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

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 13th May 2021, 2:34 pm   #1
Steve G4WCS
Heptode
 
Steve G4WCS's Avatar
 
Join Date: Sep 2004
Location: Preston, Lancashire, UK.
Posts: 632
Default Big DC power supply

I'd restored a high power 2M linear amplifier a couple of years ago, and wanted a dedicated power supply to drive it. I've been doing some judicious snaffling off eBay, and picked up a toroidal transformer, rectifier and capacitor bank, and 2 large heatsinks, each containing 4 2N3055’s and resistors. Also got hold of 2 large load resistors for testing purposes.

The toroid wasn't the correct voltage, but by removing a few turns at a time, it's OK now, and giving 23V output at 30 amps (warming the test leads up nicely).

I'm going to use the circuit in this article

https://warc.org.uk/?page_id=404

I decided to get around to starting it properly last night, and raided my sheet metal pile. A couple of hours later and a design started to emerge.

A couple of questions. are there any perceived pitfalls with the design presented that I need to be aware of?

Meter wise I'm thinking of the Chinese square ones available for a few pounds each off the internet, not after out and out accuracy, they are more for indication. Has anyone had good or bad experiences with them ?
Attached Thumbnails
Click image for larger version

Name:	D983668B-1CCA-4ED5-AB37-687F6F314265.jpeg
Views:	248
Size:	61.5 KB
ID:	233928   Click image for larger version

Name:	808955D6-AE2C-4BF6-A2C3-529E3756E32A.jpeg
Views:	242
Size:	99.1 KB
ID:	233929   Click image for larger version

Name:	6B3EB3E1-BF9D-44F8-90DF-6E68ADCD5A93.jpeg
Views:	261
Size:	124.4 KB
ID:	233930   Click image for larger version

Name:	CC86EBF8-9A5D-43A7-891D-CC731F3D5F4F (2).jpeg
Views:	264
Size:	113.4 KB
ID:	233934  

Last edited by Cobaltblue; 13th May 2021 at 3:21 pm. Reason: don't include personal details in pictures
Steve G4WCS is offline  
Old 13th May 2021, 3:23 pm   #2
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
Default Re: Big DC power supply

When building 'beefy' low-voltage power supplies it's worth designing to have a 4-terminal output; that is, two big terminals for the actual power and two smaller terminals for 'sense' - that way you can compensate for the voltage-drop in the cables between the PSU and the device it's powering.

[When you're throwing tens of amps around, 0.1 Ohms of connecting-cable resistance becomes a big deal: your 'linear' can suddenly become a lot-less-linear if its supply-voltage sags on peaks]
G6Tanuki is offline  
Old 13th May 2021, 7:29 pm   #3
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,265
Default Re: Big DC power supply

Quote:
Originally Posted by G6Tanuki View Post
When building 'beefy' low-voltage power supplies it's worth designing to have a 4-terminal output; that is, two big terminals for the actual power and two smaller terminals for 'sense' - that way you can compensate for the voltage-drop in the cables between the PSU and the device it's powering.

[When you're throwing tens of amps around, 0.1 Ohms of connecting-cable resistance becomes a big deal: your 'linear' can suddenly become a lot-less-linear if its supply-voltage sags on peaks]
If you include remote voltage sense, include a limit on how much the voltage can be increased if one of the sense lines is open circuit. Probably also worth putting overvoltage crowbar circuit on the target device.

Also watch out for ceramic and tantalum capacitors used for power decoupling when you have a beefy power supply. If a device might be powered while unattended take extra care, maybe a fusible link to the decoupling caps.
Mark1960 is offline  
Old 13th May 2021, 7:35 pm   #4
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,265
Default Re: Big DC power supply

Frequency response on remote voltage sense power supplies can also be a problem, causing ringing on the supply voltage, this used to be a problem with GSM transmitters due to pulsed current requirement of the power amps.
Mark1960 is offline  
Old 13th May 2021, 7:45 pm   #5
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,172
Default Re: Big DC power supply

Hi a common trick is to permanantly link the output and sense lines (at the terminals )with a resistor of under 100R. This prevents the sense lines floating if they go o/c

Ed
Ed_Dinning is online now  
Old 13th May 2021, 7:47 pm   #6
Silicon
Nonode
 
Join Date: Jun 2007
Location: Coulsdon, London, UK.
Posts: 2,152
Default Re: Big DC power supply

What current will the power supply provide?

