How high SWR/return loss will the QCX handle?


Jim Allyn - N7JA
 

I have seen a number of posts here about blowing up the finals in a QCX by transmitting into a high SWR.  Just how high does the SWR have to be to blow up the QCX finals?  Wide open?  Dead short?  3 to 1?  2 to 1?  Are all Class E amplifiers that SWR sensitive?  I remember I used to see in Motorola RF power transistor datasheets that some of their parts would survive a 30:1 SWR, at any phase angle.  Anybody seen any numbers on what Class E amplifiers will handle?  Anybody done any experiments?


Alan G4ZFQ
 

Anybody seen any numbers on what Class E amplifiers will handle?
Jim,

The limit could be the BS170, not class E.
From all the discussions in many previous posts nobody has experimented on the best way to destroy the BS170.
Running normally they are close to their 60 volt breakdown so any mismatch that causes that to be exceeded could be one reason for failure. Nobody has come up with another device that would not require more drive.

Class E is prone to several failure modes but protection at least doubles component count and complexity.
PA failure is often raised in this group but probably affects relatively few constructors who at least end up with some experience and confidence in fault-finding and component replacement techniques.

73 Alan G4ZFQ


Bill Cromwell
 

Hi,

I have another rig here that uses the BS170 transistors. I decline to operate it into more than 3:1 SWR and prefer to stay under 2.5:1. I have accidentally keyed up with no antenna at all, but not for more than couple of seconds. I have had (and still have some) other solid state transmitters that are okay approaching 4:1. I decline to run them over that and still work to improve to no more than 2.5:1. I have not had any failures produced bt SWR. Lightning surges - yes. And no dead shorts (as far as I know). A dead short would be from a defective/damaged feedline. I have not ever placed a dead short on a transmitter output connector intentionally! Feedline failures I have experienced are open. More usual if the antenna is not connected it is noticeable in the receiver at power up. Nothing there. This information is *MY* experience. You will not find a manufacturer telling you in writing that 3:1 or 4:1 is okay and will probably warn against even 2:1. Use you own judgement there.

I hate to consider the SWR that some of my tube finals may have seen. But those are not related to this topic nor this mail group.

73,

Bill KU8H

bark less - wag more

On 6/18/20 2:55 AM, Alan G4ZFQ wrote:
Anybody seen any numbers on what Class E amplifiers will handle?
Jim,
The limit could be the BS170, not class E.
From all the discussions in many previous posts nobody has experimented on the best way to destroy the BS170.
Running normally they are close to their 60 volt breakdown so any mismatch that causes that to be exceeded could be one reason for failure. Nobody has come up with another device that would not require more drive.
Class E is prone to several failure modes but protection at least doubles component count and complexity.
PA failure is often raised in this group but probably affects relatively few constructors who at least end up with some experience and confidence in fault-finding and component replacement techniques.
73 Alan G4ZFQ


Alan G4ZFQ
 

I have another rig here that uses the BS170 transistors.
Bill,

I've two Softrocks that have run continuously on WSPR for over 10 years. The two BS170s have an output of 1 watt class B. Several times they both have run for many hours with a broken antenna or no antenna connected.
If it were 3-4 times the power in class E I suspect they would have been rebuilt a few times:-)

73 Alan G4ZFQ


Bill Cromwell
 

Hi Alan,

I was under the impression the QCX ran the finals in class C. I had to go and look it up. It is class E (would have bet you were right). There are so many rigs to choose:) Before I look at one on the bench I look at the manual info to see what is supposed to be there. Sometimes I even look before a reply. I should do that more often.

I expect to have a QCX in my lineup and I will probably have the same attitude about SWR with that. If I have to change a lot of finals I may change my mind about SWR with that radio and then decide what I will do with the QCX. Because of the radios I already have and holding out for the QSX I don't have a lot of urgency for the 80 or 17 meter QCX.

73,

Bill KU8H

bark less - wag more

On 6/18/20 11:19 AM, Alan G4ZFQ wrote:
I have another rig here that uses the BS170 transistors.
Bill,
I've two Softrocks that have run continuously on WSPR for over 10 years. The two BS170s have an output of 1 watt class B. Several times they both have run  for many hours with a broken antenna or no antenna connected.
If it were 3-4 times the power in class E I suspect they would have been rebuilt a few times:-)
73 Alan G4ZFQ


ajparent1/KB1GMX
 

Jim,

SWR is a magnitude so usuing that form stay under 2:1.  That said 
it can mean the actual feed impedance could be 25 ohms or 100 ohms
and depenig on which you get differnt results.

Note class E doe not like wide SWR excustions.

So at 25 ohms you going to be generating more power so higher
power disssipatio and current can potentially kill parts.

At 100 ohms the currents go down but the votlages can climb
excessively.  So if the device is running say 40V peak and has a
break down of 60V (BS170 is 60V) you get closer to voltage
breakdown.

Also those cases assume resistive load.  Often the load is not
resistive and has reacance and that can interact with Class-E
in negative ways as well.

Short answer run close to 50 ohms and low SWR for best result.
If you antenna does not build a tuner to match it or better yet an
antenna that matches.

Can I speak authorively, yes.  I blew enough mosfets and many
GaN fets testing what class E tolerates.  Class E and D for that
fact are power conversion and changing the load changes the
power out and circuit dynamics.  They have limits and if you
exceed them, poof.

