Hi there,
I have just built a really simple 5867 line stage. Sounds really nice, powerful and clean. However I have a problem, its humming. I have scoped out the circuit and it would appear that the source of the hum is the AC heater I am using. I am running it off a 12.6 AC supply. I have tried various arrangements for referencing the centre tap (to ground, to 15V DC, floating), but nothing seems to effect the hum.
I understand that this valve has a tendency to hum, and most people seem to resort to DC heater supplies. Has anyone out there had any success with getting this tube quiet with AC heaters, and if so how? The problem I face is that I have shoe horned this into a PC power supply case and its got no more room left. In order to accomadate a DC supply it would require me to build a new case, which I am reluctant to do.
Shoog
I have just built a really simple 5867 line stage. Sounds really nice, powerful and clean. However I have a problem, its humming. I have scoped out the circuit and it would appear that the source of the hum is the AC heater I am using. I am running it off a 12.6 AC supply. I have tried various arrangements for referencing the centre tap (to ground, to 15V DC, floating), but nothing seems to effect the hum.
I understand that this valve has a tendency to hum, and most people seem to resort to DC heater supplies. Has anyone out there had any success with getting this tube quiet with AC heaters, and if so how? The problem I face is that I have shoe horned this into a PC power supply case and its got no more room left. In order to accomadate a DC supply it would require me to build a new case, which I am reluctant to do.
Shoog
DC'd the filament supply today. Helped a lot (about halved the hum, which I can live with). Don't think I will get this circuit absolutely quiet without a complete rebuild. I am sufficiently impressed that I might make a second version for use in my main system.
The only problem I have now is that the output is so powerful that it keeps sending one channel of my LM1875 into thermal shut down, which it never did before. I suppose this is a product of the much increased bass extension. I know there is a problem with the heatsink mounting on the dodgy channel, so I should be able to eliminate it.
5687 is a great tube.
Shoog
The only problem I have now is that the output is so powerful that it keeps sending one channel of my LM1875 into thermal shut down, which it never did before. I suppose this is a product of the much increased bass extension. I know there is a problem with the heatsink mounting on the dodgy channel, so I should be able to eliminate it.
5687 is a great tube.
Shoog
I went through this (annoying) spell quite recently with 7199's, the NOS RCA types with spiral wound heaters were brilliant... guaranted low hum level as expected.....then I replaced them with Sovteks and up it came. Desperate not to butcher the chassis I played with the 6.3V AC and got nowhere until I decided to lift the whole heater supply off deck and bias it with -10V; cap it with 100uF or so.......and the hum went. One of those weird things. However, I can't say this is the receipe for all low level tubes....but it worked on these with min component count..
rich
until I decided to lift the whole heater
supply off deck and bias it with -10V; cap it with 100uF or so.......and the hum went.
I have been messing about some more. and decided to try a very similar thing. I took the positive rail of my DC and referenced it to earth, so that the heater is no totally below the Cathode potential. I also tied the centre tap pin of the 5687 to earth via a 1000uf cap. Unfortunately the hum is still there, might be slightly better.
This is only my computer Pre, so i'am not distraught. My PCC88 cascode pre for my main system is wisper quiet, and my main amp is much more sensative. If I was using this new pre on my main system it would be unbearable.
On the original circuit the DC heater is referenced to the heater center tap pin to about +15V via a voltage divider off the B+, I think this arrangement is designed to inject a bit of noise off the B+ supply rail which cancels the noise from the heaters. I have read about such a technic.
The 5687 can't of been designed for audio use as it seems to have this design flaw, probably something to do with the huge heater current, and the inevitability of creep to the cathode.
Still looking for good solutions.
Shoog
supply off deck and bias it with -10V; cap it with 100uF or so.......and the hum went.
I have been messing about some more. and decided to try a very similar thing. I took the positive rail of my DC and referenced it to earth, so that the heater is no totally below the Cathode potential. I also tied the centre tap pin of the 5687 to earth via a 1000uf cap. Unfortunately the hum is still there, might be slightly better.
This is only my computer Pre, so i'am not distraught. My PCC88 cascode pre for my main system is wisper quiet, and my main amp is much more sensative. If I was using this new pre on my main system it would be unbearable.
On the original circuit the DC heater is referenced to the heater center tap pin to about +15V via a voltage divider off the B+, I think this arrangement is designed to inject a bit of noise off the B+ supply rail which cancels the noise from the heaters. I have read about such a technic.
The 5687 can't of been designed for audio use as it seems to have this design flaw, probably something to do with the huge heater current, and the inevitability of creep to the cathode.
