The MAX1771 SMPS Development Thread

[MetalGuy]

Well, I can't justify making an order to Banzai for just a couple of 0R050 and 0R1 1 watt resistors, so this'll have to wait for a little while.....

Farnell have all the parts for this project.

[Ice-9]

Well, I can't justify making an order to Banzai for just a couple of 0R050 and 0R1 1 watt resistors, so this'll have to wait for a little while.....

@ Mick (Ice-9): does the IRF740 in yours get very hot? I ask because the IRF740 I used in my 555 SMPS does get hot and needs heatsinking, and the site the schematic is from states "At no time did any component get hot - the FET was always cool to the touch, and the inductor warmed up only under heavy loads, but never got hot."

The mosfet (IRF740) doesnt  even get warm  at all. I have no heatsink on it either.

I'll have a look see if i have some spare 0r05 resistors if you need any. Yup just looked i have some SMD ones let me know if you want a couple.

Rick do you not have a RS components anywhere near where you live, I use mine regular for anything, from single components to tools, if there not in stock when i go down they get them there for next morning no postage costs. There not really that expensive either (only expensive on certain items)

[frequencycentral]

It occured to me that the MAX1771 smps schematic lacks the snubber that the 555 smps includes, it's the 2.2K resistor in series with the 100pF cap between the drain of the IRF740 and ground. According to Cliff:

It looks like a snubber circuit to me which goes in parallel with the FET's drain and source terminals. Or in this case, because you've got voltage-mode control, it goes from the drain to ground. These can increase efficiency in the FET by 20-30%, sometimes more with excessively ringing circuits!

As these SMPS's were designed primarily for use with nixie tubes and not audio tubes, maybe the necessary filtering isn't there and needs to be added?

[Cliff Schecht]

The filtering is really just there to kill the ringing that occurs in every (unsnubbed) boost supply. I'm sure that if you look at the drain of the FET on a scope you'll see the ringing I'm talking about. Fortunately it occurs at a frequency much higher than that of the actual switching frequency and is (almost) always completely inaudibe. But it does decrease efficiency and can reduce the life span of some of the critical components. The FET, diode and output caps all can suffer from excessive ringing whose P-P voltage actually goes above the rated value of the part. This overvoltaging only occurs for fractions of a second but it still can do its damage over time.

Snubber circuits eliminate this ringing like a champ (also cuts down P-P output noise!) but still have a caveat in the form of power dissipation. Some snubber circuits require a fairly high wattage resistor because of the amount of energy that the capacitor is absorbing can become quite substantial if the frequency is low enough. There are also different types of snubber circuits depending on where you are trying to eliminate ringing. Do some googling for snubber circuit design and see what you find there, I don't know that guys design well enough to say if it warrants one.

[robmdall]

Rick and Cliff... any new developments on the SMPS designs?

Rob

[Cliff Schecht]

Time and lack of effort have been killing me as of late. I've just been in "go" mode for so long now that I'm trying to use my free-time to relax a bit instead of getting home from school and building stuff. I've spent so long building stuff for my guitar playing needs that I wasn't even really playing guitar! Anywho, I will look into my design again later and see if I can't get my design retested and finished.

[frequencycentral]

Still not onto it yet. But I just sold my Superfly (submini Firefly) amp, I'll be building another, so it's a good opportunity to test the MAX1771 SMPS......

[Cliff Schecht]

From what I can tell the MAX1771 design is a good one. I haven't tried it or done any testing on it or anything though. Knowing what I know now, I might change my design to be of the same style (voltage-mode controller with voltage feedforward).

[robmdall]

When testing an SMPS like this how should you measure output correctly?

Simply meter to + and -?  or do you need a load across the output?

Rob

Thought of another question for the Guru's of HV. Is there a reason these designs are on small separate PCB's? Is there an issue with combining an SMPS with another circuit on the same board

[frequencycentral]

When testing an SMPS like this how should you measure output correctly?

Simply meter to + and -?  or do you need a load across the output?

Rob

Thought of another question for the Guru's of HV. Is there a reason these designs are on small separate PCB's? Is there an issue with combining an SMPS with another circuit on the same board

You can measure the output without load to test it's working, but be aware that it will drop significantly under load.

Read enough info on SMPS design and layout and it will totally put you off building one on it's own, let alone on the same board as the circuit to be powered! ie issues surrounding ground planes, using 'just the right parts' etc. My approach is more ghetto, if it works it works and f##k the science. Here's a 555 SMPS on the same board as a submini tube amp:

[robmdall]

Rick that board is so cool, far from Ghetto... Although I like Ghetto. Yeah, you are correct, the freaking more I research the SMPS, the more difficult it becomes in my mind. Your design (in the beginning of this thread) is actually very straight forward.  I guess the best thing to do is bite the bullet and build it. Thanks for the reassurance.

[nickds1]

Hi - I'm the author of that MAX1771 design - its been recently updated to use more readily available components - the original ones specified were pre-RoHS. The PCB has also been changed slightly. http://desmith.net/NMdS/Electronics/NixiePSU.html

I can say that I've used it for small headphone amps etc. and there is at least one US-based guitar effects manufacturer who is also using it. Also, over on diyaudio, several people have also used it with great success.

