I'll need to "back calculate" to find the current the transformer will need so the heaters get 12.5A. That comes to 22.5 amps or more needed.
using: Iac=Idc/1.8.. for grins I'll see what I can find on the shelf.
ack..... only up to 20 amps from hammond ( 11.11A after the rectifier / filter)
Thanks Michaael, David and Gianluca....
the adventure continues...
JD
I don't know if that formula should be taken as a constraint, but 20A is about right.
How about a 250VA 120V:12V ICT (8-10 lbs) or maybe a Hammond toroid e.g. 182T12 (5" dia, 5.5lbs)? Use a larger first cap for a 12V nominal secondary. I think the DCR will be OK on these, even may need to add a little on the primary to tweak the output down.
Attachments
I'm wondering and curious about Gianluca's recomendation against the 12.5A chokes, not sure of his reasons.
I've already recieved my $5.99 eBay SPC20976 chokes (10mh, 12.5A) and they are basically brand new in the box. New ones are in the $60 range, so I've saved a few pennies. BUT... if they are not of enough inductance and current handling (as Gianluca stated), I have a line on a pair of 15mH, 16A Chokes. They are just a "little" dear, but doable. I figure they can be used on either a Linear supply or with the SMPS.
I got ahead of myself and ordered the 18V trannys last week, I'll be returning them and just eat the return ship and restock fee.
I'm afraid to order any more iron. Knowledgable people have given good advice, problem is sometimes it's in conflict...example: Dave's comps for transformer voltage into DC conflict with PSUD, and Broskie. I can't find anywhere in a bridge reference that VDC=Vac X .9 and Dave seems to add all 4 diode voltage drops when the other formulas only count 2...etc....And Gianlucas advice about 12.5A gapped parts not working in a circuit that is only supposed to deliver 12.5A so the conflicts add to confusion and a few... aw sh*t...I just orderd a wrong part moments.... to many.
The SMPS 12v@12.5 supplies are on the way.. ordered them this AM. Not to bad pricewise... a lot less than Iron.
The formula's I've been using are from Broskie's text and a couple of electronics reference text books.
What I don't want to do (again) is get iron in house and find it not quite what it needs to be..... I have one hell-uva stash of parts just because of that reason and I really don't need to accumulate more (if possible).
I have a bad habit of following one track of thought, get the parts lined up and then when they're on the way find a better path and get stuck with parts I can't return and never use....
So.... I've got to nail this one on the head if I can instead of shooting flies with a shotgun.
Jeff
I've already recieved my $5.99 eBay SPC20976 chokes (10mh, 12.5A) and they are basically brand new in the box. New ones are in the $60 range, so I've saved a few pennies. BUT... if they are not of enough inductance and current handling (as Gianluca stated), I have a line on a pair of 15mH, 16A Chokes. They are just a "little" dear, but doable. I figure they can be used on either a Linear supply or with the SMPS.
I got ahead of myself and ordered the 18V trannys last week, I'll be returning them and just eat the return ship and restock fee.
I'm afraid to order any more iron. Knowledgable people have given good advice, problem is sometimes it's in conflict...example: Dave's comps for transformer voltage into DC conflict with PSUD, and Broskie. I can't find anywhere in a bridge reference that VDC=Vac X .9 and Dave seems to add all 4 diode voltage drops when the other formulas only count 2...etc....And Gianlucas advice about 12.5A gapped parts not working in a circuit that is only supposed to deliver 12.5A so the conflicts add to confusion and a few... aw sh*t...I just orderd a wrong part moments.... to many.
The SMPS 12v@12.5 supplies are on the way.. ordered them this AM. Not to bad pricewise... a lot less than Iron.
The formula's I've been using are from Broskie's text and a couple of electronics reference text books.
What I don't want to do (again) is get iron in house and find it not quite what it needs to be..... I have one hell-uva stash of parts just because of that reason and I really don't need to accumulate more (if possible).
I have a bad habit of following one track of thought, get the parts lined up and then when they're on the way find a better path and get stuck with parts I can't return and never use....
So.... I've got to nail this one on the head if I can instead of shooting flies with a shotgun.
Jeff
Last edited:
I wondered about Dave's 4 diodes myself... I think it should be 2 as there are only 2 diodes conducting at any one time (hopefully)
VDC = (VAC * 0.9) - 2 * diode drop at peak current (40A in the PSUD example above), rule of thumb for a bridge + critical inductance choke input filter.
Looking back at Gianluca's post, he is adding extra headroom for AC+DC assuming use as the input choke in a choke input filter. Your ebay chokes will be fine in the cap-input filter PSUD2 I just posted as long as the DCR is ~0.1 ohm. A 12V/250VA ICT should be ebay-able for mostly shipping. Experimentation and tweaking will be needed.
