Sherwood S5500 II w/7868s- take two

[oldman55]

I have another S5500 II on the bench. My first one has been playing daily for a few weeks now in its pristine original factory condition, down to the cover screws and tubes (except maybe the Tele tubes in the phono stage) and sounds fantastic. I need to pass it along to someone that appreciates that aspect. I'm more into the looks and sound, which I found in the Sherwoods.

Searched and found all the rebuild threads I could (including Gadget's mod thread) but never could find what most do for finding something close to the 120/30/40 can.

Since my bias voltages are on the money, I was thinking about not replacing that can and only replacing the HV filter cans and paper ecap.

The schematic it came with has 7591s so I am guessing that it has the same circuitry and voltages? Its a 15000s serial number.

oldman with another Sherwood

 

[gadget73]

mine has new caps underneath for the bias cap. Its the only one that was bad. The other cans are original still.

7868 and 7591 are effectively identical other than the socket they plug into.

 

[oldman55]

Given size and budget constraints in HV capacitors that were in existence when originally designed, I am sure that today those engineers would have chosen higher values in many cases.
What would be a better value(s) for the first 120uf lytic and the following two cans on the HV+ rail?

I figure that I have an issue with the second can as I only get 66v at the end of the line but it could be a resistor too. Interestingly, it was wired incorrectly with the 25/250 can fourth in line (should be the 6th/last) in lieu of the 30/400.
The voltages at the first can are in line but I dont see a point in replacing only one of them.

 

[gadget73]

Is that the voltage doubler cap, or the one that actually sees full voltage? For the doubler I might make that 150-220uf. The first cap I think is stock at 40, honestly its probably fine close to stock value. I'd probably use something in the 47-68 range. With my S-7000, I kept all the values reasonably close to stock, just making allowances for common modern values. I think my first caps are 150, the 40s got changed to 47, 30 are now 33, etc. It works quite well.

Could be bad resistors, but internal leakage in the caps will draw it down too.

 

[oldman55]

Thanks
What about this arrangement? Two cans and a radial.

 

[gadget73]

I would not replace that first 120 with a 50. The top and bottom one in the doubler also should be the same value. Other values look fine. Also keep in mind that the upper 120uf will have the negative side at 1/2 B+, some 220 volts DC. Make sure the outside of that cap is insulated from ground. It also cannot be a capacitor with a negative shared with anything else.

 

[oldman55]

Its a 50 in the can supplemented with 100uf radial. See additional info below.


Just trying to get some stock cans that would work.

 

[gadget73]

Value wise it would work, but it has to be a capacitor that does not share a negative with anything else, otherwise it will not function correctly. If you look at it, that upper 120uf cap (C61) does not go to circuit ground. The lower one does, and you can combine that capacitor with another one that shares a negative in one can. Pretty sure the stock one did.

 

[oldman55]

Had to scrap my power filter can idea. The 40/40/40/40@475 is only rated to 55 degrees. And too tall.
Usually use tubesandmore but looking for a supplier with a larger selection of multi cans. Cant find a larger selection online. Anyone?

 

[Brice]

Have you look at Authenticap? I know they have a KTL25 80/40/30/20 @ 550V

https://www.die-wuestens.de/eindex.htm

 

[gadget73]

Hayseed Hamfest may be an option. They can custom make them for you, and they use 105c rated caps inside the can.

 

[oldman55]

Hayseed Hamfest may be an option. They can custom make them for you, and they use 105c rated caps inside the can.

I have a request in to them for a 4 section can so that I can use the inexpensive 50/50@500 from JJ (with supplementing the first 50 w/parallel radial) and then one of Hayseed's to cover the last four in the chain.

On the bias supply, I was going to add a trimmer to get some adjustment. Not sure of the best way so that if it fails, tubes would be safe especially since I am leaving in the old bias multi cap can. Options:

 

[gadget73]

10K trimmer with an 18K resistor in series in place of R73 would get you there. Should allow it to go a little more negative than it does now, and if the pot failed it would just go full negative. It should end up roughly center in the pot travel to put things to stock voltages.

