Is it time for amplifier mfgrs to step back and "rethink"

Discussion in 'Amateur Radio Amplifiers' started by KM5FL, Sep 6, 2011.

Thread Status:
Not open for further replies.
ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: Left-2
ad: Left-3
ad: Subscribe
  1. K7JEM

    K7JEM Ham Member QRZ Page

    Those prices seem high, when you can get the THP 1.5KW amp for $6000. And that amp has rave reviews. It is considered top of the line for SS amps. And it's really not a lot more than high end tube amps.
    It might be, I really haven't looked at HV supplies, but a switching 48V supply for 2000 watts out should be around $500, based on the smaller supplies. The switching supply would certainly weigh less!

  2. WB0LSR

    WB0LSR Ham Member QRZ Page

    I don't know about the 50V supply being more expensive than the HV supply for tube amps.. Given today's market and the fact that there is more demand for 48-50V switchers outside of ham radio I'd say that a 50V high-current switching supply would end up costing less. Personally, for my own homebrew stuff I don't mind trying out some of the ebay asian imports (actually had a bit of luck with the ones I've had so far) to see how they are. Found a couple promising units, I could pair them up for more current and still not spend too much. This one looks particularly promising:
  3. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    I think tubes are still a good deal as long as somebody's making them.

    The THP amps, and the Ameritron amps, and many other SS amps may be the greatest thing since sliced bread but unless your antennas are close to perfectly matched, they'll need tuners. A good legal-limit tuner can easily cost $600-$700 so don't forget to add that to the acquisition cost.

    Meanwhile, my tube amps don't need any tuners under most conditions. They can easily load into a 3, 4 or 5:1 SWR without a tuner. I only use a tuner on 160m, where achieving a match across the whole band is close to impossible.

    Tube amps also withstand lightning transients (if you're transmitting while they happen) very, very well. SS amps don't.

    The high power FET modules that can run lots of power for $200 really cannot without a HUGE heatsink that has to be forced-air or water cooled, and even then if you don't have an adequate heat spreader between the transistors and the sink, you're relegated to short D/C modes like SSB. Want to run 1500W on RTTY? Copper heat spreader absolutely required, and the transistors don't come with those. The spreader has to be machined to <.005" flatness to be effective, and I don't think the mill to do that comes with them, either.

    Everything's a compromise.

    We used to say, "Price, performance, and reliability: Pick two."
  4. WB0LSR

    WB0LSR Ham Member QRZ Page

    You're certainly right about the cooling requirements, but even a precision machined piece of copper shouldn't cost thousands. For my own experimentation I'm curious if some sort of peltier arrangement could be put together to help alleviate the need for copper.. I have a large quantity of TECs on hand of all different sizes, so I'll be experimenting with that. The ruggedized LDMOS stuff from NXP and Freescale should, at least in theory, be able to handle some serious adversity in terms of mismatch. Can't say much about the transient handling capability other than to say that ESD protection is built in to most if not all LDMOS devices.
  5. WB3BEL

    WB3BEL Ham Member QRZ Page

    I don't think Thermal Electric Cooler is the solution to high power thermal management.

    All these do is move heat from one side of a surface to another while suffering operating inefficiencies. You have to take more heat off of the hot side than was on the cool side.

    They work OK for small areas at modest power levels.

    An RF amplifier operating at 60% efficiency at the 500W level is dissipating 333W. This does not take a monster heatsink if it has forced convection. But would benefit from a modest sized copper heat spreader. This copper spreader is not cost prohibitive perhaps $10 or $20. A 100% duty cycle legal limit amp with some margin does take significant engineering and obviously costs more money. But this is all doable.

    Steve is right that tubes are more forgiving of high voltage events. But many broadcast stations are using solid state amplifier pallets and the reliability in that application is acceptable.

    ESD protection is not meant for high energy events. ESD protection is usually based on a human body model which is a miniscule charge . It corresponds to minimizing device handling failures in a normal relatively static free environment. It does not even protect against situations like walking across a carpet during the dry wintertime and touching a sensitive part. It means that if someone forgets a static strap or handles parts casually they probably won't damage them.

    There have been some big improvements in RF transistor ruggedness. I think that some of the claims in the latest generation LDMOS have a strong dose of advertising hype. They are nowhere near indestructible...
    But they are displacing tube technologies in applications that have some severe mismatch conditions like laser drivers, plasma generators, MRI transmitters etc...It will be interesting to see the improvements with time and see how they can be adapted to amateur use.
  6. TF3T

    TF3T Ham Member QRZ Page

    Just touching on the DIY approach- Not being negative.

