Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.|
|9/21/2003 2:08 PM|
||input choke - voltage/warm-up question|
If I use a rectifier tube with a choke input, will the slow voltage increase on start up be OK with the minimum current needed for the choke without it shooting up to peak input cap type values?
I have a 400-0-400 PT and a 7H choke. I want to use this for a KT88 SE amp. Guessing I'm going to get around 350v B+ with a tube rectifier, minimum current needs to be around 50mA for this choke without the voltage shooting up to 564. I'm going to use 450v radial caps. That's why I'm going with a tube rectifier and no standby switch. I figure if I use a SS rectifier, it's inevitable that some day a standby switch will be in the wrong position when the power switch is thrown on and all the caps will blow.
Then I wondered whether the tube rectifier will be safe on warm up. Does the slow voltage increase (I will be using a 5AR4/GZ34) situation keep things OK? I can't wrap my mind around what happens to the current during warm up and if it stays within minimum needs for the choke without voltage shooting above the cap ratings.
|9/22/2003 12:36 PM|
I've had the exact same problem! Chokes as you may know need a minimum critical inductance - L=V/I (in ma.).
The problem doesn't seem to be the warmup time of the rectifier but rather the output tubes. They're the ones who have to draw enough idle current to stay above the ma. you cited for the choke to work.
I ended up with a bit of bleeder resistance to draw about 10 ma. Any choke will show more inductance at lower currents than the published spec and mine snapped into choke input mode at about this amount of steady draw.
Later I added a reverb to the circuit and the cathode current drawn by the driver tube came up more than quickly enough to keep choke-input voltages happening across the filters. Then I could go back to smaller, higher value bleeders.
|9/22/2003 1:01 PM|
Good question! It's cool to think about these things before the caps start popping... I think it's safe to assume that there will be a short period of time during warm-up when the GZ34 will be passing less than 50 mA; the question is, how long do your caps take to charge? They may still be charging and loading down the PS as the tubes warm up and start to pull current. But let's assume that 560V B+ exists for some period of time (with practically any period too long for 450V caps IMO)...
How about a string of four 100 volt 5W Zeners (or five 80V 5W, etc.) along with a 3.3k 10W resistor, from B+ to ground? The 100V Zeners will dissipate their rated 5W each but only for a short time during warm-up, and you'll be pulling 50 mA from the PS which should bring your choke into regulation (this circuit may oscillate briefly during warm-up as the choke goes in and out of regulation, but I don't think it should be a problem). During normal operation this circuit should do nothing, but it's still there to protect your caps if a circuit fault occurs and you lose your PS load.; plus, if it works well you can go ahead and use that SS rectifier/standby switch if you want, along with larger filter caps which I think you'll want in an SE design.
There may well be a better way to do it than this, it's just off the top of my head.
|9/22/2003 1:52 PM|
...maybe I don't understand exactly how a rectifier warms up. I'm talking about indirect heated like the 5AR4. I was under the impression that the voltage came up slowly during warm up. Is this correct or have I got it wrong? If I *do* have it right, then the critical current needed for the choke to regulate would also start out very low and increase with the voltage rise.
Wild Bill, what was the general set up for your power supply? Did you use a 5AR4? Did you still use high voltage rated caps? What type of amp? Class? The reason I ask is because your response may indicate a greater safety margin in my topology. The KT88 is going to draw close to 100mA. My driver tube (12AU7 in parallel) is going to draw about 12mA the way TubeCad figures. If the 5AR4 brings voltage up slowly, I might not need any bleeders at all.
Ray, I agree completely that if the voltage shoots up past choke regulation for any length of time, it's too long for 450v caps. This is a PT that sags to 410v with a 80mA load using a 5Y3. I'm thinking if I have to use stacked caps, I might as well go with the standard input cap pi filter arrangement and get some extra voltage for some extra wattage. The load lines tell me I can squeeze 18 watts with close to 400v B+, but only about 13 watts at 350v B+. Probably not any more clean volume with the 400v, but I'm also thinking the amp might have more dynamic range. I might also be splitting hairs, or it could just be a difference in tone. Who knows? I would like to keep the choke input if I can, but if I have to go to stacked caps and zeners, I'll just go back to the typical pi filter arrangement.
So does an indirectly heated rectifier tube bring up voltage gradually/slowly? Sorry about my ignorance regarding rectifier tubes.
How did all those HiFi amps with choke inputs of yesteryear get by?
Thanks for the help.
|9/22/2003 2:08 PM|
here's another thing you can try....
it's not terribly uncommon to put a cap across the secondary of the PT to drop the voltage somwhat. I'm not really sure how to calculate what value of capacitor you need, but I know it can be done. you could just tie it across the two plates of your 5AR4, and then nothing would see more voltage than the plates of the rectifier. I'd say just experiment with cap values until you find one that works. keep in mind with a 400-0-400 tranny. you'll need a cap that can handle 1000v or better.
|9/22/2003 4:13 PM|
Yes - but how slowly? As Wild Bill posted, the load (tube) warm-up time is what's important, and the KT88's will heat up slower than the 5AR4. The important thing that I think you want to know is, "What current will my circuit be drawing from my power supply at each point during the warm-up period?".
IMO, to compute a time-vs-voltage graph for your circuit that could be expected to have a reasonable correlation to empirical results is quite a complex task, although with a bleeder resistor it would be simpler. I had originally included the suggestion of a 6.8k, 30W bleeder as an inferior solution to the problem, but deleted it before posting; Wild Bill's smaller bleeder would be a different story, but then my calculations were based on the 50 mA figure I was given. I've designed one (1) choke-input supply in my life, so an expert on them I'm not.
The Zener stack would be inoperative during normal circuit operation, and although stacked caps would of course provide a quick-and-dirty solution, I didn't suggest them because I felt there might be a better way that would allow you to use single caps of higher value for minimum hum, and/or a SS rectifier/standby switch for greater flexibility and possibly longer tube life.
FWIW, for your SE application I would rather have the higher regulation, lower voltage, and possibly lower noise of the choke-input design, given the choice between that and cap-input (active regulation not being presented as an option ). Also, a 5AR4 can handle only 60 to 80 uF of filter capacitance directly on its output, depending on which spec sheet you use.
Probably just with small bleeders. I would try it and see!
|9/22/2003 7:53 PM|
Well, you've managed to articulate my exact thoughts which prompted this question in the first place. It's nice to know my logic in wondering was correct.
Yes, I agree about wanting to keep the choke input. As much of a pain as it seems right now, it certainly does solve a lot of the SE problems with the power supply.
I'm going to go ahead with the standby switch using a 5AR4. As I was thinking about it, I remembered kg mentioning once that even if you don't need a standby, it makes troubleshooting a whole lot easier. So the standby switch will go in and after I get the amp working I'll do a voltage reading on the first 'lytic on start up without standby. I'll be using 450v radials that can surge to 500v. I'll just keep an eye on the voltage and if it surges any bit near 500v I can just hit the power switch and know I have to use bleeders or the zener/bleeder trick.
Thanks for the help.
PS - Does it matter which side of the choke I use to insert the standby switch? Rectifier side or cap side?
|Page 1 of 3||Next>||Last Page>>|