Hi there,
I had one friend of mine build me a clone M7 Kondo preamp as per
www.diyparadise.com/ANM7.gif
This preamp (the one tube-design) sounds, at least to my ears, very good, but unfortunately it's got "a Neumann u47 complex": it's so microphonic that it can pick up and amplify a handclap 1 or 2 meters away into the room, not to mention the clangy sounds you get when you touch the volume knob, the chassis or, obviously, the very tube. The worst part is that with certain tubes it simply starts oscillating when I power it on, emitting a nasty feedback sound I can only control by pressing my fingers against the tube.
My friend has done some checking, isolating the different sections of the circuit, but the problem remains. The only concerns/clues he's got is that the tube is getting some 20 v. less than the 200odd it should be receiving (which he thinks it shouldn't be a big deal) and the fact that we installed Non Polar Muse caps on the cathode position and he doesn't know if that may have any impact on the desing. Other than that he has no clue what may be happening to the pre.
Could it be that this design makes the five tubes I tried microphonic? Could there be any other problem you could direct me to?
Thanks in advance for any help you could provide me with?
Regards,
Raul Fernandez-Garrido
Xixon-Asturies
I had one friend of mine build me a clone M7 Kondo preamp as per
www.diyparadise.com/ANM7.gif
This preamp (the one tube-design) sounds, at least to my ears, very good, but unfortunately it's got "a Neumann u47 complex": it's so microphonic that it can pick up and amplify a handclap 1 or 2 meters away into the room, not to mention the clangy sounds you get when you touch the volume knob, the chassis or, obviously, the very tube. The worst part is that with certain tubes it simply starts oscillating when I power it on, emitting a nasty feedback sound I can only control by pressing my fingers against the tube.
My friend has done some checking, isolating the different sections of the circuit, but the problem remains. The only concerns/clues he's got is that the tube is getting some 20 v. less than the 200odd it should be receiving (which he thinks it shouldn't be a big deal) and the fact that we installed Non Polar Muse caps on the cathode position and he doesn't know if that may have any impact on the desing. Other than that he has no clue what may be happening to the pre.
Could it be that this design makes the five tubes I tried microphonic? Could there be any other problem you could direct me to?
Thanks in advance for any help you could provide me with?
Regards,
Raul Fernandez-Garrido
Xixon-Asturies
Find out where the ringing is coming from; the valve or a ceramic capacitor. The valve is the most common issue, just too noisy.
Try tapping individual components with a plastic biro until you find the culprit and replace it.
Try tapping individual components with a plastic biro until you find the culprit and replace it.
Ringing is probably a symptom of instability in this case. Have you got grid stoppers on the pin and is the load resistor on the pin of the plate ?
My experience is that 5687's are not typically microphonic in the correct setup.
Shoog
My experience is that 5687's are not typically microphonic in the correct setup.
Shoog
I will second Shoog's comments, the symptoms described indicate VHF oscillation is probably occurring. The 5687 is not terribly microphonic in my experience either.
A grid stopper right at the socket and a 47 ohm resistor between the plate connection and output capacitor right at the socket will usually help. (Load resistor should be close too)
Note I fixed the link to the schematic.
A grid stopper right at the socket and a 47 ohm resistor between the plate connection and output capacitor right at the socket will usually help. (Load resistor should be close too)
Note I fixed the link to the schematic.Just out of pure ignorance...
Thanks for your time and counsel, Kevin, but I've got some questions: Would these mods affect the sound of the preamp? Why is that the designer Kondo did not include them in the design? How can I calculate the value of the grid stopper (= which I've just google and it seems to be a resistor?)? As you see My knowledge of electronics is to say the least very limited.
Thanks again.
Thanks for your time and counsel, Kevin, but I've got some questions: Would these mods affect the sound of the preamp? Why is that the designer Kondo did not include them in the design? How can I calculate the value of the grid stopper (= which I've just google and it seems to be a resistor?)? As you see My knowledge of electronics is to say the least very limited.
Thanks again.
A grid stopper need to be about a 100R and should have no effect on the overall sound.
I built this circuit as one of my first projects and it had hum pickup, I spent ages putting in DC heaters and rearranging the layout but nothing got rid of it. It turned out to be oscillation which was making it "detect" the induced electro magnetic fields in the case. I abandoned the build in the end, but have since used the 5687 on a few occasions without issues and without any problem with hum even when using AC heaters. I believe the issue must be with the relatively high transconductance of the tube and its internal design.
