How to decide what smps I need for project hybrid amp.

kees54

kees57
Hi all


In have tryed to sim some kinds of smps, forward, hald bridge.

However I think models UC3845 for example are not so well.


I need 15 volts 10 amps, 300 volts 100mA, 2 x 120 volts 250mA, 2 x 65 volts 15 amps.

Some people say make two smps, one for power 2 x 65 volts and one for the other voltages.

I did try the last, thus alone the tube voltages, however getting feedback working and such in ltspice did not work nice, I had a UB3875 who
did better.

And I think halfbridge is some to big for the voltages without the big 65 v x 2 15 amps one.

Helps is welcome also a transformer calculator I did have ExcellentIR dut is russian, and calculation did voltages get out of range.

View attachment smps-sim.zip

regardsDSCN3767.JPG
 

Silvio

Well-known member
You will need to divide them in 2 smps. well 2x65vx15A that is nearly 2000W and 15v x10A that is 150w. These you will wind on one transformer. The typology for this will be either half or full bridge. I do not know the duty cycle you need for this I do not know either if this output need to be regulated. it all depend on the type of amplifier used.

The other high voltages will be wound on another trafo. The high voltage is not so simple to make and regulate. You can still use the same bulk capacitors for both smps if they are fitted on the same PCB. You need to screen the smps from noise as if the high voltage is used for the preamp stage then you have to be careful from noise entering the preamp stage.

The ICs you mentioned are a resonant control chip and the other is a current mode flyback controller. Have you ever worked with these chips before?
 
Last edited:

MicrosiM

Administrator
Staff member
Hi all


In have tryed to sim some kinds of smps, forward, hald bridge.

However I think models UC3845 for example are not so well.


I need 15 volts 10 amps, 300 volts 100mA, 2 x 120 volts 250mA, 2 x 65 volts 15 amps.

Some people say make two smps, one for power 2 x 65 volts and one for the other voltages.

I did try the last, thus alone the tube voltages, however getting feedback working and such in ltspice did not work nice, I had a UB3875 who
did better.

And I think halfbridge is some to big for the voltages without the big 65 v x 2 15 amps one.

Helps is welcome also a transformer calculator I did have ExcellentIR dut is russian, and calculation did voltages get out of range.

View attachment 7079

regardsView attachment 7078



The most direct question comes to mind is going regulated or unregulated SMPS, Then you will figure out witch topology is needed.

However, a SMPS with those many outputs is not going to be simple project.

Hope that helps
 

kees54

kees57
Hi All


I like difficult projects, I did make also a quite low distortion class D in past a three level.

Oke I have need the 2 x 65 volts for the mosfet power stage and the 2 x 120 volts, 1 x 300 volts, and 15 volts 10 a 15 amp for the tube drive part, the 15 volts 10 a 15 amps is
filament voltage, I go do regulate these and the high voltages can be without regulation, because that will work fine.

I think using a UC3845 is not the best way, difficult to use with that current feedback stuff, special with more voltages, I can use better a fase shift or somethng like that, for the 65 volts + 65 volts
a LLC resonant version is maybe a solution.

is for the high voltages also the need for half bridge or can I use a flyback?.

thanks for advice.

PS class D multilevel, project for later in time, because of climate issues need for efficienty...

ScreenHunter_2027 Feb. 08 22.17.jpg

prepostfeedback.jpg
 
Last edited:

kees54

kees57
Hi All


I like difficult projects, I did make also a quite low distortion class D in past a three level.

Oke I have need the 2 x 65 volts for the mosfet power stage and the 2 x 120 volts, 1 x 300 volts, and 15 volts 10 a 15 amp for the tube drive part, the 15 volts 10 a 15 amps is
filament voltage, I go do regulate these and the high voltages can be without regulation, because that will work fine.

I think using a UC3845 is not the best way, difficult to use with that current feedback stuff, special with more voltages, I can use better a fase shift or somethng like that, for the 65 volts + 65 volts
a LLC resonant version is maybe a solution.

is for the high voltages also the need for half bridge or can I use a flyback?.

thanks for advice.

PS here class D but that are very much later projects, now I have not enough time for.

ScreenHunter_2027 Feb. 08 22.17.jpg

prepostfeedback.jpg
 

kees54

kees57
You will need to divide them in 2 smps. well 2x65vx15A that is nearly 2000W and 15v x10A that is 150w. These you will wind on one transformer. The typology for this will be either half or full bridge. I do not know the duty cycle you need for this I do not know either if this output need to be regulated. it all depend on the type of amplifier used.

