| john65b |
Alright, so I picked up this old receiver on ebay cheap for the torroid and chassis..nice buy btw...
Anyway, got a torroid question. The label says 40V 0V (0V) 40V and should be around 400VA. I want to try to use it for a LM3886 and get 20VAC out of it...read on...
Now here the tricky part...all four wires (blue black black blue) all have continuity and thus are the same winding. I get 40V between either of of the blue and black wires.
If I connect the 40V and 0V to the AC side of a single rectifier it should be the same as if I had 20V and 20V on the AC side of the rectifier, correct?
I tried it on a rectifier and it shows (after the smoothing caps) about 28VDC, the same as if I had 20V and 20V....am I missing something?
What would be my virtual Ground if wired this way for a dual 20V power supply? Or is this the reason why this torroid cannot be used? |
|
|
| john65b |
When I measured DC voltage across smoothing caps, I think I connected the Chassis ground as GRND between the + and - rails ....
I did this a few months back and I cannot remember exactly how I did it... |
|
|
| jerishi |
| I'm wondering if you're playing with a half-wave versus a full wave rectifier. Are you trying to reduce transformers output to please the 3886 requirements? F.W.I.W.... just got home(am a carpenter) and probably have little to offer here, none-the-less.., I'm also curious how you might use a salvaged transformer, with different specs than ones' specified (for instance) with the 3886 chip. |
|
|
| john65b |
I am using a full wave rectifier. Yes, I would like to use this transformer if I can. At 40V secondaries, it is too much for the LM3886 straight up, but when connected as I have outlined...
I would like to dig up the rectifier and caps to try it again, but my basement is getting refinished and I can't get to it at the moment. I do have the torrid within arms reach...It came from a KLH KL2400. These receivers are quite ubiquitous on ebay - very cheap, and they are supposedly Chinese "fakes", as I read, KLH never made a model KL2400...
I took apart the receiver board and it was very cheap, but the torroid and soft start board looked reuseable...
Go figure... |
|
|
| john65b |
Or how 'bout this...
Is there a way I can drop the 40V 0V 40V to a 24V 0V 24V using a couple of LM338 Voltage regulators BEFORE the smoothing caps? The torroid is 5A, so the LM338 is good for 5A.
Between the Rectifier and smoothing caps, the voltage should be lower than 40VAC rectifier input, correct? Putting it after the smoothing caps is no good, cause the 56VDC smoothing caps output is too high for the LM338, 40V is max... My guess is about 37VDC at this point, which would be perfect. Just wanna to drop it 10V - 12V.
I also could put the LM338 before the rectifier (on transformer output), as the LM338 regulates DC and AC.
But I remember some saying that you cannot use the dual LM338 regualated system on a center tap transformer, only a split winding. Is this true?
Don't see why I can't just use the centertap as grounds (0V) for both LM338 rails...
I have used the LM317's many times to regulate 20V laptop power supplies down to 13V for my Tripath amps, and think this should work here too...
Anyone? |
|
|
| Nordic |
I think those are only good for 1.5A, you will need a couple of times that....
I think wanting to build a gainclone is one thing, wasteing a perfectly good transformer that you could use for something more substantial is another... I mean it in the nicest way though... its just trannies are expensive.... |
|
|
| john65b |
The LM317 is good for 1.5A
The LM350 is good for 3A
The LM338 is good for 5A.
Correct me if I am wrong.
I have two of these 40V 0V 40V, and would like to use one of them on the "special" LM3886 I am building.
I planned on using one for the UCD180 I had, but was pushing it with 40V too, so I bought another toroid at 30V 0V 30V.
Just want to use what I have as opposed to buying another trannie...that's all... |
|
|
| Nordic |
Roger that....
Are those transformers 40v AC? If it is, you will have 56.6V - diode drop.
driving 8 ohm speakers at peak with 40V would require 5A before you start adding extra to deal with real world efficiency..
4 Ohm loads are almost not an option for you...
I'm not too familiar with the other chipamp brands, but still wonder if there isn't somthing that could use those rails with less compromises. |
|
|
| AndrewT |
Hi,
my short answer is buy the correct transformer.
