1 of 4 screen resistors dead - I have questions

[Granny Gremlin]

So, I've been enjoying my new-to-me Sunn 1200s; sounds great and working mostly fine, but occasionally I get pops on transients.  Sounds like the signal cuts out for a micro-sec and comes back sort of thing.  Some jams no problem but once it was so bad that  I was worried about the speaker and switched heads.  This is after I took her to the tech for a once over which is frustrating.

Thought it was the mid boost switch as playing with it seemed to make noise in a related way.  But then I opened her up and after discharging the caps, cleaning/testing the switch, cleaning the pots while I was in there, I was about to close her up  and noticed this:



So one of the screen resistors (47Ohm 2W from what I can tell - the smaller wirewound Ohmites should be the 47Ohm 5W plate resistors, right? **), is cracked all the way though.  After touching it (one end totally moved, the other didn't) I can't get a continuity reading like on the other 3.  So it seems like it was holding together just by lead tension but loud transients would vibrate it open (which is why I thought it was the mid boost switch or surrounding circuit - it increases the signal amplitude making more/larger transients).  I told the tech about it when I took it to him so kinda disappointed like I said (granted I also told him that the reverb was sounding like a noisy radio which  turned out to be just dirty RCA connections, and I had tested the power tubes and one looked like gas/short but apparently it's fine, so he may have just thought I was a complete tit when he didn't hear the issue at obviously lower testing volumes).  Sometimes you just gotta get in there yourself I guess.

Anyway, I am wondering if I should replace all 4 so they match (I'm not sure I can get 2W carbon resistors locally/quickly), or just the one?  Does it matter?  If I am replacing all 4, is there a better type to go with (tolerance ignored for the moment) or stick to carbons?  I originally thought these were the plate resistors which are more critical, but I don't know much about these things.

Interesting note is that the tube which I thought might be bad was not the one connected to this dead resistor.  I am not sure if it was the tube working the other half of the wave or if it was the other one on the same side....  from a look around the chassis I am going with 1st one on the other side of the wave, but like I said not certain.

** the 1200s has the same power section as the 2000s -  schem here: http://grannygremlin.com/Sunn/sunn-2000s-amplifier-schematic.pdf

[slinkp]

I can't opine about how critical is to get all 4 matching, but fun fact that applies to resistors in general:  If you can't find a 2W one, you can replace it with a pair of 1W ones.   Either get two with double the resistance and wire them in parallel, or get two with half the resistance and wire them in series.   It's exactly the same math as matching impedance and wattage between speakers and amps

[Granny Gremlin]

Yeah - that would just get messy though.  You can't see the rest of them but 2 of 4 screen resistors are packed in pretty tight on a turret board under the lip of the chassis.

[Psycho Bass Guy]

In most cases, I'd replace all the screen and plate resistors with modern resistors but the wirewound plate loads are probably fine as is. Carbon comps that crack do so because they have been hot and/or are old. Metal film or silicone resistors will take the heat and hold their value much, much better. It's not an issue of performance matching as much as it is a reliability mod. I would leave carbon comps on anything in the preamp that would affect overall tone because their inherent noise acts as a built in harmonic filter, but for things like this, output stage regulation, you want reliability. 

[Granny Gremlin]

That's what I was thinking.  I just hope the place by my work has 2-5W resistors in something other than ceramic.  I might have to mail order.

[amptech]

That's what I was thinking.  I just hope the place by my work has 2-5W resistors in something other than ceramic.  I might have to mail order.

I use 5W wirewound (rs#762-9469, cheap but reliable) and always replace all if they look iffy or too low W. Carbon comp screen grid resistors I throw away. Screen grids in tubes is a weak spot in todays tubes.

As a side note, if you suspect one gassy tube or have a bad feeling for it - just replace it. Measure the others and buy one that test similar, people sell good used ones really cheap.

[Granny Gremlin]

I was in a hurry and the local store only had 2W metal (5W was only ceramic - too bulky and kinda shit no?  Though I've seen them, perhaps better ones, on screen in Marshalls).  So I got those and put them in - needed to get the amp working in time for jam.

I'll order something better and 5W next time I'm getting parts online (extra for spare plate Rs too).

I got a single vintage GE from FrankieTBird a while back and gave it to the tech but he said it was fine and didn't install it.  Frankly, it feels like he didn't even pop the chassis.

