Just bought a 2nd hand CJ MV50 power amp. I'm surprised by the bias setting of the EL34 output tubes. Because of the 220V to 230V switch during the last 10 years or so, the tube has more than 500V on the plate while the optimal idle current has around 60mA on the meter. That means using the tubes on the edge and implying a short tube life (1000-1500 hours I guess).
So I'm not feeling relaxed about that. The bias can be adjusted by individual trimmers using the famous red led of CJ. Of course I can neglect the LED and just turn to a current of around 45 mA. The problem then is that the negative voltage supply (that gives the bias) end at around -47V which is not enough for all tubes to do the setting. So I have to mod the bias circuit.
Then the problem is: I can't use the visual inspection of the LED's anymore. To re-use that I have to mod the bias circuit yet once more.
Still another option is to bring down the input voltage of the high tension. I can put 47ohm/5Watt resistors in each AC line to lower the DC to around 470V or so (just guessing, total current for all tubes will be around 250 mA). These will become hot, being just a waste of power. But, heh, you have to bring in an offer to save some tube life.
Then again: I don't know how this all will change the sound of the (now very good sounding) MV50.
The last option maybe is using a more robust tube like the 6550 of some KTxx. This I don't like because of the nice character of the EL34 (not forgetting its lower price and fine availability) and then once again I have to change the bias circuit too.
So: I'm not sure what to do. What would you do? Any advice is welcome. Maybe some MV50/52/55 owners out there.
So I'm not feeling relaxed about that. The bias can be adjusted by individual trimmers using the famous red led of CJ. Of course I can neglect the LED and just turn to a current of around 45 mA. The problem then is that the negative voltage supply (that gives the bias) end at around -47V which is not enough for all tubes to do the setting. So I have to mod the bias circuit.
Then the problem is: I can't use the visual inspection of the LED's anymore. To re-use that I have to mod the bias circuit yet once more.
Still another option is to bring down the input voltage of the high tension. I can put 47ohm/5Watt resistors in each AC line to lower the DC to around 470V or so (just guessing, total current for all tubes will be around 250 mA). These will become hot, being just a waste of power. But, heh, you have to bring in an offer to save some tube life.
Then again: I don't know how this all will change the sound of the (now very good sounding) MV50.
The last option maybe is using a more robust tube like the 6550 of some KTxx. This I don't like because of the nice character of the EL34 (not forgetting its lower price and fine availability) and then once again I have to change the bias circuit too.
So: I'm not sure what to do. What would you do? Any advice is welcome. Maybe some MV50/52/55 owners out there.
I'm not familiar with this amp, but isn't there a way to set it (or wire it) for 240V mains?
That would take the stress off of the EL34s without hacking into the design.
That would take the stress off of the EL34s without hacking into the design.
The maximum output from an EL34 is about 25 watts. So measure the anode voltage of the amp, for this exercise we will assume that to be 445 volts. Divide the 25 by the 445 then multiply by 0.7 the answer in ma. is the 70% plate load bias setting required for that amplifier.
The 0.7 means 70% Hot anode / plate dissipation. Therefore you can use 0.5 (50%) as a Cool AB setting, 0.6 (60%) as a Warmer AB setting the one I would prefer. Or go for the 0.7 or even 0.75 (70 - 75%) if you like it hot.
The 0.7 means 70% Hot anode / plate dissipation. Therefore you can use 0.5 (50%) as a Cool AB setting, 0.6 (60%) as a Warmer AB setting the one I would prefer. Or go for the 0.7 or even 0.75 (70 - 75%) if you like it hot.
2nd round
Thanks for the replies:
- That would be a nice idea to take another winding of the transformer. There just is a bunch of coloured wires coming out of the transformer. All wires are used, no spare ones, no solder points on the outside, no markings. Maybe inside the transformer there are possibilities but I don't want to go that way.
- KT88: don't know about the heater current irt the transformer capacity. Would like to stick to the EL34.
- About calculation of the dissipation: I know, that's just the problem. There is around 520V on the anode and the current is 60 ma as recommended from the led's bias setting (1,2V on the cathode resistor of 20 ohm). So dissipation is 520 x 0.06 = 31,2 Watt!! From the beginning this would have been somewhat lower (maybe 30 Watts, from 500 plate voltage) but way too high for normal tube life.
Must be the case that all existing MV50 owners order new tubes every now and then (and/or get some fuses flushing out, these fuses from type BLS (? some professional type of fuse, large) have a price tag of around 8 dollars a piece!!!).
Thanks for the replies:
- That would be a nice idea to take another winding of the transformer. There just is a bunch of coloured wires coming out of the transformer. All wires are used, no spare ones, no solder points on the outside, no markings. Maybe inside the transformer there are possibilities but I don't want to go that way.
- KT88: don't know about the heater current irt the transformer capacity. Would like to stick to the EL34.
