hi all
i am using rca83 fullwave mercury rectifier in my 300b amps.
the 300b is a drd design, with siemens c3g as driver.
am facing a hum problem. seems like psu hum. it does not increase with volume.
psu filtering is done by 8uF/10H/55uF/5H/55uF. ultrapath cap is 45uF.
psu ground and signal ground are separated, but meets at the same point. chassis is grounded to safety earth. does not seem to be a grounding issue as when in breadboard, on a piece of wood, the hum is also there. before this, i used one single ground for psu and signal.(separate grounds did not make much difference)
i have also added 100nF cap across rca83's anode and cathode, tuned the 5H choke with a 0.22uF capacitor, but that still does not seem to reduce the hum much.
any suggestions?
thanks.
garbage
i am using rca83 fullwave mercury rectifier in my 300b amps.
the 300b is a drd design, with siemens c3g as driver.
am facing a hum problem. seems like psu hum. it does not increase with volume.
psu filtering is done by 8uF/10H/55uF/5H/55uF. ultrapath cap is 45uF.
psu ground and signal ground are separated, but meets at the same point. chassis is grounded to safety earth. does not seem to be a grounding issue as when in breadboard, on a piece of wood, the hum is also there. before this, i used one single ground for psu and signal.(separate grounds did not make much difference)
i have also added 100nF cap across rca83's anode and cathode, tuned the 5H choke with a 0.22uF capacitor, but that still does not seem to reduce the hum much.
any suggestions?
thanks.
garbage
Er, a mercury type is supposed to use small hash chokes in series with the anodes. Less capacitance the better.
Indeed, you'd have much better luck with silicon diodes, which only have reverse recovery switching issues (easily controlled). Yes, this from a person who, unfortunately, hasn't seen a type 83 in use...
Tim
Indeed, you'd have much better luck with silicon diodes, which only have reverse recovery switching issues (easily controlled). Yes, this from a person who, unfortunately, hasn't seen a type 83 in use...
Tim
I don't think the hum is caused by the PS circuit. The C-L-C-L-C 8uF/10H/55uF/5H/55uF filter circuit should provide you with excellent filtering.
The hum possibly from the filament circuitry of the 300B. Did you adjust the hum pot? If you want to further cut down the hum from the filament, it's better to use DC supply for it.
I had hum problem when using Russian made 6SN7. I need to raise the DC potential of the filament voltage to approx 80Vdc to kill the hum generated from the 6SN7 driver stage.
The 300B amp I built was based on the design from JE Lab.
The hum possibly from the filament circuitry of the 300B. Did you adjust the hum pot? If you want to further cut down the hum from the filament, it's better to use DC supply for it.
I had hum problem when using Russian made 6SN7. I need to raise the DC potential of the filament voltage to approx 80Vdc to kill the hum generated from the 6SN7 driver stage.
The 300B amp I built was based on the design from JE Lab.
One of the posts in this thread might be helpful.
"My understanding is that mercury vapour rectifiers have a negative resistance at part of their IV characteristic and that this can combine with the leakage inductance and stray capacitance of the mains transformer to produce a circuit that can oscillate if excited. Since the system is excited at a 50Hz (or 60Hz) rate, this can lead to squegging at a 50Hz rate. Ferrite beads are usually touted as the cure. The RSGB and ARRL manuals are good places to look for information on using these rectifiers."
(don't know how to quote properly on here but the above was posted by EC8010)
"My understanding is that mercury vapour rectifiers have a negative resistance at part of their IV characteristic and that this can combine with the leakage inductance and stray capacitance of the mains transformer to produce a circuit that can oscillate if excited. Since the system is excited at a 50Hz (or 60Hz) rate, this can lead to squegging at a 50Hz rate. Ferrite beads are usually touted as the cure. The RSGB and ARRL manuals are good places to look for information on using these rectifiers."
(don't know how to quote properly on here but the above was posted by EC8010)
hi all
thanks for the responses.
i've just changed my speakers from totem model 1 (87db) to an az2 (93db) in order to better match my amps. the hum is killing me now. i can still hear it from my listening position. previously with the totems, i am not able to hear it unless i go within a foot or so. will have to find some way to get rid of that hum...
the 300b heaters are dc. no hum pot, just connected the cathode resistors to the - heater supply. i tried hum pot, but no difference.
i've read that there is a max voltage when raising filament voltages. what is a safe level in my case? an when raising the filament voltages, do i connect to the + or - of the filament supply?
my driver stage is a triode strapped pentode, on ac filaments though. all filaments are floating, ie, not connected to circuit ground.
seems like most people think that it is the heater of the driver. ok. will look into that. probably go dc for it. i've tried different grounding arrangements, but still there is no audible difference in hum reduction. i'll put in sand as a last resort to isolate the problem, but would very much still like to use the 83.
i noticed that thread after i posted this. (sorry moderators...