Wow, you have leftover iron? ;-) Just think of all the future amps...
Cheers,
Michael
OOps... I had the 2.8V drop of the original bridge on my mind and didn't realize he had switched diodes. I think you are correct that only two diodes in series conduct at a time.
My question is: If there will be 12.5A of DC current going through the load, why do you need more than 12.5A of current capability from the secondary of the power transformer? Obviously you need a higher voltage, (hence VA rating) but shouldn't the current remain fixed paying attention to kirchoff?
dave
My question is: If there will be 12.5A of DC current going through the load, why do you need more than 12.5A of current capability from the secondary of the power transformer? Obviously you need a higher voltage, (hence VA rating) but shouldn't the current remain fixed paying attention to kirchoff?
dave
That is correct. 10mH/15A for choke input duties (both 10mH and 15A are minimum values); whatever mH/12.5A for cap input duties.
My experience is that you need more taps at the secondary and more taps at the primary (eg +/- 10%) to hit exactly 10V at the filament; luck can help somehow ... PSUD can be 10% wrong or parts can be 10% tolerance. Dropping resistors are quite nasty with 12.5A across them. This is true for cap inputs as well.
OH! Did I ever tell you that some 304TL even if unused are gassy or completely out of vacuum? And when you turn them on you can only see the filaments turning into grey/blue smoke.
Have fun, they are pieces of cakes ... hard to resist.
That is correct. 10mH/15A for choke input duties (both 10mH and 15A are minimum values); whatever mH/12.5A for cap input duties.
My experience is that you need more taps at the secondary and more taps at the primary (eg +/- 10%) to hit exactly 10V at the filament; luck can help somehow ... PSUD can be 10% wrong or parts can be 10% tolerance. Dropping resistors are quite nasty with 12.5A across them. This is true for cap inputs as well.
OH! Did I ever tell you that some 304TL even if unused are gassy or completely out of vacuum? And when you turn them on you can only see the filaments turning into grey/blue smoke.
Have fun, they are pieces of cakes ... hard to resist.
Hmm also may want to avoid running the filaments too long without B+ and getter action (red plate) just in case there is gas inside.
A bunch of taps on the primary would be really great. If you use the Hammond Toroid you could wind your own buck/boost winding on top... also maybe get a 120-6.3CT gives you +/- 2.5% and 5% as a separate buck/boost. The secondary rating needs to be >= the primary current of the main filament xfmr (2A would be nice) and this does add to the DCR a little.
Last edited:
gotta love those tolerances...they can sneak up and bite you..... the 10mh 12.5A chokes measure at .2 DCR, not the published .11R Changes up a few things....
Michael, playing around with PSUD using a higher voltage transformer, 18V.
If I change the first cap from the 20000uf that was in the 12v model down to 2200uf in the 18v, using .2R for the chokes DCR, voltage gets right around 10v.
If I change from .1R to .2R in the original model with 12v, the voltage drops to around 8.75 volts.... that's why I modeled with the 18V tranny and a smaller first cap.... DC becomes a bit more rippley, but the larger tranny makes up for the V drop.
Now If I get a tranny in my hands, I usually find that the Voltage RMS measure from 5 to 10% above the published RMS value. What ever that measured value is, I would then plug that number into the unloaded RMS value for the tranny in PSUD, so the 12.6v might actually be 14 volts or so, the 18v tranny mught be nearer to 20 volts or so..... by sizing the first cap I can tweak the voltage a bit, at the cost of higher ripple ( If we don't want to put the big electrolytic cap in the third position instead of the "fooler" 100uf)... we don't want no stinkin cap in the cathode loop
hmmm smps is virtually "plug and play" compared to dialing in a linear, even with an actual higher DCR of the chokes, the voltage can be tweaked with the turn of a small trim pot.... just have to deal with the RF nasties....hello Mu-metal lined box.
JD
Michael, playing around with PSUD using a higher voltage transformer, 18V.
If I change the first cap from the 20000uf that was in the 12v model down to 2200uf in the 18v, using .2R for the chokes DCR, voltage gets right around 10v.
If I change from .1R to .2R in the original model with 12v, the voltage drops to around 8.75 volts.... that's why I modeled with the 18V tranny and a smaller first cap.... DC becomes a bit more rippley, but the larger tranny makes up for the V drop.
