence the thought of dropping in a known good.

I am using a choke, standard size H wise for the system. its a wound trafo choke, very much overkill but its for my listening pleasure. I'll have to take at my B+. I knew I should have bought the 100w 8ohm bricks on my last mouser order...instead of using speakers.

tomorrow to the ohm cave!

./e

[EWBrown]

I'd look for any multiple or extra board to chassis connections first.

Next is to determine if you are using a Rev C or Rev D board, as there are some minor differences, mostly concerning the grounding scheme.

Note that I will be initially referring to a Rev C DIY ST35 board for this, the Rev D ground mounting screws are all electrically isolated. The Rev C board has all six mounting holes connected to the ground plane

With the DIY ST35, the PC Board contains the ground plane for the amp circuitry, and there should be only the connections from the (Rev C only) board's ground plane through the metal mounting screws and standoffs to the chassis' ground, and I connect the third wire safety ground (green or green/yellow) directly to the chassis, I just use one of the power transformer mounting screws and a crimp-on ring lug to make this connection.

If you are using a two-wire power cord, then there will be no "safety" ground conn ection - which is how the original 1960's were wired.

Since your board is the newer Rev D, the six mounting locations are electrically isolated from the PC board, so the SINGLE PC board to chassis ground should be made through J1, Pin 6 (called the "spare ground" in the Rev D Schematic).

The board itself has a very effective grounding scheme, and there should be no hum at all - I've built several of these from the rev A "green board" through rev D, with no hum or noise issues.

The RCA connectors should be the isolated (nylon shoulder washers) type, and I locate them as close to the input blocks (J6 and J7) as possible, in order to keep the wiring short and direct, and by doing this, shielded cable is not necessary.

If the RCAs are mounted on the rear of the chassis, then shielded cable should be used, and routed away from any AC (line, HV or filament) wiring. The speaker connectiions have the "zero" side grounded through the PC board, so they should also be isolated, which is the normal state with the "five-way" type of binding posts .

Since you are using a tube rectifier, the 5V winding connects to the rectifier tube filament pins 2 and 8 (in most cases like 5AR4/GZ34, 5U4GB, etc).

The HV DC voltage should be taken from the rectifier pin 8, and not from pin 2 or the 5V winding's center tap (if it has one) as these other connections will only serve to inject an unwanted 5VAC or 2.5VAC 60 (or 50) Hz hum / ripple onto the B+, which would create a "monster" hum stuation.

Last, but not least, , re-check all of your soldering, even one bad solder connection can cause all sorts of troubleshooting headaches.

As often stated on the Bottlehead forums, bad solder connections account for approximately 90% of amplifier problems, and wiring or component mistakes account for the other 10%



/ed B in NC

[dcgillespie]

Is the hum in both channels? That can be a clue. An even bigger clue is if you can identify whether it is 60 Hz hum (like and ungrounded input), or 120 Hz hum, which is an octave higher. The latter will definitely be power supply B+ oriented. Also, check that the heaters are properly referencing ground, as that will also introduce 60 Hz hum problems if not.

Dave

[ChrisAlbertson]

...
as long as everything is connected to the chassis I do not need to ground it, correct? ...e

You signal ground needs to be different from the chassis ground. When you say "everything", Ican'tthink of to many things you'd want grouned to the chassis, just the iron frames of the transformers and the third ground wire in the AC cord.
I hope you are not using the chassis for any kind of return curent.

[77seriesIII]

Ed, et. al;

Thanks!

no the neutral wire is not hooked to the chassis...that would hurt.

I need to re-read the posts just to make sure I got everything. Ed agree completely w/ the ground plane assessment of the board, the previous two I built and tested fired up w/ almost no issues on a slab of wood. On those I consistently cannot reach 350mv on the bias...I digress. Anyway, previous systems no hum and tested great (past two weeks). I thought I had tested this board so it is mostly likely something in what has been written above.

going to re-read, check and post.

./e

[77seriesIII]

ok just ohmed everything out and it is close enough to published specs that it should be right. Went so far as remove the 3rd wire ground, just to see. didnt work. I removed a channel's tubes and only had hum (hum is a bad word - LOUD reverberating speaker destroying terror is much better but I'll continue to call it hum) from the channel that had OP tubes. Both channels have hum otherwise.

next step disconnect the wires and re-connect.

I did order an 8 ohm 50w resistor for testing purposes, figure about 2 weeks to show...hopefully (stationed overseas w/ an APO address + Holiday season = molasses uphill in January USPS mail).

on to disconnect completely followed by a pause and a reconnect.

./e

[77seriesIII]

question.

To test the B+ voltages on this system:

Ignore the noise and test with clip attached DVM?

Disconnect all but the B+ from the OPTs? I think that may be bad.

./e

[TomMcNally]

Erick -

I am still thinking that's oscillation, not hum.
Unless you have a serious wiring error, hum
would be low, not loud.

I think the transformers are out of phase.

[dcgillespie]

Erick -- I agree with Tom. Hum is one thing. But loud speaker destroying noise is another. It does sound like the negative feedback has become positive. Check the color coding of the OPT primary wiring very carefully. Another possibility is that the plate leads may be correctly connected, but the screen leads reversed, or visa-versa. Either conditions will cause the oscillation -- and could do so even with the main feedback connection disconnected. Remember, there is screen feedback provided on the board that a misconnected output transformer could upset to become positive.

I don't believe you've indicated which OPTs you are using. If you are not using Z-565 transformers, a different brand may employ a different color code with regard to phase. That could be a possibility too.

Finally, what happens when you remove the input tubes and only have the output tubes installed? I would surely hope the noise goes away then!

Dave

[77seriesIII]

I am also using toroid trafomatics

let me go do a quick test. first w/o input tubes and I'll see if I have time tonight to change out the OPTs to tested true ones.

I am now using a known good board. I bread boarded this board and tubes last week using a trafo power supply and two trafo OPTs, all toroidals. A few weeks I confirmed from Boris the what each of the wires coming out of the OPTs.

actually having fun, a bit frustrating having to figure something out w/o knowing all that I should but still fun none-the-less.

ok w/o input tubes I still get defeaning noise. Too late tonight to change out the OPTs w/ known good ones, but I'll do that when I get home tomorrow night and report back.

Is it common for toroidal OPTs to go out of phase or is just one of those things that happens?

Many thanks

./e