Sunn Spectrum Voltages

[Hellawatt]

Hi, I'm not an amp tech or electrical engineer, but I have lots of guitar and pedal soldering experience, and I've been spending a lot of my downtime at work reading this forum and studying schematics.

So I bought an old Sunn Spectrum II because the schematic looked super simple and I wanted something loud and clean that I could use to learn about amps and also use at band practice for loud and heavy fuzz rock that would be a good platform for pedals. Its an ultralinear amp with 2 kt88s, a 12ax7 and a 6AN8. When I first got it it sounded pretty bad.

So I undid the sloppy work of a previous tech, changed the power cord to 3-prong, built a 50/50/50/50 @800v uF cap board copying the SDS Labs MK3 board, replaced all other caps except the ceramics, and tried to make everything match the 1967 schematic except that I changed the coupling cap after v1a from 0.1 to 0.022, and I changed a 1.5M resistor to a 1M on the 6AN8, which brought the voltages for that tube a little closer to spec.

The problem I'm having is that my Standby,A,B, and C voltages are still way above spec, and when I tried jamming on it today using wall voltage and silicon rectifier, it sounded awesome! but I think it blew a tube after about 3 minutes of playing, because it keeps blowing fuses if I turn it on with power tubes installed, and doesn't blow the fuse with the power tubes out.

Why did I blow a kt88? Was the voltage on the tube too high or did it just fail? I know the bias setting can affect the B+, but when it blew I just had it set to -55V like in the schematic. I was going to fine tune the bias later.

Why doesn't my choke drop the voltage by 20V like in the schematic? Should I replace the choke?

How close to the schematic do my voltages need to be? I know how to look up max values on the tube datasheets, and I know that increased preamp voltages will increase headroom, but do these values look dangerously high?

Should I run it at 117 on the variac or install zener diodes to keep the heaters at 6.3V?

I'm attaching a table of voltage readings I took with the tube rectifier, the silicon diodes, and a weber copper cap, at 120v wall voltage and at 117V with a variac. Some data points may be missing.

Thanks for reading! Sorry if any of these question are stupid.


 

[sluckey]

I think your voltages are typical. You're less than 10% high with the GZ34. I have a Sceptre which is the same amp but with reverb and tremolo. I am using a 5R4 to reduce voltages a bit more than the 5AR4/GZ34. Here are my voltages. Bear in mind that I recorded these voltages with the line set to 115VAC.

     http://sluckeyamps.com/sunn/sunn_sceptre_1971.pdf

[Hellawatt]

Thanks Sluckey. Could I use a 5R4 with my first filter cap at 50uF?

[SoundmasterG]

The wall AC voltages today are higher than they were when the amp was made, so the voltages in your amp will be higher than the schematic. It doesn't change the same as the AC voltage change though. 10 V AC means a much higher DC voltage inside the amp. A Sunn 200S I used to have was about 560 V B+, and it worked fine once the bias was set.

The problem if you set the bias voltage to -55 V is that you don't know what current the tubes are drawing, and the bias current is what is important, not the negative voltage you inject into the grid. You can temporarily lift the ground connection from the cathode on the power tubes and put a 1 ohm resistor in between each cathode and ground, then measure across the resistor and that will give you the plate + screen current the tube is drawing, which is close enough to get the bias correct. Set it to somewhere between 60% and 70% of the max dissipation for a KT88 and the amp should function unless there is some other problem. Check over your work and make sure everything is correct, especially on your board. You may have got a trace wrong or something and it would cause a problem and be easy to miss.


50uF power supply caps for the preamp is a bit large. Closer to the values Sunn used would be better for tonal purposes. The preamp filter caps are RC coupled to the preamp stages so a change in cap value changes the tonal spectrum of the stage they supply. That isn't your problem though...the amp will work with those values.

Greg

[eleventeen]

Changing to a 5R4 (or a 5V4) is precisely what I would do. A 5R4GY/A or 5R4GY/B should drop your biggest B+ by almost exactly 50 volts (relative to the drop you'd find using a GZ34) while drawing only a tiny bit more heater current than your GZ34. We would call those heater currents (1.9 A for GZ34 and 2 A for 5R4) "the same".


http://www.fourwater.com/files/fullrect.txt   All those 5V rectifiers that sit in octal sockets are pin-compatible. A 5U4 would also work for temporary test and probably be easier to get (super common tube used in TV sets) but there is a piece of concern that this tube draws 3A filament current vs the 2A of your GZ34 or a 5R4. A 5R4 is the way to go to whack your voltages some.   


With this amp and the very high voltages, IMO you want to stay with a tube rectifier and not go silicon.



