I think that a 2 amp fuse is too small. First figure out the wattage of all your tube heaters mulitply the amps times the volts for each tube. Add them all up. Multiply by 1.1 to acount for the transformer losses. If you are using DC you will need a 1.5 to 2 X multiplier due to the peak current requirements. This is the steady state power, the initial power on turn on is at least double that. The heaters are probably stressing your poor 2 amp fuse to its limits. A 2 amp fuse is good for 240 watts on 120 volts.
The surge current required to charge your HV supply is considerable. After it is fully charged you will be needing about 100 to 120 watts to feed each 845.
My 845SE amplifier uses a 6.25 amp slow blow fuse. It would blow on startup until I put in an inrush current limiter. This helps dampen the initial starting current.
The surge current required to charge your HV supply is considerable. After it is fully charged you will be needing about 100 to 120 watts to feed each 845.
My 845SE amplifier uses a 6.25 amp slow blow fuse. It would blow on startup until I put in an inrush current limiter. This helps dampen the initial starting current.
Thanks tubelab,
That is what I was thinking... Just helps to know others agree!
I imagine with all the heaters running full steam, and then the added demand of switching both B+'s and the pair of 6au4 heaters the fuse had met its match...
Oh well, I guess I should expect such things during the first full power on. Fortunately there have been no fireworks with this amp as of yet. I do get a little nervous when powering up the 845B+ for the first time without the variac.
I'll crunch the numbers and report back wth the total wattage. I imagine it is pretty high as all fils are DC (with exception of the rectifiers).
Lastly, what is a good value fuse to put on the DC side of the input/driver B+ and the 845 B+. I don't want to burn-up my tamura output and EP IT should something run away...
Thanks,
BK
That is what I was thinking... Just helps to know others agree!
I imagine with all the heaters running full steam, and then the added demand of switching both B+'s and the pair of 6au4 heaters the fuse had met its match...
Oh well, I guess I should expect such things during the first full power on. Fortunately there have been no fireworks with this amp as of yet. I do get a little nervous when powering up the 845B+ for the first time without the variac.
I'll crunch the numbers and report back wth the total wattage. I imagine it is pretty high as all fils are DC (with exception of the rectifiers).
Lastly, what is a good value fuse to put on the DC side of the input/driver B+ and the 845 B+. I don't want to burn-up my tamura output and EP IT should something run away...
Thanks,
BK
Puting a fuse in the B+ lead carying 1000 volts is not easy to do. Most small fuses are only rated to 250 volts. I have seen what happens when you attempt to use one of these at 1KV and it isn't pretty. The metal strip inside the fuse melts, causing an arc through the inside of the fuse, which raises the temperature inside the fuse enough to cause it to explode. The arc then went from what was left of the fuseholder to the chassis for a few seconds, which blew the fuse on the transformer primary.
If you can find a fuseholder and a fuse rated for the voltage you are using, then I would fuse each transformer for about 1.5 to 2 X the anticipated bias current.
I put 125 mA 250 volt fuses in series with the tube cathode on my 845SE amp. This will cut the tube off in case of a bias fault. As I found out when I was rolling a bunch of old 845's, it will not help in case of a seriously gassy tube. The defective tube glowed a bright pink glow. Plate current climbed to over 300 mA for a few seconds. A wicked frying sound came from the speaker. When the cathode fuse blew the electricity flowed into the PowerDrive board and ate up a mosfet. Fortunately a $2 mosfet was the only failure.
A fuzzy picture taken right before the fuse blew is at the bottom of the web page:
http://www.tubelab.com/845SE.htm
If you can find a fuseholder and a fuse rated for the voltage you are using, then I would fuse each transformer for about 1.5 to 2 X the anticipated bias current.
I put 125 mA 250 volt fuses in series with the tube cathode on my 845SE amp. This will cut the tube off in case of a bias fault. As I found out when I was rolling a bunch of old 845's, it will not help in case of a seriously gassy tube. The defective tube glowed a bright pink glow. Plate current climbed to over 300 mA for a few seconds. A wicked frying sound came from the speaker. When the cathode fuse blew the electricity flowed into the PowerDrive board and ate up a mosfet. Fortunately a $2 mosfet was the only failure.
A fuzzy picture taken right before the fuse blew is at the bottom of the web page:
http://www.tubelab.com/845SE.htm
Hey Hey,
Looking back at the original schematic (post #68) we can see a cap and resistor around the sceondary of the IT transformer. With respect to the capacitor, approximately how much capacitance am I looking for here, and what type of cap would be recommended? The purpose of this cap is to???
