Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Heat sink
|
OK gang
I want to try to run the PA off 24 volts. Any suggestions as to size or type of heatsink. Do I have to re-adjust the bias, etc? I was wondering if anyone has used a 7812 to regulate power to the main board, or is it better to use 2 supplies? The radio is working very well, with multiple contacts on both cw and ssb over the last week.? 73 Brent |
|
The rig draws around 150ma, so regulating down from 24 to 12v that 7812 would be dissipating 2 Watts.
toggle quoted message
Show quoted text
Could be made to work, but will need a heat sink. ?I'd go with two supplies. Or maybe a boost switcher that turns on only when transmitting when you need more power than 12v gives ? ?? I have no idea if that thing works, or if there is an easy way to enable/disable it. Or have two 12v batteries, normally operated in parallel, run in series to drive the PA at elevated power, main rig continues to run only off the bottom battery. Having at least a diode on the main power into the board is not a bad idea to give reverse power protection. On heatsink for the IRF510, just go big. ?Very poor thermal transfer from die to TO220 tab on those guys, so we want to keep the tab as cool as we can. ?/g/BITX20/message/22597 I haven't tried running mine at over 12v yet, haven't had time to play with it. Most folks here running at 24v seem to be leaving the bias set at 100ma, but that has to be clipping the bottoms off the transmitted output, to be dealt with by the transmit LPF. Though if it's clean enough, then that's the way to go as otherwise you are burning more power in the IRF510, potentially zapping it with heat. ?Would be interesting to check the IRF510 for linearity? when operating at 24v and 100ma of bias. ? And the transmissions for meeting legal requirements. Best solution would be to steal the push-pull amp? from the uBitx, as neither IRF510 there is drawing much current until you start shouting into that mike. ? ??/g/BITX20/message/33255 Jerry, KE7ER On Sat, Oct 7, 2017 at 02:07 pm, Brent Seres wrote:
I want to try to run the PA off 24 volts. Any suggestions as to size or type of heatsink. |
|
Jerry has some good ideas but the easiest solution to
clipping is to slightly increase the drain voltage. Assuming the same drive level. Most likely, clipping will occur on the higher frequency peaks (if it does). The level at which clipping begins is more or less controlled by the IF bandwidth, i.e., the xtal filter bandwidth -- the more highs that get through, the more likely it is that there is clipping. If it occurs on only an occasional high, it can usually be ignored. If it is more often, it will have to be addressed. The simplest solution is to increase the drain voltage to 28v or 30v. That allows a greater excursion of the signal in the final. The IRF510 is a 70v device; I would not try to get much beyond 60v Pk-Pk of signal out of it. Alternatively, you can get rid of the distortion in the LPF, but it is much more work and figuring to do it that way. Mind you, the LPF still has to handle the power you put through it, no matter what. Yes, you will require a heat sink. The heat transfer derating coefficient is about .30 degrees C per watt! The maximum allowed difference between the junction and ambient temperature is 3.5 degrees C per watt. That means that at 10W output the maximum temperature difference between the very small chip center and ambient is 35 degrees! You probably cannot get too big a heat sink if you are using the device for AM or SSB. Something like an Intel cpu heat sink or a large sink for a voltage converter works well. (And the cpu sink gives you a fan as well! Hi.) And use a good transfer grease between the tab and sink. It isn't rocket science. Just think about how to get rid of a lot of heat generation; it will come to you. john AD5YE |
|
A diode is a good start. ?
toggle quoted message
Show quoted text
But Q13 and R141 can exceed dissipation limits even when running at 12v, If running off a fully charged lead-acid battery at 14v you will be severely stressing those parts. I think an LM2940 between your power supply and the main Bitx40 12v power connection would be ideal. ? Unlike the older LM7812, the LM2940 will provide reverse battery protection (just like a diode). It will also protect Q13 and R141 from an incoming supply that exceeds 12v. Power to the IRF510 doesn't need protection from a couple extra volts on the power supply like that,?but could use protection from too much drain current. ? Perhaps just a 3A fast blow fuse. Better yet, a current sense circuit that grounds out the gate bias voltage when it sees trouble: ? ? Another plan is to have a dozen spare IRF510's on hand, they're quite cheap. I'll defer to John and the others here on how to avoid clipping on that IRF510 and how much of a problem it might be. ?Not something I have fully figured out. ? But my suspicion is that with only 100ma of quiescent drain current through the IRF510, it will have to have high side drain current peaks of an Amp or two in order to generate 20 Watts of RF. And since drive into the gate is AC coupled from Q14 and thus driving the IRF510 symetricaly about the quiescent gate voltage, it will spend an awful lot of time with drain current at zero when it sees the incoming low going peaks. However, most of that distortion would be harmonics of the transmit frequency, and easily filtered by the LPF. On Sat, Oct 7, 2017 at 03:06 pm, Jerry Gaffke wrote:
Having at least a diode on the main power into the board is not a bad idea to give reverse power protection. |
|
Vince Vielhaber
I use a 7812 to power the radio and send the unregulated to the PA. Take note, the cap on the PA power is only rated for 25 volts.
