Troubleshooting massive voltage loss in an "unproven" build

[SolidStateCircuitReject]

Hi Folks.

First official post (except introduction a few weeks ago) - so "hello from Belgium".

If anyone is up to the challenge (or maybe not a challenge?) I'm working on completing a build based of of a few different amps. Something of a Hot Rod Deluxe clean channel only - no ICs, transistors and such. As implied as possible.

The amp makes "sound" albeit insanely distorted and then within some 10 seconds the power tubes start to redplate - so something is screwed in the bias or power section. I have attached a pdf of my hand drawn schematic of the output and power sections, hope I didn't forget to add anything. Here is the info I know so far:

Voltage passes through the primary to the secondary okay. At the transformer side of the standby switch I am reading some 340 VAC, so the primary is transforming 230 VAC into 340 VAC - so far so good. However as soon as I flip the bypass switch, that same reading drops to about 150 VAC, so more or less gets cut in half. Reading at the bias feed is actually in the positive millivolts, obviously tubes when and if installed will redplate rather quickly.

I have rebuilt the bias section twice, even changing the circuits from a fender style bias supply to the now marshall style bias supply all with the same results... makes me feel that the bias supply section is fine and that there is an issue in the filtering or diode bridge sections but I am at a bit of a loss.

If anyone has an idea or seen something like this before, please let me know... not sure where to go from here.

Much obliged.

SSCR

[John]

Your schematic shows the Hammond 290 UX. The datasheet shows that your bias supply should be taken from the 2 brown wires. IF I'm reading it right, your bias supply should be wired with the brown-brown just like your power supply is with the red-red.  The way you have it wired now is, I think, wrong which is why you're having issues. Don't take this as gospel though, since I almost always use cathode bias. :)


http://www.hammondmfg.com/pdf/EDB290UX.pdf


See if that link doesn't answer some of your questions.

[John]

To follow up: you do have your bias diode oriented correctly in your build?

[SolidStateCircuitReject]

Thanks, John!

Let's see... I am actually using Hot Rod original trannies, I included Hammond in my schematic because they are exact replacements and I actually plan on getting this so I can restore the Fender ones to the HRD from which I harvested. 

And yes, there is a set of brown, I think 100 watt, bias tap that needs to be closed. For the HRD it supplies bias and the channel switching matrix. at first I was using this along with a Fender type bias circuit, but there was redplating, after that I switched to a Marshall style bias that taps off of the secondary AC.

I will check that diode for the right alignment, but having a few dozen builds under my belt, all marshall clones, I know it should be facing the tranny to let neg. DC pass... I will triple check that. Otherwise perhaps I will wire it up Fender style again to double check that using the dedicated bias tap.

Thanks!

[sluckey]

That bias circuit WILL NOT WORK with a bridge rectifier. You must use a cap coupled circuit like the JCM-900 amp. See page 5 of this pdf...

     http://sluckeyamps.com/misc/Amp_Scrapbook.pdf

However, since you have a separate bias winding on that PT you can use your bias circuit with that bias winding. Connect one of the brown leads to ground and connect the other to the 220K resistor that feeds the bias diode. That 220K resistor will likely have to be less than 1K to get the proper range of negative bias voltage. For 6L6s I would want the bias voltage to be adjustable from about -35Vdc to about -55Vdc.

***CAUTION*** You need to move the bias feed to the junction of the 56K and 15K resistors. As drawn you can adjust the bias down to zero volts and that's a bad thing.

I recommend using the bias winding rather than the JCM-900 circuit. Connect it as described above and test for proper negative voltage at pin 5 of the output tubes. DON'T plug in any tubes until this negative voltage issue is fixed!

[SolidStateCircuitReject]

my error - it was at the corner of 56 and 15.

Checking out the scrapbook now and will report back later.

Thanks!!!!!

[SolidStateCircuitReject]

BINGO, RINGO!

Thanks A LOT,  Sluckey!! I am 90% there, basic tone is there, few things to clean up... also the bias circuit is feeding a bit too much negative voltage (do i have that backwards?) - I have to crank the bias pot to get the bias in the zone. With B+ at 455V, bias pot fully cranked I can achieve +/- .40mA where my target should be in the high 40's (49ish mA on the "hotter" side).

It's painfully obvious that after literally months of semi-hardcore learning, studying, and using this knowledge in practice - not to mention a few dozen builds over the span of several years, there is a lack in knowledge of some of the fundamentals!

Love the scrapbook also - as a side question, I have been living in Europe for several years now and would love to build a current limiter but here in Europe I (so far) have had no success finding typical incandescent light bulbs of such high wattage - what I have found is a 105 watt halogen lightbulb - does it matter the type of lightbulb used? My "common sense" tells me it is okay... the same common sense that thought I could pull off the bias circuit in the attached drawing earlier. 

Thanks again! Much appreciated!

[HotBluePlates]

... the bias circuit is feeding a bit too much negative voltage (do i have that backwards?) ... With ... bias pot fully cranked I can achieve +/- .40mA where my target should be in the high 40's (49ish mA on the "hotter" side). ...

Make the 56kΩ a smaller value.  Maybe try bridging a 470kΩ or 220kΩ across it first to slightly lower the effective resistance in that spot.

[PRR]

> knowledge of some of the fundamentals!

Don't stick power bottles in until you verify Good NEGative bias to their grids!! (For fix-bias.)

I dunno what you did with that mix/match bias rectifier, but a quick pre-check for negative several dozen Volts before tubing-up might have avoided some wear on the tube or smoke-leaks elsewhere.

Without tube plate-cathode and heater load on the PT, the bias could be a bit more than expected; that is OK because it is in the "safe" direction and will probably fall a bit when everything sucks on the PT

[SolidStateCircuitReject]

It was a mistake of arrogance I suppose. I used a small JTM45'ish chassis... I often make my boards so they are not crowded, so cathode resistors and bypass caps will not share the same turrets for example, so when I get to the filtering and bias sections I often have to become a bit creative because of lack of room (understanding that I should do it backwards, but this is how I am used to wiring up), such as placing them on a separate board and mounting it above the PI section of the main turret board.

In this case that is what I did, but it was a mess. At this point the amp was working file. I decided to rewire it neatly, so arrogantly assuming I would have no issues with it, I turned it on and that's when the redplating began etc... from there I rebuilt the bias and changed it's design. I must have mistakingly had something hitting ground by mistake in the bias perhaps.

When building new amps I do check out voltages and negitive pin 5 voltage before installing tubes... in this case I was dumb.  BUT, the tubes biased up nicely and seem fine - they redplated for just a few seconds... Good 'ol Tung-Sol 6L6GC-STR.

Thank you!

[DummyLoad]

the ps shown simulates with the values shown with about -44V to -54V at the (-Bias) node. you can get away with 63V caps but i show 100V - personal preference.


to use EL34 change R1 to 10K or thereabouts and range should fall around -30V to -40V.


--pete

[Willabe]

Love the scrapbook also - as a side question, I have been living in Europe for several years now and would love to build a current limiter but here in Europe I (so far) have had no success finding typical incandescent light bulbs of such high wattage - what I have found is a 105 watt halogen lightbulb - does it matter the type of lightbulb used?