¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Topics

BS170 Killer?

 

Is it high voltage or dissipation during excessive AWR?
 

Is it high voltage or dissipation during excessive SWR?
 

Chris,

I assume that AWR is a typo and should be SWR.? If that is correct, then the answer is both, though we have yet to find the specific impedance of an SWR that produces a high enough voltage to take out the BS170s.? The power dissipated by the Class D amp in the QDX and QMX is susceptible to the impedance it sees from the low pass filter (LPF).? This changes with the load that is connected to the LPF.? An ATU will go through many load combinations to find the lowest SWR.? Through that process, the magic SWR that fries the BS170s may be presented.

For my QDXs (I have three operational), I have installed a Zener diode on each of the two sides of the Class D amp to protect from overvoltage.? I do not know of a way to protect from excessive power (current) in the same manner.? ?The pads are on the QMX to install the Zener diodes.? I have not done so yet; I am waiting for the SSB firmware to break out the QMX again and play with it.? I have the easier-to-repair QDXs to play with for digital modes, and I am not a CW op.

73
Evan
AC9TU
 

Evan
Thanks for your insight.
I have two QDX and two QMX and have replaced BS170s? in all of them, some operator error but some, including my latest QMX 10-20m for no obvious reason.? I see people trying TN0110 (100v) instead of BS170 (60v) but I have not seen any analytical or test verification of an advantage.
For my last failure I had the SWR protection set to 3 which would allow ~1.25W reflected back to the PA, this would add to the harmonics reflected back from the LPFs but without a good model I don't know how much is dissipated in the transistors.
Fortunately I can replace the BS170s quickly since I have a good vacuum de-soldering station.
Chris
 

On Thu, Jan 18, 2024 at 08:38 AM, Chris KB1NLW wrote:
?I see people trying TN0110 (100v) instead of BS170 (60v) but I have not seen any analytical or test verification of an advantage.

Chris,?
I thought someone had been playing with higher SWR on TN0110s??

Ahh, here it is:?/g/QRPLabs/message/102260? and??/g/QRPLabs/message/103918
With the TN060s.?

And now that the 110s are back in stock, I'd say that offers an advantage......spec sheet withstands higher voltage spikes...... and QDX has shown in testing to have spikes at the end of TX.... there's been several people running the TN0110s now, IIRC.?

My QDX Low blew BS170s for no apparent reason after ~450 FT8/FT4 QSOs (and a few WSPR cycles, too.) I added a diode across L14 when I replaced the BS170s and have had over 1250 QSOs since with no hiccups.?
FWIW, I'm building a HB QDX, and am going with TN0110s in this one along with the L14 diode. I'll be reporting what I get for power output in that build, how I'm heat-sinking, and how reliability goes. (This high band unit is likely to get used occasionally for POTA and operating while on vacation, so it needs to be more robust )

-Nate
N8BTR
 

Chris,

A potential issue with the QDX is a spike at the end of a transmission over the 60-volt specified limit of the BS170.? The TN0110 is a possible solution, EXCEPT that it must be installed backward compared to the BS170.? This reduces the amount of heat that can transfer out of the device.? That would increase the potential for power-induced failures.

There are two alternatives to the QDX potential issue.? One is to install a diode across L14 as if it were a relay to reduce the magnetic field build-up.? The other is to install Zener diodes to take the surge at the end of transmission.? Both have been reported as helping solve the problem.

The QMX has key shaping (rise and fall) in the power supplied to the PA stage.? That eliminates the spike at the end of the FT4/8 transmission.? The Zener pads were added to allow installation later.? Hans did not have conclusive evidence that the Zeners were needed at the time of manufacture.? Several designs by other designers do include Zener protection, so it was added as an option for the constructor of the QMX.

I have chosen the Zener approach for my QDXs.? The more common is the diode across L14.? You could install both, though that might be a belts-and-suspenders type of overkill.? I have not done any implementations on the QMX.? I am waiting for the firmware to settle before I return to playing with it.

I also have a solder-sucking gun, a Hakko.? I also have a hot air rework station.? You can still damage the very fine traces on the QDX board.? That happened to me when I went to clean up the excess solder with some wick after an IC5 failure.??

There have been posts on steps to protect the BS170s from failure.? The most significant is not to use an ATU.? Many of the reported failures are when using an ATU.? I must use a tuner for some bands as my HOA limits me on the size and number of antennas.? I use an ATU connected to my Icom 7300 to tune the antenna tuner and switch to the QDX.? The QDX does not see the high SWR associated with the tuning process.? Since following that process and installing the Zeners, I have not seen a failure on any of my QDX.? As stated earlier, I am currently not actively using my QMX.

73
Evan
AC9TU
 

I always use a NanoVNA to tune my antenna before plugging it to my Hi-Band QDX and I have soldered a diode across L14.
73 - Pierre - FK8IH
 

Following my previous post I also use a Hermes Lite 2 SDR transceiver. That unit uses two RF Power LDMOS transistors for its PA: AFT05MS003NT1, which are now obsolete but were sold by Digikey at 1.89$/unit and less than 1$ each for 1000 units.
This LDMOS PA has an output of 5W and is able to resist to a SWR of 65:1, which is quite amazing. Using an Alexloop magnetic antenna I already faced these very high SWR when manually tuning the loop and never experienced any damage (obviously I now also use my NanoVNA to tune the loop before connecting it to the XCVR). These LDMOS were designed for general public use, the users expecting not to break their phone or other device!
73 - Pierre - FK8IH
 

Does anyone have an idea on which soldering rework station or hot air rework station is best? Amazon has a large amount of rework stations to pick from. I have a stack of older transceivers to repair (RF relays require?a great deal of heat to melt solder)? in addition to newer fine-pritch QDX units. Is it more of a technique than a type of desoldering station?? I found the Apple laptop repair guy on youtube to be amusing as he heats?the entire?chip and just moves and picks-up with a tweezer.