I was thinking about a 200V 100uF or somehting thereabout?
Also, I've seen the resiston on the secondary omitted in certain designs.... What is the purpose of this resistor, and how is it's value determined?
Thanks,
Bryan
Looking back at the original schematic (post #68) we can see a cap and resistor around the sceondary of the IT transformer. With respect to the capacitor, approximately how much capacitance am I looking for here, and what type of cap would be recommended? The purpose of this cap is to???
I was thinking about a 200V 100uF or somehting thereabout?
Also, I've seen the resiston on the secondary omitted in certain designs.... What is the purpose of this resistor, and how is it's value determined?
Thanks,
Bryan
The glowing tube incident didn't cause too much excitement. I actually let it glow for long enough to grab the camera and take the picture. I later connected this tube up directly to an adjustable power supply and took a few more pictures of it glowing before I gave it away to a collector who just wanted to add it to his collection.
The real excitement happened when the fuseholder arced over as described above. I was given 3 HV power supplies (1500 V 1/2 amp) from some old Motorola tube type base station radios. At first I thought the fuse in the HV lead was original. It blew during one of the 833A guitar amp experiments. The resulting fireball, and wicked stink of burning plastic will get your attention real quick! I then realized that the other 2 power supplies id not have this fuseholder. I removed it from this power supply, and continued the experiments. During the excitement I had my camera on the tripod taking continuous pictures every 10 seconds. The whole thing happened so fast that the camera missed it.
In reference to the previous question, the cap provides a low impedance path for signal current to flow from grid to cathode. It is directly in the signal path and signal current will flow through it when you reach A2. Use a good quality cap here, or use multiple caps in parallel.
The resistor provides a load for the IT transformer. The value is dependent on the transformer. If you have access to test equipment, you can adjust the value to eliminate peaks in the frequency response due to the IT. If you don't, adjust for smoothest sound. It will affect the balance between the mids and the highs. Some transformers do not need this resistor. It can be left out and then added later if needed.
The real excitement happened when the fuseholder arced over as described above. I was given 3 HV power supplies (1500 V 1/2 amp) from some old Motorola tube type base station radios. At first I thought the fuse in the HV lead was original. It blew during one of the 833A guitar amp experiments. The resulting fireball, and wicked stink of burning plastic will get your attention real quick! I then realized that the other 2 power supplies id not have this fuseholder. I removed it from this power supply, and continued the experiments. During the excitement I had my camera on the tripod taking continuous pictures every 10 seconds. The whole thing happened so fast that the camera missed it.
In reference to the previous question, the cap provides a low impedance path for signal current to flow from grid to cathode. It is directly in the signal path and signal current will flow through it when you reach A2. Use a good quality cap here, or use multiple caps in parallel.
The resistor provides a load for the IT transformer. The value is dependent on the transformer. If you have access to test equipment, you can adjust the value to eliminate peaks in the frequency response due to the IT. If you don't, adjust for smoothest sound. It will affect the balance between the mids and the highs. Some transformers do not need this resistor. It can be left out and then added later if needed.
Thanks Tubelab,
With respect to the cap, when you say "low impedence path from grid to cathode" is this the same as "grid to ground" as the cathode is at ground? When chosing a value capacitor, what are the parameters to look for, and is there any low pass filter set up with this cap in any way?
I see how the resistor loads the IT, and your answer makes sense to me...
Still trying to grasp everything happening here....
Thanks,
Bryan
With respect to the cap, when you say "low impedence path from grid to cathode" is this the same as "grid to ground" as the cathode is at ground? When chosing a value capacitor, what are the parameters to look for, and is there any low pass filter set up with this cap in any way?
I see how the resistor loads the IT, and your answer makes sense to me...
Still trying to grasp everything happening here....
Thanks,
Bryan
This capacitor is in parallel with your bias supply. This supply must have a low output impedance for all audio frequencies. If it doesn't the effective bias on the tube will change when the grid draws current. This is important in an A2 amp.
Depending on what cap is already present in your bias supply, and how far away from the IT it is determines what type and value cap you need here.
If the transformer connects to the wiper of the bias pot, then you need a high quality cap. The value should be such that it is a good bypass for audio, but not too big that the bias takes a long time to change.
If you are using an active, or regulated bias supply with a low output impedance, then you may not need a cap at all.
Depending on what cap is already present in your bias supply, and how far away from the IT it is determines what type and value cap you need here.
If the transformer connects to the wiper of the bias pot, then you need a high quality cap. The value should be such that it is a good bypass for audio, but not too big that the bias takes a long time to change.
