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
Using IRF530 outputs and 2N3553 drivers in ubitx
|
Good Afternoon,
Finally received and housed my ubitx. On receiving the PCB, I replaced the IRF 510 outputs with IRF 530's. These tend to run a little cooler and are just as inexpensive. I tested the output on 40 meters as it was at 25 watts at 14v drawing about 1.9 a. I also replaced driver transistors Q92,93,Q96 and 97 with 2 2N3553's.? I socketed the drivers so I could plug in various types. You can try 2N2219a but they tend to run a little warmer.? You also want to short the 22 ohm resistors off the emitters. I used Q93 and Q96 as they were furthest away from each other and provided space for a heatsink if needed. Lastly, parallel R961(2.2 ohm) with a 1 ohm 1/2 watt.? Be careful to shield the toroids with aluminum foil when soldering so as not to short them out (like I did). After removing C261 and C262 in the feedback circuit, I came up with at 14v: 27.16 watts at 20 meters, 31.78 watts at 40, and 36 watts at 80 meters. I am running a 7812 voltage regulator for the receive and transmit drive, leaving raw B+ for the outputs.? I am also running a microphone compressor. I have made 2 contacts (one 300 miles away and the other roughly 700 miles away).? Both contacts were impressed at the level of power I was using.? I must have a reasonable attic shortened dipole antenna.? Anyway, the 2N3553's barely get warm after keying up the microphone.? I calculated that the 2N3553's were only driving the outputs with 460mw.? I was planning to drive a linear amp with it, but now I will use it as a stand alone radio.? I have not heard of anyone doing this, so I thought I would throw this one out for consideration. Regards, Bob KD9ITL |
|
Thanks for the info.? Looks like an easy way to bring the ubitx up a notch!! Joe VE1BWV On Wed, May 16, 2018 at 3:51 PM, robert_kross via Groups.Io <robert_kross@...> wrote: Good Afternoon, |
|
Has anyone yet derived a way to create a low power and a 'higher' power mode while keeping the necessary power transistors bias where they should be? One scenario I want to use my uBitX in, is for backpacking, where I'll likely only have a 7AH or 18AH 12v battery, but for times when I have a replenishable power source (such as a car alternator), I can see a good use for higher wattage? (Or for emergencies while on battery power, realizing the reduction in operating time) On Wed, May 16, 2018 at 1:06 PM, Joe <joeman2116@...> wrote:
|
|
Bob you may want to check your meters.? I just do not see how you can get 25 watts out at 14 volts when only drawing 1.9 amps.? that would be only 28 watts at the most input to the transistors.? That would make it almost 90 % efficient.?? On Wed, May 16, 2018 at 2:51 PM, robert_kross via Groups.Io <robert_kross@...> wrote: Good Afternoon, |
|
Yup, hfsignals has.? It's called RV1, and determines how hard we drive the final.
toggle quoted message
Show quoted text
To reduce power, turn it clockwise (it's backwards from most pots in this world). I suggest you mark it first so you know the setting it was at when shipped. If you increase power beyond that, you should check your signal carefully for distortion and splatter. Here's a couple threads regarding VK2ETA's scheme which moves the transmit signal within the passband of the 45mhz crystal filter to attenuate it by a controlled amount.? So this would reduce power for you without any hardware mods and without having to take the cover off. ? ??/g/BITX20/message/46235 ? ??/g/BITX20/message/46241 Jerry, KE7ER On Wed, May 16, 2018 at 01:14 pm, Walter Keen wrote:
|
|
That is consistent with my testing.
The 2n3904s are marginal for power and FT. So it shows a higher FT device that can run at more power helps, 2N3553 will do both.? Other choices would be 2n3866 and 2n5109 and along list of others. The IRF530 is an odd choice as below 10 mhz its been used by many but over 10 its input capacitance (940PF vs 135 for the IRF510) makes it a more difficult part to feed. Did you change the predriver as well? How is the power out at 28mhz? Allison |
|
Rahul Srivastava
Hi! I feel we should give BFG591 a try, they are modern devices , easily available with online suppliers and more compatible with the SMD construction of uBitx. There is plenty of headroom where Ft is concerned will work on 6 and yes I have seen it being used as driver in some designs. 73 Rahul VU3WJM?? |
|
They come up as obsolete On Wed, May 16, 2018 at 07:54 pm, Rahul Srivastava wrote:
|
|
I have used the Minipa 70 in my scratch built BitX. While I can get 70 Watts out (13.8v supply), it is short lived as a final will blow within a few seconds and short closed. 50 Watts will work for short QSOs, but taking for more than four or five seconds will destroy the IRF530s. My heat sink is 6¡±x4¡± with 2¡± fins, so I don¡¯t think heat dissipation is the issue. 25 Watts gives me no issues, however.?
|
|
Don't know enough about the minipa 70 to comment.? Is there a circuit out there to look at?