30 Amps passing through a diode will generate about 20 to 30 watts of heat in each device.

30 Amps passing through a soldered joint may also raise problems.

It is good to see a 'soft start' facility. Without it, the turn on surge would be impressive.
Silicon is offline  
Old 13th May 2021, 8:38 pm   #7
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

That looks to be the same PSU that I built back in the late 1980s. The circuit and project details were given in an amateur radio mag. The good news is that I found that the PSU was very rugged and worked well. I still have it although it doesn't get used anymore.

There are a few niggles with the design. If you turn off the PSU at the mains and then decide to turn it back on again a few seconds later then this will almost certainly blow the mains input fuse. This is because the soft start protection takes a while to drop out after turning off the PSU. So I would advise you buy a pack of 10 input fuses as it is very easy to blow the input fuse like this when setting up and first testing out the various protection modes the PSU has.

I put the current limit control on the outside of the chassis allowing the current limit to be adjusted from about 3A to 'a lot'.

When the PSU hits the current limit it will trip out the mains input relay so the PSU will turn itself off. This therefore isn't the same as a classic current limit on a big bench PSU. It does mean you have to make sure you set the current limit high enough to prevent your PSU turning itself off on SSB speech peaks when using it to power a 100W transmitter. In reality, I never had a problem with this because I turned up the limit to >25A when using it with a 100W SSB radio. I've also used this PSU many times with the current limit set to flat out when needing to start a car with a flat battery. The PSU always survives this abuse and it always starts the car!

Note that the main controller PCB has exposed mains AC voltage on it and also the mains input wiring is more complicated than on a classic power supply. This means there's lots of places to get a shock from inside the chassis. I do think the controller PCB needs a perspex screen over the mains connections and care must be taken to insulate all the mains input wiring including all the soft start wiring.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 13th May 2021, 8:56 pm   #8
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
Default Re: Big DC power supply

Quote:
Originally Posted by Mark1960 View Post
Frequency response on remote voltage sense power supplies can also be a problem, causing ringing on the supply voltage, this used to be a problem with GSM transmitters due to pulsed current requirement of the power amps.
This can indeed be a problem: trying to produce a 'stabilised' power-supply for an audio amplifier or a RF/SSB linear-amp that has to follow current-demand at an audio-syllabic rate can easily turn into an exercise in designing what amounts to another audio-amplifier!

Slew-rates, settle-times and loop-damping...
G6Tanuki is offline  
Old 13th May 2021, 8:59 pm   #9
Steve G4WCS
Heptode
 
Steve G4WCS's Avatar
 
Join Date: Sep 2004
Location: Preston, Lancashire, UK.
Posts: 632
Default Re: Big DC power supply

Thanks all. More tin bashing tonight. Ready to cut and fold the case top from the remains of a filing cabinet
Attached Thumbnails
Click image for larger version

Name:	0B5369BE-8FE6-4FD9-BF4A-17685AC44C50.jpeg
Views:	143
Size:	112.9 KB
ID:	233964  
Steve G4WCS is offline  
Old 13th May 2021, 10:27 pm   #10
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

It looks like you have the main heatsinks fitted just inside the enclosure which is a bit unconventional. My PSU heatsinks were fitted externally at the rear of the PSU and these still used to get quite hot when the PSU was tested into a continuous load of (say) 15A. Things weren't so bad when using a typical 100W HF radio on SSB though.