If you are using an active, or regulated bias supply with a low output impedance, then you may not need a cap at all.
Tubelab,
Hmm, yeah, the bias supply....
Well, as it is now it is schottky bridge - CRC - feeding a voltage divider with a 10K pot in between two resistoors 5K and 10K to ground. The R in the CRC is relatively large. I take it this is not low impedence and will cause problems when the 845 is drivin into A2...
Admittedly, this is one of the weak links in the current design... I guess I will test the circuit as is and see what happens at the bias when I give it a lot of signal... This will be another learning experience for me...
I've got plenty of space on the breadboard to add additional types of bias circuits, should anyone with more experience than me (most all of you) care to share some wisdom. The term "active or regulated" to me thus far means glow tubes or sand (neither of which I have an issue using). Glow tubes can not regulate with the current demands needed here as I believe John pointed out for me in the beginning... Do negative voltage 200 volt regulators exist, and if so are they any good?
My ignorance reveals itself
Hmm, yeah, the bias supply....
Well, as it is now it is schottky bridge - CRC - feeding a voltage divider with a 10K pot in between two resistoors 5K and 10K to ground. The R in the CRC is relatively large. I take it this is not low impedence and will cause problems when the 845 is drivin into A2...
Admittedly, this is one of the weak links in the current design... I guess I will test the circuit as is and see what happens at the bias when I give it a lot of signal... This will be another learning experience for me...
I've got plenty of space on the breadboard to add additional types of bias circuits, should anyone with more experience than me (most all of you) care to share some wisdom. The term "active or regulated" to me thus far means glow tubes or sand (neither of which I have an issue using). Glow tubes can not regulate with the current demands needed here as I believe John pointed out for me in the beginning... Do negative voltage 200 volt regulators exist, and if so are they any good?
My ignorance reveals itself
Hey Hey Guys,
Just posting the updated circuit I am working with...
I've got on hand a few options for the cap on the secondary of the IT. I have either 56uF/400V Panasonic TSHA, 100uF/400V Panasonic TSHA, electronic concepts unylytic 35uF/500V, or various large motor run POI's...
Panasonic
Electronic Concepts
Any thoughts as to which would work better in this location?
Just posting the updated circuit I am working with...
I've got on hand a few options for the cap on the secondary of the IT. I have either 56uF/400V Panasonic TSHA, 100uF/400V Panasonic TSHA, electronic concepts unylytic 35uF/500V, or various large motor run POI's...
Panasonic
Electronic Concepts
Any thoughts as to which would work better in this location?
Attachments
Just curious how hard you will puch into Class A2 ???
With a 10K plate load you limit your swing from -150 bias up to +50 and swing down to -350.... This is a 200V RMS input signal to grid.. You could go to +100 but it will be a squeeze and way heavy in grid current about 70mA.... At +50 you will be roughly at 20mA to 25mA dpending on tube quality...
What is the spec on the 300B driver transformer...
I calculate minimum of 50H for a -3dB at about 10Hz...
Assuming this is 1:1 ratio...with the 300B being a 800 ohm source..you will get a -.668dB loss when driving fully up to 200V, due to 20mA grid current....and -.827dB signal compression when driving up to 200V with 25mA of grid current....
Chris
With a 10K plate load you limit your swing from -150 bias up to +50 and swing down to -350.... This is a 200V RMS input signal to grid.. You could go to +100 but it will be a squeeze and way heavy in grid current about 70mA.... At +50 you will be roughly at 20mA to 25mA dpending on tube quality...
What is the spec on the 300B driver transformer...
I calculate minimum of 50H for a -3dB at about 10Hz...
Assuming this is 1:1 ratio...with the 300B being a 800 ohm source..you will get a -.668dB loss when driving fully up to 200V, due to 20mA grid current....and -.827dB signal compression when driving up to 200V with 25mA of grid current....
Chris
A2
Cerrum,
How much of this 70ma comes through the tx from the 300B and how much out of the bias supply???
Does the driver just voltage modulate the bias supply?
This A2 supply looks like NO place for any pots or resistors !!!
Looks to me just like an output stage HT supply!!
LCLC designed VERY close to required bias plus LOW impedance divider / C bypass voltage tuning??
cheers,
Cerrum,
How much of this 70ma comes through the tx from the 300B and how much out of the bias supply???
Does the driver just voltage modulate the bias supply?
This A2 supply looks like NO place for any pots or resistors !!!
Looks to me just like an output stage HT supply!!