73, Joe W2JEJ


On Thu, Jan 18, 2024 at 1:30?PM Evan Hand <elhandjr@...> wrote:
Chris,

A potential issue with the QDX is a spike at the end of a transmission over the 60-volt specified limit of the BS170.? The TN0110 is a possible solution, EXCEPT that it must be installed backward compared to the BS170.? This reduces the amount of heat that can transfer out of the device.? That would increase the potential for power-induced failures.

There are two alternatives to the QDX potential issue.? One is to install a diode across L14 as if it were a relay to reduce the magnetic field build-up.? The other is to install Zener diodes to take the surge at the end of transmission.? Both have been reported as helping solve the problem.

The QMX has key shaping (rise and fall) in the power supplied to the PA stage.? That eliminates the spike at the end of the FT4/8 transmission.? The Zener pads were added to allow installation later.? Hans did not have conclusive evidence that the Zeners were needed at the time of manufacture.? Several designs by other designers do include Zener protection, so it was added as an option for the constructor of the QMX.

I have chosen the Zener approach for my QDXs.? The more common is the diode across L14.? You could install both, though that might be a belts-and-suspenders type of overkill.? I have not done any implementations on the QMX.? I am waiting for the firmware to settle before I return to playing with it.

I also have a solder-sucking gun, a Hakko.? I also have a hot air rework station.? You can still damage the very fine traces on the QDX board.? That happened to me when I went to clean up the excess solder with some wick after an IC5 failure.??

There have been posts on steps to protect the BS170s from failure.? The most significant is not to use an ATU.? Many of the reported failures are when using an ATU.? I must use a tuner for some bands as my HOA limits me on the size and number of antennas.? I use an ATU connected to my Icom 7300 to tune the antenna tuner and switch to the QDX.? The QDX does not see the high SWR associated with the tuning process.? Since following that process and installing the Zeners, I have not seen a failure on any of my QDX.? As stated earlier, I am currently not actively using my QMX.

73
Evan
AC9TU

 

Joe, cheap hot air is fine for hobbyist use. I have a Yihua 3-channel, that I bought several years ago for about $45. Looks like it's gone up in price, at least for now. 3-channels is a luxury, 1 channel is fine. Any of the 858D clones in the $40 price range should work ok.? All these inexpensive hot air stations use the same handpiece, more or less, and rebuild parts are available on Aliexpress. I think the heating element runs around $15. Haven't had to replace mine yet, but if and when it goes, I know I can repair it inexpensively.

73, Rick NF6G
 

I use a cheap (47 EUR ) 858D clone I bought in my my local electronics shop with no problems, so far. It's really a big ?helper enabling to avoid mess on your pcb when it comes to removing microchips, relays, USB connectors, etc. I have just replaced, easily, a micro USB socket on my wife's tablet computer and a fried ?microchip IC05 in my Hi-band QDX transceiver . It's especially useful for removing these components , for soldering in the replacement parts I like to use a conventional soldering iron. But the hot air station would work equally good for that too. Good luck! 73! Linas LY2H
 

Thanks (again) Evan! For voltage, don't the Zeners give SWR protection that the L14 diode does not give??Do you think doing both Zeners and TN0110/TN0106 is also perhaps belt-and-suspenders?

For power/heat, I'm still working on my TN0110 heat dissipation for higher bands during operation. For?tuning specifically, isn't lowering supply voltage an easy protection?
 

Hi Peter Li,

I think the Zeners will provide some protection against SWR that results in high voltage, not in SWR that causes excessive current.? A bad SWR can cause either, depending on the impedance.? So, yes, 50% of the time, it will help ride through a bad SWR.

I do not equate the TN0110 and Zener combination to be the same.? I have measured spikes over 100 volts on the bench when testing for the Ldi/dt issue that the diode eliminates.? The spike is still there.? The TN devices are just better able to handle it.

At some point, the capability of the Zener to handle the excess energy could be exceeded.? When it does, hopefully, the failure mode is a short (most common for Zener failures) and protects the rest of the QDX.? It is easy enough to replace the Zeners.

Reducing the power to handle SWR mismatch better is a tentative alternative.? There is a minimum power the ATU needs to be able to operate.? A signal above that, yet low enough to prevent damage, is the juggling that needs to be done.? Reduced power for ATU tuning is what all of the commercial rigs do.? They have other protection mechanisms in place as well.

73
Evan
AC9TU
 

Evan
What zener do you recommend?
 

I used the 47-volt version from this Amazon assortment (I always buy assortments to build my parts bin).


73
Evan
AC9TU
 

As Evan said there are two potential failure causes with bad SWR;
1) High voltage
2) High power
High voltage will probably cause instant failure - so zener protection makes sense.
High power will probably take a few milliseconds to cause failure.? With an automatic tuner there may be a race between finding a good match and cooking a BS170.
Any thoughts?

1 - 16 of 16