ARCHIVED FORUM -- March 2012 to February 2022READ ONLY FORUM
This is the second Archived Forum which was active between 1st March 2012 and 23rd February 2022
I just finished restoring a pristine BM4000, original owner, never serviced.
It even had the Schematics inside the envelope!
It was a fairly easy and standard restoration, except for 2 issues which I'll share here.
Primary symptom: No power on - the 2 fuses were open.
The reason was a problem in the power switch, and the main B80C5000 bridge was shorted - never seen that before.
In the attached picture, see the new Vishay GBPC602 6A 200V bridge at the right rear of the transformer casing, secured with yellow nail polish, against the rear panel. An elegant, non-obtrusive solution.
The 2nd issue was the On-Off switch. It was completely corroded inside.
I removed the corrosion and pitting with a gentle round file, to restore the original concave shape on the contact rods, and then smoothed it off with a radial wire brush on my Dremel. Finally, I rubbed down all the metal parts with Brasso, and sprayed liberally with DeOxit 5.
I trust that it will work well for many more years.
Another interesting point:
This BM4000 has the extra 4x 5K potis on the output amps.
In spite of following the procedure for calibration exactly, using the 4.7uF (bipolar) cap and the 190 Ohm resistor, when injecting the 1Khz signal, and increasing it, the output sinewave remains absolutely stable and distortion-free, and will not clip or distort on either the positive or negative peaks.
Adjusting the potis produces no effect on the sinewave at all, so I reset them to the original settings of the old corroded potis (as when the machine arrived at me).
So, perhaps because of using very high-quality components in the restoration, then this calibration appears to be irrelevant, and the electronic fuse is not in danger!
To prevent switch damage and extend the service life, I'm now adding a snubber capacitor across the switch poles to all my B&O restoration projects. I've seen snubbers in almost all receivers I worked on but not in B&O, go figure.
Can you be more specific?
Photos?
Sure enough!
Beolab 5000 snubber installed on top of the switch assembly. Sparks are greatly reduced with the snubber. You can see the phenomenon lights out with and without a snubber; quite amazing the difference it makes.
My calculations show that you should be using approx 33 ohm 2W carbon resistor and a 100nF X2 capacitor (120VAC 60Hz)
I should be using a 120 ohm resistor and a 33nF capacitor (240VAC 50Hz)
Would you agree?
You are going beyond my knowledge honestly!
I've used an 0.047uF 100 Ohm snubber (60Hz 120VAC) I found into another similar spec'd unit without knowing how to calculate it.
If your calculation is right then you would need this snubber instead PMR209MC6100M033
BUT they are all rated at 0.5W dissipation factor and haven't seen another snubber with a built-in resistor...
Can you teach me how to calculate a snubber please?
Those snubbers I would NOT use - they are rated for max 250VAC.
We have power fluctuations sometimes up to 260VAC.
So, I would build my own, using a 310VAC or 440VAC -rated X2 or Y2 cap, with the correct resistor.
It's also NOT a good idea to use METALLIZED paper or film capacitors - it causes inductance.
For the same reason, use only Carbon resistors, which also handle pulses MUCH better than metal-film or -oxide resistors!!!
The Resistor's ohm < / = Voltage (off) / Current (peak on)
Voltage for you is 120VAC, Peak Current to be really generous - what the fuse is rated for.
I took 4A in USA, 2A in EU.
Capacitance nF = 100000000 x (1/(Voltage squared x Switching Frequency))
Frequency here is line frequency.
That's the quick and dirty method.
There are more exact formulae which are irrelevant here, because the line frequency is so low.
In switching PSU's with MHz frequencies, it is much more critical.
Thanks a lot for the formula, I will have to change the caps I installed in my own amplifiers; I think it helped but better go to the safe side. Mine are lower voltage but I agree that, in your case, you must be higher.
However I'm a little confused about your comment of not using film capacitors? Most of the X2 safety capacitors are film boxed...?
