For me, I'd be more than happy with air cooling, as I'd be running SSB with it 95% of the time, so my cycle duty will be pretty low. I don't need monstrous power, I'd be happy running it at 1200W PEP with good air cooling. I'll be watching, as I'd love to purchase one of your pallets when you have them completed. Thank you for another great video!
Water cooling is the best way of dissipating heat, strange enough I just converted one of my amp from air to water cool, it now runs at 38c on a 5 min key down compared to 52c . SO I WILL SAY DEFINITELY WATER COOL .
I’d say air cooling with voltage ranges from 48-56 volts would be ideal for the average user. Adding a thermal shutdown switch to the heatsink / spreader would decrease the chance of an over temp situation. Some commercial amplifiers like TPL use a to-220 like package thermal switch that would be ideal for disabling the keying circuit.
Air cooled. Do you think we really need a few hundred extra watts? Me personally I would run it at right around 1000 watts anyway. Needs to be biased for sideband.
Oh, just want to mention. The other day, I got an oven for repair. It seems the heating element has a 50 ohm resistance. And.... I connected a radio and had a 1,3 over 2 SWR. Can you believe that? So there you go... for 20 euro's, you have a 2000W dummy load! They are rated 2500 watts. I didn't investigate more, but how crazy things can be.
Air cooled sounds like the way to go. 50 volts is safe and if you push it you take the chance like any other amp. It's on the operator. Maybe you could put a temperature sensor on the unit so it can be monitored. Thanks again for looking out. 73's
Water Cooling is silent and you can place the radiator in the engine compartment, the amp under the seat with a quiet high flow temperature-controlled 220mm fan.
I do run a BLF189XRA at full power with air cooling, however, there is larger copper heatsink with fins machined in. There are 2 large bore holes that run across the heatsink, (very close to the LDMOS) and 2 heatpipes are soldered in. Then a 350 Watt fan blows directly on the copper fins. So it is possible to use air cooling at full power.
Using this single LDMOS part, you need to pull 400-500W from a very small area. This suggests water cooling. An optimal air cooled design might use multiple lower power devices for a larger footprint area. Air cooling this part in the middle of a 100 square inch heatsink will cause a hotspot in the middle that gets hotter until the system reaches equilibrium. Time to equilibrium will be variable with ambient temp, air flow, etc. Air cooled device temperature should monitored and the power derating determined at equilibrium. Semiconductor device life is reduced exponentially with a linear junction temperature increase. That's makes it easy to pop a device!
I think you should talk with Jim, DD3DT regarding the cooling. Look at his QRZ page on how he cools his pallet. I'm pretty sure he will help you out. He is an American living in Germany. No language barrier... Regarding your power supply, you can find thousands of decommissioned and refurbished cheap computer power supplies on Ebay that are sold at a bargain price (Got mine for 10 euro each!). I bought myself 5 power supplies on ebay, model HSTNS-PD18. Total cost: 50 euro + 15 euro shipping costs! You can crank them up to 13.8 volts with a simple hack. Search for hack HSTNS-PD18 power supply. They are 50 (and more) amps each @ 13,8V. All in series, without mod, you get 60V. They are very quiet, solid, no noise on the radio. I had a Maas SPS50 50 amp power supply. And I replaced that junk with the much better HSTNS-PD18 power supply. Throw that junk power supply from China out of the window. You will never look back once you used these computer power supplies. I love the HSTNS-PD18 because they are 750 Watt's @ 12V. What is the current draw of a single pallet @ 50 and 60V?
I have a question I have a brute force 350 volt Cb and ssb linear the tech who worked on it said he got it up to 400 watts it has two 2879s not red dot what do you think also it's behind a general Lee 33 HP. Supposed to be paired.
I Always Run equipment around 80% of its rated output. if you want more power go bigger not drive it harder. I also understand not everyone thinks this way and I am not judging them. so therefore the best route would be leave the option open to us the end user. Does it have to be one or the other? or can you offer us a pallet with both options? Good work I look forward to getting one of these pallets either way....
Air cooled is simpler and more rugged. I personally care more about (or am willing to spend more for) a more rugged system than a higher performance one. It might also be possible to do a hybrid to keep some ruggedness but push the limits.
Hi Greetings from Germany I think your video is very informative I have a question I bought a couple of MRF9180s and built a small amplifier with them, now I have unfortunately destroyed a few of them because they have become too hot I only run them at 14.4V Output power is around 80W with a 500mW drive how can I better cool the LDMOS ? water cooling is unfortunately not an option would it be of any benefit if I use heatpipes in the copper heatspreader?
That would be cool 💯%🎯✅, I think silent with water is next level for real,I wonder if a pure silver heatsink would cool better than just aluminum 🤔🤔.. cost wise it's probably like 300.00$ instead of 30$.
