Biasing Sunn S200

[dude]

Been a while since I've been here but I'd like to ask a question on biasing this Sunn S-200 head.

Attached is the schematic.

I'm replacing the power tubes ans would like to bias then correctly. I see the the bias voltage on the schematic is -55v. I checked the voltage on the old tubes and it was -72v, turning the biasing adjusting pot, 10k trim to get -55v I notice that this changed the B+ in the amp quite considerably about 25volts. Why does this happen and is this normal?

I don't have a scope and I usually install 1 ohm resistors off the power tubes cathodes to ground and check the current draw here to bias tubes.,for 6550s what is the correct mA's draw?

Can one just set the 10K trim pot to read -55v and be done? I wouldn't think so...? And why not?

One other question, I noticed the can cap has been replaces with four 500v/20uf filters, the schematic calls for 525v and one 600v but the amp seems to sound OK, are these 500v filters OK? The B+ is around 500dc but changes the negative voltage in the bias can raise the B+ to 525v...

[Iannone]

Plate voltage changing in response to changes in negative grid (bias) voltage is, in fact, normal.  Keep in mind that the OT primary has some resistance.  As the amount of current flowing through the plate of the tube (and, therefore through the OT primary) changes, the voltage drop accross the transformer winding will also change, as per Ohm's Law.  So, as the negative grid voltage goes up (i.e., more negative), the voltage drop will also go up and the plate voltage will come downward.

The correct plate current for your tubes will depend upon the operating conditions, i.e., the plate voltage.  The date sheet tells us that a KT88 has a maximum plate dissipation of 42W.  What you want is to adjust the bias voltage so that (a) the amp sounds good (whatever that means to you) and the idle plate dissipation does not exceed 70% of 42 W (i.e., 29.4 W.)  Plate current x plate voltage = Power dissipation.  Figuring around 5mA being drawn through each screen, at 480 V on the plate this would be about 66 mA of cathode current.  This may or may not be at -55 V of grid current.

Given that you may or may not have in excess of 500V at the plate of the first filter cap, I would not feel really comfortable about filtering that node with a 500 V cap. 

[dude]

Thanks Iannone

I have 6550's instead of KT88's so I assume I find the maximum plate dissipation for a 6550 and bias in mA's across the 1ohm R to about 70% idle current, minus the screen draw about 5mA's? and as far as the posted neg bias voltage just forget that and go with about 70% of the tubes draw at idle.

Where can I get the specs for 6550's?

Do they make filters with higher voltage ratings, 500v seems the max ?  I guess I could use two 300v filters in parallel (this correct) but wouldn't that be six caps stuffed in there? I've stayed away to long from the iron, I'm a little rusty, guess I should go back to Akiens site and read up.

[rzenc]

Thanks Iannone

I have 6550's instead of KT88's

Pretty much the same bottle...Katties having more aggressive ratings, i.e.800VDC at plate...whilst keeping G2 under tight control.

I assume I find the maximum plate dissipation for a 6550 and bias in mA's across the 1ohm R to about 70% idle current, minus the screen draw about 5mA's? and as far as the posted neg bias voltage just forget that and go with about 70% of the tubes draw at idle.

6550 data sheet:
http://www.tubebooks.org/tubedata/HB-3/Receiving-Type_Industrial_Tubes/6550.PDF

Do they make filters with higher voltage ratings, 500v seems the max ?  I guess I could use two 300v filters in parallel (this correct) but wouldn't that be six caps stuffed in there? I've stayed away to long from the iron, I'm a little rusty, guess I should go back to Akiens site and read up.

In order to add voltage ratings, caps must be serial, not parallel. Series arragement leads to less uF, i.e. 2x100uFin series leads to 50uF, whilst 2x100uF in parallel leads to 200uF and voltage rating is the same - caps rated at 500VDC will still be able to handle it... althought I wouldn't be confortable with such proximity... on the other hand, 2x100uF @ 500VDC leads to 50uF@~1000VDC...

Hope this helps
Best Regards

Rzenc

[dude]

"In order to add voltage ratings, caps must be serial, not parallel. Series arragement leads to less uF, i.e. 2x100uFin series leads to 50uF, whilst 2x100uF in parallel leads to 200uF and voltage rating is the same - caps rated at 500VDC will still be able to handle it... althought I wouldn't be confortable with such proximity... on the other hand, 2x100uF @ 500VDC leads to 50uF@~1000VDC..."

OK thanks, then to beef-up the rating I can use two 350V 40uf filter caps in series (- to ground on one 350 40uf cap, the + to the - on the next 350v 40uf and the + to my B+ line). This would give me a filter rated at approx. 700v and 20uf, correct ? No balancing resistors or anything needed?

