GDIY boost converter for HT B+

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

dogears

Well-known member
Joined
Nov 15, 2017
Messages
548
Location
Texas
Think we could put our collective minds together and come up with a DC-DC boost design? It would be really useful for tube designs to get a 12 to 250VDC design that could handle maybe 5-10W.

Anyone got a particular chip that would be a good starting point?
 
I think this would be an excellent project. If the 12V input is to be used also for heaters then a lot of care will be needed to keep switching noise out. Also 12V to 250V or more is a big step up so choosing the right topology is the first hurdle.

Cheers

Ian
 
Yes, I was thinking to use an AC-DC module that spit out 12V to drive the heaters as well, then taking that up for B+.

I’m glad to help research and prototype but I need help getting started with the basic component selection. 
 
Goal was to use an isolated AC-DC SMPS to double as the DC supply for the boost converter and the heater supply. This also eliminates the (heavy, expensive) transformer. But it wouldn’t be difficult to have two AC-DC converters, one beefy one for the heaters at 12V and a second at 24V or 48V to feed the boost converter.
 
If you're going to design something from scratch you might as well do an ideal design (or at least shoot for an ideal design and then compromise if you have to) where you use one switching IC to make both 12.6VDC and 250-ish off of another winding. The only critical part that you need to find for that is the right SMPS transformer. There are transformers that have lots of windings (like 5-10) so that you can make all sorts of voltages. So you might get lucky and find one that can make ~250VDC from one winding and then parallel all of the others together to make 12.6VDC. You just do what you have to do to make the 12.6 and use that for feedback and then just try to get something vaguely 250. It could be 230, it could be 280, it doesn't matter that much. You can always drop volts across RCs (which is a not necessarily insignificant part of a tube circuit anyway).

But then you do have a bit of an issue with size. If the SMPS is too large, it might go into hiccup mode. Maybe you can work around that with the right IC. But otherwise, you might consider some kind of differential load circuit like a capacitance multiplier that is also part LTP acting as a diverter - loading one side causes the other to decrease current.

Incidentally you can get chokes like Bourns 5900-104-RC that are 100mH, 83mA and 82 ohms. With that and a capacitance multiplier you could achieve 40dB of attenuation at mains frequencies on the HT output. Although for it to mean anything you might have to do something for the heaters as well. Maybe a common mode choke would help. They can handle a lot more current.
 
The current requirements for heaters is so much higher than HT that I’m not sure if it makes sense to combine them.

It’s simple to get an AC to 12V circuit using an enclosed and isolated supply from Meanwell or XP Power. You get all the current you need there for the heaters and then some, and a stable DC supply which is already free of mains feed through. I thought the simplest way would be to boost from that starting point.
 
dogears said:
The current requirements for heaters is so much higher than HT that I’m not sure if it makes sense to combine them.

It’s simple to get an AC to 12V circuit using an enclosed and isolated supply from Meanwell or XP Power. You get all the current you need there for the heaters and then some, and a stable DC supply which is already free of mains feed through. I thought the simplest way would be to boost from that starting point.

I agree. The other problem with trying to get HT and heaters from the same SMPS is that you can only regulate one voltage.

i did some work of this a while back. I found a few Chinese SMPS PCBs that take 12V dc in and produce somewhere in the region of 200V to 400V. I bought a couple of ready made ones to test and then found a kit version which I experimented with.  I think the chip used was an SG3525. It was attractive because it is available in a DIP package.

Cheers

Ian
 
ruffrecords said:
The other problem with trying to get HT and heaters from the same SMPS is that you can only regulate one voltage.
I agree too. the other issue is that designed an AC/DC converter is a big issue with safety regulations and certifications. Using a standard AC/DC as the interface between mains and gear solves many problems.
For the up conversion, I believe there are IC's that give predictable results without too complex issues
https://docs.rs-online.com/31fd/0900766b812524bc.pdf
Fig. 17 suggests 250V/100mA is feasible. Indeed there is always the issue with the transformer. I wonder if using five 1:1 xfmrs with the primaries in // and the secondaries in series would work.
 
Several years back when I designed an outlet tester (long story), I needed to test the insulation resistance of my gadget at 500V for UL approval. I eventually borrowed a proper insulation resistance tester, but before that I rolled my own 500V supply (actually 480V) by stringing 4x 120V transformers in series.

The first transformer had it's primary connected to mains 120V. The other three primaries were in series with the first adding their voltage on top of the mains voltage. Note: this high voltage is not isolated from the mains, actually configured as an auto-former. Adding a 5th transformer would allow the the 4 connected floating primaries to be isolated.

The output of such a rig is limited to only the primary current of the power transformers not secondary, but I only needed mA of current. This is not very elegant but works, especially if you have a bunch of the same transformers laying around like I did.

In theory you could connect transformers with different secondary voltages together generate step up voltage directly (say a 24V secondary driving a 6V secondary for 4x step up) but this gets dicey as you would be exceeding the design voltage of wiring insulation inside the step up transformer.

JR   
 

Attachments

  • 240.JPG
    240.JPG
    57.4 KB · Views: 16
Well, yes I know that connecting transformers in series works in principle. however smps are sometimes temperamental about stray capacitances. That's why I put a question mark.
I don't have time neither motivation for undertaking such a project (although I've posted a request to LT about simulation models for their LT3748), but I think someone with enough determination could use one of the eval boards as a starting point.
 
Find and study the existing thread about a 500 series tube pre.  It might be 5 years back. 
 
Wurth makes a transformer from the series recommended in the LT3748 datasheet that's a 1:10 step up with 300V rating on the secondary, 36W:

https://www.mouser.com/ProductDetail/Wurth-Elektronik/750032050?qs=sGAEpiMZZMuKmRn7rpQYPWwujborLkE%2Fae2ZURZWLQ4%3D

And a slightly larger through hole version at 75W:

https://www.mouser.com/ProductDetail/Wurth-Elektronik/750310349?qs=sGAEpiMZZMuKmRn7rpQYPWwujborLkE%2Ft1Fx0LbA1%252Bc%3D
 
I suppose this would be a starting point:

wjToqH6.png
 
I am not sure about taking the heater supply after the first filter because it has lots of HT booster noise on it. Might be better to take it direct from the IRM-30-12.

I am not sure the IRM-30-12 has enough grunt for this. It is capable of outputting 2.5A but this will only reliably power up into half that amount of heater load. The remaining amp and a bit might be enough to create 50mA at 250V but the IRM-30-12 will most likely not fire up into them both simultaneously. So I would suggest adding a slow start circuit to the HT to give the IRM time to fire up the heaters before the HT starts demanding current. Even then,it might hiccup some more.

Cheers

Ian
 
Pucho had pointed out an interesting HV module within the last year that i think was testing.  Can’t recall if it was relevant here. 
 
Could upside the Meanwell supply to 60W for a whopping $6 more haha.

I see what you mean about the heaters, that makes sense.

All I did here was grab straight off the app notes, no idea what may need to be adjusted here.
 
ruffrecords said:
I am not sure the IRM-30-12 has enough grunt for this. It is capable of outputting 2.5A but this will only reliably power up into half that amount of heater load.
One way to "slow start" and get extra filtering might be to use a simple two-transistor capacitance multiplier in place of C8. The RC of the CM can be made as slow as necessary and as a bonus it looks like a giant cap (from the downstream side anyway). I think you should be able to make it slow enough to pull the full load out of the SMPS without it's output protection mode kicking in when it starts up.
 
This is used in an Alembic preamp clone that can use a wall wart. Using TDA7293 for the power amplifier, I made a 2 channel guitar amp with it and love it.

 

Latest posts

Back
Top