Hello everyone
I've recently acquired a bewitch 6550 amplifier, that I just bought some new tubes for.
Changed the tubes after about 5 minutes it gave a very loud pop in one speaker, and I switched it off.
Then reinserted the old tubes to see if the noise disappeared, it was still there but more of a crackling sound from the same speaker that made the pop sound.
I removed the tubes and unscrewed the bottom plate to see if anything weird was going on in there, this is what I found.
As you can see the resistors in parallel is completely cooked.
How do I fix this correctly?
I know that I need to discharge the capacitors so I don't get electrocuted.
But this is my first tube amplifier and I really need some experienced user's to help me out 🙂
I've recently acquired a bewitch 6550 amplifier, that I just bought some new tubes for.
Changed the tubes after about 5 minutes it gave a very loud pop in one speaker, and I switched it off.
Then reinserted the old tubes to see if the noise disappeared, it was still there but more of a crackling sound from the same speaker that made the pop sound.
I removed the tubes and unscrewed the bottom plate to see if anything weird was going on in there, this is what I found.
As you can see the resistors in parallel is completely cooked.
How do I fix this correctly?
I know that I need to discharge the capacitors so I don't get electrocuted.
But this is my first tube amplifier and I really need some experienced user's to help me out 🙂
It would be best to measure the pair on the other side to confirm the total resistance, but the schematic (below) says 250 ohms. If they blew, then the brown capacitor next to them should also probably be replaced. Unfortunately you will have to lift the board to do that. I can't tell from the photo, but if it is bulging on top it has blown as well.
The problem remains--why did they blow?? Were the new tubes faulty? If you replace the resistor and cap, will it happen again? Hard to say...
http://lampizator.eu/amplifiers/china/bewitch 6550/bewitch.html
The problem remains--why did they blow?? Were the new tubes faulty? If you replace the resistor and cap, will it happen again? Hard to say...
http://lampizator.eu/amplifiers/china/bewitch 6550/bewitch.html
The parallel capacitor may have shorted, so be sure to replace that also.
It's best to test the new tubes in a tester, but otherwise you may have to try them in the other channel.
Fortunately, there are no adjustments to make.
It's best to test the new tubes in a tester, but otherwise you may have to try them in the other channel.
Fortunately, there are no adjustments to make.
Also, according to that web page, it's possible that higher wattage resistors would be advisable. If they are passing 40 volts at 160mA, that 6.4 watts. A pair of 500 ohm, 7 watt resistors on each side might hold up better.
There are lots of root causes for a failure like this.
1) Resistor is under rated, insufficient air flow. Resistors stacked like that will not be able to dissipate their full rated power.
2) Something wrong with grid bias. Shorted or leaking C3 or C4. Open R11 or R12.
3) Tube socket problem. Bad contact to grid terminal.
4) Shorted tube. Once the resister blew, replacing the tube won't tell you anything. Tube may only short when hot.
5) ??????
Given just how hot it got with all the burning, root cause #1 is likely only a contributor. One of the other root causes is also in play.
1) Resistor is under rated, insufficient air flow. Resistors stacked like that will not be able to dissipate their full rated power.
2) Something wrong with grid bias. Shorted or leaking C3 or C4. Open R11 or R12.
3) Tube socket problem. Bad contact to grid terminal.
4) Shorted tube. Once the resister blew, replacing the tube won't tell you anything. Tube may only short when hot.
5) ??????
Given just how hot it got with all the burning, root cause #1 is likely only a contributor. One of the other root causes is also in play.
The 6550 grid resistors are 220k.
That is very close to the maximum spec of 250k; if the tubes are run very hot (high dissipation), the tubes might go into thermal runaway.
This may, or may not, apply to New Old Stock tubes, or to tubes of recent manufacture.
There is plate dissipation, and screen dissipation. Do not forget the 6.3V 1.6A filaments (10.08 Watts).
Personally, I prefer KT88 over 6550.
I agree with Potentiallyincorrect.
Any leakage current of C3 or C4 will put positive grid voltage on the grids. Not good.
That is very close to the maximum spec of 250k; if the tubes are run very hot (high dissipation), the tubes might go into thermal runaway.
