View Single Post
  #8   Report Post  
Posted to rec.audio.tubes
[email protected] themend@gmail.com is offline
external usenet poster
 
Posts: 18
Default Restoring a Sherwood S-5000

While we wait to hear from JRC, let's go over the parts list for the rebuild.

First, let's talk about the large twistlock cap right behind the output tube section. This is the main filter capacitor for the HV+ supply, and it takes the brunt of the wear in the power supply. The reason for this is mainly because the capacitor is located in a spot which positively guarantees that it will get baked by the output tubes. Here is a picture of it:

https://www.dropbox.com/s/2lq2ltcke4gt11w/S-5000_C1.jpg

It's also peculiar in that it is mounted at an angle, just as the power tubes are, which is what gives this amplifier such a low height. There are a few schools of thought on twistlocks like these. Some people prefer to cut them open from the bottom, gut them, and stuff them with new radial or axial caps. The other notion is to simply put those same caps underneath the amp and re-wire everything to accommodate that. Some feel replacing the whole can is best when there are replacements available. Lastly, some people feel that they can magically 'reform' capacitors by doing some sort of rain dance with a variac. As you may have guessed, I do not fall into that last group of people.

While there probably are certain capacitors in certain amps that may respond to such a thing, this particular cap in this spot sees too much voltage and too much heat not to be somewhat degraded by now. Some of these S-5000s were never ran with the cover, since the cover was an additional cost; not included with the amplifier purchase. Lots of these were placed into consoles and the like and so did not build up all that much heat. These ones can often sound OK with the original cap. The S-5000 running inside its cover, especially a leatherette covered one will just cook away in there. These almost always have noticeable hum when played.

If we look underneath the cap he

https://www.dropbox.com/s/pefyhgh7l6...onnections.jpg

We see that there isn't a whole lot of room for adding caps underneath, without resorting to putting them in odd places, and adding new wires and terminal strips. So really, our best options are to either replace the can or stuff it with new caps. Since there are very good replacements available from CE, I suggest the extra money spend on this:

http://www.tubesandmore.com/products/C-EC4020X2-525

Is well worth it, especially considering the effort in re-stuffing a capacitor. Note that this capacitor is a 40/40/20/20 sectioned cap at 525v. The original Sherwood design called for a 20/40/40/5 sectioned cap at 500v for the first three sections, and 350 volts for the 5uf section. This CE cap is physically the same dimensions as the Sherwood cap, offers 25 more working volts, and gives us an extra 15uf on one section. As long as we are careful to wire the 20uf section to the 5AR4 rectifier, this cap is an ideal replacement.

Speaking of the 5AR4 and 20uf, there is a reason you must not increase the value of this particular cap. Tube rectifiers have a maximum capacitance that can be directly connected to the output. Those values are usually listed in tube manuals. If you add more here, the rectifier will likely arc over due to the current draw. After the first dropping resistor, you can literally add as much capacitance as you want, as the current rush will now be limited by that resistor. It is possible, however, to introduce a resonance by altering the R-C time constant by doing this. I have seen one S-5000 that wound up with a very low frequency power supply oscillation as a result of an odd re-cap, it showed up as very pronounced woofer excursion at around .5hz, which one might think would be beyond the capability of the output transformers, but obviously not.

So, in this case, everything will retain its original values except for the last section which will increase to 20uf from 5. I personally feel that 40uf is just adequate for the first section off the dropping resistor. Depending on how the terminals of this cap are laid out, I may decide to use the additional 20uf section to increase the first 40uf section to 60uf, and simply add a 10 or 20uf cap underneath for that last section. Since I have not used this particular cap before, I will have to wait until I have it on hand to see if the terminals and distances underneath will allow this. In the past, I have used Aero-M capacitors for this part, and was able to add capacitance easily.

For the second twistlock capacitor, I'm going to try this:

http://www.tubesandmore.com/products/C-EC20-20-450

Here's the original in the amp:

https://www.dropbox.com/s/kvxfs3grlzzpqfw/S-5000_C2.jpg

This is interesting - last time I did one of these S-5000s, there were no fresh stock 1" twistlocks available. Here we have a 20/20 @450v that is the same dimensions as the stock capacitor! The stock cap is a 20/30 @ 400, so we get a higher voltage rating with this cap, but come up 10uf short. However, there is plenty of room underneath next to this cap to add another 10uf or more. So, I'm going to order one of these to find out if they're suitable.

Lastly, we have the oddball bias supply capacitor. It has a common positive can, something that I rarely see anywhere else but on Sherwood amps. Anyone else know of amps that use these? I'm curious! As a side note, if you are ever working on a Sherwood and you see a twistlock capacitor can that is covered in cardboard, that can has high voltage on it and that cardboard is there as an insulator. Be careful with those!! Luckily, they don't exist on the s-5000 but you see them a lot in later Sherwood amps.

Anyway, the bias cap... It's a 250/150/50 @ 50v cap, and the best bet is to just leave it in place, and use axials or radials underneath to replace it. We don't need to go crazy here with capacitance, maybe a -little- more on the bias supply to lower ripple, but the vaules here seem pretty acceptable in practice.

So that's the update for now... I'm going to order the parts tomorrow and we'll see next week how they work out.


-forkinthesocket