If you install 1 ohm resistors under the cathodes of your KT88 tubes, you can (and IMO should) keep them there. Many people bring those points out to tip jacks that can be measured with the chassis installed in the cabinet because a voltage reading of volts at that junction (sitting on top of the 1 ohm R's) in millivolts translates not perfectly but very closely to the tube current in milliamps which is what you want to "keep an eye" on when setting proper bias.


Now for some subtleties: Most people would prefer not to have 50 ufd for the first filter cap, as technically, that is more than the max allowable. What's less than desirable about the oversize cap in THAT position is that on initial charge-up the current draw can be more than the rectifier wants to handle and has been known to cause flashover. On the LATER caps, the ones for the preamp stages, some folks believe this will make the amp sound too stiff. I myself am not a big zealot for either of these admonitions, I'm just mentioning them.


However, I will say this: I am not familiar with your "800 volt" cap setup, but since 800 volt caps are not available, you must have made them up using series-connected 400 volt caps. OK. BUT: It is important that you install what are called "balancing" resistors across caps so connected. Go look at how this is done on the first power supply caps (70 ufd +70 ufd) in a Fender AB763 Twin Reverb supply in Doug's schematic library. This *is* important. Usually 220K 2 or 3 watt resistors are used. This is important because small differences in those caps you used, even if they are new, can result in weird things and voltage imbalances especially on initial turn on or stby to "on". I would assume such bal resistors are on your board but I don't know that for sure. 

[sluckey]

Thanks Sluckey. Could I use a 5R4 with my first filter cap at 50uF?

I would not. I have a 20µF for my first cap and I'm cautious about that. If I was using the amp in a band, I'd stay with the 5AR4/GZ34. The Amperex tube data says the 5AR4 max input cap is 60µF.

All the tube manuals I have seen talk about using 4µF with the 5R4. Most don't say if this is typical or maximum. One data sheet does say 4µF is a max value. And the RCA manual says for values greater than 4µF it may be necessary to increase plate impedance to limit peak plate current to the rated current value (ie, use series resistors between PT and rectifier plates).

Why all the talk about 4µF? All this tube data for the 5R4 originates in the '50s. 4µF was a very common value for oil filled high voltage filter caps as found in military and industrial equipment at that time.

I recommend you stay with the GZ34 because of your 50µF cap farm (guessing you got that idea from Triode). The GZ34 also has a very slow warm up time which is a bonus. Those KT88s can handle 550V easily.

You may consider using a simple bucking transformer with this amp to lower the line voltage. Look at page 2 of this document to see the one I built. For this amp I'd use a 12VCT transformer rated for 4 amps. AES has one for a reasonable price. Here's my document...

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

[Hellawatt]

Here is a picture if my cap farm. I copied the triode schematic, but used fiberboard and some radial 100uf/400v caps that were about $2 each. It does have balancing resistors.

[Hellawatt]

I'm thinking about building a bucking transformer, but I only need to drop 3VAC to get to 117 which is where my heaters will see 6.3v. That is where I want to be for max tube life right?

I built one of those bias probes that connects to your voltmeter. I have been using that, but when I blew a tube I had been testing different rectifiers and I got lazy and only set the bias to -55 without checking the current.

I guess my question is this: if biased at 70% or less of max dissipation, my kt88s should be able to handle 120VAC at the wall and a silicon rectifier, and it will give me increased headroom and power. Right? Or will components fail or drift because of the higher voltages? I don't want to buy any more power tubes than necessary.

Thanks for all the help everyone, you guys are awesome

[Hellawatt]

50uF power supply caps for the preamp is a bit large. Closer to the values Sunn used would be better for tonal purposes. The preamp filter caps are RC coupled to the preamp stages so a change in cap value changes the tonal spectrum of the stage they supply.

Greg, I read a lot of your old posts on sunnforum that helped inspire me to take on this project. They were a huge help. How exactly does the preamp supply cap affect tone?  I've read that smaller caps were used back then because larger caps weren't available, but what benefit is there to using smaller caps? 50/50/50/50 is recommended for bass amps but not guitar amps? Is that because people want the guitar amps to break up? Or does it affect the EQ?

[sluckey]

I think a lot depends on how you use the amp and what you expect from it. The Sunns were designed to be big, powerful, loud, and clean. They do pedals well. If that's what you're after then all you need is a couple JBLs to go with it. You don't need silicon rectifiers, or a bucking transformer, and your cap farm will work well towards that loud, clean sound. But if you are looking for an amp with a growl similar to a Marshall amp, you'll probably be disappointed.
 