That won't do anything for the bias caps though. If that dies you lose voltage.

 

[oldman55]

10K trimmer with an 18K resistor in series in place of R73 would get you there. Should allow it to go a little more negative than it does now, and if the pot failed it would just go full negative. It should end up roughly center in the pot travel to put things to stock voltages.

That won't do anything for the bias caps though. If that dies you lose voltage.

Used wrong schematic.

Edit: after review, this is about the exact same schematic as my previous post other than "r73" value (now R26). Except R 26 is in a very inconvenient location. I thought that on my S5000 I has added the trimmer similar to this:

Edit no. 2: It was on my other S5500 w/7591 tubes. Advice was:
"Would need to see a good shot of the bias supply portion of the schematic, but I'd wager it looks a lot like this. This is the bias supply on my S-7000

View attachment 942254

In this case it can be done basically in two ways.
1) change R140 to a 15K, and add a 5K trimmer in series with it.
2) change R150 to 20K and add a 5K trimmer in series with it."

Guess this answers my question except I was going to go with a 16k in lieu of 18k (R60) and a 5K pot. With my proposed changes, if trimmer fails and goes to 0, would the negative voltage go closer to 0 or go more negative?

My mind is starting to go...I've done this before, and not that long ago.

Oldman

 

[crispycircuit]

My mind is starting to go...I've done this before, and not that long ago

Get a large Dunkin Donut coffee and your mind will wake up.

 

[gadget73]

If you have a choice about it, I'd put the pot on the one going to ground. That lets it fail without a tube melt down.

The fixed resistor value isn't super critical, but you want the fixed resistor + 1/2 the pot value to be fairly close to the original resistor value. If you want to use a 5K pot and the original resistor is 27K, I'd use a 25K fixed resistor. Mid travel on the pot will make a total resistance of 27.5K, which is close enough to the original value. General idea is that if you can put the pot in the middle, it will act like the stock design did but give you adjustment in both directions.

If you need more adjustment range, use a larger value pot but expect it to be more touchy. Small value pot has less range but its less touchy. I happen to have 10K on hand, so I usually design stuff around what I have. In that case it would be a 22K fixed plus the 10K pot.

 

[oldman55]

I have already installed the pot in series. The resistor to ground would have taken some rewiring due to location and this one I just stuck directly to the can tab.

Guess I could leave in place, cut off one leg of the trimmer and run a wire to a new, 25k resistor from the other trimmer leg to a chassis ground near by. Its a 5k trimmer.

While I am at it, would it be worth my while to add 10 ohm cathode resistors and a balance pot for individual tube adjustment? I have a 50k trimmer. Would I just drop the 150k resistors down to 125K?


EDIT:
Or is this a better spot as not to mess up voltage going to other tubes?

 

[gadget73]

I'd do the 10 ohm just for monitoring purposes if nothing else. My S-7000 has one trimmer and 4 10 ohm resistors. I can't really adjust it any but the set was matched and has stayed reasonably so. All I'm really concerned about is making sure its not so far out of whack that something is going to melt, which it actually was before I put the resistors and the trimmer in.

The balance pot would function better in the first location. It wouldn't give you really any adjustment after the 150K resistors. If you're going to do this though, I'd split things up so you have a bias and a balance per channel. May as well go whole hog if you're going to add adjustments.

 

[oldman55]

Pinout data sheet shows pin 6 being control grid on the 7868 but it is unused. Can I use it as a tie point for ground to run the 10 ohm cathode resistor? Or is it connected internally to something in the tube?

 

[gadget73]

2 and 6 are internally connected to grid 1. If you put the 10 ohm resistor there it will short out both the input signal and the bias supply.

On mine, the cathode was tied to ground right at the tube socket. I simply removed the jumper and put the resistor in it's place.

Pin 8 is the only one sort-of unused but the note on the datasheet says it can only be used for a tie point of stuff near plate or screen grid voltage. If you had any inkling on adding 100 ohm screen resistors, pin 8 makes a perfect tie point for those.