tennisballg said:
And what do you think about THIS MOS IRL3713 (3 pair)!!
Only one toroidal core is not sufficient???
The 44932 core can supply about 2600w at 50khz!!!
look here Datasheet at page 8...
A bias (for rhe car amplifiers) is about 100mA...
Well, the toroid may be able to handle it, but the size of the toroid makes it tough to keep all of the windings exposed. there's more room to breathe if you use two smaller ones, and probably cheaper as well. For example, my red toroid from a page or two back is #44925 off of your mag-inc page, with a similar i.d. and o.d. to yours. I am using it for a single 500W supply (2x250@2 ohm, bridged makes 500W@ 4 ohm). So I am making 40V rails, I re-wound the secondaries to fit snugly in between the primaries rather than just winding over the top, and I almost had to start layering the windings, not good for heat dissipation. You're probably counting on 55-60V rails to put 500W into 2 ohms, so that's even more windings, and probably more parallel strands as well.
Those mosfets have great characteristics, but i'm still not too sure about such a low Vdss, because I know that there is always some overshoot when the MOSFET turns on. From my experiences, long story short, you could easily have 60V turn on spikes when or before you initially tweak the dead time settings, transformer windings, snubber circuits, etc. Having the Vdss so close to the operating voltage might a)blow a few sets while you're first building b) won't allow you to clearly see any turn on spikes with your scope...but hopefully someone will back me up on this because I am not so sure of my opinion. IMO, it's a good idea to have higher ratings rather than cutting it to the edge.
Those mosfets have great characteristics, but i'm still not too sure about such a low Vdss, because I know that there is always some overshoot when the MOSFET turns on. From my experiences, long story short, you could easily have 60V turn on spikes when or before you initially tweak the dead time settings, transformer windings, snubber circuits, etc. Having the Vdss so close to the operating voltage might a)blow a few sets while you're first building b) won't allow you to clearly see any turn on spikes with your scope...but hopefully someone will back me up on this because I am not so sure of my opinion. IMO, it's a good idea to have higher ratings rather than cutting it to the edge.
Nearly all Car-Audio amplifiers use 50-60V MOSFETs for push-pull power supply since this blocking capability is required for reability reasons
A coulpe of years ago, there was a manufacturer called Rödek using a pair of 40V MOSFETs per transformer and two transformers on their power supplies. Guess?. I had to repair a dozen or so of these amplifiers with blown 40V transistors. Blown output devices were also common since single TIP 35C/36C output devices used in CFP and powered by +-30V supplies driving 2 ohms of inductive [subwoofer] loads at high temperatures are not a reliable approach. After a year, these series of amplifiers were discontinued and replaced by a different more reliable design
A coulpe of years ago, there was a manufacturer called Rödek using a pair of 40V MOSFETs per transformer and two transformers on their power supplies. Guess?. I had to repair a dozen or so of these amplifiers with blown 40V transistors. Blown output devices were also common since single TIP 35C/36C output devices used in CFP and powered by +-30V supplies driving 2 ohms of inductive [subwoofer] loads at high temperatures are not a reliable approach. After a year, these series of amplifiers were discontinued and replaced by a different more reliable design
ok, where should I start?
anyone willing to "guesstimate" how much power each core could handle?
here are some measurements........
toroid #1 (grey)
inner dia:44mm
outer dia:69mm
thickness (height):15mm
toroid#2 (black, unpainted)
inner dia:38mm
outer dia:64mm
height:25mm
toroid#3 (green, red on one side)
inner dia:35mm
outer dia:57mm
height:14mm
toroid#4 (grey, probably useless for a car amp but still worth considering for a small SMPS)
inner dia:12mm
outer dia: 21mm
height: 7mm
I'm guessing
#1 could be good for about 600-800 watts,
#2 can be up to a kW if properly built/designed,
#3 could be up to around 400 watts or so, and
#4 can be for a preamp supply or up to an ampere or two of step down/step up supply.