The building of a high power regulator (that sounds good) will be more complicated than building a good sounding chipamp.
This extra complication and cost (and risk to the sound quality) completely negates the reason for going chipamp in the first place.
Either, build a discrete amp that suits 2*40Vac or buy 2*25Vac (or 25+25Vac).
BTW 40Vac can support 120W to 150W into 8r. It will need to be about 160VA to 250VA to do one mono channel. If you regulate it down to a suitable voltage for a chipamp, then the losses in the regulator and the amplifier result in less output power but still require the same VA rating for the transformer i.e 160VA to 250VA for 50W to 60W output.
Double these VA ratings for stereo and double again for 4ohm duty. |
|
|
| JesseG |
From John65b
| quote: | | Just want to use what I have as opposed to buying another trannie...that's all... |
A couple of points
One
You can (use the trafo you have) - If you are deterimed to run at +/- 20V, you just need to get hold of a couple of 8 - 10 mH chokes - one for each side of the PS - and use LC input filter instead of RC. THis will have 3 effects... 1) it will bring down the voltage of the trafo secondaries by about 25%, 2) it will do a way better job of filtering the ripple and 3) it will provide much better current stabilization.
Two
However, you don't really need to. I am running a LM4766 chipamp in native stereo mode (not bridged or paralleled) off a 38.5-0-38.5 MONSTER conventional trafo. Because of the really high PSRR of National's audio chips (God bless 'em) mine is a simple bridge>really big caps(12,000uF)>bypass caps(0.01uF) and that's all. DC Voltage to the chip is +/- 34V. And, it sounds wonderful! A couple of very knowledegable hifi buddies (both tube nuts) have said: 'that's the best sounding chipamp I've ever heard'. This compared to Peter D's 4780s and a host of 3886s.
Seems that most people don't read the National spec sheet very well - they specify power input range of 20 to 80 volts. That is
+/- 10 V to +/-40V. Since the basic technology for all of the National Overture chips is the same, the same electrical rules and parameters apply, except for the current needed to drive them to performance.
I should add that my LM4766 rig barely warms up the heatsink running at half-throttle for hours at a time.
So, I am a big fan of running these devices at far higher voltages than seems popular. Considering that the chip costs about $5, build one a try it. It won't blow up and it will sound much better.
Lucky you is what I say. If you have any spare 40-0-40 torroids, I'll take a few ;) |
|
|
| john65b |
JesseG -->
How can you get +/-34 at your chip with a 38.5V 0V 38.5V trannie? After your 12,000uF caps you should be at approx +/-54VDC, right?
Now what am I missing here? Do I need to go back to EE101 again?
And the LM4780 max voltage was 84V (+/-42V), which is a tad more than the LM3886...not that 2V matter much here...
I have thought of taking the 40V 0 40V Tranformer connections right after the rectifier (should be a bit under 40VDC here - no smoothing caps) direct to the two 2200uf caps right on the chip (Cs - just like how the real gaincard has it - no smoothing caps)...this supply voltage would be in compliance with the LM3886 datasheet...
But I really don't wanna blow the chip, tho...kinda gettin attached to it...its my first foray into soldering all components right on the chip (exact component values of the gaincard - except no electrolitic input cap - what were they thinking...I have a 2.2uF polypropolyne cap here)
Anyone? |
|
|
| AndrewT |
Hi,
the AC voltage from your transformer is rectified by the diode bridge.
The capacitors after the rectifier TRY to charge up to the peak voltage of the AC waveform. That's where the square root of 2 comes from.
40Vac has a peak voltage just about 58V. The bridge rectifier drops about 1.4v leaving about 56Vdc across the smoothing capacitors. As mains voltage fluctuates the transformer output will also fluctuate (in the UK our tolerance is -10% and +6%)
The transformer will also produce more voltage when on a light load. A small transformer is worse here, compare the regulation, it varies from 4% for 1000VA to about 30% for 10VA.
Now you need a transformer and rectifier and smoothing capacitor to make a PSU so you are lumbered with the root 2 times Vac rule.