I sure don't want to do this unless I got the wiring down pat. :shocked: literally

dude

[Iannone]

Given the current draw from a pair of 6550's, I would want more than 20mF worth of filtering.  I'd suggest a pair of the 100mF/350V caps that Doug sells (for 50mF total capacitance at 700 V).  Balancing resistors would also be a real good idea.  As protection goes, you can't get too much cheaper than two resistors.

[PRR]

> set the 10K trim pot to read -55v and be done? I wouldn't think so...? And why not?

That IS how we did it in the old days. When tubes were very consistent, and we did not complain so much about the cost. (They are actually cheaper today, but we complain more.)

> d it was -72v, turning the biasing adjusting pot, 10k trim to get -55v I notice that this changed the B+ in the amp quite considerably

Under this 480V Plate and Screen condition, at -72V the tubes are nearly "off", near 15mA. At -55V about 70mA. For two tubes this is a 100mA increase of suck on the power supply. Real PTs have losses, the hollow-state rectifier has losses.... voltage must sag.

Assuming 480V and 70mA we have 33 Watts dissipated in a 35W-42W plate. A Dyna/Tung-Sol 6550 will stand that for decades. The GE/KenRad and Sovtek 6550s I played with seemed very happy with 40W for days.

However since we are running very close to rating, and tube consistency is dubious, and OTs (not tubes) are brutally expensive, you should wire your 1-ohms and measure. A push-pull amp does not need to be run as-hot-as-allowed, and a UL amp will tend to mask the effects of cool bias and tube unbalance. I'd aim for 35mA-55mA per tube. And I have seen other amps worked as cool as this one is with -72V bias and nobody minded.

> the amp seems to sound OK, are these 500v filters OK?

It may be OK for years. But it's awful close to the edge of a major smoke-out.

> schematic calls for 525v and one 600v

There is a wicked kick when the C-L-C filter powers-up. Lesser makers might have gone cheap with 500V caps, Sunn did the right thing.

> filters with higher voltage ratings, 500v seems the max ?

When you put aluminum in borax and make an oxide insulator, you get 400V easily and 500V if your chemistry and thoughts are very pure. So most electrolytic caps are 450V tops. A few go 500V. Some are marked 525V.

The "600V" cap which Sunn used is almost certainly two 350V caps inside one cardboard sleeve. That was once a common part; no more.

I would not stray far from Sunn's cap values around the choke. It gives excellent ripple control, and over-sizing the cap changes time-constants. A C-L-C filter is a major resonator. If you stray too far you can get in trouble. Keep the sum (actually product) of the two caps around the choke within 50% of the design values. The ones along the resistor string could be larger than Sunn used, but no reason to get crazy.

For LOTS of capacitance at other then 450V, the way to go today is "Snap Caps". Motor controllers, welders, and other heavy-DC boxes use heaps of these. Performance is excellent (cap failure in a welder is a major Warranty cost, and they hate that). Cost is low. If you get them from the same lot, and don't push the rating, they do not need equalizing resistors.

www.Mouser.com
SLP470M400A1P3 
Cornell Dubilier - Aluminum Electrolytic Capacitors - Snap In 47uF 400V +105 C
1:  $1.68   
10:  $1.30 

Two in series will be 24u, good for over 600V. Because you use them well under both voltage and temperature ratings, life should be very long, decades of typical stage-amp time.

You need eight to replace your four 525V 30u-20u caps. The 10-pack is cheaper than eight (and still may not reach Mouser's minimum order).

Diameter:  22 mm (0.87 in)   
Length:  25 mm (1 in) 
 
They are made to "snap in" a PCB. Get some blank no-copper board good for >500V. 2.5"x5" will easily fit 8 caps and leave edges for screw-holes. Drill per the spec-sheet, snap-in, then bond the series-pairs with scrap wire. I would bond the first cap-pair ground to PT R-Y, the next two pairs to power tube cathode return, and maybe take the "C" pair to the input jack, depending how Sunn originally grounded.

[PRR]

The cap at the wall-cord, the whole wall-cord, SHOULD be be changed to a proper modern 3-pin cord and direct connection from 3rd pin to chassis. Then the 0.05u 600V cap and the polarity switch should be removed. That was "necessary" in days when most wall outlets were 2-pin, but those days should be behind us. The ground-cap is a possible safety problem.

