5 watt unequal wear??
- Thread starter UKBoogieboy
- Start date
Help Support The Boogie Board:
tmac, I think your new measurements are good and pretty close to what the data sheets indicate.
I have also noticed that one tube gets a lot hotter than the other in Class A mode. This and the much higher plate voltage makes me still wonder if that tube is out of the circuit somehow. The output transformer looks beefy enough that it may be able to cope running single ended. Also, the left tube with the much higher plate voltage would mot be able to run at 40ma so there would still be a net current of 10ma or so.
I don't see how the left tube could be out of circuit, unless the cathode was unhooked, but that doesn't seem likely. Both tubes should be about the same temp.
One thing to try is to remove the left tube and see what happens. I have a non contact temp reader so I will also check the tube temps after the amp has been on, but in stand by for say 10 minutes and then after 10 minutes running.
I have also noticed that one tube gets a lot hotter than the other in Class A mode. This and the much higher plate voltage makes me still wonder if that tube is out of the circuit somehow. The output transformer looks beefy enough that it may be able to cope running single ended. Also, the left tube with the much higher plate voltage would mot be able to run at 40ma so there would still be a net current of 10ma or so.
I don't see how the left tube could be out of circuit, unless the cathode was unhooked, but that doesn't seem likely. Both tubes should be about the same temp.
One thing to try is to remove the left tube and see what happens. I have a non contact temp reader so I will also check the tube temps after the amp has been on, but in stand by for say 10 minutes and then after 10 minutes running.
ifailedshapes
Well-known member
EL84s are so cheap, why take any chances with not having a matched pair?
Ok Guys, I turned my amp on and left it to warm up for 10 minutes. Both tube plates measured about 90 degrees C after 10 minutes. I then took the amp out of standby (In 5 watt mode) the left side tube plate measured 220 degrees after 10 minutes and the right side stayed at 90. Next I set the amp back to 25W mode and both tubes ran close to 220 degrees.
Last up I removed the right side tube from the amp and plugged a guitar in to check if it would still work. Not only does it work fine like that, I could not tell any difference in sound.
I am now virtually positive that 5 watt mode only uses one tube.
Does this cause uneven tube wear ? Probably, but I don't think it is much to be worried about. It may even sound better to have one tube worn more !
For those that are worried, just swap the tubes once a year. Personally I use both power modes and plan to swap the tubes on Mesa's recommended interval, 1500 hours which will be about every 2 years for me.
Last up I removed the right side tube from the amp and plugged a guitar in to check if it would still work. Not only does it work fine like that, I could not tell any difference in sound.
I am now virtually positive that 5 watt mode only uses one tube.
Does this cause uneven tube wear ? Probably, but I don't think it is much to be worried about. It may even sound better to have one tube worn more !
For those that are worried, just swap the tubes once a year. Personally I use both power modes and plan to swap the tubes on Mesa's recommended interval, 1500 hours which will be about every 2 years for me.
Very interesting results. So one tube doesn't heat up as the amp is taken out of standby, in 5W mode.
On the other hand, my measurements indicate that plate and screen voltages are turned on for this tube.
I found the cathode current to be around 40ma for the actuve tube (in my case). After looking at some data sheets, I think that the voltage values for the unused tube would indicate that only a very small current should flow, if any. Maybe even no current... So it should be clear that the "presumably" unused tube does not neutralize the quiescent magnetic field in the OT.
The point of letting one tube neutralize the magnetic field was to get a symmetric operation in the OT, so as to not cause asymmetric OT distortion when the signal was strong. Of course, if the signal is never this strong, that is not an issue.
So why are the plate and screen voltages on the unused tube turned on, and with such strange values, if no (or almost no) current flows? Maybe there is something else is going on... Maybe they have used the same circuit in another amp, one for which the schematic is available on the net. E.g. the Lonestar schematic is available, and I gather that it also has a class A switch. I will do some snooping around.
Torquil
On the other hand, my measurements indicate that plate and screen voltages are turned on for this tube.