The maximum voltage of the chipamps is relatively low but you would expect them to perform when the incoming voltage is normal and at the two extremes I quoted. You have to ensure that after taking regulation and maximum mains that you do not destroy your chipamp.
There is one other design decision to take account of. Load impedance.
The chipamp datasheet will specify the maximum voltage for each load and you will see that as load impedance falls that the chipamp maximum voltage also falls.
Many speakers are 8ohm but some are 6ohm and a few are 4 to 8ohm due to using a low sensitivity 4ohm bass unit in conjunction with an 8ohm treble unit. Be carefull to use a PSU supply voltage that suits the load YOU will be hanging on the end.
There is a very expensive alternative, the choke regulated PSU, but it has a very major drawback, namely, if the current draw drops below a design limit the output voltage rises up towards the waveform peak. It works very well with ClassA designs that have a very high quiescent current and tends to work better when the maximum current demand is very low, ideal for Tube designs. Forget it for solid state. |
|
|
| Nordic |
Andrew any idea how to bring the dropout voltage of the actual chip into the equation too?
It seems to specify 2 and 3V, but I can't figure whicc applies when, or how to calculate mamimum supply ( not that I'd ever want to drive one at max volts).
the datasheet note says:
Note 12: The output dropout voltage is defined as the supply voltage minus the clipping voltage. Refer to the Clipping Voltage vs. Supply Voltage graph in the
Typical Performance Characteristics section. |
|
|
| AndrewT |
Hi,
if one were designing an amplifier for a customer and had to meet a particular power spec. then a first guess would be based on the various voltage drops that occur in the system from transformer to output.
Based on that one builds the prototype and tests it.
For your build the loss through the chip is unimportant. It simply makes power predictions more difficult. The other voltage drops will be much more significant.
Now, if you were to produce a model of maximum power vs all the other variables then it would be instructive, but is it any more useful when most of the design decisions are out of your hands and made by the chip designer? |
|
|
| john65b |
Thank you Andrew.
You seem most knowledgable in power supplies, as this may be your bread and butter. I work with many EE's, but they do not seem to be able to remember and be able to help much of what I ask - they are mostly involved with control systems, COGEN Units, and high voltage power transformers.
Anyway, if your read the beggining of my post, you will see the first question I raised about hooking the 40V 0V to the rectifier to get the same result as the 20V 20V. Can you comment on if this is possible?
Also, as JesseG mentioned, is there truely a way to get a 38.5V 0 38.5V trannie rectified/smoothed with a 12,000uF caps to be able to get +/-34VDC with nothing fancy inbetween (LC filter or choke)? From your post, we are in agreement and should not be possible.
And the size of the caps on the downstream side of the rectifier does not matter as far as far as outputting peak voltage - in other words, post rectifier to 12000uF or 1000uf, they both will act as smoothing caps and should output peak voltage (x 1.414)...not less with smaller caps.
Sorry if these questions are too basic, but I am sure there are a few of us out here that could use the education.
Thanks |
|
|
| Nordic |
The caps are charged at 100 or 120 times per second depending on whether you have 50 or 60Hz power...
The size and quality of the cap will influence how quickly they will charge. All that said, the moment the amp tries to draw more current than what the PSU is able to store and supply the voltage will drop... |
|
|
| AndrewT |
Hi John,
I might be a techy teacher but my bread and butter was Civil Engineering.
This simple answer assumes no regulators.
+-34Vdc from 2*38.5Vac is not possible. Ignore that poster until he can prove his assertion.
Have you considered rewinding the secondary to get your lower voltage?
Nordic,
your simplistic view on voltage available from the PSU is misleading.
Think about the emf available from the source (transformer) and then the string of resistances from source to load and all the way back to the source.
Now apply ohm's law and calculate the voltage drops when no current is flowing. Increment the current and again calculate the voltage drops. Subtract the voltage drops from the emf and the remainder is the available voltage to drive the amplifier.
You could get very sophisicated and draw a graph of current vs voltage drops and (emf-voltage drops) to see how the available voltage varies with current draw.