This may, or may not, apply to New Old Stock tubes, or to tubes of recent manufacture.
There is plate dissipation, and screen dissipation. Do not forget the 6.3V 1.6A filaments (10.08 Watts).
Personally, I prefer KT88 over 6550.
I agree with Potentiallyincorrect.
Any leakage current of C3 or C4 will put positive grid voltage on the grids. Not good.
Last edited:
I have tried measuring voltage to the grids on the left channels tubes, and they where fluctuating quite a lot. Tried to put one set of the new tubes in, and they measure stable, definitely seems like one of my tubes is bad, and that one of the new has a defect also.
Should I try the second new pair to confirm it is the tubes?
Ps. I use a quick setup with just 2 speaker drivers, so I won't blow up my main speakers 🙂
Been playing for more than an hour with no pops or anything worth mentioning.
But I'm not sure if this was merely it? The resistors measured a bit lower than the other pair - 363 ohm vs 375 ohm
Is this negligible or should I get some new resistors?
I measured the coupling capacitors, c3 c4. Which both measured 0,52uf and 2,6ohm.
I did not get to measure the other capacitors, is this a necessity?
Thank you for all the response so far.
Should I try the second new pair to confirm it is the tubes?
Ps. I use a quick setup with just 2 speaker drivers, so I won't blow up my main speakers 🙂
Been playing for more than an hour with no pops or anything worth mentioning.
But I'm not sure if this was merely it? The resistors measured a bit lower than the other pair - 363 ohm vs 375 ohm
Is this negligible or should I get some new resistors?
I measured the coupling capacitors, c3 c4. Which both measured 0,52uf and 2,6ohm.
I did not get to measure the other capacitors, is this a necessity?
Thank you for all the response so far.
It's not nessecary to measure the capacitors. As long as you see no positive DC voltage on the grids of the 6550's (pin 5 to ground), then it's probably okay. I would replace the cathode resistors, though, one looks cracked. It would be an easy soldering job, no need to lift the board. Two 750 ohm/7 watt resistors in parallel per channel would do it. And space them away from the board a bit more.
I would replace those capacitors to be on the safe side. Also regarding the cathode resistors. Possibly 2, 125 Ohm in series would get rid of heat better as they will be naturally spaced away from each other. Piggy backing in parallel is not advised unless you can put a bunch of space between them.
The grid voltage fluctuated?
Wow!
How much voltage, average and peaks? What did you get?
+/- 20mV DC is from the Power Mains Line Bounce that causes B+ voltage to Bounce, so the plates of the driver tube Bounces too. That is OK.
Any other cause of the fluctuation, and larger voltage peaks, would make me extremely nervous.
Use an oscilloscope to measure the fluctuating grid voltage, if you dare turn on the amplifier again, before finding the cause of the problem.
I should explain, line bounce of power mains comes from various loads switching on and off in your house, and from neighbors switching loads on the same power mains transformer on the power pole.
Dryers, House Heaters, Water Heaters, Refrigerators, Air conditioners, Lady's 15 second quick hair dryers, Toasters, Air Compressors, etc.
These cause very low frequency bounce in the B+, because the B+ filters are made to filter out the rectified power main frequency, and 2X frequency; and with good amplifiers they also help to stabilize the B+ when the 32.7Hz organ pipe plays.
Those B+ filter caps were not designed to filter out the Sub Hertz, and up to 3 or 5 Hz of Power Mains Line Bounce.
Measuring the voltage on the 6550 grids Is the correct way to measure any capacitor leakage . . . while they are in the circuit with operating voltage across them (probably lots more than 100V).
Measuring the resistance of a capacitor with the current limited maximum 5V into an open circuit, that comes from a DMM is Not the way to measure capacitor leakage.
When you get ready to turn the amplifier on again,
Be sure to measure the voltage across the output tube cathode's 250 Ohm self bias resistor.
Divide that voltage by the resistance V/R to get the current, I.
Then measure the plate voltage, Vp.
Vp x I = the plate plus screen dissipation of two tubes.
Example: 500V x 0.16A = 80 Watts, that is 40 Watts per tube.
Your self bias resistor is 250 Ohms, right.