LCLC designed VERY close to required bias plus LOW impedance divider / C bypass voltage tuning??
cheers,
Hey Guys,
Interesting pontification...
Chris, the specs for the IT are here...
With respect to how far I want to push the envelope into A2, I have not yet decided. I am looking for headroom of about 35 watts. I don't think I would drive the grid anywhere over +50V.
At the end of the day, I may just keep it in A1, but this is also a learning experience for me. I have a long way to go!
I'm still up in the air with the bias circuit. I've been reading bits and pieces but have not yet taken any decisive actions.
It seems LCLC to the IT may be the way to go if I can not find a negative regulator up to the task...
Hifryer, this is a good question... At +50V on the grid, I would dram near 25mA from the the grid. I too wonder how much of this is supplied by the bias supply vs. the driver/IT?
I still may play around with an 0D3 feeding -150V to the grid. That or a stack of 0A3-0B3 with a pot... I need to crunch some numbers, but me thinks I should be able to set up the glow tubes to sink 35mA under quiescent conditions, and to allow about 30mA current when the grid is drawn positive....
I still dont know if this would be adequate for supplying the grid current needed...
Interesting pontification...
Chris, the specs for the IT are here...
With respect to how far I want to push the envelope into A2, I have not yet decided. I am looking for headroom of about 35 watts. I don't think I would drive the grid anywhere over +50V.
At the end of the day, I may just keep it in A1, but this is also a learning experience for me. I have a long way to go!
I'm still up in the air with the bias circuit. I've been reading bits and pieces but have not yet taken any decisive actions.
It seems LCLC to the IT may be the way to go if I can not find a negative regulator up to the task...
Hifryer, this is a good question... At +50V on the grid, I would dram near 25mA from the the grid. I too wonder how much of this is supplied by the bias supply vs. the driver/IT?
I still may play around with an 0D3 feeding -150V to the grid. That or a stack of 0A3-0B3 with a pot... I need to crunch some numbers, but me thinks I should be able to set up the glow tubes to sink 35mA under quiescent conditions, and to allow about 30mA current when the grid is drawn positive....
I still dont know if this would be adequate for supplying the grid current needed...
I would NOT actively regulate the bias.
If you build a good low impedance supply it will be passively well regulated and similar to the B+. Output stage gain is low so hum should not be a problem id well executed.
Here is a crazy thought I just had. Never done it. Looking for comment.
Lose the cap altogether and lose the connection back to the bias supply at bottom of IT secondary. Choose a suitable value power resistor to provide a desired standing DC current through the IT secondary and connect resistor to ground.
This if wired one way this will partially cancel the standing magnetic field in the IT and may provide more current headroom for the driver primary circuit. It also loses a cap bass phase shift altogether and may provide tigher, faster bass??
In any case, I suggest try unregulated first.
cheers
If you build a good low impedance supply it will be passively well regulated and similar to the B+. Output stage gain is low so hum should not be a problem id well executed.
Here is a crazy thought I just had. Never done it. Looking for comment.
Lose the cap altogether and lose the connection back to the bias supply at bottom of IT secondary. Choose a suitable value power resistor to provide a desired standing DC current through the IT secondary and connect resistor to ground.
This if wired one way this will partially cancel the standing magnetic field in the IT and may provide more current headroom for the driver primary circuit. It also loses a cap bass phase shift altogether and may provide tigher, faster bass??
In any case, I suggest try unregulated first.
cheers
Hifryer,
Well, if I go unregulated, I will do a simple LCLC supply. What I wonder about is getting the bias to sit where I need it.... That is, with even the best planning/design, taking into account the dcr of chokes in the supply, and given the known voltage of a transformer, I will still need some "slop" factor, or "tuning" as you suggest earlier.
Would it be possible to tune the final output voltage with a small cap input set to give the desired output voltage?
Also, with a low impedance divider, how low is low....
Assuming I want to limit bias drift to +/- 2Vdc over 1-40mA of grid current, I'm looking at no more than 50 ohms of resistance in the supply or on a potential divider.... Is this possible???
Well, if I go unregulated, I will do a simple LCLC supply. What I wonder about is getting the bias to sit where I need it.... That is, with even the best planning/design, taking into account the dcr of chokes in the supply, and given the known voltage of a transformer, I will still need some "slop" factor, or "tuning" as you suggest earlier.
Would it be possible to tune the final output voltage with a small cap input set to give the desired output voltage?
Also, with a low impedance divider, how low is low....
