how to wire two smps in series...?

[zambo]

Ive read it is possible to wire 2 smps in series by isolating the outputs of each. So if I theoreticly took 2 smps like the 1364 from taylor electronics and ran the outputs of each one through a diode like 1n4007 and into a 600volt reservoire cap would I end up with a net voltage of say  ( assuming each smps puts out 150v at 50ma ) 300v at 50ma? very curiouse about this..smart guys???

[defaced]

It's something on my list to try.  I've got two sitting here, but haven't had the time to try it.  Watch out though, the input voltage ground is the same as the output voltage ground.  This means you can't run the input power in parallel and the output power in series (which was my original plan).  I haven't looked into seeing if it's possible to separate the two. 

[zambo]

so i couldnt run both supplies off of the same walwart?

[defaced]

Edit: I have the 1363 units, I don't know for sure if the 1364 are the same way.  There's one schematic, page 4, that shows two ground connections for the 1364.  I would email Taylor and ask him; being the creator of these, I'd say he's your best source of info.  I know he's posted here before, but I don't know if he checks often or what the deal is.  If you do happen to email him, please post back what you find out.  I'd certainly be interested. If this can be done with the 1364, then I'll be buying at least two of them. 

Original post:

Nope.  Check the data sheet, the standard schematics are shown at the bottom, note there's only one ground connection.  I have also verified with the units I have that all of the grounds are interconnected.  http://www.tayloredge.com/storefront/SmartNixie/DataSheets/Datasheet_1363-1364.pdf

Basically you have to float one entire input power supply and SMPS output on top of another one.  The way I plan to do this is with either two transformers, or a transformer with two identical secondaries.  I just haven't bothered to source that transformer yet.

[amptramp]

With switchmode supplies, be certain that you have some way of synchronizing the switching or you will get a beat note at the difference frequency between them.  If one is operating at 40 KHz and the other at 41 KHz, you will get noise at 1 KHz which is in the audio band.

[frequencycentral]

Yeah, you don't want two HF oscillators in the same build. The best I can offer is the idea of stacking the MOSFET, inductor, UF diode and reservior cap like this:



...which is just an idea and may not work. Or might work but have issues. The theory is that you'll have two independent B+, each at the same voltage and ma, use one for your preamp section and one for your power section. Haven't tried it yet. Who know? I'm just a solder monkey. But I do plan to try it some day, as I have a couple of Mullard ECL83 waiting to be the power section of an amp.......

[zambo]

so would those tubes take 76ma ( including plate and screen ) and put out around 7 watts in ab1 pp at 200v ? I dont think we are far off of that now... and I want one bad! Do I have the specs on those right?

[defaced]

With switchmode supplies, be certain that you have some way of synchronizing the switching or you will get a beat note at the difference frequency between them.  If one is operating at 40 KHz and the other at 41 KHz, you will get noise at 1 KHz which is in the audio band.

Damn, why'd you have to go and point out the obvious!  

I hadn't considered that.  In the big SMPS thread, the creator posted how to lock out the frequency scaling, but, yea, 1Khz difference will make the idea shot.

Rick's got a good idea.  I wonder if you "ground" the reservoir & bypass cap and the first HV out if you'd higher voltage output without letting the magic smoke out.  Probably would need a higher voltage MOSFET, but those are easy enough to source.  I bet layout could become a concern.  

[iccaros]

Yeah, you don't want two HF oscillators in the same build. The best I can offer is the idea of stacking the MOSFET, inductor, UF diode and reservior cap like this:



...which is just an idea and may not work. Or might work but have issues. The theory is that you'll have two independent B+, each at the same voltage and ma, use one for your preamp section and one for your power section. Haven't tried it yet. Who know? I'm just a solder monkey. But I do plan to try it some day, as I have a couple of Mullard ECL83 waiting to be the power section of an amp.......

If the starting voltage was say 24v, would that not almost double the output in ma to say close to 100MA?  a standard 50 watt tube amp - heaters users what? 140 ma

The only thing I can see is that the transistors must be able to handle the higher amperage.

Am I way off. This way we would not have to create a double supply?

[defaced]

a standard 50 watt tube amp - heaters users what? 140 ma

Heaters?  Those are typically 6.3v AC circuits and can be anywhere from 4 upwards to say 10 amps depending on the tube compliment. 

[iccaros]

a standard 50 watt tube amp - heaters users what? 140 ma

Heaters?  Those are typically 6.3v AC circuits and can be anywhere from 4 upwards to say 10 amps depending on the tube compliment. 

sorry that was way off from what I was asking..

Forget the heaters, I screwed up my sentence..
that was 140 ma minus heaters (- = minus)  because the heaters do not need the high voltage it is not apart of the calculation for the high voltage side. We would put the heaters in series and run them off the 24v wall wart directly.

So what I was trying to ask


IF we doubled the input voltage, we should have more amperage at the end, so If this design puts out 50ma, by doubling the voltage we should get close to 100ma (-) inefficiency.
the cap voltage values would need to go up, and possibly some others..

[frequencycentral]

I don't think it is the case that doubling the input voltage doubles the output voltage. Though I'd love to be wrong on this one.

[iccaros]

I don't think it is the case that doubling the input voltage doubles the output voltage. Though I'd love to be wrong on this one.

no doubling the input voltage, double the output current?  In case I said that wrong before..  This is already multiplying voltage

I read the data sheets for a few and they say things like
5 volts in 23ma out
12 volts in 45ma out

so its not a one for one, but like a transformer, we have 120 volts at .5 amps on the primary, we can get 12 at 5 amps on the secondary.

