EMRFD Message Archive 4695
Message Date From Subject 4695 2010-05-17 07:36:36 Tim $28.36 for fifty watts June 2010 QST has W6JL's winning entry to the 2009 Homebrew Challenge - the goal was a 50W output HF amplifier (specifically 40M) for less than $125.
Very impressive - in fact he did it for $28.36. He uses switching-class TO-220AB MOSFETs but does better than most IRF510 designs do... getting the power out with just 13.8 Vcc.
The design is broadband and still has some oomph on 15M (40W out) but is losing oomph on 10M (12W out).
An old Pentium-class heatsink and fan is modified for the amp, pretty cool. I'm very impressed.
Tim.4696 2010-05-17 11:13:58 Pat Bunn Re: $28.36 for fifty watts Unfortunatley the Mosfet is now out of production. Jameco has some but the normal suppliers are out.
Pat Bunn
N4LTA
----- Original Message -----
4698 2010-05-17 13:37:21 Tim Re: $28.36 for fifty watts Are you talking about IRF510's, or the IRFZ24?
I think the IRF510 may have reached mainstream end of life. But I see plenty of IRFZ24's at Mouser and Digikey.
4700 2010-05-17 15:19:59 ajparent1 Re: $28.36 for fifty watts There are plenty of similar parts and the ROHS part is still available but the earlier nonROHS part is going away.
While the amp works well at lower frequencies it's poor above 20M
which is typical of many of the minimum art amps that use trench
FETs. For example in the article at 10M the power out is only
12W for 5W drive.
The input C has to be dealt with usually at some loss of gain
and with care increase in efficiency and bandwidth.
As to it's power at 13.8V the IRFZ24 is a 17A 60V device and the IRF510 was a 5.6A 100V device. The latter at 28V will do in excess
of 75W in the WA2EBY design where it's still does useful power
at 10M (mine at 24V 2W does 37W SSB). The issue is matching
the input (gate) which is highly capacitive. The devices switching times due to capacitance are meaningless in non switching RF
other than it means the impedances need to be matched.
The use of a computer heatsink/fan combo is both a cost savings
and solution to a cooling problem. TO220 parts in general have
poor heat transfer (high thermal resistance) and keeping the
heatsink cool helps greatly. The penalty is some fan noise.
I've used the same trick8419 2013-03-25 10:23:40 Michael McShan Re: $28.36 for fifty watts I built this amplifier and it worked as described. Pretty straightforward to replicate. If you can find the mosfet and need an amp for your 1-2W QRP transmitter, this one is a good choice.
73,
Mike N5JKY
Oklahoma City
8420 2013-03-25 12:33:30 bobtbobbo Re: $28.36 for fifty watts Both Mouser and Digi-Key stock the IRFZ24N PBF. $1.78 and $1.45 respectively.
Bob K1AO
15210 2019-01-01 16:12:16 vk3hn Re: $28.36 for fifty watts I am finishing off a W6JL amp board I started about 18 months ago. I had it working into a dummy load on the bench recently, working as expected, on 13.8VDC, on a 5A current limited supply. Plenty of output on 80 to 30m, reducing on 20m. I had to drive it with a decent 5 watts or so to make it sit up and get going.My question is, has anyone tried running this amp (or a similar push pull amp with IRFZ24Ns) at voltages above 13.8V?I have some good 18-20V LiPo batteries and I would like to power the amp from these. I assume it will work harder, and get hotter, so the limiting factor will be cooling. So reducing the drive with a larger 50 ohm pad will be necessary.Is there anything else I should be aware of? I don't want to smoke my FETs on the first outing!Thanks in anticipation, Paul VK3HN.15211 2019-01-02 09:51:38 E-P Mänd Re: $28.36 for fifty watts Hi Paul, main restrictions are drain voltage (55 V max) and dissipation (45
W with ideal heat sink). Properly driven and matched, drain voltage may
swing from abt. 5 V (knee voltage at peak currents) to 55 V = 50 V pp or 25
V peak => (55V - 25 V) = 30 V max. supply is allowed. Just calculated, not
tested. Do not try to get much more output power as dissipation gets too
high. - Kindly, OH2NFI.15212 2019-01-02 13:01:48 Norberto Modanesi Re: $28.36 for fifty watts Has someone experimented with the amplifier published in October 2010 of QST?
73 – LU5DNM