There will be an open circuit to ground, which forces current to flow to the output. I get it now!
I found a 22000uF cap in lab today. I also dug into some books and noticed that I could use a 6V regulator, but of course they have a whopping 2-2.5V of dropout.
Sorry for asking more than once Tom. I am all over the place with my full course schedule, due date, and finals drawing near. I feel like I actually got dumber recently haha.
I found a 22000uF cap in lab today. I also dug into some books and noticed that I could use a 6V regulator, but of course they have a whopping 2-2.5V of dropout.
Sorry for asking more than once Tom. I am all over the place with my full course schedule, due date, and finals drawing near. I feel like I actually got dumber recently haha.
Oh, crap. I forgot that I was using a full winding potential and not the center tap. Yeah, my ripple is now more realistic haha.
I have other transformers, but this unit is going to weigh a little bit due to 2 chokes, a tube transformer, PCB mounted transformer, and now another transformer. It's really not a problem if it's what I need to do.
I have other transformers, but this unit is going to weigh a little bit due to 2 chokes, a tube transformer, PCB mounted transformer, and now another transformer. It's really not a problem if it's what I need to do.
Last edited:
Actually, the point you missed was that the diode between pins 2 and 4 will short out the trafo during half the cycle.
With the circuit the way it's drawn, you should see 60 Hz ripple at about 2x the voltage. When the circuit is wired properly, you should see 120 Hz ripple.
~Tom
No. In the schematic you posted in Post #58, you short out the "trafo" (V1) during half the cycle through the diode connected between pins 2 and 4 of D3 (diode bridge).
Pin 3 of the diode bridge should connect to the (-) pin of V1 and not to GND. If you connect it according to the schematic, you will blow fuses (if the rectifier or PCB traces don't die first).
~Tom
Pin 3 of the diode bridge should connect to the (-) pin of V1 and not to GND. If you connect it according to the schematic, you will blow fuses (if the rectifier or PCB traces don't die first).
~Tom
An externally hosted image should be here but it was not working when we last tested it.
It's getting there. If this transformer doesn't work, I'll have to remove it and use a chassis mount one.
You have to measure the LOADED voltage. The Vtrafo = xxx equation gives you the LOADED voltage the trafo will have to supply. Unloaded is nearly irrelevant. If the trafo outputs 10 V unloaded but drops to 7 V at 300 mA load, it won't supply enough voltage. Load the trafo with a resistor such that the current draw will be equal or slightly larger than the heater current of the tube(s). Then measure the secondary voltage. Recall that the AC mains may vary +/-5 % with sporadic drops below that, so if you measure at 120 V AC in, remember to take that +/-5 % into account. That's where the 0.95 factor in the Vtrafo = xxx equation comes from. If you are planning to use a huge reservoir cap, you'll probably end up at closer to Vcap = 1.3*VAC. But you won't hit the magic sqrt(2)*VAC.
Do the math, dude. What's the minimum loaded voltage the trafo will need to supply? You have all the numbers. You have the equation. Just do the math. Or at the very least run a sim and ensure that the voltage you've picked will guarantee 1.75 V across the regulator under load.
~Tom
Last edited:
Eh... Vout is the output voltage of the regulator. That's 6.3 V.
So Vtrafo > 6.3+1.75+0.25+2*0.85/(1.25*0.95) --> Vtrafo > 8.42 V.
250 mV ripple. That requires a big cap... But you seem to be into big caps. Hence, I've increased the RMS->Peak factor from 1.2 to 1.25. As I said earlier, the diode drop for 300 mA forward current can be found in the data sheet for the diode bridge. Vdiode_drop = 2*Vf, where Vf is the forward voltage of the diodes for the given output current (why am I repeating myself here?!)
So you'll need a trafo that will supply at least 8.42 V RMS when LOADED by 300 mA and supplied 120 VAC on the primary. I'd look for a 9 V trafo...
Good luck with your project.
~Tom
So Vtrafo > 6.3+1.75+0.25+2*0.85/(1.25*0.95) --> Vtrafo > 8.42 V.
250 mV ripple. That requires a big cap... But you seem to be into big caps. Hence, I've increased the RMS->Peak factor from 1.2 to 1.25. As I said earlier, the diode drop for 300 mA forward current can be found in the data sheet for the diode bridge. Vdiode_drop = 2*Vf, where Vf is the forward voltage of the diodes for the given output current (why am I repeating myself here?!)