I found the cathode current to be around 40ma for the actuve tube (in my case). After looking at some data sheets, I think that the voltage values for the unused tube would indicate that only a very small current should flow, if any. Maybe even no current... So it should be clear that the "presumably" unused tube does not neutralize the quiescent magnetic field in the OT.
The point of letting one tube neutralize the magnetic field was to get a symmetric operation in the OT, so as to not cause asymmetric OT distortion when the signal was strong. Of course, if the signal is never this strong, that is not an issue.
So why are the plate and screen voltages on the unused tube turned on, and with such strange values, if no (or almost no) current flows? Maybe there is something else is going on... Maybe they have used the same circuit in another amp, one for which the schematic is available on the net. E.g. the Lonestar schematic is available, and I gather that it also has a class A switch. I will do some snooping around.
Torquil
I found Lonestar schematics, but not with a class A switch on them. Maybe that is a newer version.
Measured DC resistanses from plate to OT centre tap:
30W mode:
Tube L: 222 Ohm
Tube R: 213 Ohm
5W mode:
Tube L: 2.9kOhm
Tube R: 213Ohm
This must be measured in while power is on, since the 5W engages a relay. The addition of 2.7kOhm to value for Tube L fits well with the fact that there is a 2.7kOhm power resistor bolted to the chassis near the switch... So this resistor is put in series with the primary of the OT on the side of Tube L.
Knowing this, and by measuring the voltage drop between each plate and the primary centre-tap, I calculate the following plate current.
Tube L: 36mA
Tube R: 1mA
The result for Tube L is consistent with my 40mA value from measuring on the cathode resistor. The difference could be accounted for by screen grid current.
So we can see that Tube R doesn't contribute much. This should be consistent with JJ's temperature measurements. It wouldn't get noticeably hotter with just 1mA through it. That only amounts to under 0.5W, which would not raise the temperature much, given that the heater already dissipates around 5W...
Torquil
Measured DC resistanses from plate to OT centre tap:
30W mode:
Tube L: 222 Ohm
Tube R: 213 Ohm
5W mode:
Tube L: 2.9kOhm
Tube R: 213Ohm
This must be measured in while power is on, since the 5W engages a relay. The addition of 2.7kOhm to value for Tube L fits well with the fact that there is a 2.7kOhm power resistor bolted to the chassis near the switch... So this resistor is put in series with the primary of the OT on the side of Tube L.
Knowing this, and by measuring the voltage drop between each plate and the primary centre-tap, I calculate the following plate current.
Tube L: 36mA
Tube R: 1mA
The result for Tube L is consistent with my 40mA value from measuring on the cathode resistor. The difference could be accounted for by screen grid current.
So we can see that Tube R doesn't contribute much. This should be consistent with JJ's temperature measurements. It wouldn't get noticeably hotter with just 1mA through it. That only amounts to under 0.5W, which would not raise the temperature much, given that the heater already dissipates around 5W...
Torquil
Ok, decided to drag this back up. I had been thinking that my amps volume had decreased quite bit from when it was new. I finally got it on the test bench and tested it. I was a little surprised to find that the best it could do was 13.7 watts in 25Watt mode and only 1.3 Watts in 5 watt mode :shock: Incidently, I could see the Dyna watt working. The inital peak was a bit over 20 watts which then settled back to the 13.7 watts after a few hundred milliseconds.
Iv'e had the amp for 11 months and use it for a few hours nearly every day, so I figured it was probably time for some new power tubes. Dropped a new pair of Mesa tubes in and did a retest to find that the power output was no better.
While I thought 13.7 watts was a bit low for 25 Watt mode, 1.3 is way low for 5 watt mode. This made me wonder if my amp has a fault.
After some more thought, I pulled the chassis out and started making some measurments. All voltages were very close to what Tmac had measured. Just to be sure about the tubes, I swapped them for an old set of Groove tubes I had from a Fender amp. Power increased a bit to 1.8 watts in 5 watt mode.
So, has anyone else measrued the power output of their 5:25 ? I'd realy like to know how mine compares.