Still looking for good solutions.
Shoog
Hello,
I built a simple 5687 line stage with DC filament and it's dead quiet. Definitely try DC'ing the filament. I use two 12V 2A transformer coupled together. The DC filament comes from the 6V CT and I used FW Bridge and lots of CRC filtering( 2200uf, 27Ohm, 2200uf, 3.8Ohm, 1000uf). I'm not sure if different brand of tube have any effect, but I'm using NOS RCA 5687. Have to say that it sound exceptional. I'm feeding it through a Marantz 8B and I can believe how good this thing sound with such a simple circuit. Not very microphonic either. Not like my 12AU7 which is very forward and microphonic.
ABMOGGY
I built a simple 5687 line stage with DC filament and it's dead quiet. Definitely try DC'ing the filament. I use two 12V 2A transformer coupled together. The DC filament comes from the 6V CT and I used FW Bridge and lots of CRC filtering( 2200uf, 27Ohm, 2200uf, 3.8Ohm, 1000uf). I'm not sure if different brand of tube have any effect, but I'm using NOS RCA 5687. Have to say that it sound exceptional. I'm feeding it through a Marantz 8B and I can believe how good this thing sound with such a simple circuit. Not very microphonic either. Not like my 12AU7 which is very forward and microphonic.
ABMOGGY
Hello,
You should be able to get near-hum free performance by using 6.3V on the 5687, with 50-ohm (or so) resistors from each lead, or a "hum pot" (as used with DH tubes), to AC ground. The problem with 12.6V is that each half of the 5687 will have 6.3V across it, with one end grounded and the other seeing the hum (effectively SE filament). By using 6.3V, you can get a balanced 3.15V, and thus balanced hum, across each section.
AC or DC operation should work equally well; I would use AC on sonic grounds, though this may be a matter of taste. AC is also less taxing on the PS.
Also - it is usual to reference the heater circuit to a positive (+20 - +40V), rather than a negative, DC voltage. This should reduce leakage and increase life, I believe. (?) I haven't found any significant reduction in hum (might vary?) using this technique.
The heater reference voltage divider is typically decoupled to PS ground (lower leg bypassed). I don't think it's possible to reduce total hum by injecting B+ hum into the heater, though I could be wrong. At least with AC heating, the heater hum tends to be a dark "note" (little overtones), whereas B+ hum is closer to a buzz, so they may not match in frequency content. There's also a difference in that hum across the heater is differential mode, whereas injected hum would be common mode, so I can't see it working (unless the heater feed is unbalanced - in the right direction - perhaps?).
BTW, I have also found that avoiding large impedances in the grid circuit of the 5687 helps in achieving low hum.
Best luck,
You should be able to get near-hum free performance by using 6.3V on the 5687, with 50-ohm (or so) resistors from each lead, or a "hum pot" (as used with DH tubes), to AC ground. The problem with 12.6V is that each half of the 5687 will have 6.3V across it, with one end grounded and the other seeing the hum (effectively SE filament). By using 6.3V, you can get a balanced 3.15V, and thus balanced hum, across each section.
AC or DC operation should work equally well; I would use AC on sonic grounds, though this may be a matter of taste. AC is also less taxing on the PS.
Also - it is usual to reference the heater circuit to a positive (+20 - +40V), rather than a negative, DC voltage. This should reduce leakage and increase life, I believe. (?) I haven't found any significant reduction in hum (might vary?) using this technique.
The heater reference voltage divider is typically decoupled to PS ground (lower leg bypassed). I don't think it's possible to reduce total hum by injecting B+ hum into the heater, though I could be wrong. At least with AC heating, the heater hum tends to be a dark "note" (little overtones), whereas B+ hum is closer to a buzz, so they may not match in frequency content. There's also a difference in that hum across the heater is differential mode, whereas injected hum would be common mode, so I can't see it working (unless the heater feed is unbalanced - in the right direction - perhaps?).
BTW, I have also found that avoiding large impedances in the grid circuit of the 5687 helps in achieving low hum.
Best luck,
Hi there,
Also - it is usual to reference the heater circuit to a positive (+20 - +40V), rather than a negative, DC voltage.
This should reduce leakage and increase life, I believe. (?) I haven't found any significant reduction in hum (might
vary?) using this technique.
My understanding of current is that it flows from positive to negative, therefore if the heater is at positive potential to the cathode, and has considerable current on it, then that current will be looking for a lower potential ie the cathode. I may be very wrong, but tell me why.