Assuming I want to limit bias drift to +/- 2Vdc over 1-40mA of grid current, I'm looking at no more than 50 ohms of resistance in the supply or on a potential divider.... Is this possible???
use a largish last filter cap the impedance before this is then less relevant. Model the supply.
I am against fudging psuedo choke input with little caps - cant help regulation!! You can get a tap of two on the secondary if you have a tx made and a tap on the primary for fine tuning.
I'm not an A2 supply expert. Need input from others more knowledgeable on my previous question and on sanity of my recent idea!
Maybe for bias adjustment and maintaining low enough impedance for A2 you have no alternative but use an adjustable reg. I would be very careful though and keep it a simple non feedback design as this is wel and truly in the signal path and a great place to turn the whole thing into a transistor sounding amp!!!
Divider doesnt look feasible to me due to high current that would need to flow through it. And despite perfect supply design you will want some adjustment for different tubes etc.
Hey another idea. Create adjustment via the output tube cathode with partial self bias. Will stabilise tube changes a little also.
Build supply say 20V under biased measure it and choose suitable cathode resistor?
The Japanese have done all this to death for many years now. Try to find some proven Jap A2 bias solutions!
As I said I need help from others to confirm or critique these ideas!!
cerrum??? tubelab??
cheers
I am against fudging psuedo choke input with little caps - cant help regulation!! You can get a tap of two on the secondary if you have a tx made and a tap on the primary for fine tuning.
I'm not an A2 supply expert. Need input from others more knowledgeable on my previous question and on sanity of my recent idea!
Maybe for bias adjustment and maintaining low enough impedance for A2 you have no alternative but use an adjustable reg. I would be very careful though and keep it a simple non feedback design as this is wel and truly in the signal path and a great place to turn the whole thing into a transistor sounding amp!!!
Divider doesnt look feasible to me due to high current that would need to flow through it. And despite perfect supply design you will want some adjustment for different tubes etc.
Hey another idea. Create adjustment via the output tube cathode with partial self bias. Will stabilise tube changes a little also.
Build supply say 20V under biased measure it and choose suitable cathode resistor?
The Japanese have done all this to death for many years now. Try to find some proven Jap A2 bias solutions!
As I said I need help from others to confirm or critique these ideas!!
cerrum??? tubelab??
cheers
Here is an option...
Below is an idea for my 845 bias supply. I've looked at the voltages at 2-40mA draw, and its seems fairly stable. It drops about 3 volts within the current range. Ripple is under 10mV and goes to about 6mV with the higher load.
What do you guys think?
Also, would it be possible to use the same circuit for the 300B fixed bias using traditional divider/pot/divider network pulling the voltage from the same location (C2) or should I stick with a seperate bias circuit for the driver (keeping what I currently have)... Hell, what's another 10 pounds....
If you notice, the circuit will not be cheap! Any ideas how to improve it functionally and financially?
Thanks,
Bryan
Below is an idea for my 845 bias supply. I've looked at the voltages at 2-40mA draw, and its seems fairly stable. It drops about 3 volts within the current range. Ripple is under 10mV and goes to about 6mV with the higher load.
What do you guys think?
Also, would it be possible to use the same circuit for the 300B fixed bias using traditional divider/pot/divider network pulling the voltage from the same location (C2) or should I stick with a seperate bias circuit for the driver (keeping what I currently have)... Hell, what's another 10 pounds....
If you notice, the circuit will not be cheap! Any ideas how to improve it functionally and financially?
Thanks,
Bryan
Attachments
Looks good to me. Maybe A2 supplies are just expensive? There is no free lunch.
What is first current load - a bleeder resistor??
It does need to be adustable though to allow for and match tube differences. Simple active regulation would be adjustable and cheaper.
I await a response from a mate who is more knowledgeable in these things - can bring experience from big tube transmitter modulators etc.
He might come up with something simple and less expensive!
.... Just found one!! Here is how it is done. Obvious really. Just lower impedance with a big 100uF cap after the adjustment pot!!
http://www.audiodesignguide.com/my/845se.html
Have an overall good look at this design!!
Cheers
What is first current load - a bleeder resistor??
It does need to be adustable though to allow for and match tube differences. Simple active regulation would be adjustable and cheaper.
I await a response from a mate who is more knowledgeable in these things - can bring experience from big tube transmitter modulators etc.
He might come up with something simple and less expensive!
.... Just found one!! Here is how it is done. Obvious really. Just lower impedance with a big 100uF cap after the adjustment pot!!
http://www.audiodesignguide.com/my/845se.html
Have an overall good look at this design!!
Cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.