Example:
I = 0.16A
Power = (0.16A) squared x 250 Ohms
P = 6.4 Watts.
When a resistor is right down on a PCB, it is a very good idea to use a resistor power rating that is at least 5 x the power it will dissipate.
6.4 x 5 = 32 Watts
If one or more tubes goes into thermal runaway, the resistor will cook.
Wow!
How much voltage, average and peaks? What did you get?
+/- 20mV DC is from the Power Mains Line Bounce that causes B+ voltage to Bounce, so the plates of the driver tube Bounces too. That is OK.
Any other cause of the fluctuation, and larger voltage peaks, would make me extremely nervous.
Use an oscilloscope to measure the fluctuating grid voltage, if you dare turn on the amplifier again, before finding the cause of the problem.
I should explain, line bounce of power mains comes from various loads switching on and off in your house, and from neighbors switching loads on the same power mains transformer on the power pole.
Dryers, House Heaters, Water Heaters, Refrigerators, Air conditioners, Lady's 15 second quick hair dryers, Toasters, Air Compressors, etc.
These cause very low frequency bounce in the B+, because the B+ filters are made to filter out the rectified power main frequency, and 2X frequency; and with good amplifiers they also help to stabilize the B+ when the 32.7Hz organ pipe plays.
Those B+ filter caps were not designed to filter out the Sub Hertz, and up to 3 or 5 Hz of Power Mains Line Bounce.
Measuring the voltage on the 6550 grids Is the correct way to measure any capacitor leakage . . . while they are in the circuit with operating voltage across them (probably lots more than 100V).
Measuring the resistance of a capacitor with the current limited maximum 5V into an open circuit, that comes from a DMM is Not the way to measure capacitor leakage.
When you get ready to turn the amplifier on again,
Be sure to measure the voltage across the output tube cathode's 250 Ohm self bias resistor.
Divide that voltage by the resistance V/R to get the current, I.
Then measure the plate voltage, Vp.
Vp x I = the plate plus screen dissipation of two tubes.
Example: 500V x 0.16A = 80 Watts, that is 40 Watts per tube.
Your self bias resistor is 250 Ohms, right.
Example:
I = 0.16A
Power = (0.16A) squared x 250 Ohms
P = 6.4 Watts.
When a resistor is right down on a PCB, it is a very good idea to use a resistor power rating that is at least 5 x the power it will dissipate.
6.4 x 5 = 32 Watts
If one or more tubes goes into thermal runaway, the resistor will cook.
Last edited:
I strongly belive the schematic is wrong. Based on OP's measurements and pix on the web site, the resistors are parallel 750 ohms for total of 375 ohms.
My Bewitch 6550 cathode resistor was also 375 ohm (2 x 750 ohm). There are at least two versions of this amp.
If that really is a crack in that Cathode resistor, then you should replace them...you have the posts to solder onto, so should be easy.
As stated, two 750 ohm 7 watt resistors in parallel, elevated off pcb board, with space between should do fine.
Also put your voltmeter on DC and measure on 6550 side of that fat coupling capacitor (or pin 5 on 6550 tube)... if any appreciable DC there, the coupling cap is leaking to grid of output tube.
If that really is a crack in that Cathode resistor, then you should replace them...you have the posts to solder onto, so should be easy.
As stated, two 750 ohm 7 watt resistors in parallel, elevated off pcb board, with space between should do fine.
Also put your voltmeter on DC and measure on 6550 side of that fat coupling capacitor (or pin 5 on 6550 tube)... if any appreciable DC there, the coupling cap is leaking to grid of output tube.
Just wanted to give an update here. I have completely recapped the amplifier, inclouding 4xJantzen silver coupling capacitors which I've bypassed with 0.01uf 940c 3000v capacitors.
Furthermore i stuck with the original 2x750 ohm resistors in parallel but i soldered them so they now have a decent gap between them for heat disipation 🙂
All I can say is that it went from meh to very good, minus the placebo effect off course then it ends up being good 😀
Furthermore i stuck with the original 2x750 ohm resistors in parallel but i soldered them so they now have a decent gap between them for heat disipation 🙂
All I can say is that it went from meh to very good, minus the placebo effect off course then it ends up being good 😀