[SoundmasterG]

Depending on the vintage of your 200S Sunn amp, the voltages will either be around 480-490v at the plates, or perhaps as high as 560-570v. Obviously the higher voltages exceed the 525v rating of the can cap. Moreover, can caps are expensive when compared to discrete caps, with radial, discrete caps being the cheapest and smallest these days. I've advocated replacing the can with discrete caps over at the Sunn bbs at ampage for several years now because it is cheaper, and if you are careful with your grounding in that you locate the caps and/or their grounds closer to the actual stages that they are supplying, then you can get better performance than having all of your ground points for the caps scattered around the chassis. For bass amps I've recommended going with series-connected discrete caps for the first two filter stages....two 220uf 350v caps on the first stage will give you 110uf @ 700v capability for instance.....and two 110uf 350v caps in series for the second stage will give you 55uf @ 700v capability. Using larger caps as I've suggested in a bass amp will provide for a quicker and tighter sounding bass response, and the amp will hum less. For the other two caps, I leave the value at 20uf as changing the values for those will change the tone, though the third cap should still be 500v rated for the higher voltage Sunns. The F & T 22uf 500v cap works well here. The fourth filter can be a 450v 22uf cap. The caveat with going to larger filter cap values is that you can no longer safely use the tube GZ34 rectifier and must change to solid state diodes like the 1N4007, or a Weber Copper Cap for the rectifier (my choice for my own amp is the Weber Copper Cap WZ68). The earlier Sunn 200S used the Dynaco transformers and had the lower voltages. Once Sunn changed to Schumacher for their transformers, the voltages went up to the higher value. I believe the change was somewhere around 1969, though Conrad Sundholm (Sunn co-founder) would know for sure. The other caveat with using series connected caps is that there has to be a 220k 1w resistor in parallel with each cap so the caps will share the voltage and current loads. If they are not there, one cap will try to hog all of the voltage and current and blow.

I've done the cap mod to quite a few Sunn 200S and 2000S bass amps and the customers have been quite happy with the results. Others have done this change on their own. Some others have used a replacement board that uses larger value caps for Dynaco Mk.3 amps (which is basically what the Sunns started life as). Triode Electronics sells some of these boards. I generally only recommend to go with the larger value caps on bass amps, as the guitar amps get too stiff when this is done, at least for my taste. The guitar amps often have lower voltages too, generally around 515v with today's wall voltages, so the can cap is often adequate for those. One advantage to using the solid state rectifiers or the Weber Copper Cap is that the power transformer will run cooler since it doesn't have to supply the 5v filament current, and this is a good thing because many of those power transformers will blow. The GZ34 is pretty close to it's ratings in the stock circuit too, so you run less risk of having a blowout when things are changed over, especially with today's poor quality GZ34's.

Greg

[dude]

Thanks everyone, I'm getting a good education here.

I'll bias the 6550's, 70% at idle doing the math when I check actual B+ for the correct mA's reading across the 1 ohm resistors.

I like the idea about using a higher value filter off the choke for a tighter sound (I remember using high values like 2000uf in a cathode biased 18 watt amp for biasing the el84's, made the overall response much tighter like a fixed bias amp). I will be using the 200S for bass in a recording situation so it's the perfect wattage. How high can I go on the first filter using a 5AR4?

I don't want to change or mod the amp as it's a vintage piece made in 1969 and all original except the sloppy cap job when someone put the four Spragues (500v@20uf). I'd like to clean that up. I thought I'd use two of the 500v 20uf's for the preamp where appropriate voltage wise and use "PRR's suggestion using the snap-in SLP470M400A1P3 from Mouser but I'm not familiar with making a PCB bd to hold them. Does Mouser sell anything I could use?

This is a picture of the caps http://www.mouser.com/Search/include/LargeProductImage.aspx?path=cornelldubilier/lrg/slp-slpx.jpg. I have some old board material like Fender used in blackfaces, perhaps I could silicone the caps to that and make up some kind of band to hold them on the board too, attaching the loaded board with screws to the chassis or use standoffs. Of course all inside the chassis. Sound like a plan?

OK, on the series 47uf's (or maybe higher on the first filter), I should use a 220k 1w in parallel on "each" cap? attached is a drawing, is this correct.

thanks, dude       

[sluckey]

I would just replace the caps with a multi-section can (30/20/20/20 @ 525V) available at AES. Then it will look like it did back in the '60s.

If you're still concerned because of the voltage on the B+ rail, build a separate bucking transformer device to take your line voltage back to 110 to 115VAC and the B+ should come down to safe levels.

[SoundmasterG]

Yes, when caps are connected in series, each cap needs a resistor in parallel with it in order to share the voltage between caps. Your drawing is correct. I would caution you on using any caps larger than 100uf in the bias circuit itself, as that sets up a time constant that can be too large and the bias lags, which can cause problems. Since there is so little current in the bias supply, 100uf caps are just fine there if you decide to change from the values that Sunn used. I believe the stock ones in the bias circuit are 50uf each.