Iv'e had the amp for 11 months and use it for a few hours nearly every day, so I figured it was probably time for some new power tubes. Dropped a new pair of Mesa tubes in and did a retest to find that the power output was no better.
While I thought 13.7 watts was a bit low for 25 Watt mode, 1.3 is way low for 5 watt mode. This made me wonder if my amp has a fault.
After some more thought, I pulled the chassis out and started making some measurments. All voltages were very close to what Tmac had measured. Just to be sure about the tubes, I swapped them for an old set of Groove tubes I had from a Fender amp. Power increased a bit to 1.8 watts in 5 watt mode.
So, has anyone else measrued the power output of their 5:25 ? I'd realy like to know how mine compares.
That would be great for someone else to do the check. Thanks for offering.
The way I made my measurement was to use an 8 Ohm load resistor instead of the speaker (8Ohm output socket). I used my scope to measure and look at the signal across the 8 Ohm load. The input was a signal generator set to sine wave. It was set to 1Khz, but I also changed it to 200 and 500Hz to make sure there was nothing frequency dependent happening.
The amp was set on channel 1 clean. I didn't measure the input signal level, but it was set such that the gain and master needed to be around 12:00 for clipping to be seen. I did also make sure that preamp clipping was not pressent by maxing the gain with the master lower and had no clipping pressent at the output.
The result of all this was that I was able to get 3Vrms(8.5V peak-peak) or so at the onset of clipping for 5 watt mode. The power was then calculated from power = V squared/R. This gives 1.125 watts. Pushed a little more it then made 1.3 watt. With the master maxed and a very square output wave the voltage was 3.8 Vrms (as measured by the Tektroncs scope) giving 1.8 watts.
For 25Watt mode I measured 10.48 Vrms with some clipping, which equates the 13.7 watts. My Blues junior made 18 watts using the same test setup.
The way I made my measurement was to use an 8 Ohm load resistor instead of the speaker (8Ohm output socket). I used my scope to measure and look at the signal across the 8 Ohm load. The input was a signal generator set to sine wave. It was set to 1Khz, but I also changed it to 200 and 500Hz to make sure there was nothing frequency dependent happening.
The amp was set on channel 1 clean. I didn't measure the input signal level, but it was set such that the gain and master needed to be around 12:00 for clipping to be seen. I did also make sure that preamp clipping was not pressent by maxing the gain with the master lower and had no clipping pressent at the output.
The result of all this was that I was able to get 3Vrms(8.5V peak-peak) or so at the onset of clipping for 5 watt mode. The power was then calculated from power = V squared/R. This gives 1.125 watts. Pushed a little more it then made 1.3 watt. With the master maxed and a very square output wave the voltage was 3.8 Vrms (as measured by the Tektroncs scope) giving 1.8 watts.
For 25Watt mode I measured 10.48 Vrms with some clipping, which equates the 13.7 watts. My Blues junior made 18 watts using the same test setup.
J.J said:
Hi J.J,
I agree with your calculation in the non-clipping case, but in the "almost square wave" I think the calculation should be a bit different.
In both cases, one can use Vrms in the power formula P = Vrms^2/R, but in the square wave case, Vrms is the same as Vpeak, i.e. for a square wave of voltages +V and -V, we would have Vpeak = Vrms = V.
You mentioned a 3.8Vrms for the square wave case, but I'm assuming this is Vpeak/sqrt(2) ? If so, I would say that in your square wave case,
Vpeak = Vrms = 3.8*sqrt(2) V = 5.4 V
From that, P = (5.4V)^2/8ohm = 3.6W.
Of course, please correct me if I'm wrong and/or read your post incorrectly. Also, please double-check my math!
Even with this, the wattages seem very low to me. Unfortunately I don't have a scope to check my amp.
EDIT: Hmm, I wonder if there is some clever and simple method of checking peak voltages...
