I've got a few RH-84s and I was wondering why one is very prone to hum. It is a converted Magnavox 8601 console amplifier.
The only thing that differs from the original schematic is that I had no room for a choke. This should not be that much of a bad thing right? I used a 100 ohm resistor instead.
Using my transformer values of 225v and 160 ohms DCR in PSUDII here is what I get:
6CA4
C1 = 40uf
R1 = 100R
C2 = 140uf
CCS= 100ma
Max: 263.20v
Min: 262.49v
Diff: 711.20mv
As you can see, the ripple of the main B+ is only 700mv. I confirmed this by measuring with a blocking cap on my cheap Radio Shack DMM.
If my mains ripple is so low, why does my amplifier have a low level 120hz hum when turned on?
Does the ripple really have to be even lower due to the P2PFB?
Heck, If I run the simulation of the original power supply I get about 8v of ripple, but the original circuit was dead quiet ( as are most magnavox circuits with high input ripple.) IMD is probably an issue with them, but that is another matter.
The only thing that differs from the original schematic is that I had no room for a choke. This should not be that much of a bad thing right? I used a 100 ohm resistor instead.
Using my transformer values of 225v and 160 ohms DCR in PSUDII here is what I get:
6CA4
C1 = 40uf
R1 = 100R
C2 = 140uf
CCS= 100ma
Max: 263.20v
Min: 262.49v
Diff: 711.20mv
As you can see, the ripple of the main B+ is only 700mv. I confirmed this by measuring with a blocking cap on my cheap Radio Shack DMM.
If my mains ripple is so low, why does my amplifier have a low level 120hz hum when turned on?
Does the ripple really have to be even lower due to the P2PFB?
Heck, If I run the simulation of the original power supply I get about 8v of ripple, but the original circuit was dead quiet ( as are most magnavox circuits with high input ripple.) IMD is probably an issue with them, but that is another matter.
If you take the turns ratio of the opt you can get a good idea of what your noise level will be. FWIW, I have built a pair of rh84 amps, and they are dead silent, however I used lclc filters with simulated ripple of less the 20mv, at the speakers I measure less than .2mv. If you have room to use a WE/Ultra path style cap from B+ to output tube cathode you can reduce the hum by quite a bit. The maggies, etc, usually used a global feedback network that would reduce hum and noise considerably, whereas I am not sure if the ptp feedback gives as much ripple rejection.
Thanks for that Jim. I know the original Magnavox circuit had a 150k resistor going from the B+ for the driver section to the catodes of the 6EU7's. Its effect was to inject PSR into the cathode and have it cancelled out.
The Ultrapath circuit I found has a capacitor from the B+ for the output tube running to the cathode of the output tubes. If I did this, I would think that a shared cathode resistor would not be a problem correct? It would only be referenced the difference in power supply ripple to ground, leaving little room for crosstalk.
The Ultrapath circuit I found has a capacitor from the B+ for the output tube running to the cathode of the output tubes. If I did this, I would think that a shared cathode resistor would not be a problem correct? It would only be referenced the difference in power supply ripple to ground, leaving little room for crosstalk.
I have one of my rh84 amps set up this way - I use a shared cathode r with a bypass cap, and then a WE style cap from cathodes to B+. It may not matter, but I always connect B+ for the OT to the plus side of the WE cap. WE cap should be a high quality unit such as a motor run or a solen unit. Crosstalk is not an issue using a shared cathode r that is adequately bypassed.
By we, I assume you mean western electric which pioneered the use of the Ultrapath cap? Anyway, I did some reading on ultrapath and am somewhat confused. From what I gathered you can actually make PSRR worse because you can cause the amplifier to amplify the ripple. What size cap would you reccomend? I like solens for coupling caps btw, nice caps at a budget price 
If you don't mind losing few volts by all means go for it, but don't forget to put a real (as big as you can afford) capacitor behind it. Capacitance "multiplier" doesn't really multiply the capacitance (as in: provide more energy capacity from sand and thin air), it merely chops away the tops from the ripple waveform and provides a softer start for components downstream from it. It cannot provide the energy to fill in the dips in the ripple, that's the job for the real capacitor.