So this one is not good at all? KEMET F863BC104K310R
They are different things altogether.
In Switch-mode power supplys snubbers are used to dampen ringing from transformer inductances, where one end of each of thetransformer windings are AC-grounded - or actually DC grounded (secondary windings), and the other end of the winding is free to ring ata high frequency.Small capacitances are then introduced to dampen the ringing (it typically can't be all removed).
The RC spark snubber discussed here is not at all that critical, its only job is to short spikes in the cases where thepower is interrupted by the switch when the mains voltage is far from zero (=most often).Popular speaking, the sudden voltage across the switch contacts (as it's switched off) is used to charge a capacitor rather than cause a spark.It's often more critical to place the RC network close to the contacts (as done very fine here).
47nF / 100 Ohms sound about right (lower the capacitance if the resistor runs warm) and the majority of Class X2 capacitors are indeed metalfilm types.
A snubber will provide some form of protection for the switch, but it cannot remedy a bad or already arcing switch.
Martin
The emphasis on the METALLIZED, not on the FILM.
If at all possible, use a non-metallized capacitor - it doesn't matter whether it is a Paper or Film type.
From CDE's application notes:
"The use of metallized film reduces the peak current capability to from a third to a fifth of the other high-voltage choices."
Thank you Martin for the clarification, once again, you're on top of your stuff.
I also learned that switch arcing was happening when current is flowing thru the switch when interrupted (OFF) and not when switching ON.
Given the difference of voltage and frequency between US and EU, would you say that the snubber capacitance and resistor values would be same? Based on the formulas above, it should be different isn't it?
-----
Menahem, I calculated the resistor dissipation value on my side and I'm getting way under 2 watts, actually 0.1 watts. The way I understand it is the resistor should handle the dissipation of the snubber capacitor, not the entire unit, right?
Thanks a lot guys!
Yes, the arcing occurs when switching off.Popular speaking, it's in the nature of coils (and transformers) "not to like changes".When you introduce a current through a coil, it will work against it at first, creating a magnetic field around it, maintaining it as long as thecurrent flows, and it will have a lower resistance if the current is steady.When the current is then interrupted, the coil will again work against the change and "attempt to keep the current going" by sending whateverenergy it has stored in its field back (with reverse potential) as an uncontrolled (typically quite high) voltage spike.This voltage is high enough to create a spark across the small gap of just opened contacts.This moves a little material from one contact point to the other every time, and eventually the contacts will look likerough mountains with soot and slag if watched using a microscope.
Think high voltage distributor coil and spark plugs in older conventional car engines.The current through the coil is interrupted (by the contact points in the distributor), and the resulting high voltage from the coil issent to the spark plugs (at the right timing and in the right order).The contact points wear down with use (one point ends up with a "valley", the other with a "mountain top") and they mustbe replaced regularly.
If you see a circuit where a transistor drives a relay, there will almost always be a diode sitting across the relays coil in theopposite direction of the normal current flow.This diode shorts (snubs) the reaction voltage from the coil, when the transistor switches off the current flow to the relay.Else the high voltage spike produced by coil from the magnetized material inside the relay would damage the transistor.
The mains voltage is not very critical in your case. The mains frequency only slightly more so.I suggest you start with what you already made and, as I mentioned earlier, lower the capacitance of the cap, if the resistor runs warm.
A diode connected across the coil of a doc relay is often referred to as a "flywheel diode"
Craig
DC relay even ;-)
Yes, I was just wondering whether I had missed that Bill Gates had bought Schrack, and was rewriting history to refer to DC as .doc
Yes, in fact on the BM4400, there is only a 3M3 resistor (no capacitor) from the live terminal to ground.
Am I correct in thinking that the technique of using a capacitor in series with the resistor came much later?
I was wondering the same thing Menahem if this was an attempt for switch arc suppression.
Should I remove this resistor after the snubber install or it won't hurt if I leave it there?