You will see that voltage available starts to fall as soon as even a tiny current is drawn by the load (the load includes the amplifier quiescent current). There is no point at which the voltage holds steady and then suddenly starts to fall when a threshold is reached, unless your threshold is set at zero.
I just lost my addendum on ripple voltage and the charging pulse current voltage loss. But you can work those ones out. |
|
|
| john65b |
Andrew,
That's what I thought about the previous post. Looks like I don't need to re-review my EE classes after all (I am an ME working 25 years in the Petroleum refining industry). Great hobby here BTW...
I should just get the correct transformer instead of try to get this one to work...
I thought about the Unwinding of the toroid, but I would need to remove the nice plastic wrapping...and I would also assume that unwinding the wraps on the toroid to reduce voltage would also reduce VA rating? If I had to derate (unwrap) the windings from 40V to 25V (is that almost 1/2 the winding?), that would mean I would have about 250VA? Not that it matters that much, since 400 is way more than I need, and 250 is fine...
As you can see from my questions, I have never unwound a torroid before, so a bit leary. Is it that cumbersome? Will I need to re-wrap it? If so, can I re-wrap with packaging tape? Common sense says just unwrap the 40V windings and leave the centertap windings alone?? Do I just unwind, cut wire, scrape off the coating and test the leads to the centertap until I get to 25V?
Thanks in advance for your help |
|
|
| poobah |
John,
Unwinding is not that big of a deal.
Your VA rating will remain intact (in fact, it will increase slightly).
You do not need to replace the tape... if you do just buy some polyester (mylar) or kapton.
Ideally the remaining winding should be redistrbuted evenly. Forget about it... you will get a slight increase in leakage inductance and flux... neither of which will matter for a PSU.
Both halves of the secondary are generally wound at the same time... side by side... "bifilar". It is the connections that creates the center tap... you'll see this as you remove the tape. You will have to recreate the center tap when you're through... very simple... make a sketch BEFORE you take things apart. The center tap is created by joining the START of one winding with END of the other... pretty hard to screw up, but it is entirely possible to SHORT the a winding by connecting its own start to its own end. After you break connections, ohm through and paint or label one of the windings.
;)
P.S. Don't unwind all the way to your desired voltage you should be about 5% high to allow for voltage drops that will occur when the trans is loaded. |
|
|
| Nordic |
| Andrew, I stand corrected. |
|
|
| JesseG |
From AndrewT| quote: | | +-34Vdc from 2*38.5Vac is not possible. Ignore that poster until he can prove his assertion. |
C'mon guys - lighten up a little:xeye:
We are not ADDING the two 38.5 together, this is a BIPOLAR supply. Each side is 38.5v.
Taking one side as a simple example:
38.5v AC secondary is peak voltage. into a bridge (full wave) rectifier produces RMS DC of 33.87V, when the 3% ripple is smoothed by the caps, the result is about 34.2V DC.
Respectfully, I think you fellows are confusing peak AC and RMS.
John65b - would you like a schematic or photos? |
|
|
| poobah |
John is probably stating an RMS voltage as is the norm...
:rolleyes: |
|
|
| john65b |
JesseG,
Please post pics and schematic anyway...it would be good to see you system regardless of what the nomenclature diffences are...
I will most likely unwind the trannie as Poobah instructed (Thanks Poobah).
But just for S&G, why couldn't wire up 40V & 0V to a rectifier to get same results as 20V 20V? Understand that I won't do it, but my curiosity won't let this go... |
|
|
| poobah |
John,
It can be done... but you have to create a virtual ground. You place 2 large caps in series across your rectified output, then you connect the center point between the caps to ground. It has problems... low valued bleeders across the caps for one. |
|
|
| john65b |
So this can work? Wow... remember, I have two centertaps (black wires), and two 40V leads (blue wires)... and all four wires, (blue, black, black, blue) are the same winding...voltage across any blue to any black is 40V....blue to blue is 80V, etc.
And the virtual ground (0V) between the pair of smoothing caps is made by a single connection to chassis ground and nothing else...
Don't understand the bleeder resistor issue...attached is a schematic...
So if anyone has a high voltage centertap trannie (CT needs to be split), they can do this to get something that is identical to two separate trannies at 1/2 the voltage for a dual mono power supply?? |
|
|
| poobah |
Yep...