The other high voltages will be wound on another trafo. The high voltage is not so simple to make and regulate. You can still use the same bulk capacitors for both smps if they are fitted on the same PCB. You need to screen the smps from noise as if the high voltage is used for the preamp stage then you have to be careful from noise entering the preamp stage.

The ICs you mentioned are a resonant control chip and the other is a current mode flyback controller. Have you ever worked with these chips before?

SMPS is quite new for me, but have the equipment to start, can do that with low voltages om mosfet at first to see and to get familiair with it, there are quite a lot possibillityes out there and new stuff posted.
 

kees54

kees57
You will need to divide them in 2 smps. well 2x65vx15A that is nearly 2000W and 15v x10A that is 150w. These you will wind on one transformer. The typology for this will be either half or full bridge. I do not know the duty cycle you need for this I do not know either if this output need to be regulated. it all depend on the type of amplifier used.

The other high voltages will be wound on another trafo. The high voltage is not so simple to make and regulate. You can still use the same bulk capacitors for both smps if they are fitted on the same PCB. You need to screen the smps from noise as if the high voltage is used for the preamp stage then you have to be careful from noise entering the preamp stage.

The ICs you mentioned are a resonant control chip and the other is a current mode flyback controller. Have you ever worked with these chips before?

SMPS is quite new for me, but have the equipment to start, can do that with low voltages om mosfet at first to see and to get familiair with it, there are quite a lot possibillityes out there and new stuff posted.
 

kees54

kees57
For the people behind this forum, I get errors when posting and a massage the site is not secure from google chrome, it is also very very slow and get message also that when I post I leave the site, but posting do work normally.

regards
 

Silvio

Well-known member
SMPS is quite new for me, but have the equipment to start, can do that with low voltages om mosfet at first to see and to get familiair with it, there are quite a lot possibillityes out there and new stuff posted.

Well I think an LLC 2Kw half or full bridge smps is good for the output stage I dont know if you want it regulated or not. I never find them regulated in pro amplifiers. If this amplifier is going to run at full blast for a long time then I suggest you opt for a good heatsink and also adequate cooling on the smps.

The other smps I think a flyback arrangement will be more appropriate supplying the heaters and plate voltages.

Will this all fit in a U2 standard amplifier box?
 

kees54

kees57
Well I think an LLC 2Kw half or full bridge smps is good for the output stage I dont know if you want it regulated or not. I never find them regulated in pro amplifiers. If this amplifier is going to run at full blast for a long time then I suggest you opt for a good heatsink and also adequate cooling on the smps.

The other smps I think a flyback arrangement will be more appropriate supplying the heaters and plate voltages.

Will this all fit in a U2 standard amplifier box?

What is a U2 amplifer case? or box?

I have this idea, see the pic.

Hybrid-housing.jpg

Here I did try to sim a resonant version, I did see that with a resonant the supply even without regulation is very stable, I do not professional high power audio but high end stuff.

The pic is one without and one with feedback compensation PI network calculation.

-+65 v resonance feedback simulation.jpg

-+65 v resonance feedback corrected.jpg

Now I try the forward (flyback?) zvc converter, this is a good one for the tube voltages.

Used for the new class D interesse from me is because of climate and efficienty, but it needs to soun well, I did have a version also alone prepost feedback and these sound very open, for me better the
all feedback ones, but first the circlotron hybrid has to be done, with autobias. who already work now, need thus supply, smps is very small and nice.
 

Silvio

Well-known member
Regarding the amplifier box I was referring to the standard size boxes that are rack mounted. These come in U1 and U2 and U3 sizes. The size of the box for both boxes are the same but only the height varies U1 is more or less half the size of a U2 box. A U1 height is 1.75 inch while a U2 is double the height of U1 and U3 is 3 times the height of U1. The front width is always the same but the depth can vary at times.

Getting to resonant convertors I am not so keen on this subject and is relatively new to me but other members can help you.

Regards Silvio
 

kees54

kees57
Ahh I see, The resonant is not so really complicated, much calculaters are outside so I get somewhere. I do I think using a halfbridge also for the ybe section, just with small mosfets, these are much better in EMI.

here the sim, of all the voltages, in regulation who I think is not realy needed for a amp, to slow for fast transients I think, open loop is better.

resonant voltages half bridge.jpg
 

kees54

kees57
I have the question about rectifying in the smps resonant, what topology here is best, I have 320 volts, 2 x 120 volts and 15, 150, 250 volts and 15 amps, 150ma,250ma output, are there some calculations for the ripple capacitors for smps? for rectifying, I do not use double coils special not for HV outputs, maybe it is usable for the 15 volts 15 amps output.