In any case, at moderate volume (kids asleep and home speaker can't handle more) the thing plays super nice now and is actually quite a bit quieter, esp the fx (pot cleaning probs to thank there).  True test at jam tonight.

Pics coming.

[Granny Gremlin]

47 Ohms is very low for a screen resistor (usually 470-1k) and I got some interesting information after someone suggested I replace them with 1k:

This is an ultra linear amp, rather unique with the plate stoppers which don't serve the traditional purpose that preamp plate resistors do.  As well, the 47R resistors on the screens are an added safety feature as well, but again, they don't do the same thing that screen resistors do in pentode output stages.  I wouldn't change the values.  1K will strangle the amp and make for thin, lifeless tone. ... Most ultra linear amps don't have screen resistors at all.  If they're there, it's not to limit current like pentode mode but rather to control oscillations and as a backup safety "fuse".  The tap, usually 43%, on the output transformer limits current and reduces output impedance and distortion as well as giving a slight increase in power at the optimum tapping point, which varies with different tubes.  Suggesting that higher value resistors would make the amp more controlled and useable nullifies the original designers' knowledge and intent.

[Granny Gremlin]

Update: So the Sunn went to jam (I'm on drums in this band, but we jam at my spot so the basser uses my rig). Everything was great. Band goes home, amp is turnned off and I want to fool around with it myself (also have to test a compressor pedal I built). Something is wrong. Noise. At first I only notice it on transients (no longer pops, but fuzzy distorted drone; not in the signal, but parallel to it). Then I switched guitars, cables, dumped the pedalboard and noticed the noise was always there, just quiet, and only spikes when there's a strong signal such as just about any note played or even fiddling with the cable/jacks. Also the amp was hot as hell and seemed a bit quieter hen normal. Could barely touch the power tranny. Quickly turned it off and went out for a smoke to give it cool down time and think it through.

Came back in and decided to swap out that iffy tube for the spare I got from Frankie. I know not ideal, but should be fine and prove whether that's the issue . Turns out that this iffy tube had cooled down completely while the others were still hot (the tranny, being a huge mass of metal, was also still quite warm but no longer hot). This made me more confident that the bad tube was to blame.

Put the alt 6550 in and everything was fine. No noise; great tone. Actually very surprised at the tone given that it's not properly biased with the new tube in; expected a little lopsidedness, like a bit of mild asymetric OD or something, but it was crystal clear and bell like. Played her quiet at first, then opened her up. That compressor pedal is working for me. Then I'm done and I shut her down. Go around back and the power tranny is still a bit warm, but cooler than before - didn't get hot again.

I hesitate to say problem solved, but I am hopefull. Since this is an intermittent problem, and that bad tube was in there at power for a whole jam, I will see how she goes at jam on Thursday when I'm on bass and if still good, will drop to the tech for a bias check while I am on vacay next week.

[amptech]

Might not be a problem to use tubes of different kinds, but as you experienced; they have to be in good shape! As I said earlier, if you suspect a tube have gone bad - replace it. Do not use a valueable amp with bad tubes, you don't wanna ruin it on purpose!
Have your tech measure the tubes and maybe move them around until they pull about the same current.

You could (since you are no stranger to electronics and can use a soldering iron) just install some good 1W1R1% resistors between cathode and ground (depending on the design, but I guess the cathodes here are grounded) and you can easily measure and bias yourself. 1mV = 1mA across the resistor.

[Granny Gremlin]

The new tube is actually same manufacturer (GE) and vintage (give or take) but one of those Groove Tubes tested/logo stamped ones (you can see the GE labelling underneath), feels like I lucked out on that one, cause buying a whole quartet of vintage GEs woulda been spendy (even just new prod).  I'm not one for putting too much stock in tube brands being better in amps than others, but the GEs are what was originally equipped and the diehard fans (e.g. on the Sunn forum) swear they sound best (even vs KT88s - which Sunn used to use but switched at some point, that would be the minority dissenting opinion among aficionados).

Schem says cathodes are grounded, but I wasn't looking at those when I was in there.  It's all back together now so I kinda don't want to deal with that, but thanks.  I'd have to read up more on biasing generally, and I also wonder how it works as there appears to be some sort of bias test point (see red thing middle of the chassis in pic) as well as a bias pot behind that.  I can't begin to understand how that works for 4 tubes, and I have seen restoration threads where dudes install a second bias pot.