    I have now built two SS amps - 28v (MRF141G) and 50v (MRF150) - I would consider this old tech - Tokyo HP / Yeasu Quadra etc uses these old transistors. We now see these new LDMOS transistors come to live, I think Beko has a 2m amp based on a Freescale amplifier. We have also a French amateur designing circuitry around these LDMOS. So the wave is coming.

    I have also built tube amplifiers from scratch - so I have some comparison. The biggest problem I found with the DIY approach with SS amps is:
    1. very critical in building, placing of capacitors - coil windings / transformers. I tried to built 4 "palettes" using MRF141G for 1200W total output. There was a 15-20% unexplained variance between each device.
    2. Proving that your protection circuit works! - I hope it works - but I'm not going to test it. If it does not work - 800$ of transistors will go up in smoke.
    3. RFI problems with the switch mode power supplies on 50v. That can be cured by switching them off in RX - they usually have a short "ramp-up" time . 10-15ms for full power. But I have no clue if that is good practise, e.g. QSK.
    4. You need a full set of lab equipment - spectrum analyzer - network analyzer. I solve this using a miniVNA and SDR - these SDR's are touchy (some padding needed).
    5. LPF filter design. touchy ! tricky component selection.

    I'm scared that when these LDMOS "kits" become readily available, people will skip on the LPF.

    But it's clearly the future - but I'm now building a tube amp though - and it's going to be great ;)

    - Benni TF3CY
  7. WB0LSR

    WB0LSR Ham Member QRZ Page

    You're right that TECs are simply heat pumps, moving heat from one side to the other. I've used them for quite some time to refrigerate microprocessors and also for solid-state laser stabilization, so I still think that they have a potential application here, although I do find it interesting that they haven't turned up in this usage more frequently. For this usage we're not even talking about keeping the device cold, since these things can handle some heat. All that I would require is that a good difference be maintained between the hot and cold sides of the TEC.. still seems to me like it could potentially be doable.

    You touched on a big part of why I feel they way i do about transistors vs. tubes: The broadcast industry (no, actually the ENTIRE broadcast/ISM market) is using solid-state almost exclusively. If there were inherent issues with the technology, thereby making tubes superior, these markets would be using tubes, not transistors.

    An induced ESD pulse from a nearby lightning strike won't necessarily exceed the tens of thousands of volts produced by the human body during an ESD event. I see little difference between the two, unless of course the device is struck directly by lightning.

    In my opinion these devices are ready for amateur use right now, and I know I'm not alone.

    Regarding TF3CY's statement about filters, lot of people seem scared of building low-pass filters. Having built several sets at home with VERY limited test equipment available I can tell you that it's quite possible to do and not really that hard. Designing them is best done with software (Elsie and SVCFilterDesign are very good), and tuning them sometimes requires an SWR meter and a little patience, but it's really not that bad in my experience.
    Last edited: Sep 7, 2011
  8. WB3BEL

    WB3BEL Ham Member QRZ Page

    They are similar in some ways and very different in others. The magnitude and rise/fall times of the charges involved is very different even for nearby strike events vs HBM discharge.

    These devices are specified at perhaps a few thousand volts for HBM...The Freescale part is Class2.
    The machine model MM perhaps a few hundred volts...The Freescale part is ClassB
    The charge device model CDM perhaps a few thousand volts...The Freescale part is ClassIV

    One thing to consider is the frequency of operation of the power amplifier. The spectrum of lightning discharges falls right in the HF range. As such it is harder to remove from the signal path than say a VHF or UHF transmitter where filtering and surge suppression can diminish the lightning waveform easily.

    Of course they are and experimenters are using them. My comments previously were based on the original thread which was an ultra low cost power amplifier based on these latest generation LDMOS. My point is that currently they do not enable lower cost amateur band amplifier manufacturing at the 500W level. But it still is possible for an amateur to build his own amplifier using them and possibly same money if they have the knowledge and access to some surplus and test equipment.

    TF3CY was talking about high power at VHF. I think what he was saying is that you have to be very careful of the layout and component selection etc when you are talking about VHF QRO.
    Last edited: Sep 7, 2011
  9. W1VT

    W1VT Ham Member QRZ Page

    How about an 811A/572B design with a cheap protection circuit that will protect the tubes against operator abuse?
  10. WB0LSR

    WB0LSR Ham Member QRZ Page

    I see, I missed that. That's true.. everything about layout becomes critical at VHF and above. I've never tried to build a filter for anything higher than 2m, but every UHF filter I've seen uses striplines or just bent pieces of wire for inductors and either Unelco-type mica caps or SMD RF caps. Stripline design is a mystery to me.. I know that it usually requires software to design. As far as my own interest goes, I doubt I'll ever build an amp for anything above 6m. VHF/UHF operation just isn't as interesting to me as HF, but that's just me.
Thread Status:
Not open for further replies.

Share This Page