Bryan,
Well done 🙂
Yes, you are right. Leave the series resistor in circuit and pull the shunt. The shunt will be suitable for use as a cathode resistor.
To protect the meter and resistor you might consider adding 3 silicon diodes (1N4001 or similar) in series across the shunt.
Well done 🙂
Yes, you are right. Leave the series resistor in circuit and pull the shunt. The shunt will be suitable for use as a cathode resistor.
To protect the meter and resistor you might consider adding 3 silicon diodes (1N4001 or similar) in series across the shunt.
Bryan said:
Bryan,
Why not just add an RC filter section after the second choke? You can never have too little ripple in the B+! 😉
dropping B+
You can have to high an impedance.
drop R's not a great idea on power tubes. You rely on last cap only for low impedance. Better bigger choke and bigger last cap.
And / or CLCLC. but CLC should be fine if adequately sized.
Make sure the hum is really the direct B+ usage. added RC fine for lower current draw stages. Keep R moderate dpendent on current draw.
cheers
You can have to high an impedance.
drop R's not a great idea on power tubes. You rely on last cap only for low impedance. Better bigger choke and bigger last cap.
And / or CLCLC. but CLC should be fine if adequately sized.
Make sure the hum is really the direct B+ usage. added RC fine for lower current draw stages. Keep R moderate dpendent on current draw.
cheers
Try reducing the value of the input cap. The smaller it is, the more choke input your supply becomes
Try something like a .1uf to see the effect…John
Try something like a .1uf to see the effect…John
I've thought about dropping the capacitance as a means to lower the B+, and in fact one of the reasons I went with the higher VA PS tranny was to be able to go between both topologies (CCL or LC) depending on what the final opperating points needed to be.
So, with PSUD modeling it appears that I need anywhere between 0.47 and 1.5uF depending on where I want the B+ to end up.
What type of caps should I be looking for here. The max voltage spike is about 500V. I don't want to break the bank for this position, and all other caps in the PC sircuit are ASC POI.
Thoughts?
BK
So, with PSUD modeling it appears that I need anywhere between 0.47 and 1.5uF depending on where I want the B+ to end up.
What type of caps should I be looking for here. The max voltage spike is about 500V. I don't want to break the bank for this position, and all other caps in the PC sircuit are ASC POI.
Thoughts?
BK
Be careful with the "partial choke input" approach. It is not necessarily the free lunch it seems.
I have used it before on a pre-amp and it can cause instability.
You have a single cap charging time determining your B+. Is this totally stable every cycle? It does not have the natural regulation of choke input.
I have used it before on a pre-amp and it can cause instability.
You have a single cap charging time determining your B+. Is this totally stable every cycle? It does not have the natural regulation of choke input.
Any good film/foil cap should be fine in that position. I have used ‘orange drops’, Solen and Wilma.
If you need >.47 a 630v Solen is a good choice…John
If you need >.47 a 630v Solen is a good choice…John
Hey John,
Still alive and kickin' 😉
I actually havent worked on the project over the last week as I hav been terribly busy with the PhD.
As things are now the negative rails are working, stable, and have been tested for a few long periods of use w/o the B+...
The meter is measuring current, and can be switched from the 300B to the 845 with the flip of the switch.
I'm about to wire up the actual circuitry between tubes (coupling) and cathodes etc.
I'm still trying to think about op points for the 6j5 and the 300B. Any suggestions. Also, If I want to run the 300B at a lower voltage (300V'ish) and the 6j5 near 250V, I run into a problem. With a 5K loading resistor on the 6j5 and 300V at the 300B, I have little room for the resistor value in the pi filter from the 300B to the 6j5 in the filter network. If this is the case, should I split the supply to eliminate this problem?
Finally, for the negative bias of the 845. I connect the bias circuit to the IT transformer. There should be a decouplinig cap here, right? what is the target value for this capacitor?
Hmm, thats about it. May also use the Ggary Pimm CCS to load the input tube, in which case the question above will not be an issue...
Thoughts?
Thanks,
BK
Still alive and kickin' 😉
I actually havent worked on the project over the last week as I hav been terribly busy with the PhD.
As things are now the negative rails are working, stable, and have been tested for a few long periods of use w/o the B+...
The meter is measuring current, and can be switched from the 300B to the 845 with the flip of the switch.
I'm about to wire up the actual circuitry between tubes (coupling) and cathodes etc.