BTW, there is a article on the subject of "Ultrapath" on J. Broskie's website, I recommend you to read it before making any modifications to RH84 circuit. Driving more noise into the noninverting input might not decrease the hum, but increase it ...
BTW, there is a article on the subject of "Ultrapath" on J. Broskie's website, I recommend you to read it before making any modifications to RH84 circuit. Driving more noise into the noninverting input might not decrease the hum, but increase it ...
Well it looks like I don't really have any voltage to lose, I already have a low B+ of 275v or so, equating to 250v on the plates of the 6BQ5's. What about some global NFB? Think that would clean things up? What about hum injection into the preamp, taking a portion of the B+ supply and injecting it into the cathode like the original magnavox circuit? Two ways of doing that by the way , injecting into the cathode from the preamp plate supply, or injecting into the preamp cathode from the 6bq5 B+...
.7V PS ripple is way too high, aim for sub .1V because it's SE and not PP topology.
Simply put in a low Rdc choke suitable for at least .1A Hammond has affordable models.
Disco:
I don't have any problems with adding a choke, I simply do not have room for one.
Here is a picture of your typical magnavox 8601 amplifier, not much room eh?
Now I got to thinking the other night: A ton of vintage SE 6v6 or 6bq5 amplifiers had no choke. Take a Fender champ for example. 5y3, 8uf cap, 10k resistor for the screens and another 8uf cap.
Of course ripple is insanely high for hi-fi, and introduces a lot of intermodulation distortion, but that is beside the point that even the 5F1 with its feedback network removed has extremely low hum, even through a rather high efficiency instrument speaker. It certainly has lower hum than the magnavox has right now. My speakers are about 90db acoustic suspension, so if the amp is humming audibly something really is wrong.
However: another thought occurred to me.The feedback network on the RH84 is plate to plate, or rather output tube_100k_capacitor and input to the tube again. Does this mean that any low voltage ripple on the MAIN B+ that the tube takes its supply from will be fed back into the power tube and re-amplified? If so, this makes sense to me to keep the main power supply ripple as low as possible.
I am currently listening to this :
Its an RCA RS-199B console amplifier converted to RH-84. It uses a SS voltage doubler into a 200uf cap, 100R, 80uf and the OT leads. This results in no audible hum what soever.
Unfortunatly with the magnavox, I can't use as much capacitance as I want because it has a tube rectifier. I don't know the rule for adding capacitance after a resistor or choke though, so can I add as much as I want after the 100 ohm dropping resistor? I can find rubycon 120uf /400v caps that are very small physically- I could add four for 480uf after the 100 ohm resistor if it will not kill the rectifier.
I don't have any problems with adding a choke, I simply do not have room for one.
Here is a picture of your typical magnavox 8601 amplifier, not much room eh?
Now I got to thinking the other night: A ton of vintage SE 6v6 or 6bq5 amplifiers had no choke. Take a Fender champ for example. 5y3, 8uf cap, 10k resistor for the screens and another 8uf cap.
Of course ripple is insanely high for hi-fi, and introduces a lot of intermodulation distortion, but that is beside the point that even the 5F1 with its feedback network removed has extremely low hum, even through a rather high efficiency instrument speaker. It certainly has lower hum than the magnavox has right now. My speakers are about 90db acoustic suspension, so if the amp is humming audibly something really is wrong.
However: another thought occurred to me.The feedback network on the RH84 is plate to plate, or rather output tube_100k_capacitor and input to the tube again. Does this mean that any low voltage ripple on the MAIN B+ that the tube takes its supply from will be fed back into the power tube and re-amplified? If so, this makes sense to me to keep the main power supply ripple as low as possible.