Dillen: Yes, the arcing occurs when switching off.Popular speaking, it's in the nature of coils (and transformers) "not to like changes"........ ...... The mains voltage is not very critical in your case. The mains frequency only slightly more so.I suggest you start with what you already made and, as I mentioned earlier, lower the capacitance of the cap, if the resistor runs warm. Martin
Yes, the arcing occurs when switching off.Popular speaking, it's in the nature of coils (and transformers) "not to like changes"........
......
Thank you Martin as always!
Beo_Jean:Should I remove this resistor after the snubber install or it won't hurt if I leave it there?
I would leave it, and I'll tell you why.
After I did that 4000 and had to rebuild the switch, I opened up the switch of the 4400, and found that it was clean as a whistle, no pitting or damage at all.
I simply wiped the oxidization off with Brasso, and it's as good as new.
So, the resistor is obviously bleeding to ground exactly what needs to be bled.
In this case, I would not mess with success, and I would question changing anything at all.
Just for comparison, I've just been going through some of the the Service Manuals of other more modern Japanese amps I've worked on, and across the On-Off switch is ONLY a capacitor of 10-22nF, rated at between 275-440VAC. Mostly ceramics, not film.
No resistors.
So that shows that there are many ways to kill the arcing.
On the Tek 24xx scopes, there is no protection at all.
I have restored a few of these scopes, and I have replaced the On-Off switch on 2 of them.
I never really gave any thought to this issue, until the BM4000, but now it is a definite area to look at on future repairs.
Menahem Yachad: Beo_Jean:Should I remove this resistor after the snubber install or it won't hurt if I leave it there? I would leave it, and I'll tell you why. No resistors. So that shows that there are many ways to kill the arcing. On the Tek 24xx scopes, there is no protection at all. I have restored a few of these scopes, and I have replaced the On-Off switch on 2 of them. I never really gave any thought to this issue, until the BM4000, but now it is a definite area to look at on future repairs.
I don't know if it's a matter of a coincidence or you had a unit with low mileage but switch failure is not uncommon on a BM4400 and you may ask Craig or Soren Mexico about it. I was shocked when I saw pictures of a switch failing apart severely pitted by arcing. https://archivedforum2.beoworld.org/forums/t/30179.aspx?PageIndex=8
I will certainly leave the resistor however a snubber will follow for additional protection to save that hard to find piece.
Jean,
Thank you for bringing up this whole issue, and enlightening me about its importance!
Your timing was impeccable - the BM4000 was on my shelf, ready to go back to the customer on Monday.
I took it down, and put it back on my lab table.
I have now installed a snubber - 22nF ceramic, and 2x 270 Ohm in parallel = 135 Ohm
I have a fair amount of stock of the old B&O 1W Carbon resistors, so this was a perfect application for them.
The picture shows it, BEFORE I taped it up, and secured it.
Because this On / Off switch connects BOTH the Live and Neutral lines (DPDT), I hard-wired the Neutral side's contacts (see the bridge wire at the rear of the switch), and installed the snubber across only the Live line's contacts.
Definitely a preventative measure, to ensure long life.
You are welcome but after all; this is the purpose of the forum right!
For a snubber to function, it has to sit across the set of contacts that breaks the current flow.In this case, where two sets of contacts are housed in the same switch and activated "simultanously", that'll be the set of contacts that breaks first.When the first set has opened, the only current flowing through the other set will be the current used to charge the snubber cap.
In other words, after the first set of contacts open, the other set of contacts mustn't open until all residual energy hascharged into the capacitor across the first set.If the other contact breaks before that, the circuit is open and it would be pointless to have a snubber.
You cannot possibly tell which of the two contacts in the switch breaks before the other, and it might noteven be the same one every time, so you really should put a snubber across both.
Menahem chose to short one set of contacts.You can do that, if you are ABSOLUTELY sure which set of contacts handles the live and which set handles the neutral.In many countries, two-prong mains plugs that can be inserted into the wall outlet either way around, are used.Shorting a set of contacts in such a set is not legal and could potentially be dangerous (in worst case lethal).
