Sunn 1200S Restoration project

[nevetslab]

A hazard that can arise with direct coupled stages is during startup, when V2 is not conducting and hence the grid of the cathodyne is effectively at B+. As well, the cathode of the cathodyne will be at ground. That has been known to cause arcing - and a grid-cathode diode can alleviate that.

I will look into that....many thanks for the explanation and suggestion.

I was just about to start pulling the old supply caps out and installing the new mounting points for the series-coupled supply filters on the first two stages, but stopped to look at what would happen if I just blindly used shorting input jacks in place of the open-circuit jacks Sunn used. While I could do it on the Normal inputs, making one a lower sensitivity input (attenuation) like Fender or Ampeg do, it doesn't work with the Bright input circuit. Their 'Bright' input R/C network if it gets grounded, it screws up the input response and I also loose a lot of sensitivity. Sunn used isolated Break switch contacts on those four input jacks in the later release, which shorted out the volume pot when nothing was plugged in. But, once you're plugged in, the unused input jacks remain open circuit, though effectively loaded by your guitar, so it doesn't really matter. Those jacks (Switchcraft 13A) are around $8 each. This amp doesn't have them. I'll deal with it down stream.

Sunn also has a 'brightness' cap across the volume pot's wiper-top of pot, that's not switched. I'll probably replace the volume pot with one having a Pull-switch, so it can be switched in or out.

Chassis is about as clean as I'll get it. Once I'm thru with adding terminal mounting holes and the extra bias pot, I'll apply a coating to the outside surface so it doesn't rust.



As restoration projects go, always interruptions to halt progress.

[nevetslab]

HIGH HUM level on rebuild, now trouble-shooiting and need help

After completing the replacement of all the power supply caps, other aged components and revising the grounding to bypass the 45 yr old aluminum rivets used in the original build, I finally got to powering up the amp.



Having added a second bias pot, now having one each for the top half and bottom half output tubes (6550’s), I began by verifying supply voltages and bias supply adjustment range. All looked favorable there. During the bias adjustment phase, beginning with a solid state plugin rectifier installed in place of the 5AR4 tubes, I ended up adding some 1N5406 rectifiers (two per string) across the rectifier tube sockets, and wired them up to the vacated Ground Reverse Switch, changing it to a Rectifier Mode switch (Tube/Diode)

In Diode mode, I had 535VDC at the O/T C/T winding, set the 6550’s for 35mA/20W per tube, and with the 5AR4’s it dropped to 522VDC & 30mA/18W per tube. Current balance was 3-4mA difference on the extremes..

When I gave a listen to the residual hum/noise for the first time, I was horrified. Massive amount of hum coupling, along wlth veryhigh diode commutation noise, whether in Tube or Diode Rectifier mode.

I was also hearing thump from the Tremolo circuit, which is powered off the bias tap winding. Unplugging the power to that PCB assy, the hum level dropped substantially. Unplugging the preamp tube, there’s still some hum present (16.67mS hum spikes….60Hz), and most of the diode commutation noise much lower. I haven’t yet replaced the bias rectifier diode nor added bypass cap across it Thought about changing it a UF4004 (fast recovery diode).

The schematic shows an LDR module in series with the top of the volume control from the input preamp tube’s plate for the Tremolo modulating circuit. That LDR module is now right next to the 1st stage power supply filter stack (pair of 220uF/300V caps in series, with ballast resistors). In fact, in replacing the power supply caps, I’ve added additional capacitance….1st stage is fed from the Rectifier mode switch, output of which feeds the S/B switch thru a HT fuse (1A S/B), and to a 20uF/600V cap at the input of the parallel inductors. The output of the inductors feeds the second filter stage, it being series-connected 100uF/450V caps, which feed the C/T of the O/T primary. From there, the filters are all in a new cap can, being a 40-20-20-20uF/525V, mounted onto a fiber board, and feed the driver, reverb & preamp stages.,. The C/T winding of the HT supply ties to a ground stud, along with the ground leads of the 1st 3 power supply filter caps. I had the heater C/T lead tied there two, but it is now lifted, in favor of a hum balance pot, installed nearby. A single wire from the cap can ties to this new chassis ground stud’s solder lugs. All the remaining grounds of the system tie to the cap can’s terminals.



