Sceptre: Relocating Quad Cap replacement capacitor pairs: 60 Hz Hum?

[MrCreosote]

I am concerned about hum.

I've removed the quad cap and tore the ground ring from it so I could use it over for the various ground wire connections.  Then I located a solder lug terminal strip right where the Q/C was to connect the dividing resistors and associated connections.

So all the Q/C connections are located essentially where the originals were.

My plan was to remotely locate the cap pairs where they are near their primary component.  This makes their installation clean and simple.

However, 3 or these locations are fed by wires coming off the Q/C divider resistors that pass through long tightly bound wiring bundles.

In the original design, the feed wires coming from the Q/C are filtered and essentially DC which produces no hum when present in long wiring bundles.

Although my remotely located caps will result in a DC voltage along those feeds except that now, there will be a 60 Hz current present which I am concerned about.

If I can't have bundled feeds, then I have an alternative plan to relocate the caps for separate, isolated wiring.

I'm holding up completing the project until I figure this problem out or get some some feedback from someone who has already tried this successfully.

Thanks in advance,
Tom

Pic of Q/C divider strip layout:
Pic of Feed A remote cap layout:

[mckinnon audio]

   Hi there,on all the amps I've done I locate the caps close to where the orig. quad was.You can get 2 in between the terminal strip and the tremolo board and 2 more around the bias supply area.I use a 200 watt gun and put a solder pad right on the chassis and solder the neg. end of the cap to it.Some tie wraps and some heat shrink and you can wire right to the end of the harness and resistors.Seems to work fine,I haven't noticed an increase in hum,good luck,Mel.

[Soundmasterg]

The ideal location for an electrolytic cap is close to the stage it is supplying. When you are dealing with a pre-existing design like a Sunn, often it is impractical to locate the electrolytic caps close to their respective stages because the layout of the rest of the amp compromises this. This is the case with almost all of the Sunn tube amps, and it works fine to locate the electrolytic caps close to the can. This way you save yourself time and effort, as you can use most of the pre-existing wires, or you might have to install a terminal to extend some of these wires to wher eyou locate the given cap, but it is the best way to do it for the Sunns without totally redesigning the amp. I've done it both ways and although the initial experiment was fun locating the caps next to their stages, in the end it was better to locate them by the can cap. It took less time and effort and gave an acceptable amount of hum.

Greg

P.S. If I was making an amp from scratch, I would locate the electrolytic caps close to their respective stages without hesitation.

[MrCreosote]

Greg,

When you say close to their respective stages, do you mean the various tube or transformers being supplied?

If that is the case, 3 or the 4 locations are very easy.

Tap A, the RED lead from the output transformer was very neat as you can see in the picture.

So, you're saying what I did for A was OK?

Two of the other stages, C and D, are the two 12xxx tubes, each of which has a terminal strip sufficient to hang the caps off of.

B is the 6xxx tube which is buried below the 2 main terminal strips.

____________________________

I can see locating the caps near the component that requires the power because when the component needs to get current from the caps, there is the shortest lead possible.  Best to make the long lead the one supplying the caps.  So I guess that is what you meant by keeping the caps close to the stage they supply.

[Soundmasterg]

Yah, I replied to you in the PM that you sent....but use the 12AX7 input tube as an example....the e-cap supplying that stage should ground at the same point as that stage, and should be located as close as you can get it to that stage so you have a minimum time constant to avoid oscillations. You'll find in the Sunn that grounding is the main concern because most of the circuit grounds are by the can cap...but they also ground over by the input tube, and they get away with multiple ground points probably more due to luck or experimentation than anything else. The location of the output transformer isn't ideal either, but again, it works. To locate the e-cap by that stage, you need to run a long wire supplying the DC, and you may need to run a long one back to ground near the can...it depends on how the amp responds....(oscillations or not)
All the work I've done in the past has been on the Sunn bass amps, and they have lots more room inside.

Greg

[MrCreosote]

Well, I remotely laid out the rest of the feeds.

A, C & D were very easy to locate close to the stage.

B had to be a "second story" layout basically on top of the terminal strip I located right at the cap can location.

What was neat was that A, B, and D each had an unused solder terminal right where needed for the series cap connection so it was possible to simply wire in the caps and resistors.  C required a new terminal strip that I bolted onto one of the output transformer mounting screws.

I used existing feed and ground wires so we shall see what happens when I throw the switch.

I checked resistance to ground for A-D and the numbers were around 100k give or take, so I think I have not screwed up.

I could try a slow, variac powerup w/o 6550's but I'm just going to throw the switch - let the fireworks fly (! - I sure hope not.)