You could experiment with different cap values if you like to see which values you prefer for the filters. For bass, I like the larger values. You could always put in the larger values, and then change later to stock values if you were going to sell the amp. It will record better with the larger values partly due to the less hum it will have, and partly due to the better bass response, but YMMV. The GZ34 can only take a 60uf max on the first filter, and with today's poor GZ34's, I would suggest lower than that....probably a 40uf max. The Weber Copper Cap is cool because it just plugs into the socket in place of the GZ34, so if you decide to go to larger than 60uf for the first filter stage, then I suggest the Copper Cap as a replacement. They make a WZ34 and a WZ68, with the WZ68 having twice the current capability. Either would be an equivilant replacement for the tube GZ34 should you decide to go that route. If you decide to go with discrete caps, Mouser sells a whole bunch of stuff you could use. Their Xicon radial caps (the ones with leads) are cheap and work great. I think they're little green ones like the picture you showed, but they have leads that you solder. I would go higher than 47uf on the first stage, even if you are trying to get to stock as 47/2 = 23.5uf, and the stock first cap was 30uf. Two 100uf in series would give you 55uf, which might be ok with a good GZ34 made today, and would be fine for a vintage one. I like the Copper Cap and higher filters myself, but YMMV.

As I mentioned, the problem with the can cap is that it will probably be over voltage. You can go with a seperate bucking transformer device as mentioned to get your wall voltage down to 110v AC and then your voltage on the tubes will probably be in a safe range with the can cap, but that would cost a lot more than just using discrete caps and take more time to boot. I've seen two amps have problems with those can caps not holding up to the higher than their max voltage that the amp puts out these days. Luckily for the owners, they caught it quick enough before any transformer or tubes blew, but it is a concern you should be aware of. If you just put discrete caps in and leave the can in place but disconnect it, it will look like it did back in the 60's, but will have better performance and sound, less hum, and be more reliable.

Cheers!

Greg

[dude]

Yes, when caps are connected in series, each cap needs a resistor in parallel with it in order to share the voltage between caps. Your drawing is correct. I would caution you on using any caps larger than 100uf in the bias circuit itself, as that sets up a time constant that can be too large and the bias lags, which can cause problems. Since there is so little current in the bias supply, 100uf caps are just fine there if you decide to change from the values that Sunn used. I believe the stock ones in the bias circuit are 50uf each.

I'm sorry, when I mentioned using a 2000uf cap in the bias circuit, I was referring to a different amp not the Sunn. I have used a very large cap in the bias circuit in a "cathode biased" amp like the Marshall 18 watt with two el84's to tighten the tone. 

I do hear you on going solid state and the Weber cap can, it wouldn't require any wiring changes.  The schematic calls for a 30uf @600v but I will take your advice and go to 50uf using two 100uf's in series @ 400v that way I could use a tube or the Weber Copper cap.  Thanks for confirming everything.

Not that I want too but wouldn't using some Zeners off the ctr tap lower the B+ too? I guess it might be better not to put anything between the center tap and ground and use them off the power supply...? I have used them before with good results but they do get hot.

Do those bias caps (50uf@150v) need replacement after 40 years like the big filters usually do?

I searched the forum and found a lot of good info covering how to install filter caps in series amongst other great info.

thanks,
dude     

[6G6]

I think it is safe to say that any 40 year old E cap is due for replacement. 

[SoundmasterG]

I'm sorry, when I mentioned using a 2000uf cap in the bias circuit, I was referring to a different amp not the Sunn. I have used a very large cap in the bias circuit in a "cathode biased" amp like the Marshall 18 watt with two el84's to tighten the tone. 

In cathode bias it is ok but in fixed bias I wouldn't exceed 100uf for the caps used in the bias circuit.

Not that I want too but wouldn't using some Zeners off the ctr tap lower the B+ too? I guess it might be better not to put anything between the center tap and ground and use them off the power supply...? I have used them before with good results but they do get hot.

I think the zeners would probably work fine, but they would get very hot, and it is just as easy to use series connected caps for the first two stages in the filtering. With older amps like this, when you change to a different arrangement you sometimes have to add a terminal strip to properly secure the new parts, and this may require drilling a hole. You will have to look at the amp and see if it is required to put the parts in there safely. Silicone isn't really a very good solution though it will work. The Sunn is a different setup than many amps since it is UL, and I'm not sure if that would change things are far as using a zener like you suggest, but it is worth looking into for sure.

Do those bias caps (50uf@150v) need replacement after 40 years like the big filters usually do?

Yes, any electrolytics should be replaced when that old as the other poster recommended.