So, hearing this excessive hum, and it dropping substantially when I power off the tremolo board, I located where the LDR’s Resistor cell was wired in (in series with the tone stack from the 2nd stage tube). I disconnected that, it being wired in via shielded cables, and jumpered the connection so that high impedance circuit was no longer in the power supply field present by the filters and inductors, and closer to the power transformer.

That made no difference, surprisingly. Still had the high hum and buzz. Prior to my making any changes on the amp, I did have it powered up, and listened to it, and it had a lot of hum/noise, though I didn’t document it.

The next moves I made, now wondering about the power transformer itself, I disconnected the secondary connections to the circuits, and powered it up, to see where the excitation current was at. 215mA @ 120VAC, 10.7W. That seems a bit high to me, but….I haven’t checked to see what it is on my Traynor YBA-1A’s transformer, being about the same VA rating and physical size..

Now, with it disconnected, I moved my HP 712C Power Supply to the bench, and connected up the bias, HT & heater supplies to the chassis. Set the bias to the same -81VDC source level, 550VDC in to the 1st filter stage, 6.3VAC to the heaters, and powered up the system. Minor tweak with the bias, and I had 522VDC at the C/T of the O/T primary, pulling 150mA at idle. The hum/noise level was now very reasonable to my ears.



Hmmmm. What if I disconnect the heater winding from the Sunn’s transformer, and run it from the HP supply? I’ve had excessive field radiation problems on Vox AC30’s improve by using an added heater transformer in the chassis before. So, wired that up. Didn’t solve the problem…I again had high hum & high diode commutation noise again.

That’s where I’ve stopped for the moment. Not sure if I have a major problem with the power transformer. My AR 4040AT Hypot is in storage, so I haven’t checked it for leakage. My next attempts are to try and power this up with the transformer from the Traynor YBA-1A, though by way of extended leads, not yet wanting to remove the transformers and swap it out.

I’m still pondering the changes made in the grounding, and the added filters installed, with their orientation. Also, since Sunn has two inductors wired in parallel, is there a phasing issue there? Will it make a difference with one reversed?

[trobbins]

Given you have external power supplies available, perhaps try and swap in/out some of the on board power supplies to see if any particular supply (B+, bias, heater) is a major point of hum entry. Similarly, try and work back from output stage to see where main ingress occurs (that may mean disconnecting feedback if used, and temporary disconnection of coupling caps between stages).

You may be lucky and find a silver bullet, or need to methodically wade through a few main entry mechanisms.

Any change to original has consequences - such as replacing a valve diode with ss. Perhaps if you review how well the different stages are locally decoupled to their local 0V star nodes (or not!), and then connected together. SS diode related transients on the B+ winding can glitch over to bias and heater windings, so perhaps review the smallest current rated UF diode you can use via PSUD2, and check how well the diodes and first filter cap wiring and layout are done.

[Justin Thomas]

So that would be one louder?

No, it's one quieter. It's negative volume for obnoxious singers and bassists.

Justin

[MarkusBass]

Never did understand why Sunholm used the 0-10 numbering. I guess you get "11" spaces while Fender knobs were only 10

This is actually very simple to understand: with numbers from 0 to 10, 5 is exactly in the middle (it makes easier to manufacture the front panel and to remember settings of the amp). With numbers from 1 to 10, half is somewhere between 5 and 6. With the first case it is also easier to understand setting e.g. gain to 0. Setting the knob to 0 means you get no signal on the output (which is true) while setting the knob to 1 suggest that there is just a little bit signal on the output - more than 0 (which is not true). I would say that scales from 0 to 10 are much more intuitive than scales from 1 to 10 and I wonder why Fender did not think about it.

Mark

[Enzo]

I have to admit I never thought about it. I just turn on an amp and spin the controls to where it sounds good. Numbers be damned. I leave this matter to the amp fans.

[Justin Thomas]

I wonder why Fender did not think about it.

Probably because the knobs they used had the best bulk discount. And they started using them BEFORE the "bean counters" @ CBS took over.

