Quad II fuse-blowing

[tpatton]

I just hit a wrong key and lost a lengthy reply to Deluk and Paul J, and will say it all again.

For any who want it and don't yet have it, here's where to find the circuit diagram: www.keith-snook.info >> Quad II Information >> here >>Quad II Power Amplifier

On that diagram, I see no voltage higher than 340V.  My C4 and C6 are rated at 450V, my C2 and C3 at 650V, my C5 at 50V, but no high voltage there (but lots of current: Keith S recomments higher than 3W for R12 and I have a 5W Mills).

Would a solid state rectifier need any circuit change for the amp to handle it?

I could put a 100 ohms resistor in series with each diode if I knew in physical terms where to put it.  In terms of rectifier socket pin numbers, maybe?

On the circuit diagram, there's this scary note (bottom right):

Note: The original capacitors for C2 and C3 had insulated metal bodies which were connected to ground, the small capacitance to ground controls the open loop frequency response and thus the stability with feedback; change those capacitors with caution!

This is all Greek to me.  My Obbligato Gold Premium C2 and C3 have nonconductive bodies.  Am I in trouble with them?  Is there anything I could/should do?

One last question: is the quality of C1 important to the sound?  I'm using a capacitor that came with one of my Bottlehead kits, rectangular and orange-brown.  I could use an Obbligato Gold Premium, but guessed that in that early position it wouldn't  help the sound.  Any advice on that?

Thanks to everyone who's gotten involved with this project of mine and its problems.

[tpatton]

I posted last night and it never got here--it says I posted, but no post of mine appears.  I hope this one works out better.

I didn't follow Paul J's suggestion at first, then figured out that the diodes that could have a 100 ohms resistor in series with them are, in a solid state rectifier, like the diode bridges us Bottlehead kit-builders used to mount on a terminal strip.  I see four wires going to four rectifier sockets, and am supposed to make them connect to 100 ohms resistors instead, which then go to those four sockets. 

When I feel able to take the pain of ruining my last good rectifier tube, I'll put it in, and with terrific luck the amp will play troublefree for more than 10 or 12 minutes, with no bright flashes from the rectifier tube.  But I want to change over to solid state rectifiers anyway.  Has anyone looked at what GZ32/5V4G tubes cost?  It's one of those tubes that some audiophiles go crazy over, and even the cheapest aren't anything like cheap.

I had another question that failed last night to get posted.  Is capacitor C1, which separates resistors R2 and R3, important to the sound?  If it is, I could use Obbligato Gold Premiums there.  But I assumed that it wasn't, and used a cap that came from my Paramour II kit, a F104K.  Should I upgrade there?
 

[tpatton]

I now know about Page 2 in these threads: what I wrote last night did get posted after all.  Sorry to be so dumb as a Forum contributor, maybe I'll be okay from now on.

[tpatton]

I was about to order a pair of Tube Depot solid state rectifiers, when I made the mistake of looking at the schematic for them: http://site.tubedepot.com/pdf/ssr.pdf.  And confused myself badly: there are only three rectifier tube sockets used, but I have wires going to four.  I realize that a tube needs current to allow it to work, provided by at least some of those wires.  Still, I'm back to not knowing where to put the 100 ohms resistors in series with the diodes, and it seems that there are only two diodes, not the four found in the diode arrangements used in Bottlehead kits, which puzzles me too.
Can anyone help, Paul J in particular?  I'll remember you in my will. 

[JC]

Have you looked at the link that Henry's Cat posted for you?  It shows one guy's approach.  Note the Wattage ratings on the resistors he used.  Basically, the little drop-in SS replacements for tube rectifiers just use SS diodes to replace the two diodes found in the tube version, and leave the heater power unconnected.  This generally results in, as you have learned, a significantly higher B+.  So, the conventional approach is to waste some of that extra Voltage by dropping it across a resistor or resistors.

Perhaps the first thing to do is to find out how much your rectifier tube typically drops the Voltage.  You can hopefully get that info from a tube manual.  You can then  get an idea of how much higher your B+ is going to be if you use SS rectifiers instead.  That will allow you to calculate a resistance and Wattage for the resistor you will need to drop the B+ back down to within spec. 