Place bleeder resistors in parallel with each cap such that they burn 3 to 5 Watts. The resistors should be 1 or 2% as they will ultimately determine the balance between rail voltages.
Me... I would unwind the the tranny. But I am unclear as to your tranny wiring. Can you make a picture?
What you have drawn presumes 2 isolated windings, but you describe all 4 wires as being "the same winding".
:xeye: |
|
|
| john65b |
I get 0 ohms between both black wires - there is only one winding. The two centertaps are the same.
In other words, if I had a centertap tranny (three wires) and I took the centertap wire and split it, so there are now four wires, this is what I have
I do not know why the manufacturer of this torroid split teh centertap, but it is the same wire.
Knowing this, would this still work? I have tried it a while back I remember it working.... |
|
|
| AndrewT |
Hi,
although I said rewind the toroid, you do not need to unwind all of the secondaries.
Find a point about midway along each winding. Cut the two windings at about the mid-points.
Measure the voltage in each of the four new shorter windings.
Add turns to each to bring them up to the desired voltage.
Now you can run four windings into two recifiers to give two dual polarity supplies, one for each channel or you can parallel two windings to double the current capacity and end up with one dual polarity supply.
Buy some enamelled copper wire of the same diameter as you have on the secondaries.
By adding turns the total VA rating is not reduced. Some would argue that the VA rating is slightly increased.
If you reduce the turns on the secondary then the VA rating is reduced in proportion to the reduced voltage you end up with. i.e. the current stays the same.
Do you follow all that? |
|
|
| poobah |
VA rating is perhaps not a good way to look at this. VA is really is power rating, based on an optimal design, that proiduces a given temperature rise. That is 50% iron loss, and 50% copper loss.
Reducing turns could be viewed as reducing VA, because of the obvious loss in voltage. But also winding resistance decreases, in turn heat. The flip side, increasing turns, does not increase VA. You have the same core and iron losses, but more winding resistance. Bottom line is that winding currents should be considered priority over VA. Regardless, removing turns, can be done safely without concern, as long as the winding current is not proportionately increased by presuming the VA remains equal.
John,
The decision to unwind should really be made based on whether the secondaries are wound side by side (good thing)... or one on top of the other (bad thing). To implement the drawing you have would require that the center taps be split anyway. How about pulling away some tape and checking the lay of the windings?
;) |
|
|
| john65b |
Andrew, I think I understand...I will look at the winding tonight if possible.
I suspect both black centertaps are connected at the same point on the winding. If I cut the winding at that point, I create two separate windings, and thus a dual secondary. No unwinding, or rewinding would be nessessary...see my attachment...
From there it becomes as simple as connecting the power supply with dual secondaries. |
|
|
| AndrewT |
| quote: | | as long as the winding current is not proportionately increased by presuming the VA remains equal. |
=| quote: | | If you reduce the turns on the secondary then the VA rating is reduced in proportion to the reduced voltage you end up with. i.e. the current stays the same. |
|
|
|
| poobah |
More likely you have 2 windings pulled side by side. The virtual ground technique is less than ideal... avoid it if you can.
;)
Andrew,
Therein lies the confusion, or lack thereof... as you remove turns, you shorten the winding and decrease its reisistance. Depending on the intitial current density in the winding; some increase in winding current is acceptable. My point being is that VA rating, in an optimal design, is a function of surface area and allowed temperature rise. When turns are altered, optimal goes out the window... as should VA ratings.
|
|
|
| john65b |
Would two windings pulled side by side indicate continuity between the two black 0V leads? I also have continuity between the two blue 40V leads....
I currently have one 40V 0V 40V 400VA toroid up for trade for a 22V 0V 22V 300VA or better on diyaudio. This is my first option. Next is to unwind/modify the toroid, and third is to try something crazy like I have drawn up. |
|
|
| john65b |
I guess another option is a step down trannie in front of my 40V...
120V -> 60V which would reduce my 40V 0V 40V to a 20V 0V 20V...