The Wa2EBY I have has run at 65W (40M) brick on key mode for nearly 20M before I called it good.? Only slightly larger heatsink but the real factor was stability. and avoiding overdrive. I used a 4DB pad on the amp as almost of the radios I have are about 4W to both protect the amp from overdrive, provide a stable load for the radio and input to the amp.? Also everyone that reported killing that amp was running more than 28V.? FYI that used IRF510s. I built it back in 2006 its still running the second set of transistors and has been run into shorted coax, opens (no antenna or no coax!), wrong lowpass filter (on 10M with 80m filter...oops). What happened to the fist set of IRF510s?? Testing to see what they could take, they failed at 85W out on 40M (190W DC input) and 36V and the likely failure mode was overvoltage on the gates (needed 5W drive for that power) with bias set to 200ma per device.? Previous test at 80W ran for 5 minutes without issues?though the heat sink did hit 55C and the case of the device nearly 80C.? Backing off to 28V and keeping drive under 1.2W has proven? blast proof. I have encountered enough IRF5xx mosfet amp designs and most a at best conditionally stable and if they go unstable they will oscillate and, overheat and die. Some do not like the driver (what ever radio used) and many are even less tolerant of bad SWR usually due to high? reactance of the antenna not tuned out.? At least one case it was built so open (no screening at all) that RF from the antenna was getting to the input causing oscillation.? ?Another case the conclusion was the amp was ok but the QRP rig was? unstable with a reactive load and would go into oscillation, over drive the amp and kill the IRF520s used .? I helped him build in a TX only input attenuator(3DB) into the amp and problems went away. I've seen this sort of failure and misbehaviour in bipolar and LDmosfet?and even vacuum tube amps as well.?? Allison |
|
Wow 940pf. ?You can design for it to work at 30 MHz but thats not gonna work at 3.5 or 7.0 MHz without switching parts in and out. ?
toggle quoted message
Show quoted text
? Dr.?William J. Schmidt - K9HZ On May 16, 2018, at 7:48 PM, ajparent1/KB1GMX <kb1gmx@...> wrote:
|
|
You can across the bands but you need a deign impedance of a about 3-5 ohms and enough loading?
to keep it down there which means a heavier driver and a different driver output transformer (something the 4:1 turns area).? From a system design standpoint its very doable as I make? 6M amps using MRF150? (400pf input C) and there are more than a few commercial amps using them in push pull parallel for 2-30mhz. Its just not an easy drop in.? But over 5W at 30mhz also says its far from horrific as is. I'd suspect a 2:1 transformer using BN43-6802 with a tube and strap for the low impedance side might be the winner. The driving transformer needs to have a lower? leakage inductance and impedance on the MOSFET side. All in all, if the higher bands are of less interest, then call success and move on as the results at he low end are pretty decent. Allison |
|
THe problem with grounded gate using Mosfets is twofold.
At 2A with a gm or 2-4S th input impedance would be under 2 ohms.? The second issue is the drain to source reverse diode plus the high source to drain capacitance.? Stability might be an issue. as to gain its determined by the product of inpu tto output impedances and for 10W level we are limited for both, I'd expect a gain under 5-7db. Gate capacitance is not a limiting factor.? I use BLF278 at 2M and the input capacitance is about 400PF and that is for a very high performance VHF LDMOS FET (300-400W out!). Gain up there is such that for full 350W out I need under 5W in.? How about 8 IRF510s set up as 4x4 push-pull for 6M?? The input impedance is around 6 ohms and matched? that amp?easily does over 200W at 28V.? ?The input transformer is 9:1 (3:1 turns) from 50 ohms to about 6 ohms. The problem is matching the RF impedance, if done it performs.? At low HF its not as big an issue as upper HF.? The greater than 10:1 range means the driver has to be able to push enough power into a wide range of impedances or the matching circuit between them is sufficient to over come that.? Its not as bad as it sounds.? The basics for MOSFET amps is the impedance is not infinite even though at DC it is. So the path to power is pick a fet with high Gm (most mosfets are very high gain) and load the input for stability and to force the driver to impress enough voltage? across the effective load.? Usually that means a low impedance. Allison |
|
Yes, but there is no Miller capacitance so the total is less.
toggle quoted message
Show quoted text
Quoting Brian <vk4bap@...>: Is the reason for not using power MOSFET's in grounded gate the high drain-source capacitance? Looking at the IRF510 datasheet it looks like drain-source capacitance is higher than gate-source. Have I read that correctly? |
K9HZ
to navigate to use esc to dismiss