Getting this info together is part of what Doc was referring to when he replied to you; you are essentially redesigning the B+ supply to a certain extent.  Another thing to be aware of is, when you stop using the 5 V rectifier heater winding from the transformer, your other secondary windings might creep up a bit on you.  Maybe not much, but you don't really want to exceed the heater Voltage rating on the other tubes by more than 10 or 15%, so it is important to check.

Meanwhile, is there perhaps a less expensive substitute for your rectifier tube by any chance?  It might pay you research that as well.  If there is, you could maybe get one or two for testing purposes or temporary "in a pinch" replacements.

[Henry's Cat]

I''ve not had any problems with using polypropylene MIT Multicaps for C1, C2 & C3 in my refurbished Quad II's. None have metal cases and I've not noticed any of the effects Keith Snook refers to (I am using 6L6G output tubes).

R12's power rating must be upgraded as 3 watts is far too low and it actually has more than 3 watts of power flowing through it under normal conditions!

I replaced C4 & C6 with 15 microfarad and 22 microfarad Solen polypropylene capacitors respectively rated at 630V. Because they're quite large they are attached to the base-plate using tool clips. I've not tried replacing the GZ32 with silicon rectifiers partly due to the much higher H.T voltages that would result, especially at switch-on before the output tubes start to draw current.  There's quite a big voltage drop across the GZ32 which is lost with efficient semiconductors so you're looking at a H.T supply of around 438V.  As I've mentioned in an earlier post my H.T voltage has always been higher than the stated 340V (around 370V) so with silicon rectifiers it might be over 450V!

If you can't get a replacement GZ32 easily I've used modern GZ34's without any problems.  Attached is a file from a website which I think is now defunct - Quad Pages - lots of useful information in the text though.  The images are missing, they display on my PC as the files are stored there but won't save in the archive for some reason.

[tpatton]

Last night both Quad II amps played beautifully for 30 minutes--I think I'm out of the woods, having started out with a bad rectifier tube and then, through missed solder joints, ruined another.  I have only two good rectifier tubes left, but they don't seem in danger of destruction any more.

I still want to go over to solid state rectifiers, and welcomed Paul J's idea of puttinhg a 100 ohms resistor in series with each diode in that SS rectifier.

What I need are instructions, on a level with those wonderful Bottlehead kit manuals we all appreciated so much, of precisely how to do this: which rectifier tube sockets to solder the resistors to (with the wire formerly going to those sockets now going to the other ends of the resistors).

I also wondered if 100 ohms was definitely the value to choose, or whether 120 ohms, say, might be a more conservative choice, but still allowing enough voltage to come through.

I've also asked another question once or twice, and would still like an answer: is capacitor C1, showing up close to the input on the circuit diagram, important to the sound?  Should I use Obbligato Premium Gold there instead of the F104K's that came with a Paramour II kit?

But the solid state rectifier question is the one I'd really like to get answered.  Thanks in advance for this, and a very big thanks to everyone who has helped me through the ups and downs of this long struggle to get this ancient pair of Quad II's working again--as I'm pretty sure they now are.

[Paul Joppa]

You can use a single resistor from rectifier cathode (pin to the first capacitor C6. Naturally I don't have an example in front of me, so I don't know how they are now connected, but if there is a wire that can be replaced with the resistor that would make it easy.

The point is to limit the available charging current to the capacitor, so you can use either a resistor in each diode plate lead, or a single resistor in the cathode lead. It seems likely that the power transformer high voltage lines go directly to the plates (pins 4 and 6) so there is no convenient terminal on the plate side.

The average current through the resistor will be 147mA according to the schematic, which would be 2.16 watts if it were pure DC. However the current at that point probably has peaks of 500mA, and I would estimate the RMS current at twice that, or 4.3 watts. At 120 ohms, it's a hair over 5 watts. I would use a 15-watt rated resistor; I have found a safety factor of 3 or more times the actual dissipation to work reliably with resistors. Keep the resistor from touching any capacitors or other heat-sensitive components.

You will have to experiment with the resistance to get the target voltage, which should be equal to the voltage with a tube rectifier. My guesstimate of 100 ohms is just that - a guess.

