Date   
Re: ND6T AGC implementation for uBIT-X

Kees T
 

Low price is the name of the game. ...and I for one, think it's fun.  After all, copper wire was invented by two Dutchmen fighting over a penny .......and I'm Dutch. The AGC board is ready to go for build and Don, ND6T, and I are cleaning up some details on the "RX/TX Click Noise Elimination" board and his "Polite Antenna Tuner" board. I can post a PDF of the AGC board if you want......25mm x 25mm......but I have to figure out how to "step and repeat" the design so I can get 16 on a 100mm x 100mm panel .....and NOT have the component identifiers change.

73 Kees K5BCQ

Re: Please don't make Radiuno with CH340/CH341 chipsets any more! Or at least advertise that you do use them! #radiuno

tdelozie
 

Long time mac osx user here, my issues have never been with os/firmware but can understand how frustrating any barrier can be. I’ve also never seen a CP2102 on a micro-controller, but the ch34x is on tons including the internet of things devices running on espxxx/xxxx chips. 

I believe there are signed drivers at this time which work correctly.
http://esp32.net/usb-uart/

There is so much right about the uBit, I would hate see problems arise from fixing something not related to the radio. To continue the disjointed post, I offer the following. 

Pro’s for the generic arduino w/ch34x (or whatever chip next gains popularity and support)
*cheap, cheaper than a single CP2102 chip
*multiple free tool chains
*huge community support
*use any FTDI cable to program it, including another arduino.
*current solution is more than good enough. 

Con’s
*soldered on so a bit difficult to change, but that’s not the boards fault and probably the right engineering solution. 
*usb chip is slow, again not an issue.
*drivers can be wanky and vary in quality per OS. <editorial> Not a new situation, actually is much better now. </editorial>

-Tom
kf5nwc

Re: Something different this time...

Jerry Gaffke
 

I'm genuinely curious, what IRF510 problem?
Is that from posts to this forum?
Can you point to such a post?

I have heard of counterfeit IRF510's being sold on ebay.
I'm not aware of hfsignals having trouble with bad IRF510's.

Jerry


On Sun, Apr 29, 2018 at 06:23 pm, Jim Sheldon wrote:
I firmly believe the IRF 510 problem is caused by poor quality (Fake) IRF 510's, not transmitting into opens, shorts or somewhere in between! 

Re: Something different this time...

Jerry Gaffke
 

We've been over this before.
Likely a 50 ohm BNC connector.that was slightly out of mechanical spec.
And 75 ohm BNC connectors should mate just fine with 50 ohm BNC connectors anyway.
See post   https://groups.io/g/BITX20/message/39196


On Sun, Apr 29, 2018 at 06:55 pm, William R Maxwell wrote:

it might have been a 75 ohm BNC socket?

Re: ND6T AGC implementation for uBIT-X

Jerry Gaffke
 

My apologies to anyone who did not find that humorous.
Here's a smiley face    ;-)  to show I was not intending to insult.

That post was triggered in part by Kees' kind offer to provide kits
at what looks like might be a ridiculously low price.

Re: Something different this time...

William R Maxwell
 

it might have been a 75 ohm BNC socket?


On 30/04/2018 11:23 AM, Jim Sheldon wrote:
It isn't really odd at all.  It's just that their supplier has gotten some very poor quality BNC connectors.  I've had 3 of them with the larger center pin that would not make good contact.  Replaced them all with Amphenol BNC connectors from Digi-Key.  It appears the Chinese can't measure things right when they make these things!  Fortunately, it was readily noticeable on receive.  However, I have transmitted into opens a number of times, high SWR antennas several times and still have intact IRF 510's.  I firmly believe the IRF 510 problem is caused by poor quality (Fake) IRF 510's, not transmitting into opens, shorts or somewhere in between! 

Jim - W0EB

------ Original Message ------
From: "Brian L. Davis" <brianldavis@...>
Sent: 4/29/2018 5:23:30 PM
Subject: [BITX20] Something different this time...

I've seen some odd things but this is one of the oddest.
I have experienced intermittent receive and transmit, just momentary that seemed to be cable related.
Traced it to the BNC jack.
Seems the female socket in the jack is slightly oversize and the male pin was a loose fit and occasionally showed an open circuit.
Changed the jack and the problem went away.