Shoog
I built this circuit as one of my first projects and it had hum pickup, I spent ages putting in DC heaters and rearranging the layout but nothing got rid of it. It turned out to be oscillation which was making it "detect" the induced electro magnetic fields in the case. I abandoned the build in the end, but have since used the 5687 on a few occasions without issues and without any problem with hum even when using AC heaters. I believe the issue must be with the relatively high transconductance of the tube and its internal design.
Shoog
The 5687 is a high gm type. High gm types are quite prone to parasitic oscillation. The suggestions to use "stoppers" on both the grid and plate socket terminals make sense. Use a carbon composition part on the grid and a carbon film part on the plate. Mount the resistor bodies as close as can be to the socket lugs. Metal that forms an antenna is highly undesirable.
Carbon composition is both non-inductive and non-metallic, but it can be noisy when current is present. Current is all but non-existent in the grid circuit and noise is not an issue. OTOH, substantial current is present in the plate circuit and noise IS an issue here. Carbon film is non-metallic, but the laser trimming during manufacture introduces some inductance. The tiny inductance of a 47 Ω part should not be troublesome.
Carbon composition is both non-inductive and non-metallic, but it can be noisy when current is present. Current is all but non-existent in the grid circuit and noise is not an issue. OTOH, substantial current is present in the plate circuit and noise IS an issue here. Carbon film is non-metallic, but the laser trimming during manufacture introduces some inductance. The tiny inductance of a 47 Ω part should not be troublesome.
Last edited:
RAFG, probably safe to assume to that there might be a few minor details missing from that schematic. Construction approach in the Kondo gear also played a significant role. Adding the suggested changes should fix the noted problems and will not change the sonic character of the pre-amplifier.
Can you post some pictures of your build, possibly something will leap out at us.
Can you post some pictures of your build, possibly something will leap out at us.
I'll try to get around to taking some pics of the innards of the preamp and posting them. I'm afraid I won't be able to do it until the weekend.
Regards,
Raul
Regards,
Raul
Hi, Mach1, thanks for your help. I hope not to put my foot in it -I'm illiterate as far as electronics is concerned- but I believe that what you refer to is if we built that part of the circuit which goes to the 8 pin of the 5687. In that case, the answer is yes.
The first cap should be 10uF not 100uF. The 6X4 datasheet is very clear about maximum capacitor value for the first cap in a CRC or CLC filter.
200V FW rectified is about 282 V.
V-drop on the rectifier is about 20V.
This leaves 262 at the first cap.
If you are running 3mA for two triodes in parallel you only have 1.5ma/plate(Typical operation at 120Vak is 36mA!).
At 3mA (X2 for two channels), the voltage drop across the 5K is 30V, leaving 252V for the tube, and two resistors.
The drop across the anode resistor is 0.003 * 50000 = 155V. This only leaves 97V for both the cathode resistor and tube.
At 3ma, the drop on the 2K cathode resistor is only 6V which leaves 91V across the tube, which is pretty close on the plate curve which puts VAK around 100V.
I think this is not the best place to be biased, and will result in higher output impedance and distortion than if biased closer to manufacturers specifications.
with 72pF miller capacitance per tube, you have about 142 plus stray for the input capacitance. Say 150pf.
1/2 PI 10,000 * 150e-12 = 106,103Hz
10K grid-stop will result in an upper f3 of over 100KHz, so I recommend a larger grid-stop than 100R. 4,7K or 10K would be good. You could even go up as high as 47K before entering the primary audio bandwidth (20hZ to 20KHz).
Plate-stop of 100r would be good. A ferrite bead might be a better solution.
A 1uF film cap would be better than the 100uF electrolytic for the filament elevation filter cap. I usually use 0.15uF film caps in that application.
Ignore - two tubes will bring it in range.(I recommend you measure the voltage on the filament. I suspect it is close to 15V and you need to increase the resistor value in the CRC filter.)
200V FW rectified is about 282 V.
V-drop on the rectifier is about 20V.
This leaves 262 at the first cap.
If you are running 3mA for two triodes in parallel you only have 1.5ma/plate(Typical operation at 120Vak is 36mA!).