and I do have twelve dual 16A 200V ultrafast diodes,
ten 80V 95A MOSFETs, (IFR1312)
and eleven 55V 110A MOSFETs. (IRF3205)
sounds like I could make monster SMPS's.
and I need a SMPS that can put out a regulated 15V, 2-3A max output from 10V to 15V input. it has to be very simple and can be made to fit in a board area of about 4-5 sq inches max. Isolated output is not required but a plus.
anyone willing to "guesstimate" how much power each core could handle?
here are some measurements........
toroid #1 (grey)
inner dia:44mm
outer dia:69mm
thickness (height):15mm
toroid#2 (black, unpainted)
inner dia:38mm
outer dia:64mm
height:25mm
toroid#3 (green, red on one side)
inner dia:35mm
outer dia:57mm
height:14mm
toroid#4 (grey, probably useless for a car amp but still worth considering for a small SMPS)
inner dia:12mm
outer dia: 21mm
height: 7mm
I'm guessing
#1 could be good for about 600-800 watts,
#2 can be up to a kW if properly built/designed,
#3 could be up to around 400 watts or so, and
#4 can be for a preamp supply or up to an ampere or two of step down/step up supply.
and I do have twelve dual 16A 200V ultrafast diodes,
ten 80V 95A MOSFETs, (IFR1312)
and eleven 55V 110A MOSFETs. (IRF3205)
sounds like I could make monster SMPS's.
and I need a SMPS that can put out a regulated 15V, 2-3A max output from 10V to 15V input. it has to be very simple and can be made to fit in a board area of about 4-5 sq inches max. Isolated output is not required but a plus.
To reduce the spike i've found this: TOSHIBA
What do you think about this???
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
What do you think about this???
You should be able to get your supply from having spikes much above 30V by just playing with the ocillator timings and making sure that your transformer is wound properly. I think that if you NEED to use these, then there is a big problem with your SMPS, and they are just masking the problem, if you can even get these to work how you want. Trust me, I've done a lot of whining about MY spikes on the forums in the past, and after many hours at the bench I found that it's really not so hard to get it right, you just have to know what you're doing.
Even so, the beads in the top pic could be a good idea for optimization purposes, but get your power supply working correctly first, then the beads will only make it better.
Just go for the higher voltage rating, they'll last longer anyway. Imagine a dam that is constantly filled to the top, there's a lot more pressure and stress on it than one only half full. Well, your MOSFETs have to "dam" the voltage in the same way.
Even so, the beads in the top pic could be a good idea for optimization purposes, but get your power supply working correctly first, then the beads will only make it better.
Just go for the higher voltage rating, they'll last longer anyway. Imagine a dam that is constantly filled to the top, there's a lot more pressure and stress on it than one only half full. Well, your MOSFETs have to "dam" the voltage in the same way.