I'm still trying to think about op points for the 6j5 and the 300B. Any suggestions. Also, If I want to run the 300B at a lower voltage (300V'ish) and the 6j5 near 250V, I run into a problem. With a 5K loading resistor on the 6j5 and 300V at the 300B, I have little room for the resistor value in the pi filter from the 300B to the 6j5 in the filter network. If this is the case, should I split the supply to eliminate this problem?
Finally, for the negative bias of the 845. I connect the bias circuit to the IT transformer. There should be a decouplinig cap here, right? what is the target value for this capacitor?
Hmm, thats about it. May also use the Ggary Pimm CCS to load the input tube, in which case the question above will not be an issue...
Thoughts?
Thanks,
BK
5K plate load resistor is way too low for a 6J5, something on the order of 20K is reasonable. You will need a higher voltage supply, although 20K @ 5ma would put you around 200V at the 6J5 plate.
Kevin
Kevin
Right, but that's without another RC filter after the 300B.
I guess the problem is how to set up the PS filter when the input and the driver have a similar plate voltage. Splitting the supply is the only thing that comes to mind. I want a good amount of decoupling between the two stages...
I need to have a better look at the curves for the two valves, as I can probably find a few good op points where I can drive the 300B with a little higher plate voltage giving me plenty of room to drop voltage with an RC filter and then give adequate resistance after the RC for loading the plate....
Time time time....
I've just started training for an Olympic Triathlon with "Team In Training", and between working out, trying to raise funds for the Leukemia and Lymphoma society with Team In Training, along with finishing my degree (not to mention keeping the girlfriend and family happy and myslef sane) there are not enough hours in the day. The irony is that the solder iron is one of the things that helps me relax and unwind. I'll be back on it this week!
Oh, and I'll have to post some photos of the beast. They look great, if I do say so myself
😀
Keep Smiling
I guess the problem is how to set up the PS filter when the input and the driver have a similar plate voltage. Splitting the supply is the only thing that comes to mind. I want a good amount of decoupling between the two stages...
I need to have a better look at the curves for the two valves, as I can probably find a few good op points where I can drive the 300B with a little higher plate voltage giving me plenty of room to drop voltage with an RC filter and then give adequate resistance after the RC for loading the plate....
Time time time....
I've just started training for an Olympic Triathlon with "Team In Training", and between working out, trying to raise funds for the Leukemia and Lymphoma society with Team In Training, along with finishing my degree (not to mention keeping the girlfriend and family happy and myslef sane) there are not enough hours in the day. The irony is that the solder iron is one of the things that helps me relax and unwind. I'll be back on it this week!
Oh, and I'll have to post some photos of the beast. They look great, if I do say so myself
😀
Keep Smiling
Brian,
The 300B can take quite high voltages and so is no problem. The important thing is to keep its supply low impedance i.e. NO resistors and well filtered. It is common on large amps for all power tubes to have their own supply. Dropping only works well for pre-amps and other lower current voltage amp applications.
There are a huge range of tubes which will then work at below this voltage as a driver. You do NOT need a lot of decoupling. Poor decoupling will just give you a small amount of potentially advantageous positivc feedback!! BUT the decoupling cap is sonically critical and needs to be of high quality and not too large. Often around 100uF is good.
For a plate loaded triode keep the load to at least 3 x rp e.g. 20K for a 6SN7. This would provide minimum drive.
A lot of the sonics comes from the driver and power tubes need good hefty high current well coupled drivers. 300B's are often driven by other power tube types or cathode followers etc and transformer coupling is popular to provide low impedance.
Most things will work, it depends how far you want to go. Take a look at famous 300B designs. For a single stage something like a choke loaded 6C45pi will do it well, otherwise you may ultimately want to consider separate voltage amp and driver stages.
Another solution is to get voltage gain from an input transformer and use the driver more for just current gain.
cheers
The 300B can take quite high voltages and so is no problem. The important thing is to keep its supply low impedance i.e. NO resistors and well filtered. It is common on large amps for all power tubes to have their own supply. Dropping only works well for pre-amps and other lower current voltage amp applications.
There are a huge range of tubes which will then work at below this voltage as a driver. You do NOT need a lot of decoupling. Poor decoupling will just give you a small amount of potentially advantageous positivc feedback!! BUT the decoupling cap is sonically critical and needs to be of high quality and not too large. Often around 100uF is good.
For a plate loaded triode keep the load to at least 3 x rp e.g. 20K for a 6SN7. This would provide minimum drive.