At 3mA (X2 for two channels), the voltage drop across the 5K is 30V, leaving 252V for the tube, and two resistors.
The drop across the anode resistor is 0.003 * 50000 = 155V. This only leaves 97V for both the cathode resistor and tube.
At 3ma, the drop on the 2K cathode resistor is only 6V which leaves 91V across the tube, which is pretty close on the plate curve which puts VAK around 100V.
I think this is not the best place to be biased, and will result in higher output impedance and distortion than if biased closer to manufacturers specifications.
with 72pF miller capacitance per tube, you have about 142 plus stray for the input capacitance. Say 150pf.
1/2 PI 10,000 * 150e-12 = 106,103Hz
10K grid-stop will result in an upper f3 of over 100KHz, so I recommend a larger grid-stop than 100R. 4,7K or 10K would be good. You could even go up as high as 47K before entering the primary audio bandwidth (20hZ to 20KHz).
Plate-stop of 100r would be good. A ferrite bead might be a better solution.
A 1uF film cap would be better than the 100uF electrolytic for the filament elevation filter cap. I usually use 0.15uF film caps in that application.
Ignore - two tubes will bring it in range.(I recommend you measure the voltage on the filament. I suspect it is close to 15V and you need to increase the resistor value in the CRC filter.)
Last edited:
Not nearly as simple an issue as it would first appear.
Will a 6X4 handle 100uF on the filter input?... - Steve O - Tube DIY Asylum
Why risk it though. Why do something that the datasheet specifically warns against, and given that all 6x4 are not made equal its only a matter of time before you encounter a poor one which can't stand the abuse.
I have mine delivering about 50mA at 400V (which is probably slightly over spec) and have initial cap of 10uf followed by a fluro ballast and then about 440uf of capacitance. It has worked without issue for a few years at this stage. These are rugged and reliable rectifiers - but why abuse them unnecessarily.
After all you can get away with constantly driving at 100mph without incident - but its hardly advisable.
Shoog
Last edited:
You appear have a great talent for putting words into the mouths of people.
All I said was that the issue is not nearly as simple as it would first appear. If you read the posts you would realise that there is actually a divergence of opinion on the matter.
People are free to research the issue, choose whose opinions they find most credible and decide how they use their rectifiers.
All I said was that the issue is not nearly as simple as it would first appear. If you read the posts you would realise that there is actually a divergence of opinion on the matter.
People are free to research the issue, choose whose opinions they find most credible and decide how they use their rectifiers.
Last edited:
Honestly, guys, I've got a bit of mixed feelings. On the one hand, a good row over my preamp is good to keep this thread alive and invite this way other contributions; but on the other, I hate to see two people who are giving me a hand and I'm so grateful to entangling in a heated discussion. I know that I'm risking this peacekeeping mission may have as an end you both turning on me. But anyway, call me a Blue Helmet of the United Triodes Organization... Going back to my first intention on writing this post: Thanks you both, your help is invaluable. I'm trying to take in all the technical aspects -which is no easy for me- but they will eventually help to take the decision as to what course of action to take.
Thanks again.
Raul
Xixon-Asturies
Thanks again.
Raul
Xixon-Asturies
Kevin asked me for some pics... here they are, hopefully...
An externally hosted image should be here but it was not working when we last tested it.
OK, good you posted the photo. The way the schematic is drawn, I interpreted it to use two sections of the dual triode in parallel. Two tubes would be used for two channels. Looking at your photos, I see you have one triode section per channel. So my estimates will be off somewhat. You will be running less current per triode than I expected.
Check your filament voltage. I think the filament supply was meant for two tubes. You probably have higher than recommended voltage on the triode filaments. You may need to increase the R in the CRC filter to get the voltage below 13.23V, and might as well adjust it to 12.6 for optimum tube life.
No issues with dissipation of any components as they are all way over-rated power wise.
Otherwise nice clean layout.
Check your filament voltage. I think the filament supply was meant for two tubes. You probably have higher than recommended voltage on the triode filaments. You may need to increase the R in the CRC filter to get the voltage below 13.23V, and might as well adjust it to 12.6 for optimum tube life.
No issues with dissipation of any components as they are all way over-rated power wise.
Otherwise nice clean layout.
Last edited:
- Status
- Not open for further replies.