[Hellawatt]

I think a lot depends on how you use the amp and what you expect from it. The Sunns were designed to be big, powerful, loud, and clean. They do pedals well. If that's what you're after then all you need is a couple JBLs to go with it. You don't need silicon rectifiers, or a bucking transformer, and your cap farm will work well towards that loud, clean sound. But if you are looking for an amp with a growl similar to a Marshall amp, you'll probably be disappointed.

Yes I want that loud clean sound! I'm going to run pedals so I want all the headroom I can get. I just like the idea of silicon rectifiers because they are more reliable than glass, and I wouldn't have to carry a backup 5AR4.

[SoundmasterG]

Greg, I read a lot of your old posts on sunnforum that helped inspire me to take on this project. They were a huge help. How exactly does the preamp supply cap affect tone?  I've read that smaller caps were used back then because larger caps weren't available, but what benefit is there to using smaller caps? 50/50/50/50 is recommended for bass amps but not guitar amps? Is that because people want the guitar amps to break up? Or does it affect the EQ?

You're welcome....glad I could be of help. Smaller value power supply caps were used back then because caps were very expensive. Nowadays they are cheap so you can use larger value caps if you want and it won't cost much. Sometimes you can pick a value of cap that is too high and it causes the stage that supplies to develop too much bass response, which can cause motorboating and other issues. In the preamp and phase inverter you don't need much bass response, and if the amp has much distortion then reducing bass response is a good idea as it goes from the power amp to the preamp. Sunn amps have a TON of bass stock, so increasing that...well I don't see why you would want to, but it is up to you. If the amp functions fine and you like the sound of the larger caps for the preamp, then it is correct for you.

I primarily use my Sunn amps for bass, not guitar. In fact I don't really like Sunn amps for guitar at all.    So in that respect I want loud and clean and tight and punchy bottom end. So that is why I recommended the values I did on the Sunn Forum....110uF, 50uF, 20uF, 20uF.....notice I left the ones supplying the phase inverter and preamp at the stock values....because since they are RC coupled (resistive capacitive---more later), changing the value of the e-cap there changes the tone of the stage those caps supply. I'm not sure where you got the all 50uF values from? Using the larger caps at the plates and screens gives a tighter and more punchy feel for bass, but it doesn't change the tone directly...it just changes the feel of those stages. Using the larger value cap for the first filter isn't safe to use a tube rectifier anymore however. The 50uF that you chose will work with a GZ34 tube rectifier, but that is about the only one that can handle that value, and with today's poor quality GZ34's, they may not like that value. Vintage ones said they could handle a 60uF cap on the input and certainly the old Mullards will take it but with today's tubes it is anyone's guess. The best quality GZ34 today is probably the Ruby tubes one according to Stan at Ampage. He would likely know as he is pretty knowledgeable with all things tubes. I use solid state or the Weber Copper Cap WZ34 or WZ68 (solid state rectifiers that emulate the voltage drop of the tube GZ34) in my Sunns and don't use tube rectifiers in them anymore.

RC coupling refers to the way the caps and resistors are connected around the preamp and phase inverter stages. The power amp is transformer coupled...so the plates of the power tubes have a transformer wire attached to them and any signal the power tubes put out goes directly into the output transformer and to the load. The phase inverter and preamp have resistors attached to the plate, and a capacitor connects between stages to couple the signal from one stage to another. With the resistor connected to the plate, the output signal is developed across this resistor from the current flowing through it. For the preamp nodes in the amp....there isn't much actual filtering left to be done at this point in the amp so the capacitor, along with the dropping resistor feeding it, establishes how the amp reacts (time constant). A smaller cap value will discharge more quickly but also charge more quickly, giving the amp more apparent punch and touch sensitivity...the pick attack will stand out more. There is also some bass cancellation between the preamp stages through the power supply that smooths the bass more with a lower cap value. Higher cap values reduce this bass cancellation and can make the tone more boomy...this may be a good thing in a larger amp sometimes, but other times not....too much bass and too abrupt bass will make it muddy. A larger cap value will also de-sensitize the amp somewhat. As I said above, I prefer to leave those preamp and phase inverter nodes stock on my Sunns but I change the plate and screen to higher values to get tighter and quicker bass. With the values you chose, you would get tighter and quicker bass in the first two stages, and then quite a bit more bass response out of the phase inverter and preamp. That said, as I was saying earlier, the values that you picked won't make the amp non-functional or blow a KT88. (unless you had severe motorboating and the tube ran away as a result, but you would likely have heard the motorboating and seen the tubes red plate.)

Greg