Here are 2 more pictures completing my installation:

[MrCreosote]

Oh No.

Just learned about "caps forming" and I simply must do this since this amp has been OFF for about 15 years except for when I turned it ON a year ago and smoke my quad cap in seconds.

If I had CAPS FORMED my amp, that may not have happened.

I'm going to leave this topic here and start another thread - actually, Caps Forming should be a sticky since we are all dealing old electrolytic capacitors many of which have not been powered up for many years.

[Soundmasterg]

If you use new electrolytic caps instead of old ones for your replacements, then you don't have to worry about forming the caps. If you use old electrolytic caps and try to reform them....well you get what you deserve there if they blow up.


Use new electrolytic caps is the moral of the story....

Greg

[MrCreosote]

I had read on this forum that there has been a lot of trouble with the quad cap, however, I bought my Sceptre from a piano player from a popular local band who used the amp with his Fender Rhodes on the job for years.  I used it for at least 5-10 years after that.  The quad cap only failed after sitting inactive for about 15 years.

What is interesting is that part of the design requirement is that the device should be able to sustain periods of non-usage.  I believe this is the harshest "load" a capacitor ever sees.  I suspect that if a capacitor is kept in operation continuously, its voltage rating could be reduced.

I agree that with properly selected new parts from Digikey, you shouldn't have to worry about anything, however, anyone who buys one of the new, reproduction quad caps, it might not be a bad idea for form it.  Especially when you don't know how long it has been sitting on the shelf.

I'm just wondering what problems I may have with the smaller electrolytic caps in my amp...  so I'm definitely going to bring up the voltage nice and slow.

[Soundmasterg]

Most electrolytic caps, including the cans, have a date code which signifies when they were made. Given that the manufacturers know all about how electrolytic caps degrade with non-use, I would suspect that they try to control their production to make sure they are making just enough for their sales so they don't have to throw out caps that get too old from sitting on a shelf. Any amp that sits for 15 years will have suspect electrolytic caps...its just the nature of the beast. Sometimes you get lucky and they are ok, but in most cases they aren't. This would go for older paper coupling caps too as they get moisture in them and degrade. The old Astrons that were used in tweed Fenders are famous for this. Any time I get an amp in for repair, I try to find out the history of the amp because this will tell me a lot about it's use or non-use, and then I get a better idea of what repairs it might need before I even get into the amp.

The problems with the quad cap have been due to the increased wall voltage these days, and that the cap is only rated for 525v. The amps are usually 560v and up, so that is not good to go over the ratings of those caps. The other problems with the quad cap have been that in the beginning of the production of the new one by CE, they didn't make a very good product. There are still some people that have had problems with one section in that particular cap going bad prematurely. Personally, I prefer to use discrete caps as that gives a better and cheaper selection, and better voltage handling....but you can do a search and find my recommendations in a few places so I don't need to go into it here again. The quad cap will work, though it is over voltage and stressed beyond it's specs, so who knows how long it will last.

Greg

[MrCreosote]

Can anyone be slower than me in finishing off a "completed" project.

My Sceptre has been sitting on the bench since the previous post with all the reworking done.

All kinds of worrysome thoughts about how to power up an old amp lead to no real decision and so it just sat there.

Well, finally, it was sink or swim time.  Either **** or get off the....

Switch to STANDBY
POWER ON
2-3 min later...
STANDY OFF

It works.

Noisy pots.

Clean pots, now OK.

Now I don't have any kind of musical instrument or speaker cabinet to test so I've basically hooked it up to a junky 12" audio speaker in an open back box and used audio from the home theater receiver.

My question right off is what would be the tone settings for flat?  I looked at the schematic and it looks like flat would be a jumper wire bypassing the little cluster of caps and resistors for the BASS, TREB, and CONT.

What I did notice was there seemed to be insane amounts of treble if I turned up the treble.  With treble = 2-3, I could get reasonable frequency response.  I could get bass only, treb=min, bass=max, but could not make the crappy speaker in box "thump."  It is a very crappy speak with a very high fs so mayby I have no idea what amp is outputting.

Maybe I should hook it up to a actual high fidelity stereo speaker?

Anyhow, it does look like the relocated caps are working fine - I didn't notice any hiss or hum however, I couldn't get past 2-3 on the volume due to the capacity of the crappy 12 I was testing with.

If I learn anything else, I'll post it.