)
where should the ferrite beads be placed? along the heater supply of the 83?
ps. i think there is no method of cross-post quoting. you are not alone there.
thanks for the responses.
i've just changed my speakers from totem model 1 (87db) to an az2 (93db) in order to better match my amps. the hum is killing me now. i can still hear it from my listening position. previously with the totems, i am not able to hear it unless i go within a foot or so. will have to find some way to get rid of that hum...
kmtang said:
the 300b heaters are dc. no hum pot, just connected the cathode resistors to the - heater supply. i tried hum pot, but no difference.
i've read that there is a max voltage when raising filament voltages. what is a safe level in my case? an when raising the filament voltages, do i connect to the + or - of the filament supply?
my driver stage is a triode strapped pentode, on ac filaments though. all filaments are floating, ie, not connected to circuit ground.
analog_sa said:
seems like most people think that it is the heater of the driver. ok. will look into that. probably go dc for it. i've tried different grounding arrangements, but still there is no audible difference in hum reduction. i'll put in sand as a last resort to isolate the problem, but would very much still like to use the 83.
martinab2 said:
i noticed that thread after i posted this. (sorry moderators...
)where should the ferrite beads be placed? along the heater supply of the 83?
ps. i think there is no method of cross-post quoting. you are not alone there.
garbage said:
The heater must be within 120V of the cathode - or so I think... the datasheets are in German, and I don't read German
You connect the centre tap of the 6.3V winding to a potential divider to raise it 40V or so above the voltage at the cathode. If you have no centre tap, connect two 100 ohm resistors in series across the winding, and connect the centre of the two resistors to the potential divider.
garbage said:
Floating heater potentials are generally bad. You would prefer to determine it rather than let the other what-nots in the circuit and the valve determine it for you. It may also be the source of the hum.
garbage said:
As Martin (and Tim) has already said, in series with each of the anodes.
A question please, the title of this thread is "rca83 hash" - does this mean that you are getting a low frequency - 50Hz or 100Hz or so with few higher harmonics, or buzz/hash with low frequency stuff but also more audible (and more annoying
) higher frequency components?
audiousername said:
The heater must be within 120V of the cathode - or so I think... the datasheets are in German, and I don't read German
You connect the centre tap of the 6.3V winding to a potential divider to raise it 40V or so above the voltage at the cathode. If you have no centre tap, connect two 100 ohm resistors in series across the winding, and connect the centre of the two resistors to the potential divider.
Floating heater potentials are generally bad. You would prefer to determine it rather than let the other what-nots in the circuit and the valve determine it for you. It may also be the source of the hum.
...
A question please, the title of this thread is "rca83 hash" - does this mean that you are getting a low frequency - 50Hz or 100Hz or so with few higher harmonics, or buzz/hash with low frequency stuff but also more audible (and more annoying
hi audiousername
i've already made a dc rectifier for the c3g heaters.
rectifiers followed by 4700uF/4700uF/4.7uF/0.1uF, followed by lm317 with 100R and 1000R trimmer.
An externally hosted image should be here but it was not working when we last tested it.
will put this in some other night...
i will ground my c3g heaters to circuit ground to see if the hum goes away first. the 300b filaments are floating, that's ok right? cos i did try to ground it, but i think it causes some shorts... 300b is DHT.
forgot to get the ferrite beads when i was at the store...
oh, the hash is low frequency. only audible in the woofers, and does not increase with volume. doesn't help when i've got 8" woofers.
garbage said:i will ground my c3g heaters to circuit ground to see if the hum goes away first. the 300b filaments are floating, that's ok right? cos i did try to ground it, but i think it causes some shorts... 300b is DHT.
Quite right. Don't ground the 300B's filament, that would be bad. By doing so you remove bias
my power trans for c3g winding is 3.15-0-3.15.
the CT is grounded.
currently i am using ac with 1R2 on each filament pins to drop the voltage to slightly below 6.15vac.
i am not sure if i am able to ground one of the pins, so i did not try grounding my supply, but instead i put in the rectifier and regulator.
with the regulator putting out 6.10vdc, i fired up the amps, but the hum/hash on the woofer is still there. next, i tried grounding the ground filament to circuit ground with a wire in hand, but the hum/hash did not reduce. this wire however, became hot, so i removed it.
the anode and cathode of the 83 have a ferrite clamp on it, but it does not seem to make any difference to the hum/hash levels.
yet to try is to raise the heater voltage. as of now, i have reverted to ac on the filaments. to raise the heater voltage, do i put in dc or ac into the CT of the filament trans winding?
short of using dc on the 83 filaments or changing rectifiers, i've almost tried everything...
is there an easy drop-in rectifier for 83? a ux-4, that can take 450-0-480 of HV from the power trans?
rgds
garbage
the CT is grounded.
currently i am using ac with 1R2 on each filament pins to drop the voltage to slightly below 6.15vac.