Re: Something different this time...

Ryan Flowers
 

This is the very reason I switched to PL259/SO239 connectors. 

--
Ryan Flowers - W7RLF
MiscDotGeek.com
Multi Band BITX40
The BITX40 FAQ

Re: Something different this time...

Jim Sheldon
 

It isn't really odd at all.  It's just that their supplier has gotten some very poor quality BNC connectors.  I've had 3 of them with the larger center pin that would not make good contact.  Replaced them all with Amphenol BNC connectors from Digi-Key.  It appears the Chinese can't measure things right when they make these things!  Fortunately, it was readily noticeable on receive.  However, I have transmitted into opens a number of times, high SWR antennas several times and still have intact IRF 510's.  I firmly believe the IRF 510 problem is caused by poor quality (Fake) IRF 510's, not transmitting into opens, shorts or somewhere in between! 

Jim - W0EB

------ Original Message ------
From: "Brian L. Davis" <brianldavis@...>
Sent: 4/29/2018 5:23:30 PM
Subject: [BITX20] Something different this time...

I've seen some odd things but this is one of the oddest.
I have experienced intermittent receive and transmit, just momentary that seemed to be cable related.
Traced it to the BNC jack.
Seems the female socket in the jack is slightly oversize and the male pin was a loose fit and occasionally showed an open circuit.
Changed the jack and the problem went away.

Re: ND6T AGC implementation for uBIT-X

Skip Davis
 

Jerry you wrote “Don is well aware how cheap this crowd is.” I take offense to this remark. Just because we purchased this rig doesn’t mean we are all cheap. If you have paid much attention to the board offerings many have requested more than one. These are great learning platforms for experimenting with hardware and software without a large investment and if it gets fried or firmware bricked it doesn’t cost a fortune to fix or replace.

Skip Davis, NC9O

On Apr 29, 2018, at 16:41, Jerry Gaffke via Groups.Io <jgaffke=yahoo.com@groups.io> wrote:

Don is well aware how cheap this crowd is.

Re: ND6T AGC implementation for uBIT-X

Jerry Gaffke
 

Michael,

The Q90 and RV1 you refer to are definitely in the uBitx schematic.
I'm not finding R1 except in the Tandem Match here:  http://www.kitsandparts.com/bridge1.4.php
Is that the R1 you are referring to? 
Probably not, as it is not fed with "weak RF from the first amplifier of the transmitter"
An attenuator around RV1 is certainly a good choice for adjusting output power, effective for both CW and SSB.

For SSB, this attenuator could be most anywhere, all the way back to the mike. 
Perhaps an audio AGC circuit or speech compressor between mike and rig?

Have you found Swan's ALC to be a useful feature?
In what way? 

Jerry



On Sun, Apr 29, 2018 at 05:32 pm, Michael Shreeve wrote:
Looking at the circuit, I see R1 simply is fed with full RF at the top, but relatively weak RF from the first amplifier in the transmitter. I'll measure the voltage, but its simple,   the wiper sends however much RF you need on to the rest of the output circuit. If this control could be mounted on the panel of the radio, you would have a transmit power control. But, it might not be good to take it off the board. 
 
This is done, if you look , right at the output of the first rf amplifier Q90 and RV1 passes the rf along to the input of a couple of dc direct coupled 2n3904s . Some sort of RF attenuator after rv1 would do the trick, and control voltage from a bridge swr detector could control that. I'm guessing the Circuit Don uses, or part of it, the AGC, could be modified To control RF when it detects a lot of voltage from a swr bridge. Then you would have what swan calls a "ALC" !
 
 

Re: uBitx heatsink temp? #ubitx

john@...
 

Thanks. I appreciate the input. This is my first HF radio, so I don't really know what to expect on some of these things.

Re: uBitx heatsink temp? #ubitx

Jerry Gaffke
 

If them heatsinks ever get too hot to comfortably hold your finger on them, that's too hot.   
A sign that you have the bias current too high, the PA-PWR voltage too high,
you are driving them too hard, and/or you need a bigger heatsink.
Or perhaps some sort of fault condition, like a shorted antenna or parasitic oscillation.
The smell of molten plastic could prove to be a handy warning flag, consider it a feature.