By using 6.3V, you can get a balanced 3.15V, and thus balanced hum, across each section
I tried 6.3AC on the heaters, still had hum. As I said I am now using DC heaters (but not very smooth DC, about 2V ripple out of the 12.6V - which was all I could manage in the very constricted space), and this still produces about about 15mV hum at the output. My conclusion is that the only really satifactory way to get this tube quiet is to use absolutely smooth DC heater supplies.
I intend to do this because I was just comparing it up again my two PCC88 line stage preamps, which should be a much superior sounding circuit, and the 5687 is great compared. It has incredably good bass and very clean presentation. I put this down to the fact that my PCC88 line stages are only biased at 7mA, but the 5687 is biased at 25mA. Bass obviously benefits from a bit of current grunt behind it. My Gainclones never usually break a sweat , but this amp gets the heatsinks roasting hot.
I have the bits and pieces to build an even better version than the little test one I have currently built, with chokes. I will decomission one of the PCC88 line stages and use its case to build the new 5687, which should make it nice and easy. Its a pain because I spent a good deal of time and effort getting the last PCC88 line stage up and running, and I fully expected it to become my standard. I was only building the 5687 as a bit of fun.
Shoog
Also - it is usual to reference the heater circuit to a positive (+20 - +40V), rather than a negative, DC voltage.
This should reduce leakage and increase life, I believe. (?) I haven't found any significant reduction in hum (might
vary?) using this technique.
My understanding of current is that it flows from positive to negative, therefore if the heater is at positive potential to the cathode, and has considerable current on it, then that current will be looking for a lower potential ie the cathode. I may be very wrong, but tell me why.
By using 6.3V, you can get a balanced 3.15V, and thus balanced hum, across each section
I tried 6.3AC on the heaters, still had hum. As I said I am now using DC heaters (but not very smooth DC, about 2V ripple out of the 12.6V - which was all I could manage in the very constricted space), and this still produces about about 15mV hum at the output. My conclusion is that the only really satifactory way to get this tube quiet is to use absolutely smooth DC heater supplies.
I intend to do this because I was just comparing it up again my two PCC88 line stage preamps, which should be a much superior sounding circuit, and the 5687 is great compared. It has incredably good bass and very clean presentation. I put this down to the fact that my PCC88 line stages are only biased at 7mA, but the 5687 is biased at 25mA. Bass obviously benefits from a bit of current grunt behind it. My Gainclones never usually break a sweat , but this amp gets the heatsinks roasting hot.
I have the bits and pieces to build an even better version than the little test one I have currently built, with chokes. I will decomission one of the PCC88 line stages and use its case to build the new 5687, which should make it nice and easy. Its a pain because I spent a good deal of time and effort getting the last PCC88 line stage up and running, and I fully expected it to become my standard. I was only building the 5687 as a bit of fun.
Shoog
Hi there.........I frigged about with plus/minus polarity on the heater off ground for some time, and I found that I had to go much higher with the +off ground to get the same result compared with -10V. A space charge mystery..........not all tubes behave the same. The best results I get with nothing at all is with NOS tubes. These had the heater properties reseached into thus I get none of these problems.
At -10V (wrt heater to cath) the effect on tube life is nil..the data sheet of the NOS Syvania 7199 states -200V permitted !That was probably with a spiral wound heater. One can only presume that recent makes Sovtek/Svets would be lower. Unfortunately much data isn't published anymore.
From having the heater biassed up and lower didn't make a hoot of difference to distortion nor to any DC shift. As the heater on some tubes is physically screened from g1, any ion shift would see a low Z cathode, the cathode current would swamp any effect. That's probably why NOS are better thought off.
rich
My understanding of current is that it flows from positive to negative, therefore if the heater is at positive
potential to the cathode, and has considerable current on it, then that current will be looking for a lower potential
ie the cathode. I may be very wrong, but tell me why.
Thinking about it that statement is wrong isn't it. Electrons are negatively charged and are generated by the heater, they flow up to the anode. Therefore to prevent flow to the cathode the heater would have to be positive. Still I don't think it makes a heap of difference with the 5687 (or at least the one i've got)
Shoog
potential to the cathode, and has considerable current on it, then that current will be looking for a lower potential
ie the cathode. I may be very wrong, but tell me why.
Thinking about it that statement is wrong isn't it. Electrons are negatively charged and are generated by the heater, they flow up to the anode. Therefore to prevent flow to the cathode the heater would have to be positive. Still I don't think it makes a heap of difference with the 5687 (or at least the one i've got)
Shoog
Hi again.