- About calculation of the dissipation: I know, that's just the problem. There is around 520V on the anode and the current is 60 ma as recommended from the led's bias setting (1,2V on the cathode resistor of 20 ohm). So dissipation is 520 x 0.06 = 31,2 Watt!! From the beginning this would have been somewhat lower (maybe 30 Watts, from 500 plate voltage) but way too high for normal tube life.
Must be the case that all existing MV50 owners order new tubes every now and then (and/or get some fuses flushing out, these fuses from type BLS (? some professional type of fuse, large) have a price tag of around 8 dollars a piece!!!).
recommended
By the preferred calculation of Jon Snell I would get: 25 / 500 = 50 mA. Times .7 of .6 gives bias current of 30 - 35 mA. I would be happy with 40-45 mA which is a reduction of plate dissipation from 31 Watt to roughly 20 Watts.
By the preferred calculation of Jon Snell I would get: 25 / 500 = 50 mA. Times .7 of .6 gives bias current of 30 - 35 mA. I would be happy with 40-45 mA which is a reduction of plate dissipation from 31 Watt to roughly 20 Watts.
That you have to ask the designers of CJ. The only reason I can come up with is that they think the sound quality is better with this ludicrous high setting. Another possibility is that they have shares is some tube-factory and hope to sell a lot of tubes each year.
For sake of clarity I will repeat my issue: what best to do for a longer tube life (and therefore lower current and plate voltage, preferably so that the sacrifice of sound quality is minimal).
Hi,
I haven't check the schematic of the MV50 so I'm not so sure but I'm working on a MV-55 with similar circuit (I think) so here are my opinions:
- LED bias checking circuit uses the voltages across the 20R cathode resistors of the EL34s to compare with a reference voltage. This ref. voltage is set up by a voltage divider across the output of a LM7805 around the +/-12V supply for the opamps. Look around and you will find a small trimmer for this job. Adjust this trimmer so that the LEDs go off at your desired current (800mV for 40mA...)
- C.J uses neg. regulator for the bias supply, if you want more (not too much) negative supply you can tweak the reference zener diode of this circuit (higher voltage zener for more negative supply)
- KT88/6550 might be able to use in this amp since heater requirements are just 7% higher (1.6A vs. 1.5A at 6.3VAC) so you can try as you like but with not too high anode current. Still, you need to watch out if Power transformer overheat happens.
Cheers,
Duong
I haven't check the schematic of the MV50 so I'm not so sure but I'm working on a MV-55 with similar circuit (I think) so here are my opinions:
- LED bias checking circuit uses the voltages across the 20R cathode resistors of the EL34s to compare with a reference voltage. This ref. voltage is set up by a voltage divider across the output of a LM7805 around the +/-12V supply for the opamps. Look around and you will find a small trimmer for this job. Adjust this trimmer so that the LEDs go off at your desired current (800mV for 40mA...)
- C.J uses neg. regulator for the bias supply, if you want more (not too much) negative supply you can tweak the reference zener diode of this circuit (higher voltage zener for more negative supply)
- KT88/6550 might be able to use in this amp since heater requirements are just 7% higher (1.6A vs. 1.5A at 6.3VAC) so you can try as you like but with not too high anode current. Still, you need to watch out if Power transformer overheat happens.
Cheers,
Duong
Last edited:
solved so far
After some experimentation I solved the problem so far.
1. I inserted a zener diode in the bias circuit so that the range of adjustment at the downside is broader (now -55v in stead of -48V).
2. I changed a resistor in the bias circuit so that the range at the upside is broader (was -36V, now -30V).
3. So now I can experiment freely with more tubes and currents and voltages.
4. I tried to change the kathode resistor from 20 ohm to 24 ohm so that less current still made the LED glowing red. For some reason this seemed not to sound as good. So I reset to 20 ohm again.
5. I lowered the high voltage from 530V to 460V by adding 2x 0.68ohm before rectification. These become rather hot and will not do much good for the electrolytics that sit in the neighbourhood. Maybe I try for .47 of .33 later.
So now I listen to various setting during the next weeks.
After some experimentation I solved the problem so far.
1. I inserted a zener diode in the bias circuit so that the range of adjustment at the downside is broader (now -55v in stead of -48V).
2. I changed a resistor in the bias circuit so that the range at the upside is broader (was -36V, now -30V).
3. So now I can experiment freely with more tubes and currents and voltages.
4. I tried to change the kathode resistor from 20 ohm to 24 ohm so that less current still made the LED glowing red. For some reason this seemed not to sound as good. So I reset to 20 ohm again.
5. I lowered the high voltage from 530V to 460V by adding 2x 0.68ohm before rectification. These become rather hot and will not do much good for the electrolytics that sit in the neighbourhood. Maybe I try for .47 of .33 later.
So now I listen to various setting during the next weeks.