Nah, don't wanna throw more money at this issue - someone trade me!!! |
|
|
| poobah |
| No... you can't tell how it is layed up with out pulling off some tape. The blues will give continuity... although you should be able to see 2 the resistance from blue to blue as blue to black. |
|
|
| john65b |
| OK, I took some pics of the toroid with Plastic wrap removed - red primaries, all others secondaries... |
|
|
| john65b |
| And here is a pic of the blue and black windings - you can see the two blacks (0V) are tied to the same point at centertap...a dual centertap - why? I dunno... |
|
|
| poobah |
OK... now you'll see how a center is created from 2 coiuls wound
at the same time.
Look closely and you will see one blue and one black that start/end at the same point. The other black and blue will lead off to another point. All you have to do is find the pairing and unwind both coils at the same time. Separate the coils. take some measurements... derive a volts per turn and go to town.
Remember to leave you voltage about 5-10% high. of target.
BTW... what are all the other windings? |
|
|
| john65b |
| Why can't I split the single winding at the black (0v) connection point to get two separate independant winds like this... |
|
|
| john65b |
| Sorry, I mean this... |
|
|
| john65b |
The other winds are
5.2V 0V (Dark Green, Green)
and
16V 0V 16V (Brown, Black Brown)
There was a preamp section in this receiver for the 16V... |
|
|
| poobah |
Oh sorry... of course you cut all the wires! If you look at your first photo, it is pretty clear how the wires are paired.
Remeber the center tap is formed by the start of one winding and the finish of the other...
Just keep track of them. What you want is 2 coils each producing your target voltage.
Don't do the the virtual ground thing unless you're forced to...
Send a PM if you please! |
|
|
| john65b |
OK....I will make sure the correct black wire pairs with the correct blue one which is on the same winding.
Now instead of unwinding a trannie of double the voltage you need, you could split them at the centertap and get two separate windings of the voltage you want....and have 1/2 the total VA rating on each secondary winding...
Is this correct? |
|
|
| poobah |
NO... not quite. You have 2 windings there, side by side. These are joined to make 40-0-40.
You could simply parallel the 2 windings to make a high current 40 VAC winding and then the virtual ground technique.
What is you target AC voltage? |
|
|
| poobah |
If this is for a 50 watt chip amp you really want about 28-30 VAC... so you're only pulling away about 25% of the winding.
;) |
|
|
| Bearman |
John,
With the gas prices in Utah, I hope that it somehow translates back down to you :) The trickle down effect...you know??? :)
Get a good transformer that suits your needs. Here's a good source at a good price.
http://stores.ebay.com/antek-inc
I have bought several and they are good transformers.
If your around SLC.....Backer's Bakery........ Me.... |
|
|
| john65b |
I am actiually targeting +/- 20V rails since it would be easiest...One modded toroid for a special "Gaincard style" LM3886 and another modded toroid targeted at +/- 20V rails too for an Amp 5..
This is what I was thinking... |
|
|
| john65b |
Bearman,
Gas prices in Utah? I MAKE more than 25% of the Gas and Diesel in Utah!! I work for the biggest Petroleum refiner in the valley,
Gas and Diesel prices are finally going down here....but don't get used to it...one hurricane, or North Korean Missle test, or cheap shot from Hugo will make the prices skyrocket again...no tellin where they will end up...
Thise Antec torroids are great, but pricey. I usually buy from Steve at Apexjr.com |
|
|
| poobah |
Oops... sorry John, disregard post my last post... big error.
20-22 rail voltage is right
Perhaps the best thing to do is to parallel the windings and use a cap stack to make a virtual ground.
Or use the windings seperately and use 4 caps as you have drawn... this will still give you about 28 +/- rails though. That's ok though gives headroom for transients... common sense and the built in temp limit provide the rest.
Are you interested in keeping the 5 and 16 volts windings and are they on top... final layers?
;) |
|
|
| john65b |
The 5V winding will power the power LED, and the 16V 0V 16V windings, well, I donno. I may just put in a TDA1543 DAC I have (regulated dwn to 12V via LM317)
So the Amp will have Optical Digital input -> DAC -> Stepped Attenuator -> LM3886 Amp-> Speakers....