Shoog wrote:
"I tried 6.3AC on the heaters, still had hum. As I said I am now using DC heaters (but not very smooth DC, about 2V ripple out of the 12.6V - which was all I could manage in the very constricted space), and this still produces about about 15mV hum at the output. My conclusion is that the only really satifactory way to get this tube quiet is to use absolutely smooth DC heater supplies."
Strange. Though 2V ripple is quite a lot (esp. if rms value). Sorry if I missed this, but did you balance the heater feed with small resistors and provide a low-Z AC path to ground?
I have mainly used 12.6V AC with this tube (wanted to use what I had in the way of transformers). I get a slight hum with very (overly IMO) sensitive power amps (like 30dB+ gain) together with efficient speakers (like 95 dB), but even here, the hum is hardly audible in the listening chair.
What brand of 5687 do you have? I've had the same (small) hum from the common Philips WBs, Ratheons (same thing, I think), and GE Five Stars.
Shoog wrote:
"I intend to do this because I was just comparing it up again my two PCC88 line stage preamps, which should be a much superior sounding circuit, and the 5687 is great compared. It has incredably good bass and very clean presentation. I put this down to the fact that my PCC88 line stages are only biased at 7mA, but the 5687 is biased at 25mA. Bass obviously benefits from a bit of current grunt behind it. My Gainclones never usually break a sweat , but this amp gets the heatsinks roasting hot."
I really don't think you need to use that much current. I would recommend cutting down to 10-15 mA, unless you're driving a near-short. (IMO, 90Vak and 12mA is a sweet spot for pre:s.) I think it could be worth trying a lighter op point to see what happens to the hum. What are the load conditions when driving your Gainclones? How much swing does it need? That bass should require more current capacity to sound good seems unlikely, except under heavy load (< 5k or so); motivation for high quiescent current is usually a better ability to drive cable and power amp input capacitance, which increases with increasing frequency. Sure, there's more amplitude in the bass than the treble, but still, it's just line level (?). 25 mA would be chosen e.g. to drive a 300B's grid +/- 70V. But - YMMV.
BTW, a sure way of "de-humming" 5687s is to feed them current-regulated and balanced 0.9A DC (resulting is about 6V across heater). I do this in my RIAA ('Asphodel MkIV'). It is utterly noise-free.
Description:
http://www.eng.umu.se/personal/morgan/Valvular Pages/pages/projects.htm#Asphodel
(Text isn't updated to say that a 5687 replaced the 6N30P - not that there was anything wrong with the latter, just an experiment!)
Schematic:
http://www.eng.umu.se/personal/morgan/Valvular Pages/images/Asphodel3.gif
(Not the latest version, but heater cct the same.)
Hope my thoughts on this have been of some help.
Anyway, I wish you good luck with your pre! Like you, I think the 5687 is a fine-sounding valve, and worth the effort, despite the somewhat quirky heater demands.
Shoog wrote:
"I tried 6.3AC on the heaters, still had hum. As I said I am now using DC heaters (but not very smooth DC, about 2V ripple out of the 12.6V - which was all I could manage in the very constricted space), and this still produces about about 15mV hum at the output. My conclusion is that the only really satifactory way to get this tube quiet is to use absolutely smooth DC heater supplies."
Strange. Though 2V ripple is quite a lot (esp. if rms value). Sorry if I missed this, but did you balance the heater feed with small resistors and provide a low-Z AC path to ground?
I have mainly used 12.6V AC with this tube (wanted to use what I had in the way of transformers). I get a slight hum with very (overly IMO) sensitive power amps (like 30dB+ gain) together with efficient speakers (like 95 dB), but even here, the hum is hardly audible in the listening chair.
What brand of 5687 do you have? I've had the same (small) hum from the common Philips WBs, Ratheons (same thing, I think), and GE Five Stars.
Shoog wrote:
"I intend to do this because I was just comparing it up again my two PCC88 line stage preamps, which should be a much superior sounding circuit, and the 5687 is great compared. It has incredably good bass and very clean presentation. I put this down to the fact that my PCC88 line stages are only biased at 7mA, but the 5687 is biased at 25mA. Bass obviously benefits from a bit of current grunt behind it. My Gainclones never usually break a sweat , but this amp gets the heatsinks roasting hot."