toggle quoted message
Show quoted text
Vince. On 10/07/2017 05:07 PM, Brent Seres wrote:
OK gang --
Michigan VHF Corp. |
|
I do this too.? I'm running a large 2"x2"x1.5" chipset heatsink with good thermal grease.? It doesn't even get warm. On Oct 7, 2017 7:33 PM, "Vince Vielhaber" <vev@...> wrote: I use a 7812 to power the radio and send the unregulated to the PA. Take note, the cap on the PA power is only rated for 25 volts. |
|
I relocated the IRF510 to the other side of the board and have it screwed to the metal case I have the Bitx40 in. I currently use a 19v laptop supply to power the rig with 19v on the IRF510 and a small DC DC converter set to 12v to run the rest of the rig. This allows me to power the rig from any voltage above 12v to a max I have tested of 28v without having to make adjustments or mods to have the correct volts on the board and raduino.
I have 2 rigs set up this way, 1 on 40 and the other on 80 meters,neither have any issues with getting too hot at least on SSB. -- Mike VK3XL |
|
I only wonder whether these laptop psu and dc -dc converter shoul not be a source of internal noise or interference. All the bestregards Sarma ? On Mon, Oct 9, 2017 at 10:00 AM, vk3xl via Groups.Io <vk3xl@...> wrote: I relocated the IRF510 to the other side of the board and have it screwed to the metal case I have the Bitx40 in. I currently use a 19v laptop supply to power the rig with 19v on the IRF510 and a small DC DC converter set to 12v to run the rest of the rig. This allows me to power the rig from any voltage above 12v to a max I have tested of 28v without having to make adjustments or mods to have the correct volts on the board and raduino. |
|
I've got a cheapie $10 buck regulator off Amazon that generates some serious noise to my BitX, and I tried 2 different brands of mini DVM's (as a simple battery indicator) and both kicked off even more noise, but I also have an even cheaper boost converter built into the case to get my finals up to 20V that is nearly silent - Only a slight quiet hiss can be heard.? It seems to be rather hit or miss.
|
¿ªÔÆÌåÓýI installed a RF choke and decoupling capacitors on the input and output of mine and it helped quite a bit. I also use a DC Boost Converter from Amazon. ? ? From: [email protected] [mailto:[email protected]]
On Behalf Of KE0GYC
Sent: Monday, October 09, 2017 2:36 PM To: [email protected] Subject: Re: [BITX20] Heat sink ? I've got a cheapie $10 buck regulator off Amazon that generates some serious noise to my BitX, and I tried 2 different brands of mini DVM's (as a simple battery indicator) and both kicked off even more noise, but I also have an even cheaper boost converter built into the case to get my finals up to 20V that is nearly silent - Only a slight quiet hiss can be heard.? It seems to be rather hit or miss. |
|
WERNER G VAVKEN
¿ªÔÆÌåÓýJohn,How much power do you get out of the rig, now, on 40M with 20V in the TX PA? I just ordered a DC Boost Converter from Amazon, too. I ordered this one: SUNKEE LM2577 DC-DC Adjustable Step-up Power Converter ModuleIs that the one you got? Seems to be good for 2amps output. Thanks, Werner WB6RAW
|
|
Jerry had some comments and concerns with several of the stock bitx components being stressed at voltages over 12 volts. I'm currently using a supply running about 13.5 volts. Has anyone had any problems with things failing when using a similar supply or fully charged lead acid batteries? I know the PA is fine, just wondering about the rest of the board.