I purchased the same toroidal transformer as shown in the original article and this was a very pretty transformer. However, it can produce quite a bit of low frequency hum noise. During the day this wasn't a problem but back then this power supply was in my bedroom and I could certainly hear it humming away at night if I forgot to turn the PSU off. It might be worth a few experiments with your chassis/lid/vents to make sure you don't let the enclosure act as a speaker for the hum noise from the transformer.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 14th May 2021, 1:22 am   #11
Terry_VK5TM
Nonode
 
Terry_VK5TM's Avatar
 
Join Date: Oct 2010
Location: Tintinara, South Australia, Australia
Posts: 2,324
Default Re: Big DC power supply

The only thing I would add is a noise filter after the on/off switch on the cabinet and before the rest of the circuit - 723's get quite annoyed with spikes on their supply input and tend to give up the ghost easily.

I used to make 723 based PSU's for a company here in Oz many moons ago and although the on/off switches they supplied were nice big chunky things, the did tend to put a nice spike on the line when switched.
__________________
Terry VK5TM
https://www.vk5tm.com/
Terry_VK5TM is offline  
Old 14th May 2021, 7:24 pm   #12
Steve G4WCS
Heptode
 
Steve G4WCS's Avatar
 
Join Date: Sep 2004
Location: Preston, Lancashire, UK.
Posts: 632
Default Re: Big DC power supply

Case is finished. Plenty airflow around the heatsinks, they are mounted on stand offs so plenty airflow over them Not even thought about teansformer hum, will have to see how it goes, its only cost my time
Attached Thumbnails
Click image for larger version

Name:	8F814EFD-8BE6-4C2E-9E1C-91DA570796D1.jpeg
Views:	125
Size:	117.2 KB
ID:	234034   Click image for larger version

Name:	2533B21D-CFCA-4696-8098-D2A568F158ED.jpeg
Views:	125
Size:	142.3 KB
ID:	234035  
Steve G4WCS is offline  
Old 14th May 2021, 8:59 pm   #13
g4wim_tim
Pentode
 
Join Date: Nov 2018
Location: Freckleton, Lancashire, UK.
Posts: 232
Default Re: Big DC power supply

Hi Steve,

I built practically the same design in the mid 80's and its still working well even now.

Mine was a little different and rated to a conservative 50 amps at 13.8V 100% duty cycle.

It uses six 50amp motorola darlingtons each on thier own heatsink and all driven from the the 723 output directly, but the 723 is powered sepearately from a PSU voltage a few volts higher than the main high current power supply.

Doing this makes it easier to drive the series pass devices closer to saturation without having an excessively high voltage across them.

The design as drawn has to have enough raw volts to take into the account the 723 and series pass devices voltage drop plus head room resulting in a larger than necessary drop across the series pass devices.

Mine also has two 50 amp bridge rectifiers and 200,000uF of smoothing to reduce raw ripple voltage again to minimize power loss / heat in the series pass devices.

Building a big high current power supply takes some doing !

73 Tim
__________________
Building and tinkering for over 50 years, from 807's to digital amateur TV.
g4wim_tim is offline  
Old 15th May 2021, 12:34 am   #14
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

Assuming ~240V AC mains I think the original transformer typically produced 16.5Vrms at the secondary. This would mean the (smoothed) bridge rectifier would give out about 22V DC at light loads. This would mean there would be about 8V across the pass transistors.

At higher load currents this 22V DC voltage at the pass transistor inputs will drop due to losses in the bridge and the various connections and there will be some ripple here as well.

I used similar 1.1degC/W rated heatsinks as the original on my build and I made up an adjustable and switchable load. This was made from some huge 0.47R and 0.22R and 10R metal clad resistors mounted to a big aluminium plate and I still have this load up in the loft somewhere. I recall switching the load in and out and measuring the dynamic regulation from the supply. It was very good even with a harsh test like this. To prevent transient sagging at the output it did require the current limit setting to be set quite high and I think I could rapidly switch between load currents of 1.4A and about 20A in this test.

I do remember that it got very hot at a continuous 15A load current so I'm not sure how well it would perform on a 20A continuous soak test. I'm not sure the original heatsinks are big enough for this test and my old test load certainly wasn't up to it.

