First Stage Filtering

[jazzyjoepass]

Gil,

I'm talking about voltage drop across the caps only - not referenced to gnd (except the bottom cap of course).

I switched to another set of caps - still the same. But I've not checked the dropping resistors if they're OK.

How about the rest? Any diff in voltage drop ?

Mike.

[ayan]

Gil,

I'm talking about voltage drop across the caps only - not referenced to gnd (except the bottom cap of course).

I switched to another set of caps - still the same. But I've not checked the dropping resistors if they're OK.

How about the rest? Any diff in voltage drop ?

Mike.

Mike, then your totem-poled resistors are bad... [BTW, I call dropping resistors the ones that go in series between the various power supply nodes in order to drop the voltage, and which play no role whatsoever in equalizing the voltage across totem-poled caps]. If anyone else should say they have a 160VDC difference between the voltage across a pair of totem-poled caps, as you do, they would have a problem in their amp as well. This really is not a subjetive thing, as any person with some good electronics experience under their belt will be able to confirm. I encourage you to take the caps out and measure the resitor values, as I said in my previous post, and I am sure you will find they read differently.

Cheers,

Gil

[BobW]

I agree with Gil it sounds like your pair of totem pole resistors are not the same or even close. A common mistake I have made myself is to select a pair that are off by a factor of 10. eg, Thinking you've installed a pair of 220Ks but really installed a 220K and a 22K.

[jazzyjoepass]

Surprise, surprise ... you're both wrong. The resistors were fine - KOA SPEER 1W 470k 1% ....

I checked and swapped a few caps in and realised something that all of us have taken for granted ... the tolerance of the elec caps .... and that's dangerous ....

All the caps seemed to be quite diff in terms of charging characteristics (not sure which parameter - the capacitance "C" or coulombs "Q" parameter) though both were marked 220uF/400V. Thus the ability to charge evenly in a totem pole config was now compromised. Both are now competing for the rectified output. The one with poorer charging characteristics now displayed lower voltage across its terminals. The better cap now charged to a higher voltage.

So that can be dangerous ... e.g. the B+1 is expected to be 475V, but if my caps (say 100uF/250V) in my totem pole network is not matched and if one of them has poorer charging characteristics, then the other cap will be heavily loaded and voltage across it may even exceed its rating .... we have an electrolyte mess of a bomb.

Maybe you guys really want to get your hands dirty and do a tryout test yourselves and give me your input.

Thks for the sharing.
Mike.

[ayan]

Surprise, surprise ... you're both wrong. The resistors were fine - KOA SPEER 1W 470k 1% ....

I checked and swapped a few caps in and realised something that all of us have taken for granted ... the tolerance of the elec caps .... and that's dangerous ....

All the caps seemed to be quite diff in terms of charging characteristics (not sure which parameter - the capacitance "C" or coulombs "Q" parameter) though both were marked 220uF/400V. Thus the ability to charge evenly in a totem pole config was now compromised. Both are now competing for the rectified output. The one with poorer charging characteristics now displayed lower voltage across its terminals. The better cap now charged to a higher voltage.

So that can be dangerous ... e.g. the B+1 is expected to be 475V, but if my caps (say 100uF/250V) in my totem pole network is not matched and if one of them has poorer charging characteristics, then the other cap will be heavily loaded and voltage across it may even exceed its rating .... we have an electrolyte mess of a bomb.

Maybe you guys really want to get your hands dirty and do a tryout test yourselves and give me your input.

Thks for the sharing.
Mike.

With all due respect, Mike, my hands are dirty with thirty years worth of tests in electronics. Make no mistake, if your resistors were OK under load, then your caps are bad -- or else there is something wrong with your power supply connections. One note on the example you picked: if B+ is supposed to be 475 VDC, one should not use 250V caps totem poled because it would be easy for them to exceed their power rating; 300V or 350V rating would be my choice. Speaking of your real amp, you said you measured 360V on one of the caps... was that one rated at 250V? If so, you might as well throw it away, unless you used it for a few seconds only.