Thanks Brent |
|
Raj has seen several boards with R141 blown: ?/g/BITX20/message/32096
toggle quoted message
Show quoted text
That resistor and the mmbt3904 at Q13 have quiescent currents that exceed their dissipation ratings.? Q13 is connected to the antenna, can also blow if a nearby QRO transmitter causes Q13's Vbe to get excessive. Transmit LPF caps will see around 100v pk-to-pk of RF (centered on ground) when driving 20 watts into 50 ohms, at high SWR they will see considerably worse. ? But many in the forum are successfully running with 24v into the IRF510. I suspect most reports of damage are due to heat in the IRF510.? That's how you learn. ?Good thing the IRF510's are so cheap. Jerry On Mon, Oct 9, 2017 at 05:53 pm, Brent Seres wrote: Jerry had some comments and concerns with several of the stock bitx components being stressed at voltages over 12 volts. I'm currently using a supply running about 13.5 volts. Has anyone had any problems with things failing when using a similar supply or fully charged lead acid batteries? I know the PA is fine, just wondering about the rest of the board. |
|
Hi, ?I have 3 bitx40 v3 , all running 13.5v.? The have been running perfectly ok... for nearly 24 hours a day... for weeks at a time. I use the same voltage for both rig and PA - output around 11 watts. I did try 19v at the PA , but only got around 5-6 watts more... Not enough for me to keep it at that level.. However, I do have a large computer cpu heatsink and a low noise (can barely hear it)? fan for the PA. I run hours of straight none stop digital modes and at the 13.5v for rig and PA... It is cool, hardly detectable it is on even on Joe VE1BWV On Mon, Oct 9, 2017 at 9:53 PM, Brent Seres <brentseres1@...> wrote: Jerry had some comments and concerns with several of the stock bitx components being stressed at voltages over 12 volts. I'm currently using a supply running about 13.5 volts. Has anyone had any problems with things failing when using a similar supply or fully charged lead acid batteries? I know the PA is fine, just wondering about the rest of the board. |
|
Thanks Jerry
I may just use the radio at 5 watts for now, and may put a 7812 on the main board power for insurance. As long as the supply is far enough above 12v to allow the regulator to work, it should relieve some of the stress. My one supply runs as high as 14v, so that probably stresses r141 an q13 quite a bit. I'm making lots of qsos on 5 watts and have several other radios if I want to go qro. 73 Brent |
|
Werner WB6RAW wrote...
SUNKEE LM2577 DC-DC Adjustable Step-up Power Converter ModuleI haven't had time to take it out of the packaging yet, but I ordered this one from eBay. It is rated t 6A (9A max) so it's a bit of over-kill, but I'll give it a try. It didn't take much more than two weeks to arrive. -- 73 Keith VE7GDH |
|
6w sounds about right.
toggle quoted message
Show quoted text
11w sounds really high for 13.5v, I'm curious how that was measured. Jerry On Tue, Oct 10, 2017 at 05:57 am, Brent Seres wrote:
Did you make any mods to get 11 watts at 13.5v? I only get about 6 on cw. |
|
No? mods but have adjusted PA? for more output.. It started as an experiment to see what? I could get out of the Final. I started with lower values and went up over time .. I had purchased a half dozen finals from digikey --- just in case... but never used one yet I have all 3 of my bitx running from 9-11 watts. I have large heat sinks and fan on each . using digital modes - jt65 and ft8 - they stay very very cool. They have been running? at the above levels for hours of continual working the digital modes and no issues... Measurements with 2 power meter - and swr 1.0 Also same level recorded into to a dummy load. Joe ve1bwv On Tue, Oct 10, 2017 at 1:52 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote: 6w sounds about right. |
|
You made power output readings using three different measurement techniques
toggle quoted message
Show quoted text
and they all pretty much agreed? ?Guess I won't argue with that. So, what adjustments? ?R136 fully CCW? Could be you are driving that IRF510 into non-linear class C operation. That could be fine for some digital modes, not so good for digital modes that have more than one simultaneous tone. ?And not so good for phone. Though even with R136 fully CCW for class C, I doubt most of these rigs would give 11 Watts from 13.5v. Jerry On Tue, Oct 10, 2017 at 10:01 am, Joe wrote: No? mods but have adjusted PA? for more output.. |
Brent Seres/ VE3CUS
to navigate to use esc to dismiss