Torquil
Hi Torquil. It's aways good to get a reaility check from someone else. As you have said, for a square wave Vrms = Vpeak. My scope gives a reading directly in Vrms, so I did not do the conversion mt self. As far a I know, the scopes reading is true rms and calculated from internally as a real average. So the rms voltage of 3.8 would have been pretty close the the peak for the sqaure wave. I seem to remember it being about 4Vpeak, but will have to go and make another measurement to confirm as I didn't record it.
As for the wattage, my Fender Bjr put out almost 18watts at the onset of clipping. Fender rated the output of that amp at 15 watts with I think 5% harmonic distorsion. I think that is fairly standard industry way to rate the output of a tube amp. I then expected to see something fairly close to 5 watts at the onset of clipping for my 5:25. Even with the same power tubes that made 18 Watts in the Bjr I could only get a little less than 14 watts. By comparision the Bjr only runs about 325V on the plates. As we know the express is up around 380V.
I didn't check the bias for 25 Watt mode, but in 5 watt it was a little over 36ma. I made that measurement by first measuring the DC resistance of the output transformer and then measuring the voltage across the winding with the amp on. I think this was 217Ohms and 8 Volts. Anyway, the 36ma equates to about 10.5 watts of idle plate disipation which shows that the tubes is working relativley hard.
It should be possible to make a measurement with just the dummy load, a diode, capacitor and bleed resistor. I'll give it a try and see if I can get a good result. You could probably just use your guitar, but an audio signal generator would be better.
I have also observed the voltage swing on the tube plates and that all looked good too. I am starting to suspect my output transformer of being faulty, but don;t want to take my amp to the dealer until I am sure something is wrong.
As for the wattage, my Fender Bjr put out almost 18watts at the onset of clipping. Fender rated the output of that amp at 15 watts with I think 5% harmonic distorsion. I think that is fairly standard industry way to rate the output of a tube amp. I then expected to see something fairly close to 5 watts at the onset of clipping for my 5:25. Even with the same power tubes that made 18 Watts in the Bjr I could only get a little less than 14 watts. By comparision the Bjr only runs about 325V on the plates. As we know the express is up around 380V.
I didn't check the bias for 25 Watt mode, but in 5 watt it was a little over 36ma. I made that measurement by first measuring the DC resistance of the output transformer and then measuring the voltage across the winding with the amp on. I think this was 217Ohms and 8 Volts. Anyway, the 36ma equates to about 10.5 watts of idle plate disipation which shows that the tubes is working relativley hard.
It should be possible to make a measurement with just the dummy load, a diode, capacitor and bleed resistor. I'll give it a try and see if I can get a good result. You could probably just use your guitar, but an audio signal generator would be better.
I have also observed the voltage swing on the tube plates and that all looked good too. I am starting to suspect my output transformer of being faulty, but don;t want to take my amp to the dealer until I am sure something is wrong.
I'd shoot off an email to Mesa Boogie support while waiting for responses here. They did bother to forward one of my earlier requests to the technicians, and I got a good answer from them. They might be able to confirm that your output is too weak.
Since you seem very competent with tube amps and you have a scope, you could try to check the OT by using a signal generator and your scope to analyze the turns/impedance ratios? I think this is possible without even disconnecting anything, right? Although since the primary winding is connected to the plus-side of an electrolytic cap in the power-supply, you would have to make sure you did it in a way that doesn't put any negative voltage on the cap, since that might damage it.
Hope you get it sorted out soon! You mentioned playing it a lot, so its not so cool if this takes long.
PS: Stay tuned! I just checked the manual for my multimeter, and it says that the AC setting is good for up to 400Hz. So I'll try to generate 400Hz from my computer soundcard and see if I can get a reading on the output of the amp. I have a powerful L-pad to use as a dummy load, and it allows me to hear a little bit of the signal. That way I can hear if it is clipping or not. I'll definitely do it in 5W mode. Possibly also in 25W mode if I trust my dummy load, but a 5W mode reading should be enough to confirm that your amp is damaged (if it is). I'll be able to do it this evening. I'm in Norway and the local time is 20:00 now.