However, a typical 100W CW/SSB radio will probably run at a much lower duty cycle even though it will probably gulp current at 20A peak. I don't recall the PSU heatsinks getting very hot when transmitting 100W SSB or CW with an old Yaesu FT707 or a Trio TS430S.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 15th May 2021, 12:54 am   #15
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

The (rebadged?) Palomar 2m amplifier looks a bit like an old KLM 2m amplifier. These were available in 70W/140W/160W flavours. The heatsink looks very small on the amp but it does look like it has some sort of fan as well.

Looking closer, yours looks like a 140W or 160W amp and I think these had a pair of MRF24x BJTs as the final stage. Assuming 150W peak output this would mean 11A peak current if it was 100% efficient. However, the efficiency is going to be in the ballpark of 50% so the amp might draw just over 20A on peaks.

It probably wouldn't be wise to run that amp at a high duty cycle so the PSU design you have chosen should perform very well if run at 50% duty. I think your larger PSU heatsinks would barely run 'coffee cup' warm at a continuous 12A load current for example.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 15th May 2021, 2:33 pm   #16
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

I dug my old homebrew 13.8V PSU out today and cleaned the dust off it. See the image below. This PSU design is the same as the one linked to in the original post. It hasn't been used in a long time but it still works! It produced 13.69V on a DVM and the electronic load which isn't bad considering it hasn't been adjusted in nearly 35 years.

I did a few tests on it with a modern electronic load and looked at it with a thermal camera. I found a couple of hot connections in a terminal block that needed tightening and one of the pass transistors was running a bit hotter than the others. The TO-3 fasteners needed a quick nip up to cure this. I did a 1 hour soak test at 12A load current and then looked at the temperature of the two big heatsinks and the metal case of each 2N3055 pass transistor.

Each heatsink reached 68degC when viewed right next to each 2N3055 and the hottest 2N3055 case temperature was at 80degC. The coolest 2N3055 was at 78degC.

I also modified an old excel spreadsheet to include the option of multiple heatsinks and pass transistors and this predicted 65degC for the heatsinks and 75degC for the case of the 2N3055 pass transistors. This was quite close to reality. I measured 6.5V drop across the pass transistors and guessed the heatsinks at 1.1degC/W. I'm pretty sure I bought heatsinks with the same rating as the original design but they look quite small. Each has a volume of about 135mm x 135mm x 31mm but they have lots of fins so this helps with the performance.

See below for an image of my PSU. As is often the case, I never really finished this PSU project. I never got around to adding a meter at the front and I never made any decent labels for the front panel. It also has an aux 5.0V supply output and I can't remember why I added this. I think it is only rated to 250mA max. There's also a screenshot of the excel spreadsheet.

If I increase the load current to 20A on the spreadsheet then it indicates that the junction temperature of the 2N3055 transistors would be quite worryingly high if soak tested at full duty. However, I only really built it to power a typical 100W transmitter and this would never be run at full duty like this.
Attached Thumbnails
Click image for larger version

Name:	13v8_view.jpg
Views:	148
Size:	76.0 KB
ID:	234088   Click image for larger version

Name:	13v8_soak.jpg
Views:	82
Size:	99.0 KB
ID:	234089  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 15th May 2021, 8:34 pm   #17
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

Here's a transient response plot where the current is briefly ramped from about 2.5A to 15A. It shows about a 50mV sag in the 13.7V output when the current briefly ramps up to 15A.

The scope timebase is 500us/div and the green trace is taken from the current sampler in the electronic load.

There is very little overshoot in the plot below. However, when I decrease the risetime of the current pulse (to get a faster edge) there is some ringing that has a 20us period. This might be due to the way I wired the connections to the control board. However, in normal service it would never see a step in current that would be this fast.

