Heater wiring approaches

[jrenkin]

I use Visio. I created my own stencil for schematics and layout. I add new items to the stencil as I need them so not all components are represented, but it works from me. If people are interested, I will find a place to post it for download.

The layout items are all drawn by hand to scale. For example, I traced an engineering drawing of a Kiwame resistor for the resistors, then scaled the height and width based upon the manufacturer's specifications. I did the same with the socket.

This lets me rotate and place the virtual components first to find a good layout. Of course, I have yet to try to convert a virtual layout to a real one.

Ouch! considering my competence with computers, I may as well rebuild an amp three times. Saves effort...

 

[bluegti]

Should the resistors forming the virtual CT be 100R or 100K R?

I wrote 100K because of what Tomchr wrote above, but Tom Bavis said 100R. I'm not sure.

 

[chrish]

100 Ohm

 

[marvinbrando]

Do I simply attach 100K ohm resistors from pins 2 and 7 to pin 8 on the last tube socket?

I'm building a 5e3 Deluxe clone and am currently wrestling with a similar question. My interpretation of the kit company's schematic is using 100 ohm resistors from 2 to 7 and 7 to 8 on the 6V6 sockets (both on one socket, or one on each 6v6 socket). As far as I can tell, the pin 8 of one of the 6V6 sockets is tied to ground so tying to pin 8 piggybacks on that ground connectoin. This seems to be an alternative to flying resistors to ground as I've read elsewhere online. I'm very new at this and don't want to steer you in the wrong direction.

 

[garymlyons]

Making a virtual ground

I have been seeing allot about using 100 ohm resistors from each heater winding lead to ground when you don't have a center tap. My question is this: On the Dynaco amps with a center tap heater winding they place a .02uF cap from the center tap to ideally the star ground point. A .02uF capacitor has an Xc at 60Hz of 132K ohms. Why not just use a couple of 1% metal film 1K ohm resistors from each lead to ground? I read the hum.pdf article and they recommend 50 to 100 ohm pot with center tap to ground. I see this in fender twins as hum balance and it is like a 5 watt pot. Can any one give me the reason if all you are doing is referencing the heater windings to ground(close to cathode potential) to keep big potentials from floating heater windings to ground(and cathode) and very little current should flow through the resistors why the low ohm resistors that consume power?
Regards
Gary

 

[DF96]

There may be capacitive coupling from HT secondaries to heater secondaries. This means that rectifier switching noise can get onto the heater wiring. Low value resistors can reduce this. A cap has the same effect. High value resistors merely establish a DC potential but do not reduce switching spikes.

 

[Bas Horneman]

bluegti, I'd be very interested in your visio stencil.

Thanks in advance.

Kind regards,
Bas

 

[garymlyons]

Making a virtual ground

DF96, So the the capacitor with the inductance of the heater winding form a LC low pass filter to attenuate the rectifier switching noise? The suggestion of using .1uF caps from each heater winding side to ground is a good filter but capacitors are hard to get in high precision and so the AC virtual ground caused by the capacitors can wind up offset by the tolerance of capacitors. Don't know if a 5 to 10 percent offset matters much.
The low value resistors form a low pass filter sort of a LR low pass filter and they are cheap in 1% tolerance to provide a well balanced virtual ground at low frequencies. Has any one looked at the power spectrum of the rectifier switching noise to see the frequency content? It must be harmonics of 60Hz. I will have to measure the inductance of the heater windings to get an idea of the filter characteristics.
Thanks for your input on the rectifier switching noise.
Regards
Gary

 

[DF96]

No, the resistor forms the lower part of a potential divider. The upper part is the stray capacitance between secondaries in the transformer. The lower the resistor the better, but it wastes power. A capacitor will actually work better. No need for high precision, as it is not a narrow band filter.

I'm sure many people have looked at the rectifier spike spectrum, but every one will be different as it depends on charge storage in the rectifier and exact details of circuit capacitance, inductance etc. It will be harmonics of 50/60Hz, but this fact is not very useful except as a diagnostic tool (the regularity shows up easily on a spectrum, so you can be sure it is not coming from some other source). Harmonics can easily go up to low radio frequencies.

A good way to spot this to put a scope on your heater line, and look for narrow spikes - usually near the voltage peaks.

 

[garymlyons]

DF96,
So you think that .1uf caps from each heater lead to the star ground would work as well as 100 ohm resistors and reduce power consumption?
Regards
Gary

 

[garymlyons]

DF96,
I measured the capacitance from the HT winding to the 6.3V winding and with my LCR meter I get about 300pF at 60Hz and 1KHz. With .1uf from heater winding to ground it looks like a voltage attenuation of about 116dB assuming a voltage divider with top leg the 300pF and the bottom leg .1uf. Seems pretty darn good for a first order approximation.
It is starting to make sense!
Regards
Gary

 

[devonhaas]

.Hey guys/gals...On a RELATED note...I have just finished a tedious turret board- built Redd 47 , that 'photographs extremely well', but using this 6k49VG-460 volt transformer with a 6.3 volt secondary(no CT), fails to supply 6.3-7. volts to 'Two' tubes', namely, the E88cc and the EF86. My wiring of AC heaters using pentodes and dual triodes from an AC source of 2.5 amps/6.3volts has- not been able to reconcile how to terminate- this' connection to yield 6.3 volts on BOTH tubes!( I am now getting 3.5 volts..).I build Microphone systems with OB2's that sound better than Neumann.. but when it comes to providing "AC" 6.3 voltage to -two 'differing tubes' such as is found on the REDD 4 7, I am not experienced.. I have only done DC discreet choke fed shunt regulation for filament .1. First attempt - simple parallel- ( load to pin 4 and 5- both tubes..)no ground ref.. gave me a steadily changing '5-7volt' on the E88cc and '4- 5.1 volts' on the EF86(receiving current last..). 2. Second shot, I changed to the Guitar Amp approach where you short 4 to 5 and then use "9" on the E88cc as the other terminal.. with the same on the EF86- except using pin "7".3. Third attempt, I -now have 3.5 volts --across load , with E88cc pins 4 and 5 tied together, other lead to "9" , with the EF86 simply going to pins 4 and 5- however with a ground reference.There is no center tap, so I built out two 5 watt Ohmite resistors to signal ground and took my 7 volts from their. Also, can I not 'clamp' this with some simple voltage clamp at 6 volts? Across the load?..The transformer is "6K49VG" Hammond(La2A) and is capable- more so than I (ha..) with a 460 volt secondary @50ma/6.3v@2.5A WITH CT for B+. I have great B+ voltage at 294volts and I'm using an AMI Tab Funkenwerk BV8r at .6.5:1 for input and output- since they are around.. Input secondary load resistor is 63k. Output is 10k over secondary. Any help with this Multiple Tube AC Heater dilemma would be appreciated!

 

[DF96]

Using normal punctuation conventions would help us read what you have written; a necessary precursor to helping you.