UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Reply
 
Thread Tools
Old 22nd Jun 2018, 10:12 pm   #21
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 4,416
Default Re: Switched Mode Power Supplies

Quote:
Originally Posted by kellys_eye View Post
Is there any way to get information on the ferrite cores that I'd salvage from ex PC PSU's? most of which are unmarked.
Despite the proliferation of core material grades that are available, they all tend to be much of a muchness. If you assume the ferrite saturates at 300mT, you won't go far wrong. Dimensions you can easily measure, to find the all-important cross-sectional area. Permeability doesn't really matter. Some grades of ferrite offer lower losses than others under certain conditions, but this really is the icing on the cake. The only thing missing is an air gap - if it's a smoothing choke, or a flyback transformer, with an air gap ground into the ferrite, you need to know this - but here, by winding 10 turns or so on the core and measuring inductance, you can work out the core's Al value.

Some toroids in SMPS's are not ferrite, but powdered alloy (iron, moly-permalloy, etc) mixed in a resin binder to give quite low (but predictable) inductance factors Al. Again, a test winding will determine this. You also need the dimensions, but again, a few moments with a vernier calliper will get you what you need. Knowing the inductance you want, and how much current you are pushing through your choke, you can calculate the flux in the core and if this is below 500mT, you won't have a problem.
kalee20 is offline   Reply With Quote
Old 22nd Jun 2018, 10:39 pm   #22
Oldcodger
Octode
 
Join Date: Oct 2014
Location: West Midlands, UK.
Posts: 1,107
Default Re: Switched Mode Power Supplies

Another chip to look at is the TL494. A bit old fashioned, although they are still available at a low cost. GEC used to use them to change voltages on telecoms kit on a shelf basis. I've copied their circuit and it's basically as per the data sheet. Only problem was frequency and finding a transformer and a suggested cure on here (for my use as a 12-240 convertor) was to use the transformer out of an old laptop PSU. GEC solution was to wind Ferox cubes as required. I did find details of a suitable cube and windings, but that went with a HD failure.
Oldcodger is offline   Reply With Quote
Old 22nd Jun 2018, 11:07 pm   #23
MrBungle
Nonode
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 2,419
Default Re: Switched Mode Power Supplies

Can still get new TL494's from RS! I've seen UCC3895 used a lot in recent kit that exploded that I took to bits.
MrBungle is offline   Reply With Quote
Old 22nd Jun 2018, 11:12 pm   #24
Argus25
Octode
 
Argus25's Avatar
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 1,520
Default Re: Switched Mode Power Supplies

Quote:
Originally Posted by TonyDuell View Post
Quote:
Originally Posted by Argus25 View Post
One of the early SMPS's for computers was the one in the IBM5155 computer from the early 80's . Recently I documented it, (as IBM didn't), there are remarks and schematics & waveforms
That's not that early.
Yes that is right, the SG3524 IC it is based on had been available and used for the application in many other supplies for a decade before that, but by the time the 5155 supply was made the ideas behind SMPS's were maturing and well tested. For example, just to cite some of the numerous innovations:

There is a method there to efficiently power a small 115V transformer primary from a 230V source for dual voltage inputs.

Since, at the time, Mosfets were not as commonly used in SMPs's, there is a method to enhance the base drive to the main switching transistors with increasing load, by adding the current in the current sensing loop to the driver transformer, which increases the efficiency of the supply. Some but not all BJT based SMPS's use this technique today. Also it contains an independent current sensing & shutdown (separate from the current limiter inputs of the IC) and a clever current inrush limiter based on a PUT.

So for anyone wanting to design/build a SMPS, an early circuit like this is a very good place to start for a first attempt. And while it is easy enough to program a PIC uP for the job, there are unexpected pitfalls (vs using the SG3524) for example when they initialize at power up you can have both of your output devices turning on simultaneously (with obvious consequences) and they are more sensitive to RFI.
Argus25 is offline   Reply With Quote
Old 22nd Jun 2018, 11:46 pm   #25
trobbins
Hexode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 282
Default Re: Switched Mode Power Supplies

I'd also recommend using only power mosfet(s) and a control ic that can drive it. Simplicity comes with the impost of a good pcb feedback of smt current sense, and good gate drive current spike control, and good output or other indirect feedback into the control ic - each loop requiring care with pcb layout so as not to corrupt each other.