Justin

[Mick Bailey]

Quite recently I'd finished work on a Fender Supersonic and gave the owner a quick demo. I thought it was one of the best sounding amps I'd ever heard when it was finished, and he agreed.

But then he said, "No...hang on - that's all wrong" and adjusted the controls to give a fairly average sound. He asked me why I'd set it the way it was, my reply was that it sounded good that way. He then informed me it was set incorrectly, and there is a mathematical 'rule' that he'd read, and that all Fenders had to be set a certain way. It's a numbers game. Anyhow, that was that.

The way I had it set was described by Matt Schofield;

“You’ll find that every pot has an obvious cusp point in its rotation – every single one – where it goes from not doing much, to where it really kicks in. So I sweep through the pots one by one and find those spots. All amps will do this – bass is especially easy to hear... nothing, nothing, nothing, then bang, there it is! If you do that, the amp is set as evenly as it can be; not too trebly, not too dark. With everything set right on the cusp, you can push it over by digging in, or pull it back by playing softer. It makes everything really responsive and sensitive to what you do. The rest is up to you!”

[Justin Thomas]

and there is a mathematical 'rule' that he'd read, and that all Fenders had to be set a certain way. It's a numbers game. Anyhow, that was that.

I've read 1,000 posts that say treble on 10, mids on 6, Bass on 0. That's the most-often cited "rule" I've heard about Fenders: don't turn the Bass above 2. Tried it once on my Bassman. Friggin' OUCH.

Justin

[Enzo]

All this, and considering that the numbers on the knobs have zero relationship to the circuit. They are merely there so the user has a reference point. They could just as easily have labeled the knobs a-b-c-d-e-f-g...

[nevetslab]

Additional details on Power Supply Revisions, Hi Hum level

The multiple postings on the merits of 0-10 vs 1-10 marked control knob legends is interesting, but at the moment, down-stream to the present state of the high hum level I now have, having installed fresh power supply caps and revisions in grounding to this 45 yr old Sunn 1200S Guitar Amp.

I've included some labeled original images of how it was wired up, and the present approach taken to add new filter caps inside the chassis, using the available space around the Tremolo PCB Assembly.



The original wiring had the 20uF/600V Axial Lead cap grounded to a chassis solder lug, mounted to one of the power transformer mounting screws, along with the C/T of the Heater winding. The C/T of the HT winding is connected to the Ground terminal of a terminal strip (aluminum rivet), along with the ground lead of the 1st bias stage filter. The output of the dual 5AR4 Rectifier tubes feeds the S/B switch (routed close to chassis floor), and output of it ties to the 20uF cap, and input to the pair of parallel inductors. In the second image, we see the output of those inducors (brown wires) connect to the 40uF filter section of the 40-20-20-20/525V Cap Can, which is the main filter section of the amp. It's also the ground collection tie points from the Driver Stage, Cathodes of the power tubes, ground from the Tremolo circuit and ground from the bottom of the tone stack. The rest of the circuit grounds of the preamp & reverb driver stages get to chassis locally on their respective terminal strips underneath the output transformer. Secondary of the O/T ground wired to grounded phone jacks, input grounds for the 1st preamp stage made at the 4our input jacks.

I've presently lifted the local grounds of the preamp stage circuits, and brought them to the new Cap Can ground, along with the bias filter caps. The 1st three filter supplies, along with the HT C/T lead connect to the new Ground Stud adjacent to the new Cap Can.



The ripple levels at 120V/60Hz mains are shown below.

Stage -----Tube Rect---Diode Rect

1st---------8V P-P------7V P-P
2nd--------700mV P-P---500mV P-P
3rd--------<10mV-------<10mV
4th--------<10mV-------<10mV
5th--------<10mV-------<10mV
Bias 1st----2V P-P
Bias splits--1.5V P-P
Bias wiper--25mV P-P

The rest of the details are in the post I made on 1/9/16, which seems to have been lost in all the dialog around control knobs. I'm not certain that I've made some bad decisions in the placing the new caps surrounding the tremolo board, instead of mounting them to a plate above the floor (underneath the power transformer) or inside an enclosure on the top of the chassis (like Fender does underneath their chassis) and keep the loop area of the filter wiring well contained).