i am not sure if i am able to ground one of the pins, so i did not try grounding my supply, but instead i put in the rectifier and regulator.
with the regulator putting out 6.10vdc, i fired up the amps, but the hum/hash on the woofer is still there. next, i tried grounding the ground filament to circuit ground with a wire in hand, but the hum/hash did not reduce. this wire however, became hot, so i removed it.
the anode and cathode of the 83 have a ferrite clamp on it, but it does not seem to make any difference to the hum/hash levels.
yet to try is to raise the heater voltage. as of now, i have reverted to ac on the filaments. to raise the heater voltage, do i put in dc or ac into the CT of the filament trans winding?
short of using dc on the 83 filaments or changing rectifiers, i've almost tried everything...
is there an easy drop-in rectifier for 83? a ux-4, that can take 450-0-480 of HV from the power trans?
rgds
garbage
garbage said:
If your filament winding is providing 3.15-0-3.15 under load, there is no reason to use dropping resistors. It is disappointing that grounding the centre tap did not solve your problem... it was the most obvious thing wrong there.
garbage said:
Ground what pin?
garbage said:
What do you mean by grounding the ground filament?
garbage said:
The ferrites should be in series with each of the anodes of the 83. By the way, with directly-heated rectifiers, the input to the power supply filter should come from the centre tap of the filament winding, unlike indirectly heated rectifiers like 5AR4/GZ34 which should take it from pin 8.
garbage said:
It should be 6.3V AC RMS or 6.3V DC. Significantly raising or lowering the heater voltage (say more than 5%) will reduce the life of the valve.
You should never allow DC into a transformer which was not designed for it.
garbage said:
Using DC on a rectifier filament is a waste of time because you have lots of ripple voltage there as well, which is far more significant than the little variations caused by AC heating.
garbage said:
When it comes to hard vacuum rectifiers, a 5Z3 (essentially UX-4 based 5U4) or an 83-V might be suitable for replacement, but neither will have as good regulation and as low a voltage drop as the 83 because of its mercury-vapour construction.
Note that there are many causes of hum, quite possibly it has nothing to do with the valves themselves. It may be an earth loop, magnetic coupling between the power transformer and output transformer, or a whole variety of other things. I recall your posting of a picture of the internals of this amp a little while ago, and it appeared quite crowded inside.
Konnichiwa,
And THERE is likely your problem. Why?
Your DC supply is not regulated. It will likely have a few 100mV Hum on it. Consider adding some filter chokes and additional capacitors to it and make a virtual centertap from two 10R Resistors.
Also, connect the centertap of the C3g heater to around +20V, this reliably reverse-biases the parasitic cathode/heater diodes.
You might want to have a look at a few of my recent (300B) amp's and pick up on a few of the tricks I use.
Sayonara
garbage said:
And THERE is likely your problem. Why?
Your DC supply is not regulated. It will likely have a few 100mV Hum on it. Consider adding some filter chokes and additional capacitors to it and make a virtual centertap from two 10R Resistors.
Also, connect the centertap of the C3g heater to around +20V, this reliably reverse-biases the parasitic cathode/heater diodes.
You might want to have a look at a few of my recent (300B) amp's and pick up on a few of the tricks I use.
Sayonara
audiousername said:
ok, think i get it. grounding of the filaments is only thru the CT.
if i want to float the filament supply, how do i do this?
audiousername said:
my rectifier filaments winding is just 0-5vac. no CT. the input to psu supply filter is taken from pin 4.
Kuei Yang Wang said:
hi KYW
hmm... initially i had a regulated 300b filament supply via a lm338. it was mounted to a large heatsink. unfortunately i think it blew. both channels within seconds of each other, sending my 83 to glow super blue. i switched it off immediately. ever since then i just have rectifiers, caps and dropping resistor to get correct voltage for the 300b.
during my prototype stage, the filaments were regulated, but i still experience the same hum. at that time, i thought it was the proximity of the drd choke to the psu chokes(drd choke was in between both the psu chokes).
i'll check out your circuits to see how you wire your 300b filaments and the floating method.
Konnichiwa,
Resistor divider from +B or from the 300B Cathode, capacitor across the lower part. And note, I strongly recommend that you do bias/float up the heater of the driver. I remember a pair of very hummy paraglos of a friend, Bias the heaters of the driver, quiet. The hum was 50Hz BTW, very funny wavform and quite audible.
That should be fine.
Is the hum fundamental 50/60Hz or 100/120Hz?
Sayonara
garbage said:
Resistor divider from +B or from the 300B Cathode, capacitor across the lower part. And note, I strongly recommend that you do bias/float up the heater of the driver. I remember a pair of very hummy paraglos of a friend, Bias the heaters of the driver, quiet. The hum was 50Hz BTW, very funny wavform and quite audible.
garbage said:
That should be fine.
garbage said:
Is the hum fundamental 50/60Hz or 100/120Hz?
Sayonara