Now If I get a tranny in my hands, I usually find that the Voltage RMS measure from 5 to 10% above the published RMS value. What ever that measured value is, I would then plug that number into the unloaded RMS value for the tranny in PSUD, so the 12.6v might actually be 14 volts or so, the 18v tranny mught be nearer to 20 volts or so..... by sizing the first cap I can tweak the voltage a bit, at the cost of higher ripple ( If we don't want to put the big electrolytic cap in the third position instead of the "fooler" 100uf)... we don't want no stinkin cap in the cathode loop
hmmm smps is virtually "plug and play" compared to dialing in a linear, even with an actual higher DCR of the chokes, the voltage can be tweaked with the turn of a small trim pot.... just have to deal with the RF nasties....hello Mu-metal lined box.
JD
Last edited:
cold? hot? how did you measure it? Jeff, you really need to breadboard it to see how things go. PSUD will not tell you all the truth.
I guess cold.... I took my trusty Fluke and measured across the leads.
Breadboarding is a catch 22....you need the parts to build it, but once the parts are set up and used they can not be returned. If they prove not to be wokable then I'm stuck with a lot of expensive iron. In this case, for the surplus 5.99 parts no big deal, but I will have to buy a transformer to test them hot on breadboard ( I have some huge 300 watt .8 ohm and .4 ohm load resistors for testing the supply by itself), and if that transformer proves not to be the right one, well... that's why I mentioned earlier about have a "pile-o-stuff". Then I have to buy everything once more for the other channel.
Using PSUD "before" I bread board..and yes breadboarding has always been my plan, gives me an idea of the value and rating of the transformer (or other bits in the chain) that need to be obtained, so I can avoid, if possible, an expesive cycle of trial and error.
JD
Breadboarding is a catch 22....you need the parts to build it, but once the parts are set up and used they can not be returned. If they prove not to be wokable then I'm stuck with a lot of expensive iron. In this case, for the surplus 5.99 parts no big deal, but I will have to buy a transformer to test them hot on breadboard ( I have some huge 300 watt .8 ohm and .4 ohm load resistors for testing the supply by itself), and if that transformer proves not to be the right one, well... that's why I mentioned earlier about have a "pile-o-stuff". Then I have to buy everything once more for the other channel.
Using PSUD "before" I bread board..and yes breadboarding has always been my plan, gives me an idea of the value and rating of the transformer (or other bits in the chain) that need to be obtained, so I can avoid, if possible, an expesive cycle of trial and error.
JD
Last edited:
In my jouneys, I have yet to come across the mythical beasty.
At least not off the shelf.
A custom one would be fairly costly? and lead times might be painfully long?
Would you happen to have a source? What would be the voltage incriments of the secondaries? 1 volt? 1/2 volt? Voltage range?
JD
At least not off the shelf.
A custom one would be fairly costly? and lead times might be painfully long?
Would you happen to have a source? What would be the voltage incriments of the secondaries? 1 volt? 1/2 volt? Voltage range?
JD
I got mine in Italy. I'd suggest 6 secondaries 2.5V each and primary with -10/5 and +5/10 % taps. Mine is winded this way and allows me to play with quite a number of tubes.
If you plan to use the OPT only for 304TL one option would be to have just a secondary as predicted by PSUD and add as many primary taps as practical (2.5% steps maybe): cost and winding constrains apply.
If you plan to use the OPT only for 304TL one option would be to have just a secondary as predicted by PSUD and add as many primary taps as practical (2.5% steps maybe): cost and winding constrains apply.
Update; FINALLY sps showed up; 5v at 50!amp. Yet still has hard time starting tube, will start if filiments are staged, haven't tried resistor in series yet for starting. However do have some hash specs for you. With both filiments going and at 5v, filiments paralleled; at 27.3khz there is a 8mv spike and at 54khz, 3mv. The sps switches, they say at 20khz. I only ran the tube (first tube one filiment died, RATS) for about 5min to stablize, but the sps didn't even get warm, it does have a cooling fan which didn't turn on... YET. I don't have any experience with DHTs as far as filiment noise, 8mv audible? Cheers Ronp
update on up date, 1 ohm resistor , then bypassed will allow tube to start. Just used a 25w one cuz that is what was handy, not on long enough to get hot. And one choke, or hash filtre in this case knocked down the 3mv at 54kh, spike at 27 still there. Choke dropped voltage to tube from 5.v to 4.89. ronp
Ready to bread board...almost, just waiting for the Output transformer which is due next week. I've decided to use Yamamoto amp meters and the schematic reflects their installation per Yamamoto vis-a-vis their internal impedances.
I'll be bread boarding the power supplies first, using some bigazz 300 watt load resistors sized to what the actual load should be, just to make sure the voltages are as they were calculated (or very close), and to see the difference between PSUD and real world measurements.
JD
I'll be bread boarding the power supplies first, using some bigazz 300 watt load resistors sized to what the actual load should be, just to make sure the voltages are as they were calculated (or very close), and to see the difference between PSUD and real world measurements.
JD