It would seem to me that individuals that are looking for a quality amp kit/pallet that is capable of 100% duty cycle is pretty much a no-brainer. Personally for myself I would be more into the 100% duty cycle amp so as not to have to worry about limited key down time / thermal shutdown or destruction of the device. Anybody that is into rag chewing over the air would more than likely be interested in something that is 100% duty cycle. don't get me wrong key down boxes are fun but very limited as far as transmit time in my opinion you cannot stomp on the loud pedal for 15 to 30 minutes at a time ! And as I stated once before I am a long winded MoFo I enjoy my key down time ! That being said I personally think that water cooled 65 volt LDMOS seems to be the future. I think you have done one hell of a job with this project.
You could sell the board complete, then sell the custom parts required for water cooling, leave it up to the consumer for the cooling method. They buy the board, solder it to a recommended spreader size and heatsink, or they could buy the CNC cooper piece and the recommended water cooled components. I didn't realize that you were making the copper for the liquid cooling that is neat. I think the 50v power supply is much easier to find and purchase. I just ordered the 62v from PCS this weekend. Currently, I use a MeanWell which only goes to 57-58 volts for the MRFX1K80H.
I might end up eating my hat on the power supply thing... just found some 63v linear unregulated supplies that might do the job. Ordered a couple, will test soon.
I do believe I've seen another comment on here about this amp being built for 160 through 6 meters, something like that I would love to see. Just curious if that is another project you may be working on in the near future. And I said it once and I'll say it again I think you have done one hell of a job with this project.
air cooled is every were. their are no kits availible for water cooled. i have built many air cooled LDMOS boxes and im ready to step up to water. !!!!!!!!!!!! please offer a water kit, anybody can figure out how to put a big old heat sink and fan in their box. personaly im looking to step it up on the reliability and 'cool factor' . maby you could offer some simple guid lines for an air cooled set up for the guys on a low budget. and i will take a two port out put combiner and two of your cool low pass boards for for 11 meter. for experimental purposes only. WESTSIDE STATION
Why do you need to run a half inch tube for the water supply couldn't you use a 1/4 " tubing and make a small coil underneath the transistors .I just confused on why your cooling lines are so big
What would be cool for me if you built a self contained system that was all plug and play for the guy who was not very well familiar with building this stuff for himself. Or who did have the time, now that would be cool for a guy like me.
Well, me personally, if I had extra money to spend on hard to find power supplies and then filtering components, I'd buy a 4 tube Z box. I would be way more likely to go with air cooled and 50v. But that's just me.
hey just make a block that would fit to a CPU water cooler. this would fit in a case and be install on the bottom of the amp. oh yea almost forgot Mr bbi sent me
cant share an ebay link to youtube for some reason..... search for these on ebay 12V 60W TEC1-12706 Cooling Peltier Plate Thermoelectric Cooler Heat Module
Good Q. TLTs are very good at passing a wide range of frequencies including unwanted harmonics of the fundamental. If an amp is only intended for a narrow range... a conventional transformer can (if carefully selected and tuned) significantly filter the unwanted harmonics.
Huh? Those transformers you're using are wideband-transformers. Those transformers are low-Q devices... that's why they have a wide bandwidth. The TLT transformers are also low-Q. I was just interested why you weren't using them. I don't use them either... for several reasons. They're certainly not right for every application. Both types have their problems... albeit they're different problems.
@@TelstarElectronics "Good Q" meant good question. And no, as tuned on these pallets, they are certainly not wide banded. The leakage inductance alone makes tranformers like this reject higher frequencies, not to mention the shunt capacitance used to hit the target impedance.
"Q"=question... OK... LOL. You can see where I got that confused. My experience has been that trying to get greedy and going above a 9:1 impedance ratio on the conventional transformer gives poor results. The leakage in those transformers is not the greatest. I notice that you don't seem to exceed the 9:1 ratio... that's good. But with the very high power devices you're using, don't you really require a higher transformation ratio? I would think you'd need like 25:1 or so. That's not easy... a problem I'm wrestling with now.
@@TelstarElectronics haha yeah my bad. Things definitely get funky past 9:1... I generally avoid situations that would require anything more. The mrfx1k80h that I've been using needs about 6ohms J2 on the drains so 9:1 is perfect. There are a handful of ways to get lower without adding more turns. A bunch of capacitance on the drains/collectors brings the impedance down a LOT and can make a 9:1 act like a 25:1. There a couple other ways. Using a 9:1 to go from 1.4ohms to 12.5, and then an L network to get to 50. I suppose you could even cascade a 9:1 and a 4:1 to get the same effect. Haven't tried the last two ideas so results may vary.
Air cooled not too much money and less power and key down time. Water cooled is the best top of the line for the money.. and key down time. Self explanatory question bud.,