On Sun, Apr 29, 2018 at 04:42 pm, <john@...> wrote:
I am thinking about putting my ubitx in a pelican clone. How hot do the heatsinks get? I will mount a fan to pull air through, but want to make sure this won't start melting the case

Re: ND6T AGC implementation for uBIT-X

Michael Shreeve
 

Looking at the circuit, I see R1 simply is fed with full RF at the top, but relatively weak RF from the first amplifier in the transmitter. I'll measure the voltage, but its simple,   the wiper sends however much RF you need on to the rest of the output circuit. If this control could be mounted on the panel of the radio, you would have a transmit power control. But, it might not be good to take it off the board. 

This is done, if you look , right at the output of the first rf amplifier Q90 and RV1 passes the rf along to the input of a couple of dc direct coupled 2n3904s . Some sort of RF attenuator after rv1 would do the trick, and control voltage from a bridge swr detector could control that. I'm guessing the Circuit Don uses, or part of it, the AGC, could be modified To control RF when it detects a lot of voltage from a swr bridge. Then you would have what swan calls a "ALC" !


On Sun, Apr 29, 2018 at 2:44 PM, Arv Evans <arvid.evans@...> wrote:
Jerry

Instead of using AD8307 Log Detectors, have you considered just adding a bit of forward
bias to the diode detectors?  Conventional SWR bridges do not do this because they
usually do not have access to DC power, but in this application there is power available
inside the BITX case.

Arv
_._


On Sun, Apr 29, 2018 at 3:06 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
Not mad.
Just didn't see much point in continuing that particular conversation.
 
John, VK2ETA has a very elegant method for varying output power by shifting where
the signal hits the 45mhz filter:  https://groups.io/g/BITX20/message/46235
Elegant in that there is zero additional hardware. 
I can think of a couple ways this might not be absolutely optimal (slightly distorted 
transmit bandpass, might strengthen some nearby spur), but doubt those are
any real issue.  Might use it someday for constant power out to drive an amp.

I have one of Diz's Tandem Match kits, will listen to it on A6 and A7, display
swr and forward/reflected power to the second line of the LCD so I can constantly
monitor what's going on with the antenna while transmitting:
    http://www.kitsandparts.com/bridge1.4.php

Could then shut the transmitter down automatically if reflected power gets out of hand
through firmware, no additional hardware beyond that tandem match.
Doubt that's necessary, as my opinion is that a uBitx on 12v is bulletproof at any SWR.
Could use a couple $0.30 ebay ad8307's instead of the diode detectors,
allowing transmit levels of just milliwatts when tuning up.

Jerry, KE7ER


On Sun, Apr 29, 2018 at 01:39 pm, Michael Shreeve wrote:
Hey Jerry ! Don't go away mad. Many of us are thinking, but thinking is a process, for some of us its like melting a glacier ? A little slow. My thoughts are, and this is something I've been thinking about for a while but I admit haven't "searched" for in this forum reflector , How about a way to control output. Start with a control. But, to do that you should have a point in the circuit board where you can control the output in an efficient manner. Some would say "Why not just control the voltage to the finals ?" . That would be easy enough, but, seems like a more elegant method should be around, one that could be used with a type of automation. One that would perhaps allow a voltage from some sort of SWR sensing bridge and be coupled back into the radio, High SWR would cause lower output. I have an old radio that does this to protect the Solid State 200 watt amplifier on its back ! Been done for quite a while. I think they call it ALC. In fact, I pulled out the 70s early solid state transceiver docs and this is what it says. ALC DIRECTIONAL COUPLER - the directional detector serves several functions. First it provides the control voltage to protect the transistor linear amplifiers. Either full rated power in a reflected direction causes the directional detector to activate the ALC and reduces the drive power. In this way the transceiver is fully protected for antenna shorts, opens or severe mismatch. Second, the directional detector provides the control voltage to the ALC so that the ALC can be adjusted (at the factory) to limit maximum rf output. ... From a swan ss-200/100/and SS-15 manual. 

 

 





--
Michael Shreeve N6GRG

Re: U1 fried

Howard Fidel
 


I used an LM380.

Howard

On 4/29/2018 7:19 PM, Kd4epg@... wrote:
Love to see some pics of this adapter. Also, did you look at the NE5532? May be an easier one to make.
--
72 and God bless
KD4EPG


uBitx heatsink temp? #ubitx

john@...
 