I really don't think you need to use that much current. I would recommend cutting down to 10-15 mA, unless you're driving a near-short. (IMO, 90Vak and 12mA is a sweet spot for pre:s.) I think it could be worth trying a lighter op point to see what happens to the hum. What are the load conditions when driving your Gainclones? How much swing does it need? That bass should require more current capacity to sound good seems unlikely, except under heavy load (< 5k or so); motivation for high quiescent current is usually a better ability to drive cable and power amp input capacitance, which increases with increasing frequency. Sure, there's more amplitude in the bass than the treble, but still, it's just line level (?). 25 mA would be chosen e.g. to drive a 300B's grid +/- 70V. But - YMMV.
BTW, a sure way of "de-humming" 5687s is to feed them current-regulated and balanced 0.9A DC (resulting is about 6V across heater). I do this in my RIAA ('Asphodel MkIV'). It is utterly noise-free.
Description:
http://www.eng.umu.se/personal/morgan/Valvular Pages/pages/projects.htm#Asphodel
(Text isn't updated to say that a 5687 replaced the 6N30P - not that there was anything wrong with the latter, just an experiment!)
Schematic:
http://www.eng.umu.se/personal/morgan/Valvular Pages/images/Asphodel3.gif
(Not the latest version, but heater cct the same.)
Hope my thoughts on this have been of some help.
Anyway, I wish you good luck with your pre! Like you, I think the 5687 is a fine-sounding valve, and worth the effort, despite the somewhat quirky heater demands.
Hi there again,
Did some more work on the preamp today. Stacked on another 3,000uf of filtering on the heaters. 1000uf each side and 2000uf across the whole heater. Pin 8 is tied to earth now. Ripple is down to 1V pp now.
I also did some other things. I originally had the star ground attached to the case and then linked to the socket ground, again on the case. I have now isolated the star ground off the case. This seemed to tidy up the hum somewhat.
Since reducing the heater ripple had only slight effect i started looking for other sources. I noticed that it was only on the positive going wave form. There was the first 8uf filtering cap and a 220R resistor rather near the anode leads, which I decided to move (god only knows how I found the room), and hey presto, with just the power on and no leads attached - down to a few mV of ripple on the output. That cap must have been transmitting like a radio station.
Still all is not yet fixed. With the leads attached to source and amp and no power, there is still a significant hum. Earth loops between the amp and the pre. I am about to isolate the socket earth from the star ground and hopefully this will be enough to eliminate this hum. All been well by tonight I will have a quiet amp.
The valve I am using is a black plate TungSol. I have built this circuit as per the circuit diagram supplied- I am reluctant to mess about and alter the bias point on this one. My gainclone is fairly standard, and shouldn't be hard to drive.
When I build the improved version I will have learn't enough to get it right the first time, and then I may alter the bias point.
Maybe its just this valve that produces the prodigious bass output, what ever it is, its great. It gives up nothing in detail.
Thanks everyone, I have learn't a few more valve details along the way.
Shoog
Did some more work on the preamp today. Stacked on another 3,000uf of filtering on the heaters. 1000uf each side and 2000uf across the whole heater. Pin 8 is tied to earth now. Ripple is down to 1V pp now.
I also did some other things. I originally had the star ground attached to the case and then linked to the socket ground, again on the case. I have now isolated the star ground off the case. This seemed to tidy up the hum somewhat.
Since reducing the heater ripple had only slight effect i started looking for other sources. I noticed that it was only on the positive going wave form. There was the first 8uf filtering cap and a 220R resistor rather near the anode leads, which I decided to move (god only knows how I found the room), and hey presto, with just the power on and no leads attached - down to a few mV of ripple on the output. That cap must have been transmitting like a radio station.
Still all is not yet fixed. With the leads attached to source and amp and no power, there is still a significant hum. Earth loops between the amp and the pre. I am about to isolate the socket earth from the star ground and hopefully this will be enough to eliminate this hum. All been well by tonight I will have a quiet amp.
The valve I am using is a black plate TungSol. I have built this circuit as per the circuit diagram supplied- I am reluctant to mess about and alter the bias point on this one. My gainclone is fairly standard, and shouldn't be hard to drive.
When I build the improved version I will have learn't enough to get it right the first time, and then I may alter the bias point.
Maybe its just this valve that produces the prodigious bass output, what ever it is, its great. It gives up nothing in detail.
Thanks everyone, I have learn't a few more valve details along the way.
Shoog
Cracked it, all is now quiet at normal listening levels on my main system. At high setting of the volume pot I have a strange motorboating noise, but I will never be using it at those levels so thats OK. Maybe its the source, as that the satelite, which goes through the video and has the DVD on a scart socket box (a hum waiting to happen).
Thanks all
Shoog
Thanks all
Shoog
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.