So you'll need a trafo that will supply at least 8.42 V RMS when LOADED by 300 mA and supplied 120 VAC on the primary. I'd look for a 9 V trafo...
Good luck with your project.
~Tom
The 3N249 datasheets I looked at (both on ALLDATASHEET.COM - Datasheet search site, Datasheet search site for Electronic Components and Semiconductors and other semiconductors.) had a graph that showed Vf as function of Iout. So I looked at the worst case for 300 mA. I believe it was about 850 mV. That's why I used that number rather than the 1~1.5 V that was on the front page of the data sheet.
Just letting you know where I got the number from.
Have fun. I'm looking forward to seeing your calculations for the worst case power dissipation in the LM317... (yes, that's a hint...
)
~Tom
Just letting you know where I got the number from.
Have fun. I'm looking forward to seeing your calculations for the worst case power dissipation in the LM317... (yes, that's a hint...
)~Tom
Haha, I do have it strapped to a heatsink, but I doubt it'll be cool.
Like you mentioned before; I don't need adjustability, but using a 6V regulator needs 2V of dropout. Is that any smarter of a path?
Unfortunately, I can't get into this transformer to add windings either. I have some 10Vrms transformers in the lab, but they have no specs on them and I can't find them online. Due to the small size (1.5"x1.5"x1.75" or so), I could infer that it can't handle much current.
There are hundreds of transformers that I can sift though, so I'm not worried... yet.
Like you mentioned before; I don't need adjustability, but using a 6V regulator needs 2V of dropout. Is that any smarter of a path?
Unfortunately, I can't get into this transformer to add windings either. I have some 10Vrms transformers in the lab, but they have no specs on them and I can't find them online. Due to the small size (1.5"x1.5"x1.75" or so), I could infer that it can't handle much current.
There are hundreds of transformers that I can sift though, so I'm not worried... yet.
I doubt that as well. That's why I ended up with a switching regulator. But that's beyond the scope of this project. But for future reference, check out the LMZ1200x parts from National. Cool stuff (in more ways than one)...
Certainly. There's a group of regulators optimized for this kind of application -- Low Dropout Regulators (LDO's).
LDOs @ National Semiconductor
Linear Tech LDOs
LDO's @ Digikey
Find something that'll handle the input voltage, supply the output voltage, has a current limit beyond 300 mA (I'd probably go with at least 500 mA), and a lower dropout than the LM317. They're not that expensive and options are plentiful. Also... Pick one that can be hand-soldered... Many of the newer regulators come in uSMD or BGA packages that require expensive PCB tooling and are impossible to solder by hand.
~Tom
That 7.5-0-7.5 V trafo has another advantage. You can use the center tap for ground and connect the two 7.5 V taps to the positive terminal on the reservoir cap through a diode -- just as is commonly done for B+ with tube rectifiers. That would give you full-wave rectification with only two diodes.
In this setup, you only one diode conducts at a given time, hence, Vdiode-drop is 1.0 V rather than 2.0 V. That gives you a little more headroom for the regulator. Not enough for an LM317 probably, but something like an LM1086 might work...
~Tom
In this setup, you only one diode conducts at a given time, hence, Vdiode-drop is 1.0 V rather than 2.0 V. That gives you a little more headroom for the regulator. Not enough for an LM317 probably, but something like an LM1086 might work...
~Tom
Beautiful. I have an aluminum heat sink for the LM317, but man that thing would cook an egg with 15VAC dropping to 6.3V.
I forgot to mention the transformer went from 9VAC to 7.97VAC @ 300mA.
This would yield 11.273Vp before the rectifier, dropping to 9.573V ripple DC @ 120Hz, which will then charge the capacitor.. Let me figure the rest out, then I'll post my results and see if it is high enough to feed the LM317.
I forgot to mention the transformer went from 9VAC to 7.97VAC @ 300mA.
This would yield 11.273Vp before the rectifier, dropping to 9.573V ripple DC @ 120Hz, which will then charge the capacitor.. Let me figure the rest out, then I'll post my results and see if it is high enough to feed the LM317.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.