Thanks for all help.
Tom
__________________________

added 3-11-2010

A little more time I have spent with it and the high level inputs from the HTPC made hum/buzz levels irrelevant since output maxed at volume set at 1.  With terminated inputs, I ran the volume all the way up and there is hum and buzz.  Not insane levels, but more than I think should be there.  I want to take the amp to someone who can hook it up to a guitar and cab and just see what happens.  I have a feeling I will be running DC and ground leads to my remote cap locations.  Having the now unfiltered DC running in the wiring harness just can't be a good idea so I will basically leave the harness DC lead unconnected at both ends and find some open space to run the new leads.  Frankly, having all those wires in the harness so tightly bundled seems like asking for trouble.  Parallel wires have capacitance between them.  With spark plug wires in an engine, you never want parallel wires, you want perpendicular crossings.  If I had to bet, i would think that breaking open those harnesses and spreading out the wires as best as possible would reduce hum and hiss, however, I'm not going to do it because once that is done, there is no going back and I'm not sure if my theory is correct anyway.

Thinking back, it might have been a good idea to find even a single can cap that had adequate voltage.  If such existed, it could have been mounted where quad cap was, the result would look like a stock quad cap which is good for aesthetics and would eliminate the relocation of a cap.  I don't know if such a can cap exists, though.

added some more

I'm thinking about my DC lead routing.  I have the voltage divider located exactly where the quad cap was so these leads will be coming off that added terminal strip.  Since this will be the raw, pumping DC, I see no reason why 3 of these leads could not be wrapped together on their way to the 3 remote caps which are basically in the same area.  That way I would only need a single routing path which would make things easier.

I think I will do these 3 remote DC's first and leave the grounds where they are.  Also, I'm thinking of using wire I have harvested from a PC power supply.  I will have a nice assortment of colors and hopefully the voltage rating of that wire will be high enough.  Any comments?

and still more thinking

Talked with a friend of mine that built radios when he was younger and laughed when I explained my using the DC leads in the wiring harnesses.  He said (as others have) that I need to run isolated wires directly from the resistor ladder to the capacitors.  He said he would use shielded cable to boot.  ALso, ground everything to the same lug.  So now I'm trying to find 600v shielded cable.  I have lots of audio cables laying around - maybe I'll get lucky and find one with lettered cable and a 600 deg rating.  I guess not finding that, a twisted pair, DC and Ground would be better than nothing.  I'm also wondering what will happen if I use 300v hookup wire (unspec'd wire) and not 600v?

Thanks

[Soundmasterg]

If you are going to locate your caps closer to the stages that they supply, that is a good thing usually, but sometimes it can be a bad thing too. Such is the case in the Sunns because of the wires running in the bundles among other things. The amp works fine with the bundled wires as long as the grounds are grounded where the stock amp is. Once you move your caps closer to their individual stages, then your grounds are probably grounding locally, which tends to cause more ground loop issues than when they are all grounded at the can. The other thing going on is that you now have noisy DC going around the amp from stage to stage whereas before that noisy DC was more localized. You would do better to try to shield the grid wires rather than shielding the plate wires anyway. That is always going to be quieter and safer. Apex Jr. sells good shielded cable, but it probably shouldn't be used for plate wires when the plate is probably up around 560v and you're going to shield that with a ground connection in such close proximity.

I modified a 200S and isolated all grounds from the chassis. I added a ground buss that only touches the chassis through a ground lift switch, and then a resistor can cap to ground. These recommendations came from Kevin O'Connor's series of books. All the caps are locally grounded for the stages that they supply, and all grounds go to the buss in the order of the stages in the amp. It made the amp very quiet as far as the hum level, but was a LOT of work. In order to make it work, all jacks had to be isolated from ground too. I've since fixed quite a few Sunns, and no one has wanted to go to that level or expense. In the most recent one, I used series connected caps for the first two stages to get a higher voltage rating (700v vs 525v from the can). I had to add some terminal strips to secure the added caps, and I made sure to ground them where the original amp was grounded. That amp happens to be almost as quiet as my 200S that I spent a ton of time on.

By the way, you will never find a can cap that has anywhere close to the voltages that amp is running. Back when they were new, with lower AC wall voltages, the can cap was close to it's max voltages. Now with today's higher wall voltages, it is over spec. That is why the recommendation for series connected, discrete caps. The added benefit is that there is a much wider selection and discrete caps are cheaper.

Greg

[MrCreosote]

Greetings Greg!

I was patiently awaiting your comments on my current progress.

FYI, the shielded wire I was/may still use is a Belden http://www.belden.com/pdfs/03Belden_Master_Catalog/04MultiConductor_Cables/04.15.pdf.  It is 600v rated.  I was originally thinking of the 22 ga which I can buy locally (25 min drive).  If they have the 20 or 16ga, I will use those simply because of thicker insulation thickness.