I am currently listening to this :
Its an RCA RS-199B console amplifier converted to RH-84. It uses a SS voltage doubler into a 200uf cap, 100R, 80uf and the OT leads. This results in no audible hum what soever.
Unfortunatly with the magnavox, I can't use as much capacitance as I want because it has a tube rectifier. I don't know the rule for adding capacitance after a resistor or choke though, so can I add as much as I want after the 100 ohm dropping resistor? I can find rubycon 120uf /400v caps that are very small physically- I could add four for 480uf after the 100 ohm resistor if it will not kill the rectifier.
Use an ultrapath capacitor, you may need to iterate to get to the right value.. The value isn't all that critical in any event, to do a better job you actually need to know the output stage's effective voltage gain relative to the cathode at which point the ultrapath cap just equals the value of the cathode bypass divided by the gain at the cathode. (mu+1 for a triode, complicated here by the pentode RH84 topology, you need to know the closed loop gain of the output stage with feedback applied, but spice should give you a number close enough.)
Basically just start with something of around 1/10 - 1/20 of the cathode bypass capacitance value and see what happens. Add or remove capacitance and note the effect.
Hope this is not hopelessly garbled, I'm writing it pretty late..
Basically just start with something of around 1/10 - 1/20 of the cathode bypass capacitance value and see what happens. Add or remove capacitance and note the effect.
Hope this is not hopelessly garbled, I'm writing it pretty late..
Last edited:
You've got a luvly wooden frame there, very nice.
5Y3 may be presented with a limited amount of capacitance, depending on DC current, supply voltage and secundary resistance. Exact values can be found in the 5Y3 datasheet.
In order to make most out of the small space use a solid state power supply. If you have the extra voltage you can implement a high voltage regulator in the powersupply. Maybe the Swenson or the Salas Simplistic HV shunt.
5Y3 may be presented with a limited amount of capacitance, depending on DC current, supply voltage and secundary resistance. Exact values can be found in the 5Y3 datasheet.
In order to make most out of the small space use a solid state power supply. If you have the extra voltage you can implement a high voltage regulator in the powersupply. Maybe the Swenson or the Salas Simplistic HV shunt.
Last edited:
ultrapath/we cap
Kevin's suggestion on cap value is a good place to start. I have generally used caps in the range of 1/4 to 1/3 of the cathode bypass cap with good result. This arrangement usually results in lower hum/noise. With 700 mv hum in psu, you are going to get hum in a se design. I am guessing the maggie power supply was probably powering a tuner as well, may be the hum is higher without the extra load. The simulators will give you a pretty good idea of what you can expect, so might be a good idea to enter your present values, and see what changes might give better results. Just throwing more capacitance into the circuit won't always result in lower hum. If the caps are the originals, they may need to be upgraded as they may only present a fraction of their original value.
jim
Kevin's suggestion on cap value is a good place to start. I have generally used caps in the range of 1/4 to 1/3 of the cathode bypass cap with good result. This arrangement usually results in lower hum/noise. With 700 mv hum in psu, you are going to get hum in a se design. I am guessing the maggie power supply was probably powering a tuner as well, may be the hum is higher without the extra load. The simulators will give you a pretty good idea of what you can expect, so might be a good idea to enter your present values, and see what changes might give better results. Just throwing more capacitance into the circuit won't always result in lower hum. If the caps are the originals, they may need to be upgraded as they may only present a fraction of their original value.
jim
I'm late to the party, but my RH84 hummed somethin' fierce until I cobbled up a center tap on the 6.3V heater winding using matched resistors and grounded it. All three tubes (12AT7 + 2 x EL84s) are fed from the same winding.
Now, it's *totally* silent with my passive TVC preamp and 96db Fostex speakers, and I haven't switched back to my 45 SET amp since.
That said, I'm sure the two 10H chokes in the PS don't hurt...
Now, it's *totally* silent with my passive TVC preamp and 96db Fostex speakers, and I haven't switched back to my 45 SET amp since.
That said, I'm sure the two 10H chokes in the PS don't hurt...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.