A lot of the sonics comes from the driver and power tubes need good hefty high current well coupled drivers. 300B's are often driven by other power tube types or cathode followers etc and transformer coupling is popular to provide low impedance.
Most things will work, it depends how far you want to go. Take a look at famous 300B designs. For a single stage something like a choke loaded 6C45pi will do it well, otherwise you may ultimately want to consider separate voltage amp and driver stages.
Another solution is to get voltage gain from an input transformer and use the driver more for just current gain.
cheers
Part of the learning experience for me is seeing and hearing why things work and don't work. I've been told that the simple 6J5 may have trouble driving the 300B. I am fairly sure that the 300B will drive the 845... Never the less, building the simple initial version (on a relatively large breadboard) and seeing why it doesn't work (as optimally as it could) will be what I learn from. A lot of this is new to me, so the learning experience is half the trip!
Three independent power supplys would be a lot of weight. As it stands now I have a CLCLC supply for the 1000VDC supply for the 845, and a CLCLCRC supply for the niput and driver (6j5 and 300B). Breaking up the input and driver supplies may be more than I want to do right now... Time will tell if I need to.
Also, keep in mind that my current preamp has a gain of about 5 (simple but very nice 12B4A design). And, at the end of the day, I would still like to be able to drive the 845 into A2 for transients and headroom.
BK
Three independent power supplys would be a lot of weight. As it stands now I have a CLCLC supply for the 1000VDC supply for the 845, and a CLCLCRC supply for the niput and driver (6j5 and 300B). Breaking up the input and driver supplies may be more than I want to do right now... Time will tell if I need to.
Also, keep in mind that my current preamp has a gain of about 5 (simple but very nice 12B4A design). And, at the end of the day, I would still like to be able to drive the 845 into A2 for transients and headroom.
BK
hifryer,
you should have seen the look on my girlfirends face when I asked her to help me move the breadboard last week. She looked at me like I was nuts!
Off topic question. What is a good wire type for connecting the ouptut tranny to the biding posts?
Thanks,
BK
you should have seen the look on my girlfirends face when I asked her to help me move the breadboard last week. She looked at me like I was nuts!
Off topic question. What is a good wire type for connecting the ouptut tranny to the biding posts?
Thanks,
BK
Well, the output secondary is heavy gauge, solid core, enamelled copper wire. Why not just use similar? My view is that a few extra inches of the same can make no difference.
I used to use this for speaker wire as well but found you can do better. Speakers wires have me totally confused. I just use what I have found to sound good and no longer try to understand why!!
I used to use this for speaker wire as well but found you can do better. Speakers wires have me totally confused. I just use what I have found to sound good and no longer try to understand why!!
Closer and closer to music.
Hey Guys!
Things are getting there. I am about to make the last few connections and finally amplify some signal. Before I do so, let me ask what type of dummy load you recommend for testing? I've always used an old speaker, but I owuld like to purchase something a little smaller/compact specifically suited for the task.
Also, with respect to the 6J5's drive capabilities, how does one measure (presumably with a scope and sig. gen) the ability to efficiently drive the driver tube. That is, what happens when there is not enough current to the 300B? What also does this sound like.
With past projects I've usually just listened and determined what I like. I would like to go with a more scientific approach here. I have most of the test gear I think I will need, with exception of a distortion analyzer (is this a good piece of uquipment to own?).
Thanks,
Bryan
Hey Guys!
Things are getting there. I am about to make the last few connections and finally amplify some signal. Before I do so, let me ask what type of dummy load you recommend for testing? I've always used an old speaker, but I owuld like to purchase something a little smaller/compact specifically suited for the task.
Also, with respect to the 6J5's drive capabilities, how does one measure (presumably with a scope and sig. gen) the ability to efficiently drive the driver tube. That is, what happens when there is not enough current to the 300B? What also does this sound like.
With past projects I've usually just listened and determined what I like. I would like to go with a more scientific approach here. I have most of the test gear I think I will need, with exception of a distortion analyzer (is this a good piece of uquipment to own?).
Thanks,
Bryan
distortion analyzer
Hi Bryan,
you can use the soundcard in your computer as a distortion analyzer...
I did the following with my amp: 8Ohm resistor on the output of the transformer (preferably a 50W type on a heatsink).