metal :
You need to know the number of 12V winding turns and the oscillator frequency to calculate the required number of turns for the voltage and frequency of your application
Te number of turns on each 12V winding is usually specefied for a 48V-pp square wave [+24V / -24V] operating at half the TLT494 oscillator frequency
For 12V car SMPS applications you will be appliying +-16V [32Vpp] to each winding of the transformer so you have to scale down the number of turns proportionally
Also, you may prefer to use a different oscillator frequency than the usual 70Khz found on PC SMPS [35Khz effective], so you have to scale up or down the number of turns knowing that decreasing frequency requires increasing proportionally the number of turns and vice-versa
12V secondary windings usually have 7 turns each. Scaling this for car-smps operation we get 4.666 turns. Scaling again for 40Khz operation [80Khz oscillator] we get 4 turns per winding
You have to do the calculations for your transformers since values may be different
You need to know the number of 12V winding turns and the oscillator frequency to calculate the required number of turns for the voltage and frequency of your application
Te number of turns on each 12V winding is usually specefied for a 48V-pp square wave [+24V / -24V] operating at half the TLT494 oscillator frequency
For 12V car SMPS applications you will be appliying +-16V [32Vpp] to each winding of the transformer so you have to scale down the number of turns proportionally
Also, you may prefer to use a different oscillator frequency than the usual 70Khz found on PC SMPS [35Khz effective], so you have to scale up or down the number of turns knowing that decreasing frequency requires increasing proportionally the number of turns and vice-versa
12V secondary windings usually have 7 turns each. Scaling this for car-smps operation we get 4.666 turns. Scaling again for 40Khz operation [80Khz oscillator] we get 4 turns per winding
You have to do the calculations for your transformers since values may be different
metal :
Power handling is determined by core dissipation and magnet wire dissipation. In your case core dissipation would be negligible so power handling depends only on how much current can flow through your windings without excessive heating
You should find out optimum magnet wire gauge and number of wires for each winding so that : All windings opetare at similar temperatures [inner windings would have to dissipate a bit less than outer windings], all windings must fit inside the winding area and maximum magnet wire temperature should be reasonable
I use to find out wire number an gauge empirically
dissipation in a winding = Irms^2 * 0.017 * L / S [where 'L' is lenght in meters and 'S' is total wire cross-section in milimeters^2]
For these core sizes I wouldn't allow more than 2W dissipation for each winding [4 in total]
Power handling is determined by core dissipation and magnet wire dissipation. In your case core dissipation would be negligible so power handling depends only on how much current can flow through your windings without excessive heating
You should find out optimum magnet wire gauge and number of wires for each winding so that : All windings opetare at similar temperatures [inner windings would have to dissipate a bit less than outer windings], all windings must fit inside the winding area and maximum magnet wire temperature should be reasonable
I use to find out wire number an gauge empirically
dissipation in a winding = Irms^2 * 0.017 * L / S [where 'L' is lenght in meters and 'S' is total wire cross-section in milimeters^2]
For these core sizes I wouldn't allow more than 2W dissipation for each winding [4 in total]
Hi,
has anyone ever considered balun type transmission line transformer instead of toroid. I think these type transformers are a little easier to wind and have very little leakage inductance. Primary is usually made of copper tube onto which toroids are stacked, secondary is then wound through copper tubes. Because of large ferrite cross section only one turn (tube) primary is necessary, for push pull one can use even half turn by using fork type winding (right side of picture).
Best regards,
Jaka Racman
has anyone ever considered balun type transmission line transformer instead of toroid. I think these type transformers are a little easier to wind and have very little leakage inductance. Primary is usually made of copper tube onto which toroids are stacked, secondary is then wound through copper tubes. Because of large ferrite cross section only one turn (tube) primary is necessary, for push pull one can use even half turn by using fork type winding (right side of picture).
Best regards,
Jaka Racman
Attachments
Car amplifiers
Eva: Would you be interested in designing car amplifiers?
I would like to manufacture (yeah I know it's a low profit market) car amplifiers.
Also I've been looking for a monster smps, just in case you had one
lying around gathering dust
Drop me a mail if you're interested: bruce.bushby@wol.co.za
Eva: Would you be interested in designing car amplifiers?
I would like to manufacture (yeah I know it's a low profit market) car amplifiers.
Also I've been looking for a monster smps, just in case you had one
lying around gathering dust
Drop me a mail if you're interested: bruce.bushby@wol.co.za
Aha it's very simple. I used the glued double toroid (rectangular profile) in the my DIY car audio class D amp .5Kw, 84% total efficiency (93-95% are amp efficiency, and 89-87% SMPS) and cloned it about 10 times- never seen any problems. Stacked toroids with ellipsoid profile, i've seen in the MTX 500D car audio class D amplifier in the output filer coil, but for transformer these will be worse vs rectangular, however not so much. Actually, art component is pretty poor in the car audio SMPS (<1Kw) theme, and i'm impressed by posts quantity&quality on this thread
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.