The design switches at about 55-60kHz using the components specified - a nice ceramic output filter and GOOD filtering on the input side (to prevent those switching transients appearing elsewhere in the circuitry) makes it very usable for small tube designs. I tend to enclose/screen mine in a small box made from bits of FR4 PCB soldered together - a common radio ham trick.

There is no snubber for the FET as its completely unnecessary in this design - the MAX 1771 internal driver is explicitly designed for this job - a 555's is most certainly NOT. Further, no feed-forward capacitor is needed on the top of the feedback network. The design as it stands is very stable and there are many 100s of these running happily. Several of the commercial designs being sold on eBay and elsewhere are direct lifts.

What IS very important is to take note of my comments in the design notes re: component choice and layout. These are very very important. The MAX1771 is a bit of a Ferrari in the SMPS world - it's very efficient but can be temperamental - there is definite method in my madness!

[nickds1]

When testing an SMPS like this how should you measure output correctly?

Simply meter to + and -?  or do you need a load across the output?

Rob

Thought of another question for the Guru's of HV. Is there a reason these designs are on small separate PCB's? Is there an issue with combining an SMPS with another circuit on the same board

You can measure the output without load to test it's working, but be aware that it will drop significantly under load.

Read enough info on SMPS design and layout and it will totally put you off building one on it's own, let alone on the same board as the circuit to be powered! ie issues surrounding ground planes, using 'just the right parts' etc. My approach is more ghetto, if it works it works and f##k the science. Here's a 555 SMPS on the same board as a submini tube amp:

Small point - a 555-based SMPS will always be very inefficient - typically 70% at very best, more typically around the 55% mark in my experience.

Also, the comment about sagging under load is simply wrong, so long as you stay within the design parameters of the supply and your design is regulated - your 555 SMPS may well not be regulated - there are many regulated 555 designs out there (it only takes a few extra components - see the files section of the Yahoo neonixie-l group for several examples). The MAX1771 design IS regulated and delivers up to 100mA quite happily at 87% or so and 180VDC. I tested the loading by using non-inductive and then wirewound Al load resistors - its all documented on the original design page.

The MAX1771 design has been built on the same boards as the rest of the circuitry by many users - you just have to be a bit more careful. If you are happy with an unregulated inefficient design, that's fine, but it really is not that difficult to build an MAX 1771 design - I've built loads of different SMPS designs, and if you obey the simple rules, its not a problem.

It all comes down to what you want. If you want an efficient, quiet, stable, predicable & small design then use a specialised SMPS chip (like the MAX1771). If you just want to get something working from bits you might have lying around, then use a 555 design - the two are apples & pears. But don't kid yourself the interference & other general issues about using SMPSs in audio kit will magically go away if you use a 555 - all the the problems will still be there.

[Ice-9]

Thanks for the update Nickds1, i had originally used the older circuit from the site you mention and included the SMPS into PCB that has the rest of the circuit as well. I will check out the newer version you mention.

[Cliff Schecht]

Nick, welcome to DIYStompboxes and thank you for your posts regarding the MAX1771 design! I've worked on similar switchers that were inspired by this design, albeit I went with current-mode control and a Unitrode controller as I find Maxim to be too expensive and usually too specialized. Then again, I didn't know until recently that Maxim has such an extensive power management portfolio, so I guess my spite for Maxim was more out of ignorance than anything.

My real goal in designing a miniature HV switcher was to meet or beat the capabilities of your switcher. I did manage to do so, making a slightly smaller board that could put out a little more power (about 40W at 250V) but I'm still yet to be able to finish the "production" board design because of school and work and what not. But thanks again for the inspiration, I soon hope to be "competing" with you in the DIY community with my HV design .

Also, I think I've made similar comments about the 555 timer based supplies. While they aren't complete crap, nothing beats a well designed switcher!

[frequencycentral]

Finally got this on the breadboard. Not having any 10K trimmers in stock, I had to sub in a 1k trimmer and fiddle with the voltage divider values to get it to work. I'm using 180k/1k trimmer/1k, and am able to dial in voltage between 159v and 317v. At one point when fiddling with the voltage divider values I mis-read a resistor value and got over 500v, which let a little of the magic smoke out........Anyway, I've had it running at a steady 250v without load for the last hour or so. There is no heat being generated anywhere - very cool (in both senses of the word) compared to 555 SMPS's. Now to build an amp around it............

[frequencycentral]

Guess what going on the rest of the board?

[defaced]

At one point when fiddling with the voltage divider values I mis-read a resistor value and got over 500v, which let a little of the magic smoke out.

What fried?  I've got a need to get up into the 400-450v range for some high gain tube preamps, and I think I know where to start, but knowing what failed on your build would be valuable information.   

[frequencycentral]

At one point when fiddling with the voltage divider values I mis-read a resistor value and got over 500v, which let a little of the magic smoke out.

What fried?  I've got a need to get up into the 400-450v range for some high gain tube preamps, and I think I know where to start, but knowing what failed on your build would be valuable information.   

I ended up using a 220k resistor/1k trimmer/1k resistor arrangement, I mistook a 22k for a 220k.......now that let the genie out of the bottle - I now have a blackened 22k resistor. 

[nickds1]

You are actually quite lucky it worked on a proto board - MAX1771s are pretty sensitive to stray capacitance and you may have a few unexpected high frequency artifacts in there - if you have a 'scope, I'd check it carefully... (a good start would be to check the efficiency - it should be around 85% for a reasonable layout...)

Cheers

Nick