As to the SWR claims from Motorols/ST/Freescale/Macom is for
a totally different class of device that will not run at class E and
for nominal Class-C or linear modes they are SWR resistant but
still not blow proof.  See above as some one provided MRF137s for 
the system under test and at 5W it didnot take much to kill them.
They proves less durable than IRF510s in that application and that
was less than optimum.

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting


Steve in Okinawa
 

I have lost count of all the ways I accidentally abused my QCX40 and my KD1JV Tribander, with no damage to the finals.  


ajparent1/KB1GMX
 

Mine too but I also have it well tuned for efficiency which means I do
not have to crank up the DC to get power.  That helps.

Like any transmitter most are not fragile but some things do make
for issues.

I have a bunch of different QRP to 50 and 100W class amps I've built
up over the years and people report blowing them up likel they are fragile.
Some have suffered unspeakable thing and still work fine.  Best example
was the 100W pair of linears for 6 and 2 built for FD.  I ran bot into the
opposing antenna and was windering why it was only meg performance.
Needless to say both were seeing terrible (10:1 or more) SWR and did
so for over an hour without fail.  OTher report they "stank, were fragile,
died every time I used it", and so on.  Beats me whaat I do they didn't fail.

So the answer stands run a resonable SWR and try to to get over 2:1.
If it fails then maybe something else you did was a problem.


Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting


Jim Allyn - N7JA
 

On Thu, Jun 18, 2020 at 07:23 AM, Bill Cromwell wrote:
You will not find a manufacturer telling you in writing that 3:1 or 4:1 is okay and will probably warn against even 2:1.
Perhaps not any manufacturer of transmitters or amplifiers, but the manufacturers of the transistors used in those devices will.  Motorola used to guarantee their transistors would function into a 30:1 SWR, at any phase angle.  And some of the new stuff is rated to withstand an SWR up to 65:1, at any phase angle.  I haven't seen any of the manufacturers of equipment using those devices make any claims about their transmitters, amplifiers, etc. but I wouldn't be surprised if some of them do, or will at some point.  All it will take is for one major manufacturer of equipment to guarantee their equipment will withstand a certain SWR, and then everybody else will have to also.


Jim Allyn - N7JA
 

On Thu, Jun 18, 2020 at 10:16 AM, ajparent1/KB1GMX wrote:
SWR is a magnitude
SWR is a magnitude, but it also has an angle, which most of us don't pay much attention to most of the time.  I have an application in mind where not much power is required, but the SWR could be quite high at times.  The BS170s would easily deliver the power I need, and I was hoping somebody might have some idea just what SWR they will withstand.  But apparently not.  I have a handful of BS170s here, maybe I'll just do some experiments.


Bill Cromwell
 

Hi,

If the transistors don't cost a lot and are not to difficult to change....

Go for it.

73,

Bill KU8H

bark less - wag more

On 6/18/20 10:47 PM, Jim Allyn - N7JA wrote:
On Thu, Jun 18, 2020 at 10:16 AM, ajparent1/KB1GMX wrote:
SWR is a magnitude
SWR is a magnitude, but it also has an angle, which most of us don't pay much attention to most of the time.  I have an application in mind where not much power is required, but the SWR could be quite high at times. The BS170s would easily deliver the power I need, and I was hoping somebody might have some idea just what SWR they will withstand.  But apparently not.  I have a handful of BS170s here, maybe I'll just do some experiments.


Alan G4ZFQ
 

The BS170s would easily deliver the power I need, and I was hoping somebody might have some idea just what SWR they will withstand.
Jim,

SWR is just a number telling you how close you are to a match. It does not tell you why there is a mismatch, what to do to correct it.

If you run an amplifier at low power then it is less likely that voltage or current will destroy the finals. That is the principle of ALC feedback from a SWR detector as used in most ham transmitters.

So if you run your BS170s at relatively low power they will probably last for a long time. As I said my Softrocks have tried to put 1 watt into no load and bad loads many times with no failure - yet..

73 Alan G4ZFQ


Hans Summers
 


Hi all

QCX was conceived in Feb 2017 and after 6 months of development went on sale in August 2017. To date I still haven't destroyed any BS170s during various kinds of antenna abuse. Neither during development nor since. 

The 5W PA kit http://qrp-labs.com/pa ... a lot of IRF510 died during development, but quietly, and I eventually diagnosed a floating ground as the cause (soldering iron tip was grounded and there was a potential that kept quietly frying the MOSFET gate). No IRF510s blew up otherwise. 

During development and operation of the 10W Linear http://qrp-labs.com/IRF510s i don't recall losing any IRF510s either. That has been subjected to open load, short circuit, double the rated power, continuous 100% duty cycle, various other torture. 

During my early experiments with the 50W PA kit http://qrp-labs.com/50wpa I did explode a lot of MOSFETs! Some of that was IRF510 torture (single transistor, demanding too much from it) and some was me trying various other junkbox MOSFETs that didn't turn out to be suitable. Several spectacular flashes and bangs. Since the final circuit was settled I have only lost one pair of IRF510 and that was a power supply glitch; that failure mode is prevented now by a 1-resistor change that was added in the QCX+. 

So really, it beats me, hearing reports of transistor destruction. Maybe it's a small percentage of the overall QCX population out there, and "mailing group bias" makes it seem more. 

My BS170s just don't blow up here! 

73 Hans G0UPL