EDIT: Just did it in 5W mode with a 400Hz signal. I wasn't able to surpass 3.8V on my multimeter at 400Hz. If that was a true RMS reading, I guess that corresponded to P = Vrms^2/R = (3.8*3.8)/8 W = 1.8W. At that level I could hear quite a bit of clipping.
Maybe nothing is wrong with the amp? Although I would be interested in hearing from Mesa how they got 5W out of it... Btw, the power tubes in mine are not stock, they are JJ's.
Maybe "Mesa Boogie 2:15" doesn't sound as good
Torquil
Since you seem very competent with tube amps and you have a scope, you could try to check the OT by using a signal generator and your scope to analyze the turns/impedance ratios? I think this is possible without even disconnecting anything, right? Although since the primary winding is connected to the plus-side of an electrolytic cap in the power-supply, you would have to make sure you did it in a way that doesn't put any negative voltage on the cap, since that might damage it.
Hope you get it sorted out soon! You mentioned playing it a lot, so its not so cool if this takes long.
PS: Stay tuned! I just checked the manual for my multimeter, and it says that the AC setting is good for up to 400Hz. So I'll try to generate 400Hz from my computer soundcard and see if I can get a reading on the output of the amp. I have a powerful L-pad to use as a dummy load, and it allows me to hear a little bit of the signal. That way I can hear if it is clipping or not. I'll definitely do it in 5W mode. Possibly also in 25W mode if I trust my dummy load, but a 5W mode reading should be enough to confirm that your amp is damaged (if it is). I'll be able to do it this evening. I'm in Norway and the local time is 20:00 now.
EDIT: Just did it in 5W mode with a 400Hz signal. I wasn't able to surpass 3.8V on my multimeter at 400Hz. If that was a true RMS reading, I guess that corresponded to P = Vrms^2/R = (3.8*3.8)/8 W = 1.8W. At that level I could hear quite a bit of clipping.
Maybe nothing is wrong with the amp? Although I would be interested in hearing from Mesa how they got 5W out of it... Btw, the power tubes in mine are not stock, they are JJ's.
Maybe "Mesa Boogie 2:15" doesn't sound as good
Torquil
To double check my multimeter reading, I also measured my AX84 SEL do-it-yourself amp in the same manner. I have one EL34 tube in it. The seller claims that the output should be around 9W. Assuming that my multimeter measures Vrms:
Across 8ohm dummy load
Vrms = 8V -> P = 8W
This was for a distorted signal. So that gives me at least some confidence in my multimeter measurement technique. If we assume that Vpeak is close to 9V since the above is probably not a perfectly square wave, that gives a clean Vrms at around 6.4V. Thus a max of P = 5W for a clean signal out of that amp.
I think other 1xEL84 amps can have around 5W output (e.g. AX84 P1 and Hi-Octane). Maybe Express OT is less efficient in 5W mode because it is not specially designed for single-ended amps and/or perhaps the impedance matching is not optimal in 5W mode?
Across 8ohm dummy load
Vrms = 8V -> P = 8W
This was for a distorted signal. So that gives me at least some confidence in my multimeter measurement technique. If we assume that Vpeak is close to 9V since the above is probably not a perfectly square wave, that gives a clean Vrms at around 6.4V. Thus a max of P = 5W for a clean signal out of that amp.
I think other 1xEL84 amps can have around 5W output (e.g. AX84 P1 and Hi-Octane). Maybe Express OT is less efficient in 5W mode because it is not specially designed for single-ended amps and/or perhaps the impedance matching is not optimal in 5W mode?
Thanks for checking that out. Your measurements seem to compliment mine.
Do Mesa have an email address ? I have not been able to find one on their web site. I would definately like to ask Mesa about this.
I did do some checks on the output transformer including injecting a signal into the speaker output and measuring the primary. From memory the impedance calculations worked out fairly close to ideal for an EL84. I don't think there is any issue with injecting a signal into the speaker. The center of the primary is connected to the power supply as you say, but the outside winding is only connected directly to the tube plates and when power is off that makes the windings pretty much isolated.