Great, it seems like you got everything sorted. Kind of odd the 24 Ohms cathode current sense resistor has such impact on the sound... anyway, please let us know what you end up with in terms of the various voltage readings after making the changes.
posted
I will keep you posted.
With respect to the cathode resistor: I added 2 paralell small 8 ohms resistors (1/2 watt a piece) to the existing 20 ohm (10 Watt or so). Maybe such a match is not ideal (big high quality resistor with 2 small cheap ones in series). And maybe it just is in my mind, didn't try it for long.
Another question for the MV50. I have high sensitive speakers (95db) and can hear a soft humming with my ear in the speaker. I guess with normal sensitivity speakers it is hard to hear at all. Suppose I want to get rid of this hum. What is more likely to work:
1. Make the filaments of the input and driver tubes 6,3 V DC (5751 and 6FQ7). Not a big deal, I could use the existing 6,3V AC for that.
2. Or could it be that some hum is injected via the bias supply for the output tubes (maybe one electrolytic 50uF/150V has to be renewed (can never do any harm)?
I hesitate about turning the filaments to DC because I read that the best to do is using AC (for tube life and for sound reasons). On the other hand: the more expensive CJ premier 4 uses DC filament supply. Maybe I'm just nit-picking (well I've built tube amps before that were dead silent in all respects).
Is there an existing MV50 owner who can confirm that this amp is dead quite on high sensitive speakers?
I will keep you posted.
With respect to the cathode resistor: I added 2 paralell small 8 ohms resistors (1/2 watt a piece) to the existing 20 ohm (10 Watt or so). Maybe such a match is not ideal (big high quality resistor with 2 small cheap ones in series). And maybe it just is in my mind, didn't try it for long.
Another question for the MV50. I have high sensitive speakers (95db) and can hear a soft humming with my ear in the speaker. I guess with normal sensitivity speakers it is hard to hear at all. Suppose I want to get rid of this hum. What is more likely to work:
1. Make the filaments of the input and driver tubes 6,3 V DC (5751 and 6FQ7). Not a big deal, I could use the existing 6,3V AC for that.
2. Or could it be that some hum is injected via the bias supply for the output tubes (maybe one electrolytic 50uF/150V has to be renewed (can never do any harm)?
I hesitate about turning the filaments to DC because I read that the best to do is using AC (for tube life and for sound reasons). On the other hand: the more expensive CJ premier 4 uses DC filament supply. Maybe I'm just nit-picking (well I've built tube amps before that were dead silent in all respects).
Is there an existing MV50 owner who can confirm that this amp is dead quite on high sensitive speakers?
Last edited:
If you can hear hum, there's two likely culprits:
1. Grounding. Proper grounding practice is rarer in the "high end" than you would expect. Ruthlessly analyze it and rework.
2. Radiative pickup from the power transformer and/or choke (if there is one). Fixing that may involve some major physical changes, unfortunately.
1. Grounding. Proper grounding practice is rarer in the "high end" than you would expect. Ruthlessly analyze it and rework.
2. Radiative pickup from the power transformer and/or choke (if there is one). Fixing that may involve some major physical changes, unfortunately.
PM sent...
I once modified two MV100, originally equipped with EL34.
They both kept on blowing the HT fuse due to exceeding the max screen voltage.
Once I changed to 6550 in triode mode, changed the biasvoltage as Duong explanained and some other tweeks they never blew a fuse anymore and performance increased.
I once modified two MV100, originally equipped with EL34.
They both kept on blowing the HT fuse due to exceeding the max screen voltage.
Once I changed to 6550 in triode mode, changed the biasvoltage as Duong explanained and some other tweeks they never blew a fuse anymore and performance increased.
This thread reminds me of the story of how in the 70s, the US importer of Marshall switched the EL34s to 6550s to reduce the number of warranty claims. (presumably without changing the NFB nor the bias setting...)
I haven't had problems with my old plexi running EL34s (@520V) but you need good ones: JJ lists the max screen voltage @ 450V, while a Philips datasheet shows 500V (800V cold). Add to that the usual spec busting done in those days and you know why you're in trouble with current production valves.
I'll never forget the Marshall Major amp I once saw in a second hand shop: they had put EL34s in there instead of KT88s. The shop owner was very annoyed when I pointed that out and told me Marshall always used EL34s... Not pushing 200W out of a quad running around 650V-UL they didn't!
Blackened as they were, those EL34s survived.
Conrad Johnson - MV 100 double triode 5731 I can't find the pedinautra of the double triode 5731 on the net… ..; is there anyone who knows her? Thank you so much GP Treviso
Hello, I have a MV55 and really want to use TS 7581A’s, ideally I’d like to possibly Install a switch that lowers B+. Do you have any pictures of the mods you have done to lower the B+ and the bias mods.?
Cheers.!