Yes, wise words.
If this machine had the 2-prong AC plug, 2 snubbers would be appropriate.
But here, on this machine, it has the 3-prong AC plug - Live, Neutral and Ground, so one snubber is just fine across the Live line.
Now, if some idiot in the future decides to install a new AC plug on this machine, and doesn't have any knowledge of the importance of wiring the plug correctly, and chooses to not ask anyone for help, then his fate is upon his hands.
I didn’t mean to bring your thread into another direction but wow, amazing stuff learned from this snubber mod!
Thanks a lot to both of you!
Ok.....so now that you boys seem to have reached an accord.....what should I be fitting to the power switch on my latest project? ;~)
And what exactly is your LATEST project? ;-)
Well another eBay BM4400 of course............what else ;-)
I think you need Ritalin!
That's very constructive.....however I suspect I will take the option of a snubber circuit on both poles as opposed to relying on the integrity of unknown future users.
"tune for maximum smoke" perchance?
Take care in there sir!
Nick
do I sense a suggestion that this is not a good idea after All?
craig
Here is a link to arc suppression that I was sent. It might help the discussion.
-sonavor
I think that this AC snubber is a great idea. The idea didn't exist when these Beo's were made.
The Littelfuse article is interesting - being an AC circuit, only Figure 3 applies to our discussion.
Figure 1 is what we see in all the Beo's, where there is a relay - the diode reduces the inductive ringing across the coil load. But that is a DC circuit, and is not relevant to the AC on/off switch.
Anything that prolongs the working life of components in these old classics must be worth considering,
Interestingly,snubbers are employed in later equipment such as Beomaster 5000/5500/65/7000,across the power supply relay contacts.
My earlier comment was not suggesting "bah humbug",but merely caution,as the components will need to be insulated as at mains potential,and there is not much room in there of course with a '4400.
Keep us posted on progress.
This started a journey of discovery for me:
In summary, there are TWO types of Snubbers
1.
A single Capacitor (usually 4.7nF – 10nF) mounted across and bypassing the two contacts of the On-Off switch.
The capacitor should be DC-rated for a few thousand volts and AC-rated for about 2 x 1.41 x the AC RMS line voltage. It dampens the arcing across the switch contacts.
For example, Onkyo uses a 10nF across the switch contacts.
2.
A “R+C” snubber (across the Mains Live and Neutral lines), which comprises a Non-Inductive Ceramic or Carbon Composition resistor rated for about 700V-1000V, in series with an AC-rated Class X/Y capacitor, the values of which are calculated according to the previously-discussed formula.
The R+C snubber is usually placed BEFORE the transformer, to be of benefit to the On-Off switch, and reduces the ringing effect of the transformer.
For example, Quad used a R+C 100Ω + 100nF across the Live and Neutral lines, BEFORE the Power Transformer.
But on the BeoMaster 5000 (1980’s), the unusual setup shows the R+C 10Ω + 100nF AFTER the transformer, across the On-Off switch contacts – very unconventional for this setup – as shown in Charles’ June 5th diagram above.
The capacitors can be Film or Ceramic, and Vishay, which is easily sourced here, has a wide range of suitable components.
Unless I’m missing something, on Charles’s BeoLab 5000, the Rifa R+C 100Ω + 47nF R+C is mounted ACROSS the Live and Neutral lines BEFORE the Power Transformer, so there is no need to mount another R+C snubber across the actual switch On-Off contacts.
In fact, on the BL5000, there is NO snubber across the actual On-Off contacts.
I hope that this brings some clarity to the issues, and IMO it’s well worth installing snubbers on all restorations.
Menahem
Very informative Menahem as usual,
In addition to your findings, I found schematics of Dual 704 and 721 turntables where both snubbers are present.
One 47nF + 100 Ohm R+C across the AC line and one 10nF across the switch contacts.
So your conclusion makes perfect sense to me and serve two distinctive beneficial fonctions.
Thank you very much!