I think both windings are on top - yes... |
|
|
| Bearman |
John,
With the gas prices in Utah, I hope that it somehow translates back down to you :) The trickle down effect...you know??? :)
Get a good transformer that suits your needs. Here's a good source at a good price.
http://stores.ebay.com/antek-inc
I have bought several and they are good transformers.
If your around SLC.....Backer's Bakery........ Me.... |
|
|
| john65b |
So where's Backer bakery? I am on the North end of Town by the Gateway...
I think I solved my Toroid issue, and will not be buying a new one...I still have one more I may trade if the correct one comes along...
Now instead of having a higher voltage toroid I can't use, I have a correct dual voltage toroid I can use... A little inginuity I guess....Viva la DIYAudio Spirit!! |
|
|
| Bearman |
Good for you John. Power is always a little spooky for me, so I choose not to mess with things too much on that end. ZAP!
South Temple, between 4-5th East, South Side of the street. Been there 60 years.
Here's to DIY. |
|
|
| john65b |
| OK, I correct myself once again...these two are equivalent... |
|
|
| AndrewT |
Sorry John,
but they are not.
One is a centre tapped 40-0-40 and the other is a twin secondary 0-20, 0-20.
You can join the twin secondaries to give the equivalent centre tapped form, but it will only be a 20-0-20 NOT 40-0-40.
Have another look. |
|
|
| john65b |
Yes!! This is exactly what I was looking for from the beginning of this post.
Turning a 40V 0V 40V into a 20V 0V 20V toroid I want for a "Special" Gainclone and an Amp5 -
I know I have to join the post-rectified/smoothed pos and neg output to make the new 0v plane... |
|
|
| BrianDonegan |
Nope. This is what you will have...
The tie between the two windings just creates a common point (0V). Seperating the two windings does not affect each windings voltage, it just removes the common voltage point.
So, looking closely at the windings, do they appear to be layered over each other? The other low voltage windings may make it difficult, as they are probably right over some of the stuff you want to unwind.
If you go for the virtual ground approach, DO NOT tie the virtual ground to your chassis ground. If you want to know why, keep them seperate and measure the voltage between them... |
|
|
| AndrewT |
Hi John,
I said| quote: | | One is a centre tapped 40-0-40 and the other is a twin secondary 0-20, 0-20. You can join the twin secondaries to give the equivalent centre tapped form, but it will only be a 20-0-20 NOT 40-0-40. |
What do you mean by saying what appears contradictory| quote: | | Turning a 40V 0V 40V into a 20V 0V 20V toroid |
It cannot be done easily. |
|
|
| john65b |
Hello Brian..
I noticed earlier that you also have been getting these KLH2400 receivers and reusing the toroids...this is one of those toroids and I want to use it as a dual 20V winding tranny or a 40VCT single secondary tranny.... have you read up on this post?
Will this work? Notice I took off the chassis ground... |
|
|
| BrianDonegan |
| I think that would give you +-56V rails, not 28V. |
|
|
| john65b |
If you have a 20V - 20V toroid on the AC side of a rectifier, you get the same as a 40V - 0V on the AC side of a rectifier, which is +/- 28VDC.
I tried it... |
|
|
| AndrewT |
NO!
If you use 20-0-20 then you get +-28Vdc because the 0V gives a reference.
The 40-0 gives 56Vdc and there is no 0V reference to split the 56Vdc in half.
You have been advised (by others) to avoid using a virtual earth for the power amp.
I now concur with this advice.
The reason I arrive here is that you would require a high current power amp to generate the virtual earth and having generated it artificially there is no guarntee that it will sound nice or even work stabily. A lot of development time and cost to avoid rewinding or buying a transformer. |
|
|
| john65b |
| Very well. I will leave it alone. |
|
|
| john65b |
Yes, after reading AndrewT's response again, I see my mistake. The 40V-0V secondary has no center reference to get the values I am seeking.
To total voltage across the AC pins of a rectifier may be the same from 20V 20V (centertapped case) versus a 40V 0V (standalone secondary case), but output voltages are referenced from different points...
Thanks again. |
|
|
|