I am surprised you could repeat the same results over and over again, using different caps, since it really means they are all bad. Perhaps they needed to be slow formed before being subjected to high voltage, and neither the manufacturer nor you did that, or their useful lifetime has expired. But there just is no way a 160VDC difference in volatge between the two totem poled caps is something anyone should expect to see, even factoring in the capactiros' tolerance, as I will try to show below.

I always check all the voltages in the power supply in amps to make sure everyhting is working as it should, ensure no cap is connected backwards, etc. In all these years, I have never encountered a pair of totem poled caps that measure more than a few (say up to 5) volts aways from each other when properly totem-poled, and if I did, I would send them back to whoever sold them to me and ask for my moeny back.

Out of curiosity, could you tell us what type of caps you are using and where you got them from? Might they be surplus items?

Since you mentioned this, the basic formula for caps is Q = C*V, where Q = charge in Coulombs, C = capacitance in Farads and V = voltage in volts. The formula works as-is in DC circuits, and you can use it to analyze your case. When two caps are totem-poled, they will have a common terminal which will force them acquire the same charge. The total voltage applied to the pair, V, will be, using the formula above:

(1) V = V1 + V2 = Q/C1 + Q/C2

diving both sides by Q,

(2) V/Q = 1/C1 + 1/C2 = 1/Ceq

So this is the equivalent capacitance for the series is C eq = 1/[1/C1 + 1/C2] -- same formula as for resistors in parallel; caps in parallel, othe other hand, add up like resistors in series.

You can use Equation (1) above to see how bad off your caps are: if you said one cap measured 360V and the other 100V, it means that the ratio of their voltages is 3.6 and therefore so is the ratio of their capacitances. 360% off from one another cannot possibly fall within the tolerance of the caps.

I'm not trying to beat this horse to death just because, Mike, but I want to encourage everyone who might read this to be very careful. If anyone finds results like yours, they need to either use diffferent resistors, different caps, or fix their wiring if that's what's wrong. But in no case is this to be considered to be OK, or within tolerance, etc., because it is not.

Cheers and out,

Gil

[tonelab2]

Wow! Thanks Gil for the insight on totem caps and the formulas. I'm
always sus on electrolytic caps as you can never tell how long they've been on the shelf. I tend to put them through the cap testers I have checking capacitance and leakage. Kind of a pre-build break in. With my limited skills and knowledge I need all the help I can get to minimise debugging. A good cap tester is a handy tool, the older ones with leakage test of 500v are easy to pick up cheap and only need recapping and calibrating to work as new again. Worth it to save a headache later on.

[BobW]

Maybe you guys really want to get your hands dirty and do a tryout test yourselves and give me your input.

fwiw, jazzyjoepass, I'm glad you found the problem, but assuimg anyone's technical capability on this board is wrong and presumptious. Gil Ayan and I offered our advice to help you find the problem not dog you because you asked for advice.
There are a lot of very qualified people here on this board offering help. As an example I was a military equipment tech for 12yrs and have been a hands-on mil power systems engineer for the last 15, have repaired over 400 amplifiers over the years as a part-time hobby. There are also a few beginners here as well,but it doesn't matter, we are ALL here to have fun and discuss/learn about our favorite passion.

Electroyltics, as Gil suggested do required reforming, since over a period of time the electrolyte will crystalize. During a discussion with a Nichicon rep a few years ago, he mentioned they usually reform electrolytic stock usually over a year old by placing them into an oven or environmental chamber at a elevated temperature over a period of time, (the rep gave no temp or time numbers).
I'm not suggesting you do this but, again as Gil suggested, you caps are obviously bad, and may have been old stock. You might want to try reforming the caps if you can't get your money back. Gerald Weber discusses a method in one of his books for preforming caps already in an amp using a variac by initially setting the line voltage to approx. 40 volts. After an hour, and after every hour, raise the line voltage an additional 10 volts so that after 8 hours, the line voltage is at the standard 120 volts, and the caps are then at the correct secondary voltage. At this point the caps will be reformed. I personally don't use this method since it takes 8 hours to reform one amp. I use a variac and step the voltage every 5 min. while observing the input current.

peace

[jazzyjoepass]

Nope, the caps were rated 400V so I was fortunate it didn't explode on me. I took them out and wired directly with crocodile clips to ensure my soldering wasn't the cause. In fact I switched them top and bottom positions and could tell that the voltage drops consistently corresponded to the changed positions.

