|
I seems the QCX is northappy if you disconnect your antenna, it will blow the PA transistors (probably will do the same for a short circuit antenna).
As this is a good rig for portable operation I think it is a high risk that you connect it to a faulty antenna (perhaps intermittent).
How can you protect the PA, any ideas or proven circuits ?
Jan
|
Hi Jan.
Simple protection by a small led swr meter.
Max swr 1:2
Idea ?
73 John
|
Good idea yes.
But I think of protection when you by mistake transmit without an antenna or you have a loose contact in the antenna.
|
I've seen other QRP rigs with a 36V or 39V zener diode in parallel with the antenna.? Any experts here see a downside to adding one to at least keep the MOSFETS from going over-voltage?
Bill
W2EB
|
I like W2EB, Bill’s ?Zener diode idea but was also wondering if a resistor across the antenna terminals would also clamp unwanted voltages.? ? For example, say our load impedance is 50 ohms and we parallel a 1.2 K resistor, the impedance drops to 48 ohms.? Assuming the source impedance is also 50 ohms this gives an SWR of 1.04:1 – hardly worth fretting about.? Would this work? ? I have 15V, 16V and 24 V Zener diodes in my stock.? I measured peak output voltage with around 2W out (still have not tweaked T1) at ~15 V.? How about a low value resistor (50 ohms?) in series with the Zener ?to dissipate the energy ? ? 73 ? Earl, 4Z4TJ
|
What is it that kills the PA overvoltage or overcurrent ?
Jan
|
Putting in a SWR bridge will not keep the PA from blowing, just let you know why it blew.
On Dec 22, 2017 7:50 PM, "John PD7MAA" < pd7maa@...> wrote: Hi Jan.
Simple protection by a small led swr meter.
Max swr 1:2
Idea ?
73 John
|
Putting in a SWR bridge will not keep the PA from blowing, just let you know why it blew. Jess, You will not know until you've repaired it:-) But as the devices can be blown by overvoltage or overcurrent I'd think the only way to be safe is to have a SWR bridge with feedback permanently connected. It is the only way I've seen, is there another? The only other alternative is to use less power or a well over-specified device(s) so it copes with a bad match. For example 5V to a U3 BS170 seems to cope with any state. Push it harder, be careful. Many of the CQX failures seem to be unexpected TX, dummy load or matched antenna at all times power is applied! 73 Alan G4ZFQ
|
This very good Class E tutorial and design information by?Dan Tayloe, N7VE??have circuits both for over-voltage protection and over-current protection.
Jan
|
Aren't both symptomatic of each other ?
In which case, choose the easiest parameter to monitor and deal with.
73 de Andy
|
This is very common in high power amps. ?Across the device, not the antenna and a TVS or “Transzorb” type device not a zener. ?Just make sure the clamp voltage is calculated right.?
Dr.?William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ ? Owner - Operator Big Signal Ranch – K9ZC Staunton, Illinois ? Owner – Operator Villa Grand Piton - J68HZ Soufriere, St. Lucia W.I. Rent it:
email:??bill@... ?
toggle quoted message
Show quoted text
On Dec 23, 2017, at 6:23 AM, w2eb@... wrote: I've seen other QRP rigs with a 36V or 39V zener diode in parallel with the antenna.? Any experts here see a downside to adding one to at least keep the MOSFETS from going over-voltage?
Bill
W2EB
|
This very good Class E tutorial and design information by?Dan Tayloe, N7VE circuits both for over-voltage protection and over-current protection. Jan, Thanks, interesting that a bigger device is not really suitable for class E QRP. But protection needs many more components. I'm not sure whether I would bother provided just the BS170s blew. 73 Alan G4ZFQ
|
Aren't both symptomatic of each other ?
Andy, The way I see it is low output impedance could be high PA current. High output impedance could be high PA voltage. But what I do not know is how much antenna impedance affects the PA devices seen through the LPF. 73 Alan G4ZFQ
|
Can be either but each if from different causes. ?I read about all of the failed PA devices in the emails and it makes me wonder why design and calibrate the output for perfect conditions (perfect SWR, limited voltage and current)? ?Its much more reliable to design the PA with a device and associated supporting circuitry with enough power dissipation, voltage excursion, and maximum current so that it works no matter what happens (eg no antenna connected). ?This has me curious enough to start a project in the new year for exactly that... simple, cheap, fool-proof, and plugs into an existing BS170 holes on the U3. ? ?
Dr.?William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ ? Owner - Operator Big Signal Ranch – K9ZC Staunton, Illinois ? Owner – Operator Villa Grand Piton - J68HZ Soufriere, St. Lucia W.I. Rent it:
email:??bill@... ?
toggle quoted message
Show quoted text
On Dec 23, 2017, at 7:30 AM, SM7ETW Jan < sm7etw@...> wrote: What is it that kills the PA overvoltage or overcurrent ?