Note that the extra 4.3 watts increases the heat under the chassis; that extra heat is normally dissipated by the tube above the chassis. Again, I don't have one of these in front of me so I can't assess how well such under-chassis heat is vented.

[tpatton]

Thanks very much, Paul, for all this.  I can easily put a resistor between rectifier pin 8 and C6.  It's clear from the circuit diagram, isn't it, that the power transformer high voltage does go directly to pins 4 and 6 (now that I know, from knowing where pin 8 is, which pins these are).

Experimenting with 15 watt resistors would be both expensive and call for voltage measurements, which I'm chicken about, only doing them when absolutely necessary (as with the Bottlehead kits) and with my wife there ready to pick me up off the floor and call an ambulance.  So despite the fact that I've already ordered a pair of solid state rectifiers, I might go back to looking for 5V4G tubes that don't cost an arm and a leg.  It boggles the mind what lots of people are willing to pay for them!

I've become neurotic about rectifier tubes and their ability to last--less than 12 minutes, in a couple of cases--but maybe I'll get over that if the present ones continue not to fail or to cause fuse blowing.

[Henry's Cat]

Re. C1's effect on the Quad's sound.  I've not had experience of the what effect it may have on the sound simply because I've not tried different brands or dielectric type.  You may have to experiment.

120 ohms as a dropper resistor seems rather a low value to reduce the H.T voltage.  If 147mA is flowing through then it only drops 17.6V DC and about 25V with A.C current. 

Re. the Tube Depot SS rectifier.  It's a full-wave rectifier equivalent of the GZ32. The anodes connect to the A.C windings of the HT transformer and the third connection is the equivalent of the cathode and supplies the rectified DC.  The centre tap of the transformer is the ground connection.  You could replace the wires from the HT transformer to each anode connection on the tube socket (pins 4 & 6) with dropper resistors and perhaps another dropper resistor from pin 8 to the smoothing capacitor. Like Paul Joppa says you will have to experiment with resistor values but remember, the higher the value the more power is dissipated and heat generated.  You could always try GZ34's which are cheaper than GZ32's.
 

[Paul Joppa]

Mouser has Xicon 100 ohm, 15 watt wirewound resistors for 64 cents each. Antique Electronics has 5V4s for $15-$20.

[JC]

If these 100 Ohm resistors are for use in front of the anodes of the diodes, then I think from a look at the schematic that 5 Watt resistors would be just over double what they will see in operation.  In the past, I have found that putting these resistors in that position when using a SS replacement rectifier package is a good idea, both for dropping some B+, and also for  current limiting for the diodes during turn-on and resisting any little diode switching transients on their way back to the high Voltage winding of the transformer.  Additional resistance can be added between the junction of the cathodes and the first filter to tweak the B+ further, and similar attention must be paid to the Wattage of the resistor.

IIRC, the last time I bought a 5V4 from Groove Tubes, it was around $20.

[Doc B.]

Antique Electronics has 5V4s for $15-$20.

Yeah, I was thinking "since when are 5V4s that expensive?" I'll stand by my suggestion to replace all the stock caps with similar values - and any resistors that are out of spec, and keep the tube rectifier as the best course of action for you.

[JC]

Well, the only thing I have been able to find is that Dan Torres has sold a lot of NOS ones, and has gone to Sovtek now.  I haven't bought any of that type for a while, and tubes can be like the commodity markets sometimes, I guess.

Any progress reports on the Hammond?

[tpatton]

Doc B, I've replaced ALL resistors and capacitors, but am leaning toward your recommendation of sticking with rectifier TUBES, to avoid complications I'm not competent to handle, not the least of these being the heat Paul J warned about.  I'd guess that the Quad II chassis is especially bad in that respect.

I guess I'll have to experiment with a "boutique" C1 instead of the Paramour II's F104K.  I thought it would be easy to tell from the circuit diagram whether it would affect the sound quality--guess not. 

Paul J, your efforts to tell me how to handle a solid rectifier, with its several cautions, and your latest on where to buy, are much appreciated, even though I may not go for a solid state rectifier in the end.

Will a GZ34 work perfectly well in these amps?  Has anyone tried them?  I would have thought probably not, since they don't come up as equivalents.