Its also pretty arduous to find a recycled magnetic part, identify it, re-engineer it and use it just once. Many would, as recommended near the start, define up front a goal and design around a commercially available core set and former with appropriate pinouts. The goal spec may change a bit due to practical operating levels and winding constraints.
trobbins is offline   Reply With Quote
Old 23rd Jun 2018, 1:28 am   #26
Refugee
Dekatron
 
Refugee's Avatar
 
Join Date: Apr 2012
Location: Worksop, Nottinghamshire, UK.
Posts: 3,140
Default Re: Switched Mode Power Supplies

Would this sort of circuit work for powering a small valve circuit.
https://threeneurons.wordpress.com/n...ply/uc3843_ps/
The chip is still available in thru hole and there is a vast selection of FETs that can be substituted. There is an ST FET that is vastly cheaper than the IRF types.
Refugee is offline   Reply With Quote
Old 23rd Jun 2018, 7:28 am   #27
Radio Wrangler
Dekatron
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 10,190
Default Re: Switched Mode Power Supplies

It's a pretty standard boost converter with current-mode control. It should work quite well.

If you try different FETs you may need that filter in the current sense line. It prevents the gate drive current surges affecting the tripping point of the current mode controller.

It'll be one hell of an interference generator. A screened box and added filters will be needed

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 25th Jun 2018, 12:25 pm   #28
brightsparkey
Tetrode
 
Join Date: Sep 2017
Location: Cambridge, Cambridgeshire, UK.
Posts: 85
Default Re: Switched Mode Power Supplies

Whilst I admire the can-do attitude on this forum, I would caution that boost converters are works of the devil, especially if they have a high output to input voltage ratio.

The control of the output voltage is non trivial - you will likely end up with strange behaviour or blown devices if you dive in without simulated design first. I whole-heartedly recommend the LTSPICE simulator available from Linear Technology (now owned by Analog Devices). Its easy and intuitive to drive and there are lots of resources that come with it. Its also free for non-IC design use, even commercial.

Their own chips, some of which are excellent and easy to use, are all modeled accurately, so you should be able to get something going easily. That way you can experiment without burning pound notes or fingers.

90V 10mA is an unusual output these days, so many modern IC's will struggle to control the output voltage without mis-behaviour. The Unitrode IC are general purpose and can be made to work in these applications as long as you aren't too ambitious about size and efficiency and keep the switching frequency low (i.e. 10's kHz, not MHz). I would avoid Bipolar switching devices - MOSFETs are much easier.

EMC is another matter - boost converters make junk on their output (the output current is dis-continuous) and a extra L-C filter and some screening are likely necessary for radio use.

Kevin.
brightsparkey is offline   Reply With Quote
Old 26th Jun 2018, 9:21 am   #29
trobbins
Hexode
 
trobbins's Avatar
 
Join Date: Oct 2009
Location: Melbourne Australia
Posts: 282
Default Re: Switched Mode Power Supplies

Quote:
Originally Posted by Refugee View Post
Would this sort of circuit work for powering a small valve circuit.
https://threeneurons.wordpress.com/n...ply/uc3843_ps/
As well as Kevin's advice on step up ratio extremes, I'd suggest you also work through how you would purchase/make/design and then test the inductor for adequate performance as a standalone part.

If your aim is for a B+ supply then I recommend starting at the other end and buy in a suitable inverter (eg. as per http://www.diyaudio.com/forums/power...step-smps.html ) and modify and probe and play around with it, and then tear down the transformer and then make up a transformer variant etc etc. That may allow your exposure and experience advanced further and faster, without getting bogged down in the hassles of parts and pcb acquisition.
trobbins is offline   Reply With Quote
Old 26th Jun 2018, 1:00 pm   #30
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 4,416
Default Re: Switched Mode Power Supplies

Boost converters can do funny things, and as brightsparkey says, can be 'interesting' to stabilise.

Although increasing the duty cycle causes the steady state output voltage to increase, when you do increase the duty cycle, the output voltage actually DECREASES briefly before it recovers and increases. This can make the feedback loop oscillate very readily. It has to be quite slow acting. Luckily, this only matters if the load (or input voltage) change rapidly. If they don't - and I'm talking tens of milliseconds - then it is not a problem.

If you are aiming to drive a battery-portable radio with 90V HT, be prepared to be very demoralised by the squeals and whistles breaking through. It really will need careful design plus lots of filtering, and then you'll probably have to put it in a screened box to stop radiated interference getting out. But you'll have a lot of fun.
kalee20 is offline   Reply With Quote
Old 26th Jun 2018, 3:29 pm   #31
kellys_eye
Heptode
 
Join Date: Jan 2014
Location: Oban, Scotland, UK.
Posts: 530
Default Re: Switched Mode Power Supplies

I have magazine articles (I'm referring to an Elektor design here) that illustrate high(er) step ratios (0-30V DC input, 0-300V DC output) and the transformer is of a 'hand-wound' construction. The circuit seems to be pretty standard as far as SMPS are concerned and as high an output as I'm ever likely to require anyway.
kellys_eye is offline   Reply With Quote
Old 27th Jun 2018, 1:48 am   #32
Refugee
Dekatron
 
