Need advice on HV tap blowing fuses

[luthierwnc]

Hi All,  I am putting what I hope are the finishing touches on a Traynor VoiceMaster to Hiwatt conversion.  I checked the circuit against the schematic (attached) and smoke-tested it with a bulb-limiter.  The bulb did light but not crazy.  Voltages came in close to spec.  I'm using all lightly-used 1968 Mullard tubes and the negative voltage on the power tubes was set at -32 and -35 ma respectively with an eight-ohm power resistor in the 8-R jack.  While I was checking voltages the 1-amp fast-blow fuse on the HV center tap blew.  I replaced it and slowly brought the amp up to voltage and checked the signal with a 1k sine wave on the scope.  Those came out really well also.

So far so good but then the HV fuse blew again.  Not a loud ZZZZZ.  It just failed.  The power tubes didn't flash.  They could be faulty but I doubt it.  Amp levels were set well under 50% and the sine wasn't all that hot either. 

I'm kinda stumped for how to proceed.  I've got a few more 1-amp fuses but don't want to just slap them in hoping for a different outcome.  How would you check for amperage surges?

BTW: The amp was in pristine condition with shiny black transformers.  No arcing on the sockets.  It smelled warm but not hot.  All components are high-quality.  I'm using a FWBR rather than the rectifier in the schematic but all else is the same.   

Any help would be appreciated.  Cheers, Skip

[philipdfisher]

Hi Skip, just looking at the schematic.  The bridge rectifier for the 12V supply does’t look correct.  The diodes attached to ground should be pointing the opposite direction.  If you are wiring this way you will have a short

[sluckey]

Set the bias voltage to -40v on pin 5 of the output tubes. Fuse hold?

[luthierwnc]

Thanks philipdfisher,  I have a four-pin component in that spot so it is fine on the build and now corrected in my first post.

Steve; I'll try that tonight and report back.  sh

[luthierwnc]

-40 and it held steady.  No input.  Signal looked good but I didn't test the overdrive.

To keep going I pulled the dummy load and plugged the speaker in.  It sounded like a wounded water-buffalo so I've got a serious phase issue.  Still, it didn't smoke the fuse.  I'll check back, sh

[luthierwnc]

Blew the fuse again.  The negative voltage comes from a separate winding so with a slow warm-up there is always bias voltage on the grid when I flip the stand-by.  sh

[pdf64]

I don’t think that HT fusing arrangement has much to recommend it.
It won’t protect from a rectifier diode short, and at 1A, it may allow other failure modes to continue without blowing. And a F type fuse may not cope with the current required to charge the reservoir cap when the standby switch is closed, if the HT winding’s equivalent series resistance is fairly low. I assume that’s when the fuse is blowing?
I suggest to try a F0.5A and a CL80 inrush surge suppressor in series with the HT winding.

Regarding the preamp, when switched to clean, the overdrive gain stage V2 grid - cathode pin7&8 will clip the upper half of the signal going into the next stage V2 pin2 at 0.7V, thereby hamstringing the max clean output signal.

[luthierwnc]

I replaced the individual fuses with a FWBR unit and it works fine -- well, that works fine.  I'll try the CL80 inrush suppressor when I get a chance to spend some time in the shop.  Cheers and thanks, sh

[luthierwnc]

PDF64; I wondered about the OD.  It was built bone-stock except for the rectifier and the relay supply.  From my Dumble-build days I always tried to completely separate channels.  The OD is way too hot so I halved the pot.  What do you think of this?  sh

[pdf64]

… The OD is way too hot so I halved the pot…

I don’t think that will have much effect, sorry.
I suggest to lift C5, reduce R11 to 100k, maybe increase R12, eg good old 1k5, 2k2 …
With an unbypassed cathode, the higher its value, the lower the stage gain. Bear in mind that a classic 2204 Marshall uses an unbypassed 10k cathode resistor on its second stage.

[luthierwnc]

More like this?  I'm thinking using both sides of the relay switch isolates the two circuits better.  sh

[pdf64]

Looks good 
Pin2 should probably have a hardwired grid leak, just in case.

[luthierwnc]

thankee

[luthierwnc]

Got the new parts and it sounds thin and weak.  The schematic above has been amended with the changes.  Attached is a pretty sine wave picture at at V4 Pin 2 and another on the other side of a 1meg resistor on pin 7.  At the power tube grids the wave-form is almost square -- same at the speaker jack.  This was really loud and I haven't changed a think on the inverter wiring.  Tried a different 12AT7 in the 4-hole -- same result.  Voltages are close to the schematic on the plates and grids/cathodes too.  OD is off.

Any ideas what I did wrong?  Thanks, as always, Skip

[sluckey]

Decrease the signal level until you have 1V_pp at V4-2. Do the other waveforms get better?

[pdf64]

Inside a feedback loop, waveforms can get weird, especially so if it’s being overdriven. Because the feedback will be modifying the signal to attempt to improve linearity.
So yes, reduce signal level to below the onset of clipping.
But I wouldn’t bother with the waveform at V4 pin7 anyway, that section of an LTP is in a common grid arrangement, it’s also handling a local feedback loop.

[luthierwnc]

Not sure what it was but it is loud now!  Still have a problem getting the OD to engage and will be working on that over the weekend, depending on the weather.  If it is partially sunny, we'll be leaf-peeping.  Cheers, sh