I suppose I could take it apart and try and improve the transient response. The 723 regulator has a wide loop bandwidth so maybe it isn't a surprise to see some slight ringing at 50kHz when severely provoked. I think this level of performance is already very good for a low cost PSU like this and the plot below should give you some idea of what to expect in terms of dynamic regulation.
Attached Thumbnails
Click image for larger version

Name:	13v8_tr.gif
Views:	81
Size:	13.2 KB
ID:	234129  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 15th May 2021 at 8:41 pm.
G0HZU_JMR is offline  
Old 15th May 2021, 9:39 pm   #18
Steve G4WCS
Heptode
 
Steve G4WCS's Avatar
 
Join Date: Sep 2004
Location: Preston, Lancashire, UK.
Posts: 632
Default Re: Big DC power supply

Evening all, thanks for the interest. Tim, the 736 is still performing faultlessly but not yet on 23cm .

Found this thread from when I unwound the transformer.

https://www.vintage-radio.net/forum/...ghlight=Mc3423

The warc design used 4 pass transistors but acquired the 2 heatsinks with transistors already fitted for less than the price of bare heatsinks.

Yes the amplifier is the 160 watt version with a driver and 2 finals. It was bought off ebay from a forum member as a kit of parts so was a bit of an unlnown. I rebuilt it and tidied the wiring, and added a small 12v fan and ventilation holes as parts inside have the potential to get warm and there was no means of airflow into and out of the case. Ive given it some brief testing and it was pulling 20 amps on a whistle test, but thats as far as I got with it before putting it back on the shelf

I do have a drae 24 amp supply but wanted to keep that hooked up to the HF gear ( I have a sommerkamp badged ft-707 that is a superb radio to use ) and have a dedicated one for driving the linear. Its probably totally overkill given the duty cycle on ssb and the use it will get given how quiet the bands are , but its something to keep me mentally stimulated through the drudgery of working from home and not getting out anywhere at the moment. Not got too long to go now though, decided to finish next april

Last edited by Steve G4WCS; 15th May 2021 at 9:50 pm.
Steve G4WCS is offline  
Old 15th May 2021, 11:04 pm   #19
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Big DC power supply

It is a nice PSU project and I think you will be pleased with the performance it offers. I couldn't resist looking inside mine and I spotted that the wiring could be improved. It now shows about a 10mV dip in voltage (from 13.70V to 13.69V) when stepping the load current from 2.5A to 15A.

Sadly, the ringing is still there if I speed up the edges on the 15A pulse. This may be something to do with the circuit design. I don't have any experience with the 723 regulator but it may be possible to improve the transient response with some subtle circuit changes.

However, the result below is now looking very impressive. It might be possible to improve it still further but there probably isn't much point as I rarely use this old PSU anymore. Having done these performance tests this weekend I'm even more impressed with it. Back when I built it in the late 1980s I think I only had an old Tek 585 scope, a few load resistors and a basic DMM to test it with.
Attached Thumbnails
Click image for larger version

Name:	13v8_tr2.gif
Views:	67
Size:	12.8 KB
ID:	234144  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 16th May 2021, 5:20 pm   #20
Steve G4WCS
Heptode
 
Steve G4WCS's Avatar
 
Join Date: Sep 2004
Location: Preston, Lancashire, UK.
Posts: 632
Default Re: Big DC power supply

According to my “memories” on social media which are far better than my memory, it was 2 years ago when I finished the amplifier. Id used a riding helmet bag for the fan filter
Attached Thumbnails
Click image for larger version

Name:	5D59616A-D638-49F4-A3C2-534B86FBCB89.jpg
Views:	98
Size:	51.8 KB
ID:	234211   Click image for larger version

Name:	AE973D9C-4EC9-423B-B245-BE3DA901D9DF.jpg
Views:	94
Size:	48.0 KB
ID:	234212   Click image for larger version

Name:	43EB9B53-7EEE-4CBD-81A8-79E110AED7D9.jpg
Views:	95
Size:	74.5 KB
ID:	234213   Click image for larger version

Name:	9F8A97EC-6FFD-4E70-B460-82212590330A.jpg
Views:	99
Size:	41.0 KB
ID:	234214  
Steve G4WCS is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 11:25 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 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.