One thing I'm sure is that investment in good caps is important - they were surplus and have been stored longer than a year

Yes, I agree Gil and Bob are experienced people - no need to give me a portfolio!!! Your sharing tells me you have much stored up in those grey cells.

Pls guys no faulting anybody. You guys have been great buds to talk amps with ... keep them coming ....

Yeah tonelab, I'd think a cap meter is handy ... I'll source for that on the E(vil)Bay.

I'd do some reforming of the caps while I go grab new ones from the stores.


PEACE too.

[tonelab2]

Jazzyjoe

I have bought all mine from evilbay (sold a couple as well after restoring
them). A good (calibrated) digital cap tester is a must for accurate readings. It's surprising how far out some new caps are. A old Good model tester is the Heathkit IT-11 as it has up to 600v leakage test which is good for reforming caps as well. If you do get one make sure you check
all resistors and replace any that have drifted ( I like to replace the CCs with CFs or MFs). When replacing caps keep all values the same, that is to say if a pico cap is over/under by 1% make the nano over/under by 1%, same with the micro caps ( this is where a digital tester comes in handy). That way the readings will be accurate ( to 1% or even less depending on the tolerences you aimed at when selecting caps) across the dial. You can restore one in less than a couple of hours. Hope that made sense.
They are good machines for reforming NOS caps that have been on the shelf to long. Other good ones are the Eico 950b ( be warned the chassis is Hot (live) while doing a leakage test or reforming caps, So don't touch unless you like the feel of 500v surgeing through your bones) or any of the Sprague models (TO-5, TO-6) ect. I haven't paid more than $20 for a tester over the years (most of them $10) although people are starting to catch on of late and the prices are slowly rising.
Good Luck and Cheers

[jazzyjoepass]

Found an article on the totem pole network.

http://www.nmr.mgh.harvard.edu/~reese/electrolytics/

Fast forward to the part on series capacitance.

[jazzyjoepass]

Found an article on the totem pole network.

http://www.nmr.mgh.harvard.edu/~reese/electrolytics/

Fast forward to the part on series capacitance.

[HiGain]

Hi Guys,

On the subject of B+ filter caps and stiffness:

I was reading up in TUT1 and TUT2 recently, and tried out a trick.

I modified my deluxe reverb to have a dumble preamp in channel one. In addition to a bunch of mods (that I will not detail here), I increased the B+ filter caps. The result was a stronger and punchier bass. However, the negative consequence was too much stiffness. It felt like the amp was working against me, and soloing felt like wading through molases.

Solution. Place a 100 ohm (I might try a little higher) resistor in series between B+ and the OT centertap. WOW! Much more of a rubbery feel, while retaining the crispy low end on the bass. Makes the amp easy to play clean, or with overdrive/distortion pedals.

Jake

[pedro]

Jake,

what wattage did you use ?

interested - I have TUT1 - will read up !

cheers
Pete

[HiGain]

10W. I believe he discusses this topic in the PS section at the beginning. His point was that by inroducing some sag in the power amp section only you enable the preamp to push the power amp around some more. In this way, it FEELS like the preamp is doing some of the work FOR you. It really works.

BTW, in reference to the screen resistor topic, he does go into some detail about his preference for 1K screen resistors in TUT2, but his reasoning was a bit over my head.

Jake

[Bob-I]

...........Solution. Place a 100 ohm (I might try a little higher) resistor in series between B+ and the OT centertap. ..........

Jake

Sounds like a great solution. I need to try this on my amps. I think I have some 130ohm 10 watters at home. My bet is that the stiffness that I'm feeling will be softened up this way.

Thx a bunch.