By the way in open loop the 15 do run to 35 volts, with low current and 15 volts with max 15 amps output, maybe we need aregulation here because this is the filament output, I use a lm317 here so maybe it is not needed then, maybe I use regulators for all voltages so I have supply rail noise filtering.

For the halfbridge resonant type I see 2 Kv on the resonant capacitor, that is a lot, happens with full resonant so maybe I need to shift the frequency so it drops blow 1 Kv, what it does, giving lower output voltages, things are tricky here.

The coil inductions are very low, so I have to use a proper coil vorm who allow for resonant maybe someone here now what best to use, but I go search myself also or ask a manufacturer for that, she are specialised in cores..

I gu use for the regulated one the NCP1395 controller, I have a bunch of them.

resonant 65+65+120+120-320-15-volts.jpg

Transformer is only 120+120+320+15 volts

DSCN3767.JPG

thanks in advance.
 
Last edited:

res_smps

Member
hi,

there are 4 common types of resonant converters
1. SRC: used by most pro-audio, using IGBT working on ZCS (for example crown xti series), unregulated
2. PRC
3. LLC: widely used today for LCD / LED TV, regulated
4. LCC: from theory suitable for step up (Vo> Vin)
resonant.jpg

for the calculation of the voltage at the resonant capacitor you can see the infineon appnote (AN-1160) https://www.infineon.com/dgdl/an-1160.pdf?fileId=5546d462533600a40153559a85df1115

regards,
res
 
Last edited:

kees54

kees57
hi,

there are 4 common types of resonant converters
1. SRC: used by most pro-audio, using IGBT working on ZCS (for example crown xti series), unregulated
2. PRC
3. LLC: widely used today for LCD / LED TV, regulated
4. LCC: from theory suitable for step up (Vo> Vin)
View attachment 7090

for the calculation of the voltage at the resonant capacitor you can see the infineon appnote (AN-1160) https://www.infineon.com/dgdl/an-1160.pdf?fileId=5546d462533600a40153559a85df1115

regards,
res


Thanks res_smps for your help and pic,s.

Is the IGBT a forward or a half bridge? I have halfbridge with smaller fet,s or igbt or wat is best. mosfet
blows sometimes in resonat versions I did read, but the newer ones has better body diode, faster.

I go look at the LCC for the stepup because I need HV for the tubes.

here some examples I did simulate, I go build one to test, carefully protect myself afcouse with a isolation transformer and a leakage switch. Maybe you have the idea hw to do that best
for protection, however the cap is then still charged, so I need always to watch out. But i did repair tv in the old days, so I am used on HV, but get older and as such maybeless carefull.

regards

LCCresoant.jpg

LCCresonant-out.jpg

LCCresonant-primairy-sinusoidal.jpg
 

res_smps

Member
from what I observed most audio pro use igbt and igbt is suitable for zcs, works under resonant frequency.
and i don't know why they choose zcs over zvs

Now LLC seems to be more popular on TV sets, using mosfets that are suitable for ZVS
and LLC can also be used for step-ups, in this appnote the output voltage is 400V (see attachment) View attachment AN2492 - Wide range 400W L6599-based HB LLC resonant converter for PDP application.pdf
View attachment AN2509 - Wide range 400W L6599-based HB LLC resonant converter.pdf

I already made a halbridge LLC converter in this thread http://www.diysmps.com/forums/showthread.php?1001-Resonant-SMPS-using-L6599

I think for output with large wattage ( 2 x 65 volts 15 amps) would be better to use halfbridge LLC (regulated) or LC series (unregulated)
and for this (15 volts 10 amps, 300 volts 100mA, 2 x 120 volts 250mA) you can choose multi-output flyback
 
Last edited:

kees54

kees57
The world is ever changing on this technonoly, I have read somewhere about reason to uses ZVS of ZCS has to do with above or below resonance regulation, and stresses, ZVS is better
it seems. Maybe ZCS is better for igbt.

I do not like flyback, it is not recomended also, I need 15 amps not 10, because I try to get enough for filaments, some draw much.

I do have seen a ZVS forward converter somewhere on paper, maybe this is usable for hv, resoance is less emi and less disturbance into the amp itselfs.

thanks for thinking for me, appreciated.

regards
 

kees54

kees57
This is interesting, because also for welding (20-250 amps) LCC or LLC is used, problem is low load conditions, here a paper who discuss that.

file:///C:/Users/Gebruiker/Downloads/electronics-08-00205%20(1).pdf

regards
 
Top