If you want to measure at 1Watt output, then voltage over an 8Ohm resistor is already more than high enough to feed into a line-in of your soundcard, so you could use a voltagedivider inbetween (=simple potentiometer or 2 resistors)
A couple of protection zeners on the soundcard input also wouldn't hurt. And a soundcard that's not integrated on the motherboard is preferred...you don't want to fry it with a misplaced connection to the amp 🙂
Do a search on the internet for audio analyzer, there a several programs that can do a real time spectrum plot.
Before you turn on the high voltage, make sure you resolve the groundloops between amp and computer, otherwise you get a big hump on 50Hz.
If you want to measure very low amplitude signals, you can build a small signal amplifier with a couple of opamps 🙂 🙂 to do a differential measuring.
But for an amplifier output, not really necessary. Even a cheap soundcard gives enough headroom to see enough for a normal amp.
It's very interesting what you can see....
but it can't replace your ears & your taste of sound.
Filip.
Hi Bryan,
you can use the soundcard in your computer as a distortion analyzer...
I did the following with my amp: 8Ohm resistor on the output of the transformer (preferably a 50W type on a heatsink).
If you want to measure at 1Watt output, then voltage over an 8Ohm resistor is already more than high enough to feed into a line-in of your soundcard, so you could use a voltagedivider inbetween (=simple potentiometer or 2 resistors)
A couple of protection zeners on the soundcard input also wouldn't hurt. And a soundcard that's not integrated on the motherboard is preferred...you don't want to fry it with a misplaced connection to the amp 🙂
Do a search on the internet for audio analyzer, there a several programs that can do a real time spectrum plot.
Before you turn on the high voltage, make sure you resolve the groundloops between amp and computer, otherwise you get a big hump on 50Hz.
If you want to measure very low amplitude signals, you can build a small signal amplifier with a couple of opamps 🙂 🙂 to do a differential measuring.
But for an amplifier output, not really necessary. Even a cheap soundcard gives enough headroom to see enough for a normal amp.
It's very interesting what you can see....
but it can't replace your ears & your taste of sound.
Filip.
Hey guys,
Questions...
So, yesterday I tried powering the entire mono up connected directly too the wall w/o variac.
The power-up sequence is that with initial switch on the following circuits are energized:
ALL audio fils.
input/driver fils of the 83
cooling fans for the fils
bias curcuitry
After a variable delay the remaining curcuits are energized:
HV 6au4 heaters
HV B+
input/driver B+
Without the 845 B+ and 6au4 heaters and only the input/driver B+ energizing after the delay everything powers up (initail setup for testing w/o 845 B+ applied) just fine.
When I added in the 845B+ and heaters and fired up the amp the fuse blew. As currently configured there is a 2A fuse on the mains. I plan to add an additional fuse to the 845 B+.
Also when I isolate the 845 HV Tranny and bring it up slowly with a variac there are no problems, so I know the 845 B+ circuitry is sound.... Is it possible that the total current demand on the mains is exceeding the fuses capibility and causing it to blow as a function of powering up the 2 6au4 heaters and charging the 845 B+ caps...?
Repalcing the fuse and removing the 845 B+ from the curcuit results in everything working again.
This leads me to another question. How does a 2A fuse allow near 8+A (all the fils) to pass without blowing in the first place? The fuse is infront of the mains AC switch...
Thoughts?
BK
Questions...
So, yesterday I tried powering the entire mono up connected directly too the wall w/o variac.
The power-up sequence is that with initial switch on the following circuits are energized:
ALL audio fils.
input/driver fils of the 83
cooling fans for the fils
bias curcuitry
After a variable delay the remaining curcuits are energized:
HV 6au4 heaters
HV B+
input/driver B+
Without the 845 B+ and 6au4 heaters and only the input/driver B+ energizing after the delay everything powers up (initail setup for testing w/o 845 B+ applied) just fine.
When I added in the 845B+ and heaters and fired up the amp the fuse blew. As currently configured there is a 2A fuse on the mains. I plan to add an additional fuse to the 845 B+.
Also when I isolate the 845 HV Tranny and bring it up slowly with a variac there are no problems, so I know the 845 B+ circuitry is sound.... Is it possible that the total current demand on the mains is exceeding the fuses capibility and causing it to blow as a function of powering up the 2 6au4 heaters and charging the 845 B+ caps...?
Repalcing the fuse and removing the 845 B+ from the curcuit results in everything working again.
This leads me to another question. How does a 2A fuse allow near 8+A (all the fils) to pass without blowing in the first place? The fuse is infront of the mains AC switch...
Thoughts?
BK
- Status
- Not open for further replies.