NOTE:  Regarding voltage rating.  The spec is RMS which is AC.  Now granted peak AC will be 1.4 times RMS but DC is hell on insulation because it does not alternate and causes a continuous buildup of statically charged dirt.  But then, in my favor is 105 deg C which that wire will never see, so there is a kitty there for voltage.  I'm guessing I'm ok with the 600v rating.  If I could find some 1000V, I would get in a heartbeat.

Shielding is probably overkill if I route new DC wires up against the chassis away from the bundles.  I guess I could even do a twisted pair where one was grounded which I seem to recall doing in a Dynakit.  I remember black and green wires(!)

NOTE:  I did determine that PC power supplies use 300v wire but CRT monitors have pieces of 600v and 1000v wire in them.  But no HV shielded wire.  Apex Jr. seems like a good place for wire, etc.  however I wish he stated voltage ratings.
_________________

I might as well run grounds back to the quad cap area where there is now a terminal strip for easy grounding.
_________________

The thought of using the two lowest caps in the can is a good idea - especially if you have a can that is still intact.  While it would eliminate 2 cap relocations, the $40 cost of a new can is probably still a good idea since any kind of mod takes lots of time.  In particular, you would still have the can there for aesthetic purposes which I like.
_________________

Shielding the grids is not on the table at this point.  I took a quick look at the schematic and it wasn't clear exactly where I would need the shielding.  I guess I could just about shield everything.
_________________

Do any of the simpler models have the 600v Atom cap that is upstream of the quadcap?  I left mine alone since it seems to be OK.

Thanks so much,
Tom

Will update with progress which should flow quickly now(!)

[Soundmasterg]

Hi Tom,

I'd suggest to email the guy at Apex Jr with any questions you have. He has become my go-to source for shielded wire as it is such good quality. For other wire I get it at a place I used to work at for cheap that is local. You can go through the trouble of shielding the high voltage wires, but what I was trying to say is that if you just shield the grid wires, you will get better noise reduction from doing that. The grid wires are the sensitive ones so if you shield them, then they won't pick up the dirty DC and that noise won't be amplified. Given that the DC on the wires that you are running to the caps is noisy and such high voltage, you will not get full shielding of that no matter what you do, and the grids will still pick up noise and amplify it. Incidentally, in my 200S that I modded, I did not use shielded cable anywhere in it, and it is very quiet.

On the Sunns that I work on, I leave the can cap in place for cosmetic reasons but disconnect it electrically. If you have an old can in place, there is no reason to use any of the sections in it, even the lower voltage ones. If one cap in the can is bad or going bad, the others will soon follow. Electrolytic caps have a definite lifespan, and I would replace every electrolytic in the amp if the amp is over 15 years old or hasn't been used much. I'm not sure where you're getting the idea that using any part of the can is good. Can caps are more expensive than discrete caps, and are a compromise and detriment as far as the shared ground. I didn't learn this until after I had bought some can caps when I first started and have been trying to use them up in projects and/or repairs ever since. I even had to reform some of them since they had sat on a shelf unused for years, and now I won't use those in customers amps. Can caps were a convenient and cheap way for a manufacturer to install caps in their products. They required less labor and that means money. Can caps were cheap and that is why they were used. You could go cheap because that was done originally, but since it is your labor, you may as well upgrade a bit and go with a better product in place of the can. Your amp will work better and it will be easier and cheaper to replace 15 years down the road when those caps need to be replaced again and there aren't any more cans available. Unfortunately, can caps are not made in many of the values that they were made in years past, and they are often over-voltage these days too. I actually prefer to use radial caps if I can as they are WAY cheaper and available in many more values and voltage ratings than axials. If you pay attention to other forums, you'll also see that Weber and others have been hit with the fake Rubycon and Elon cap problem lately. Chinese manufacturers have done bait and switches and stuck radials inside axial cans, and relabed many times to where the original cap is under voltage or off spec from what it should be. All the more reason to avoid axials and go with radials, and can caps were never a great solution from a noise perspective anyway.

The best place for shielded grid wires is anywhere those wires are in the loom, or even better, on the input to the first stage. Since Sunn located the first stage close enough to just go through the resistors directly to the socket, then anywhere after that where there is a grid wire in the loom could be improved by using shielded cable. Keep in mind there is a capacitance when using shielded cable that can affect things, rolling off the highs....which in the Sunn could be a good thing. It has plenty!

The proper way to check an electrolytic cap is to test it for ESR and leakage at working voltage on a meter such as a Sprague Tel-Ohmike. You can find them on ebay. Just looking at the cap isn't sufficient to tell whether it is good or not. Not all bad caps bubble on the positive side. If it were me I would just replace it.

Greg