The impedance of one for just one side of the primary seems to be just about right for sinlge tube class A operation too. I would not have thought a transformer the size of the Express 5:25 would be saturating by 36ma, in fact if it was it would proably sound quite bad. That is another test I guess, to see how much current it can take before saturation.
One other thing I did was change the dummy load to 4Ohm. I was suprised to find that the output voltage just about halved making the power output fairly constant. Increasing the load also maintains the power similarily. Initally I expected that the power may nearly double witht he 4Ohm load, but that just doesn't happen.
2:15 , thats quite funny ! I want to test the dynawatt envelope at some stage to see exactly what it does. So far I know it hits about 20 watts so that would make the Express a 2:20.
Do Mesa have an email address ? I have not been able to find one on their web site. I would definately like to ask Mesa about this.
I did do some checks on the output transformer including injecting a signal into the speaker output and measuring the primary. From memory the impedance calculations worked out fairly close to ideal for an EL84. I don't think there is any issue with injecting a signal into the speaker. The center of the primary is connected to the power supply as you say, but the outside winding is only connected directly to the tube plates and when power is off that makes the windings pretty much isolated.
The impedance of one for just one side of the primary seems to be just about right for sinlge tube class A operation too. I would not have thought a transformer the size of the Express 5:25 would be saturating by 36ma, in fact if it was it would proably sound quite bad. That is another test I guess, to see how much current it can take before saturation.
One other thing I did was change the dummy load to 4Ohm. I was suprised to find that the output voltage just about halved making the power output fairly constant. Increasing the load also maintains the power similarily. Initally I expected that the power may nearly double witht he 4Ohm load, but that just doesn't happen.
2:15 , thats quite funny ! I want to test the dynawatt envelope at some stage to see exactly what it does. So far I know it hits about 20 watts so that would make the Express a 2:20.
- Joined
- Jan 12, 2008
- Messages
- 1,469
- Reaction score
- 10
There's an American flag on their contact page. Click on it, I believe that was how I got to their e-mail address.
A
Anonymous
Guest
1-2w is about what you get out of an el84 when running it strapped as a Triode... Is it configured as such in the "5w" mode?
I have a 5:50, but I was always curious as to why mesa rated the two amps at 5w when running SE. I always figured the 6l6's were run Triode-strapped and the EL84's as Pentodes. Maybe we'll find out now
I have a 5:50, but I was always curious as to why mesa rated the two amps at 5w when running SE. I always figured the 6l6's were run Triode-strapped and the EL84's as Pentodes. Maybe we'll find out now
ducati said:1-2w is about what you get out of an el84 when running it strapped as a Triode... Is it configured as such in the "5w" mode?
I have a 5:50, but I was always curious as to why mesa rated the two amps at 5w when running SE. I always figured the 6l6's were run Triode-strapped and the EL84's as Pentodes. Maybe we'll find out now
I have also wondered how they made the 6L6 in the 5:50 only put out 5 watts. While I don't know about the 5:50 I can tell you that the 5:25 is not running in triode mode, but the sceen voltage is quite low at only 200V. The plate voltage however 300V which is 50V over what the data sheet (For EL84) says will deliver 5 watts. The screen voltage is 50V under.
The 200V screen has also be observed by Tmac, so it appears that is not the problem with my amp.
Just managed to hook my 5:25 to my CompuBias. Since I usually stick to the 5W mode I never bothered to test the darn thing. For one thing it turned out that the so-called perfectly matched Mesa tubes are not matched at all. You'll see below:
Readings in 30W mode :
Left - plate 339V, Ik 31, W10.5
Right - plate 337, Ik 23, W7.7
Readings in 5W mode:
Left - plate 151V, Ik 70, W10.5
Right - plate 330, Ik 0, W0
Readings in 30W mode :
Left - plate 339V, Ik 31, W10.5
Right - plate 337, Ik 23, W7.7
Readings in 5W mode:
Left - plate 151V, Ik 70, W10.5
Right - plate 330, Ik 0, W0
Similar threads
