Topics

Voltage for finals

John Faivre
 

Has anyone tried using a buck boost module to raise the finals voltage supply? I’m worried about the noise from the supply ending up causing spurs on my output.

How high have you successfully run the final voltage, and with what heat sink setup.

thanks for you help 
--
John Faivre WA9SGD

Evan Hand
 

Did not use it for the finals, though did try one for the full system.  Generated some noise, so you do have less sensitivity compared to the normal QRN.  If only used on the finals, may not be an issue.

Since there are 30+ amp switching supplies for other rigs, I would think that if you found the right unit the noise would not be significant.  It is just that the units that I tried did generate noise.  Might also be able to filter out the noise with active or passive filters.

My 2 cents
73
Evan
AC9TU

 

My plan is to use them only on the battery circuit, and possibly to power something that needs a stable 5v supply with proper filtering.

Matthew N8OHU


On August 7, 2019 1:31:55 PM EDT, John Faivre <wa9sgd@...> wrote:
Has anyone tried using a buck boost module to raise the finals voltage supply? I’m worried about the noise from the supply ending up causing spurs on my output.

How high have you successfully run the final voltage, and with what heat sink setup.

thanks for you help 

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Don - KM4UDX
 

John -- do it the reverse.  Start with an old 15-20v laptop power supply. Put a 2a fuse and drive the finals direct from the laptop brick voltage.  Use the buck converter to generate 12v for the board from whatever the laptop brick squrts out.  I added a simple LC ripple filter (used in RC drones) to the 12v output, but it made no detectable difference to anything I could observe. With 17v, I get ~28 watts on 80m, and the normal 3 watts on 10m.  Good enough for all the digital modes, and yes, marginal for SSB DX, says me. Others disagree. 

Increase the size of the heat sinks, or go active if you up the final voltage. Not that the finalls need it in normal operation, but to cover a likely bone-head move by the operator. Don't ask me how I know this. 

In anycase, it is (largley) over driving which generates spurs and uggly bits.  Not the PSU ripple.

Don
km4udx

Don - KM4UDX
 

Here is a pic of the dc-dc which makes 12v for the board (and a few other accessories).  The "native" power from the old laptop brick goes straight  through a fuse and to the finals. Note the LC filter after the dc converter.  I'm not sure it makes any difference, but it looks cool.  

_Dave_ K0MBT
 

I have a supply of lm317 linear regulators. On one of my uBITX radios I used an old laptop supply 18v. I split it off the regulated power at 12.6v for the receive control circuits using a 317 and an auxiliary transistor and let the tx go to 18v. The case on this radio is quite large so I built it all in

MVS Sarma
 

Nice Dave,
 I too did tbat as ì am informed that a small group did such mod. I wòuld like to know the final result àfter mod. Perhaps you need to reduce by adjusting the bias for the finals.

I get Tx audio oñ speaker aftèr such increase.. i got better qso repòrts than earlier, from my rooftop dipole for 40m .
All the best of the improvisation.
Sarma vu3zmv

On Fri, 9 Aug 2019, 5:48 pm _Dave_ K0MBT, <davesters@...> wrote:
I have a supply of lm317 linear regulators. On one of my uBITX  radios I used an old laptop supply 18v. I split it off the regulated power at 12.6v for the receive control circuits using a 317 and an auxiliary transistor and let the tx go to 18v.  The case on this radio is quite large so I built it all in



Don - KM4UDX
 

I vote for DaveK_K0MBT's approach. You get a good 12v for the board, and 18v for the separate "final's wire" on the board's power plug.  Dave used a lm317,and  i used a cheap DC-DC buck converter. Both get the same goal accomplished. Reuse a old brick style laptop (or whatever) power supply, down-convert to get 12v for the board and whatever is native to the finals.  You get a great jump (okay a small jump) in output, and you can feel great about what you did to your hotrod. 

Use a fuse on the power supply line. The finals will work fine without a fuse, but if you are like many of us (well me anyway), you may (will) screw something up and the fuse will save your bacon. hahah.

Don
km4udx

Arvo W0VRA
 

On Fri, Aug 9, 2019 at 12:26 PM, Don - KM4UDX wrote:
You get a great jump (okay a small jump) in output, and you can feel great about what you did to your hotrod. 
How did you adjust the gain on the amp to use the extra voltage?

MVS Sarma
 

Thanls for the photo. Perhaps you might observe increased noise, presumably created by switching regulator.
Let us imagine a linear regulator with lower noise level. At the cost of some power wastage, being linear.
Regards
Sarma vu3zmv

On Thu, 8 Aug 2019, 8:18 pm Don - KM4UDX, <dontAy155@...> wrote:
Here is a pic of the dc-dc which makes 12v for the board (and a few other accessories).  The "native" power from the old laptop brick goes straight  through a fuse and to the finals. Note the LC filter after the dc converter.  I'm not sure it makes any difference, but it looks cool.  

Don - KM4UDX
 

Arvo -- there is no desired adjustment of the gain due to the extra voltage.  That is the goal of the few extra volts...to get a few extra watts.  You most defiantly do not want to adjust anything...just let the extra volts do their magic thing to the finals and bask in the glow of accomplishment. 

Having said that, I did adjust the PA bias per instructions when I first got the rig and used a 12v wall wart for the the whole radio.  Did I need to do that? Na...it was set fine from the factory, I just wanted to hook up my meter and see what every body was talking about. I've never set the bias of anything in my life, never mind two little plastic/metal thingies in the back of the board.  So I used the meter and twisted the little driver pot till I got the 100ma. I even did it a few times just to show off to myself. hahahah. I'm easily impressed. hahaha.

Anyway, there MAY be a case for adjusting the PA basis when you (me) use a higher voltage, I have no idea and didn't look it up.  But I'm not going to mess with it regardless. Having flown high to the sun with my first successful PA bias adjustment, I have zero motivation to try my luck again. hahaha. [but typically I will mess with something until I break it, then I figure that I now understand what is going on...well sort of..hahah]

Now, I did bolt on larger heat sinks (hack sawed from a ancient Pentium CPU heat sink), and I'm glad I did -- given that I left the TX on full power for 15 minutes while in stupid mode (a new digital mode for the inattentive and easily distracted).  Things got all Chernobyl on me, but everything recovered.

So the heat sinks and fuse were the only adjustment I made to accommodate the higher voltage. 

Don - KM4UDX
 

Yes Sarma -- I'm sure there is some extra noise someplace due to using the switching regulator. 

On the other hand, I also gained an extra voltage display, which I love, and I got to solder a new board into the mix, which I love.

If I ever run on battery power, I will feel really smart for using a switching regulator and not a power hungry linear regulator. (actually, I may never do that, but just in case...I can say I planned ahead. Which is totally untrue, it just worked out that way.)

And as an extra bonus, nothing blew up -- which I also love.

Don
km4udx  

_Dave_ K0MBT
 

MVS Sarma
Definitely the LM317 is more power hungry. Have a good sized heat sink on it. It has to adsorb 5 volts at maybe 150-200 milliV But a linear is quiet and I thought that the having the receive quiet was worth it. Not measured however. The computer power supply is linear I am pretty sure.

I was not sure whether to change the settings the bias so I left it alone. The uBITX puts out more than 20w on 40 and higher on 80 which I hardly ever use. It is a v3 radio.

It also has large computer heat sinks At least 4x heavier than the stock heat sinks. a 80mm low speed fan and a printed shroud. The output transistors don't go above ambient in digital mode at 20 watts with 1:1 SWR.

My technique is to touch the transistors. I figure the less heat the less degradation there will be.