Jan
|
On Sat, Dec 23, 2017 at 08:26 am, Alan G4ZFQ wrote:
The way I see it is low output impedance could be high PA current.
High output impedance could be high PA voltage.
I think you may well be right, so I guess one has think of both monitoring and reacting to high voltage or current. Options. (1) Use a current limiting PSU (2) Use a voltage sensing device to provide a foldback arrangement. (3) Use a zener diode for across the antenna terminals feeding a power dumping resistor in series to dampen high VSWR effects. Old CB's used this trick and called it an Antenna Warning Indicator, when an LED was included in the mix. (4) Better heatsinking of output devices to cope with temporary SWR faults. BS170's aren't easy to heatsink, but that's the price you pay for using them ;-) (5) Use a better device that you can heatsink, and is over rated for the job (IRF510's or RD06HHF1 again). 73 de Andy
|
Hello Andy,
Saturday, December 23, 2017
With my high power Class D amps used with a U3S I have looked into
protection quite a lot. Whilst this VHF circuit initially probably
looks irrelevant it, when adapted to our needs, shows a pretty unique
trait. Using an SWR bridge means monitoring the voltage level of the
return side, Above X volts something happens to protect things (PA or
drive to it is shut down). All well and good. But a high SWR at one
Watt produces a far lower return voltage than a high SWR at say 50
Watts. This circuit used with a rotary switch allows you to preselect
at what return voltage level from the bridge a fault should be
assumed. Possibly a bit OTT for QRP but it's a useful circuit that
fixes an inherent issue with fixed return volts tripping the fault
circuit and affords tunability.
toggle quoted message
Show quoted text
On Sat, Dec 23, 2017 at 08:26 am, Alan G4ZFQ wrote:
The way I see it is low output impedance could be high PA current.
High output impedance could be high PA voltage.
I think you may well be right, so I guess one has think of both
monitoring and reacting to high voltage or current. Options. (1) Use a current limiting PSU (2) Use a voltage sensing device to provide a foldback arrangement. (3) Use a zener diode for across the antenna terminals feeding a
power dumping resistor in series to dampen high VSWR effects. Old CB's used this trick
and called it an Antenna Warning Indicator, when an LED was included in the mix. (4) Better heatsinking of output devices to cope with temporary SWR
faults. BS170's aren't easy to heatsink, but that's the price you pay for using them ;-) (5) Use a better device that you can heatsink, and is over rated
for the job (IRF510's or RD06HHF1 again). 73 de Andy _._,_._,_ -- 2E0ILY Best regards, Chris mailto:chris@...
--
Best regards, Chris Wilson (2E0ILY)
|
This has some discussions on protection circuits for Class-E PAs:
www.norcalqrp.org/files/Class_E_Amplifiers.pdf
PeterO
G0DZB
--
Please don't email me "off list" about "QCX low output" or my filter experiments.
|
Tayloe's article talks about using a higher current CMOS gate to drive
the class E amplifier. But many designs get more drive capability
another way: using multiple CMOS gates in parallel. 74x00 NAND gates
come four to a package, so it's often simple to use two or three of
the gates to get more current.
toggle quoted message
Show quoted text
On Sat, Dec 23, 2017 at 4:52 PM, G0DZB <g0dzb@...> wrote: This has some discussions on protection circuits for Class-E PAs:
www.norcalqrp.org/files/Class_E_Amplifiers.pdf
PeterO
G0DZB
--
Please don't email me "off list" about "QCX low output" or my filter experiments.
|
I read the Dan Tayloe article with interest and was thinking about how
protections could be added in the QCX with minimal additional hardware.?
With regard to operating into an open load, one possible way with no hardware
addition (but with a software addition) would be to make use of the RF power
meter function (diode detector). Under no load condition while transmitting, the
output voltage will be abnormally high, and this higher than normal voltage
could be used by software to inhibit transmit and display a warning say for 5
seconds.? I’m suspecting a “no load” condition occurs more commonly than a
low impedance or shorted load, so this might be a workable overvoltage
protection scheme with no additional hardware.? Overcurrent is a bigger
problem to solve with minimal hardware additions.? One thought I had would
be to modify the key shaping circuit at Q6. This is already controlling power to
the BS170s. Now if that circuit could be modified to limit current, you would
have a potential solution.? There are several circuits around to limit
current in a switching transistor.? This would require, as a minimum, a low
value resistor which could be used to sense current, and an additional
transistor and resistor to limit switch turn-on if current is exceeded.? An
example is shown at this link:
?
Steve K1RF
toggle quoted message
Show quoted text
Sent: Saturday, December 23, 2017 10:58 AM
Subject: Re: [QRPLabs] PA protection
?
This
very good Class E tutorial and design information by Dan Tayloe, N7VE have
circuits both for over-voltage protection and over-current
protection.
Jan
|
John PD7MAA
Earl
J68HZ
Shirley Dulcey KE1L