The ideal is that voltage is relatively consent, so if you rise the input voltage where does that potential go?
the inductor would also need to be able to handle the current, the one most people use can only handle 65ma.. but this one http://parts.digikey.com/1/parts/664661-inductor-100uh-3a-50khz-clp-mnt-swc-3-0-100.html is good for 3A

That is the theory.. I have everything but the transistors and the inductor to test.. 
may have to watch what voltage the 555 gets..

[amptramp]

Yeah, you don't want two HF oscillators in the same build. The best I can offer is the idea of stacking the MOSFET, inductor, UF diode and reservior cap like this:



...which is just an idea and may not work. Or might work but have issues. The theory is that you'll have two independent B+, each at the same voltage and ma, use one for your preamp section and one for your power section. Haven't tried it yet. Who know? I'm just a solder monkey. But I do plan to try it some day, as I have a couple of Mullard ECL83 waiting to be the power section of an amp.......

Be aware that in this design, only the lower output is regulated because the feedback comes from it.  The upper output is "anti-regulated" - a higher load on the lower output will increase the duty cycle to the point where the lower output regulates, but the upper output will also be subjected to the same change in duty cycle, so as you load the lower output, the voltage at the upper output will increase.  Many inverter chips have a method of synchronization, but the MAX1771 does not appear to be one of them. 

[iccaros]

So I found this in a search


not much difference

but the owner states that they could run 4 double triodes with no issues.
They are using a bigger inductor and oscillate the 555 faster.

the post
http://www.freestompboxes.org/viewtopic.php?f=28&t=12195
http://www.freestompboxes.org/viewtopic.php?f=17&t=12255

[frequencycentral]

Be aware that in this design, only the lower output is regulated because the feedback comes from it.  The upper output is "anti-regulated" - a higher load on the lower output will increase the duty cycle to the point where the lower output regulates, but the upper output will also be subjected to the same change in duty cycle, so as you load the lower output, the voltage at the upper output will increase.  Many inverter chips have a method of synchronization, but the MAX1771 does not appear to be one of them. 

Thanks Ron. My assumption was that the regulator would regulate both outputs. But I see the obvious now that you explain it. So, instead, what about removing the reservoir cap from the upper section and connecting the cathodes of both UF4004 together, so both upper and lower sections are charging a single reservoir cap. What's the obvious reason that I've missed why this wouldn't work? 

So I found this in a search



but the owner states that they could run 4 double triodes with no issues.
They are using a bigger inductor and oscillate the 555 faster.

That's pretty much the exact same SMPS I use in Superfly. It runs 4 triode sections plus a PP transformer nicely at ~190v. DarkRain is also running 4 triode sections (not 4 dual triodes), but as he's not using a transformer, there's more juice available for the triodes, hence the extra voltage.

[iccaros]

@FC
Sorry Linkhttp://www.freestompboxes.org/viewtopic.php?f=17&t=12255

For some reason cut and paste did not work...

I knew this looked familiar.. The differences I saw were the ratings on the Inductor and the caps.. He is using a 1.8 amp inductor..

from about middle page "The trimmer value is 1k in my case, but anything between 1k - 5k will be excellent. I've test it with 4 double triodes @ full blast for about 6-7 hours... the voltage droped from 300v @ about 280-275v but that i find more then acceptable. I I used ferrite shilded inductor (100uH @ 2.1A), the voltage ripple is 150mV @ full blast (more the enogh for any tube applications). If you need anymore help please do not hesitate to write me."

I was just thinking, if you upped the rating of the inductor so it could take the load, then a power supply with more amperage available would be able to push, say 12v @ 1 amp, then the tubes are going to pull what they need.  Then there would be no need to double everything else.. 


I just need to build this out and see

[amptramp]

Thanks Ron. My assumption was that the regulator would regulate both outputs. But I see the obvious now that you explain it. So, instead, what about removing the reservoir cap from the upper section and connecting the cathodes of both UF4004 together, so both upper and lower sections are charging a single reservoir cap. What's the obvious reason that I've missed why this wouldn't work? 

This would work.  But you are talking about a relatively small amplifier.  Do you need a pair of IRF740's?  Due to the inductors, the current will be shared between them, so that will not be a problem.  My assumption is that you had one output in the original design to supply the output stages and another to supply the preamp.  The preamp should only be a few milliamps, so there is not much reason to separate them, especially if you output stage is biased for Class A which exhibits no change in current drain with changes in sound level.

[zambo]

I am actualy trying to figure out the smallest lightest way to get about 330v net and 80 ma if possible without using a power trans and rectifier. It may not be doable easily I just figured if someone knows how then they are probably on this forum. My dream is a 7 to 8 watt tube poweramp that fits in a guitar case or pedal board easily. Smps seems like a good solution. I have heard of them being run in series but not in a hv situation. usualy 24v versions with several amps. I just thought you could put a diode on the output of each and run them together into a reservoire cap. No?

[zambo]

So I just realized I am actualy tunning them in parallel in my hypothetical situation  ...ingnoramus i swear...anyway, So i realize the voltage will stay the same, but can I get double the ma if running them in parallel ? that would be enough voltage at 200v if I could get more ma. Could I run 3 or 4 in parallel ? a 1spot walwart puts out 1.7 amps so i could probably power at least 3 of these together .