When I had first begun, after slowly charging the original filters over the course of a day, and getting basic measurements of the system, I did give a listen to the amp, and it had high hum level, both 60Hz & ripple, with dried up 45 yr old filter caps. I didn't make detailed measurements then.

At any rate, in it's present state, the amount of hum coupling, coupled ripple and diode commutation noise is overwhelming, so I"m now looking for solutions to that. Amongst these to try is substituting the power transformer with that from my Traynor YBA-1A, being nearly the same VA rating and size, though it lacks a bias tap.

The excitation current on this Sunn 1200S transformer is 215mA @ 120VAC/60Hz, 10.7W. That seems high to me, but, I don't yet have numbers from my Traynor's transformer to compare it to.

[trobbins]

The C/T of the HT winding is connected to the Ground terminal of a terminal strip (aluminum rivet), along with the ground lead of the 1st bias stage filter.

Does the C/T of the HT winding go directly to the first filter 20uF cap negative terminal, as close to the cap body as possible? Given that you then go through a choke before powering any circuitry, and the choke goes to your can cap, then the can cap negative tab should be your star point for all major current loops - including a negative wire coming from the first filter 20uF cap negative terminal; and the output stage cathode returns. But I'm a little lost with the photos and descriptions of all the various can and axial capacitor connections and multiple chokes etc - are you able to sketch up a simple power supply diagram and photograph it?

[nevetslab]

I'll have to draw up the circuit, but for the moment, I've marked up a Mesa Dual Rectifier schematic to show the HT Power Supply as it's currently wired up.

The Original wiring had the HT C/T tied to chassis at a riveted terminal strip, while the negative lead of the 20uF/600V axial cap was a few inches away. The inductor output went to the cap can, which was the collection point of the grounds of the power amp stage. it is no longer wired like that.

The present 1st stage filter, being a pair of 220uF/300V caps in series, the C/T joins that ground at the chassis stud between the pair of inductors, so no, it's not as close as I could get it to that cap. Those are the Gray caps near the middle of the chassis, along the bottom edge of the Tremolo PCB assy.

I'll sketch out a better drawing shortly........in the middle of another repair at the moment.

[trobbins]

It may not be a silver bullet, but I can suggest that the key current loops are the PT secondary HT taps and CT going through the diodes to the 2x 220uF series cap. The interconnect wiring should be as short and twisted as practical, and end up connecting as directly to the cap terminals as possible. This is of most importance for the ss diode pathways, as they have much higher peak current pulses. The 2x 220uF cap terminals then would normally become the take off points for the rest of the powering circuitry. It may appear pedantic, but without that commonly applied method to minimise mains rectifier related noise, then you may spend time chasing other red-herrings.

As the circuit uses a standby switch with caps on both sides of that switch, then I suggest that is pretty ugly for the switch, and some series resistance should be with the switch contact to minimise the peak current. If you add 100 ohm then you limit the peak current to 5A, which is still fairly high, but possibly ok. As that 20uF cap is also going to take diode current pulses, it too should be separately wired to the 2x 220uF cap terminals by short twisted cabling.

The original ss diodes were 1N4007. The replacement 1N5406 have a lower PIV, and will let through more reverse recovery current and more primary winding voltage spike, as they are 3A rated. Perhaps series UF4007 would be a better option.

[nevetslab]

Sounds like I'm back to creating a power supply module to mount the caps and rectifiers, getting the input and gnd terminals next to each other (distance just over 1 diameter of the 220's and 100's. I had considered that before....should have gone that route it appears. No worries. Easy salvage job, though mounting the plate may be interesting, since the P/T mtg hardware & P/T has precedence over chassis floor in that area. Love to get the inductors onto that plate, though I doubt I'd have the space. I'll give this some thought.

Perhaps giving up the 20uF/600V cap at the input of the inductor would be wise. Use instead a high voltage film cap, around 47nF? Looks like that was what Mesa used in that circuit diagram I marked up. I'll check to see what the DCR of the inductors is.

I hadn't considered the slower speed of the 1N5406 3A rating letting higher primary voltage spikes thru. Using the UF4007's sounds like a better solution there too. I was seeing a lot of commutation noise in the Tube Rectifier mode.