I am thinking about putting my ubitx in a pelican clone. How hot do the heatsinks get? I will mount a fan to pull air through, but want to make sure this won't start melting the case.

Re: U1 fried

Dennis Yancey
 

Love to see some pics of this adapter. Also, did you look at the NE5532? May be an easier one to make.
--
72 and God bless
KD4EPG

Re: ND6T AGC implementation for uBIT-X

Jerry Gaffke
 

I'm thinking that's not much of an argument.

I really don't give a rip what the IRF510's were designed for.
They work.
And they will do fine with SWR's well in excess of 20:1 on a uBitx powered from 12v.

If using automobile turn signal switch FET's makes you feel like less of a ham,
then by all means invest in some RD16xxx's.
And I'm willing to be convinced that moving my $109 uBitx to RD16xxx's is a good idea.
But have yet to hear a vaguely compelling argument as to why.

Jerry, KE7ER
 


On Sun, Apr 29, 2018 at 03:38 pm, John Backo wrote:
Think, Jerry. The IRFxxx devices were made to be power supply
switches. The RDxxx devices are made for RF.

What do you think are the die differences?

Incidentally, the RD16xxx devices are made to withstand
an antenna SWR of about 20:1.

john
AD5YE

Re: ND6T AGC implementation for uBIT-X

Jerry Gaffke
 

Arv,

I'll first try the raw design as presented by Diz, read those voltages from the Nano,
Then add some scheme to bias the diodes.
If I want better results, may yet try the AD8307's.

The diode bias scheme won't give the same dynamic range as an AD8307,
but it this case it will probably be good enough.  And much easier.

All tandem match diode compensation circuits I've seen thus far use op-amps with a diode in
the feedback loop to do that voltage level shift.  Maybe I'll find out why.

Given the schematic here:  http://www.kitsandparts.com/bridge1.4.php
the obvious thing would be to cut the ground away from the two 50 ohm resistors
but keep the resistor bottoms joined at a new node I'll call VBIAS.
Then add a new 1n5711 diode from VBIAS to ground, also a 0.1uF bypass cap in parallel with that new diode. 
Then add a 50k resistor from a 5v regulated voltage rail to VBIAS, for a constant roughly 100uA into the bias diode.
Since this only matters when transmitting, we could steal that 5v from the IRF510 bias voltage circuit.

I think we want the roughly 100uA  bias current into the diode to be roughly equal to the average current
through the two signal diodes when at the intended lowest end of the scale.  I'm not really sure what that will be.
Will be some non-linearity as we go up from there, but that could be corrected for with a table lookup.
 
R3,R4, C3,C4 go away.
The junction of D1,C1 goes to A6 of the Nano (CW key and paddle firmware moved to A0, A1 digital inputs?).
The junction of D2,C2 goes to A7 of the Nano (currently unused.
If the signal voltage will exceed the 5v max into the Nano, then adjust the transformer windings.

Something like that.  
Anybody see of a better way?
Any obvious errors?

An advantage of the diode bias scheme is that reading straight voltages makes the SWR calculation
on the Nano easier than if using the dB readings from AD8307's.  

Jerry, KE7ER


On Sun, Apr 29, 2018 at 02:44 pm, Arv Evans wrote:
Instead of using AD8307 Log Detectors, have you considered just adding a bit of forward
bias to the diode detectors?  Conventional SWR bridges do not do this because they
usually do not have access to DC power, but in this application there is power available
inside the BITX case.

Re: ND6T AGC implementation for uBIT-X

John Backo
 

Think, Jerry. The IRFxxx devices were made to be power supply
switches. The RDxxx devices are made for RF.

What do you think are the die differences?

Incidentally, the RD16xxx devices are made to withstand
an antenna SWR of about 20:1.

john
AD5YE

Something different this time...

Brian L. Davis
 

I've seen some odd things but this is one of the oddest.
I have experienced intermittent receive and transmit, just momentary that seemed to be cable related.
Traced it to the BNC jack.
Seems the female socket in the jack is slightly oversize and the male pin was a loose fit and occasionally showed an open circuit.
Changed the jack and the problem went away.