Refugee's Avatar
 
Join Date: Apr 2012
Location: Worksop, Nottinghamshire, UK.
Posts: 3,140
Default Re: Switched Mode Power Supplies

I did use a boost converter many years ago to power the "head end" of a high powered car stereo back in the early 1980s.
It worked well.
On test during construction it produced 60 volts but it tamed down to 24 volts quite nicely.
It used one leg of a Motorola push pull SMPSU chip.
It is long gone now.
Refugee is offline   Reply With Quote
Old 27th Jun 2018, 7:03 am   #33
Radio Wrangler
Dekatron
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 10,190
Default Re: Switched Mode Power Supplies

I've designed a couple of boost converters over the past few years. They run 8.5 to 60v input and 50v regulated output, so a SEPIC architecture is used... it amounts to a plain boost but without the dc offset from the input voltage. They power both pulsed radar transmitters (a very unsmooth load!) and AM speech transmitters. They are in the same unit as the associated receivers. Switching frequency is around 600kHz. Getting the control loop dynamics right was fun and the filtering/ screening was double fun.

The radio shack at home runs Icom IC765 and IC7700 transceivers. both have internal off-mains switching supplies and very sensitive receivers.. The screening is um, let's say, somewhat extreme. The HP spectrum analyser on the bench covers a few hundred Hz to 1800Mhz with no gaps and its switcher causes no problems. it too has massive screening and filtering.

It can be done, but for a wooden cabinet radio and an antenna nearby, think of a metal box for the PSU with feedthrough capacitors and multiple turns on ferrite cores filtering every input and every output. Doing a circuit to convert the voltages is only the first part of the job.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 27th Jun 2018, 7:48 am   #34
MrBungle
Nonode
 
MrBungle's Avatar
 
Join Date: Jun 2016
Location: London, UK.
Posts: 2,419
Default Re: Switched Mode Power Supplies

Some of them are quite noisy. I found my MyDel unit spews all over 20m. No lower bands for some reason. The joy is it has a noise offset control so I can keep it out of what I’m working. Also the switcher in tek 22xx’s is noisy as anything.

Have decided to buy a big scary HP 30A 20V linear supply when one comes along. Missed one the other day unfortunately.
MrBungle is offline   Reply With Quote
Old 17th Jul 2018, 6:17 pm   #35
martin_m0nxp
Diode
 
Join Date: Jul 2018
Location: Southampton, Hampshire, UK.
Posts: 7
Default Re: Switched Mode Power Supplies

Did you get anywhere with your SMPS design?

I'd quite like to build a bench supply capable of around 250V @ 250ma DC (so around 50W) - with all the usual facilities you associate with a bench supply (variable voltage and current limit / constant current modes) as I hate the thought of powering any equipment under test without a programmable current limit (and display).

I appreciate that a 50W HV SMPS is no trivial task.... I'll be playing in ltspice when I get some spare moments I think....
martin_m0nxp is offline   Reply With Quote
Old 17th Jul 2018, 6:27 pm   #36
evingar
Octode
 
evingar's Avatar
 
Join Date: Oct 2003
Location: Newbury, Berkshire, UK.
Posts: 1,503
Default Re: Switched Mode Power Supplies

Quote:
Originally Posted by MrBungle View Post
I was running mine off a 12V SLA which can kick a few amps out

Mains I’d kill myself in two minutes thus I will only use linear supplies I built myself there.

Most of my experiments were deriving typical linear and CRT circuit bias from 12V supplies so I could run home brew designs off grid (ie ground referenced -15v, +15v, 5v, 200v, +2kv, -2kv). Managed to destroy about about 25 quids worth of coil formers learning about oscillator stalling

I designed a switching supply many years ago for an automotive smoke meter. There was a need for a high current LT AC supply for use elsewhere, so it made sense to use that for the regulator I/P as well - It makes things loads easier when you are a neophyte power supply designer not having to switch HV and to actually fix the things in a Production environment.
__________________
Chris
evingar is offline   Reply With Quote
Old 17th Jul 2018, 6:30 pm   #37
kellys_eye
Heptode
 
Join Date: Jan 2014
Location: Oban, Scotland, UK.
Posts: 530
Default Re: Switched Mode Power Supplies

The Elektor SMPS will do what you want albeit via a different process. It is a design that takes the output from a standard 0-30V regulated (and current limited) supply and 'upping it' to 0-300V (so a x10 capability).
kellys_eye is offline   Reply With Quote
Old 18th Jul 2018, 11:05 am   #38
martin_m0nxp
Diode
 
Join Date: Jul 2018
Location: Southampton, Hampshire, UK.
Posts: 7
Default Re: Switched Mode Power Supplies

My Google foo clearly isn't what it should be - I can't find such a device / design.... do you have a link?
martin_m0nxp is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 4:54 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Copyright ©2002 - 2018, Paul Stenning.