Accidental Calibrate #ubitx #calibration


I've just assembled my uBitx into the enclosure and powered it up with no problems. The audio quality was excellent on receive. I was scrolling thru the Settings about to set the BFO and accidentally clicked on the Calibrate mode, it went straight into a note, ("Tune to exactly 10Mhz") I turned the Encoder and pushed the button to try and get out of the mode. Since this disaster my excellent audio quality has deteriorated into a squeaky garbled nightmare.
How can I fix this and restore my uBitx to its former glory.??

Re: PCB Artwork hint

Clark Martin

Yeah, I had to do that.  One time I was working over a large backplane board.  I, and the board layout person, missed one TINY detail.  The daughter board connectors were swapped, left for right, oops.  We ended up building the board with the connectors on the back side for one of the prototypes.  It worked.

Clark Martin

On Jul 27, 2018, at 11:14 AM, AA9GG <paul.aa9gg@...> wrote:

I didn't do the tape-up, but I did have to verify it to the schematic!

On Tue, Jul 24, 2018 at 6:16 PM, Robert Ogburn <ogburnrobert2@...> wrote:
You dinosaur, me pond scum...
Red Tape, Blue Tape, Black Tape, DIY pads.....6x8 foot tapeups...  Ah, the xxxx old days!

Paul Mateer, AA9GG
Elan Engineering Corp.
NAQCC 3123, SKCC 4628

Re: End Fed antennas w/ uBITX #ubitx

Chuck, N1KGY

On Fri, Jul 27, 2018 at 11:00 AM, iz oos wrote:
Where could I get the schematic and the hints to make an acceptable copy of a good design?

From AA5TB's website -
Recommended reading for anyone who intends to build or buy an EFHW antenna setup.


Re: uBITx PCB power connector

Chris Clarke G3SQU

On Fri, Jul 27, 2018 at 03:26 PM, Ian Reeve wrote:
Bright Components
Thanks for the info Ian, and for the offer. I've ordered a 5-pack from Bright's - it's good to know they're the right ones!

73 Chris

Re: End Fed antennas w/ uBITX #ubitx

Chuck, N1KGY

On Thu, Jul 26, 2018 at 03:32 PM, Jerry Gaffke wrote:
Here we have a forum with perhaps thousands following the conversation,
trying to make sense of all this.  For some readers, it will be the loudest or most
persistent voice that carries the argument.  

<b>If I see something of consequence proclaimed here that I know is wrong,
I will not just let it ride.  That's a disservice to everyone else here who might choose
to act on that information.</b>  If it bruises an ego somewhere, that's arguably better
than a hundred forum readers wasting time and/or money trying to follow through.
Completely agree.  <b>Emphasis added</b> is my own. 
YMMV, and please mind the gap.

Re: End Fed antennas w/ uBITX #ubitx

Chuck, N1KGY

On Thu, Jul 26, 2018 at 12:42 PM, <m5fra2@...> wrote:

I have a homebrew EFHW  tuner that works on 40m-30m-20m. <b>I use

a different wire for each band, a half-wave length.</b>

This is the best case scenario.  The geometric mean between 7 and 14 Mhz is 12.125Mhz, and assuming an optimum design frequency for the transformer at ~12Mhz is done, then losses anywhere between 7 and 15Mhz should be very manageable with Type 43 or Type 31 (better) core material, so long as limits on Flux Density and inter-winiding voltages for the chosen core size/geometry are observed with care.  With careful construction, you could achieve a loss of  1.5dB or less across these three bands.  15M is 'within reach' at a slightly higher loss figure, using Type 43 material, without changing the design or materials at all. 

You could certainly "get away with" just two antenna wires - one for 40/20 meters, and another for 30/15 meters (losses slightly higher on 15M,  but still a very serviceable configuration when 15M is open).  Similarly, 80M should be well within 'reach' using Type 31 material - in this instance you might get 5 bands - 80/40/20 Meters on one wire, and 60/30 meters on a second wire, with careful compromises on the wire lengths (and heights) - the VSWR at the feed point should be below 3:1 on all 5 bands in this case, so feedline losses will be low enough that "touching up" the VSWR at the shack end of the line as needed will not incur punitive losses.

On Thu, Jul 26, 2018 at 12:42 PM, <m5fra2@...> wrote:

can never understand the need to make one antenna to

cover a wide frequency range, it will always be a compromise.

I couldn't agree more.

Re: uBITx PCB power connector

Ian Reeve

Hi Chris they are standard 0.1" connectors type KK.I get mine from Bright Components as per the attached photo.

If you would like to give me your details and address I will send you one.They are not expensive from Bright Components if you wanted to order a few,the folk are on eBay. Regards. Ian M0IDR

From: <> on behalf of Chris Clarke <csclarke@...>
Sent: Friday, July 27, 2018 11:01:05 PM
Subject: [BITX20] uBITx PCB power connector

Been searching around and can't seem to identify the type of 3-pin connector used to supply the 12v and 15v power to the uBITx main board.

I want to make up a new power lead (so I need the female socket wired for preference). Can anyone tell me what type of connector it is?

Chris G3SQU

Re: End Fed antennas w/ uBITX #ubitx


If you need help sing out.  Antennas can be fun and not over complicated.

Sorry bout the signal to noise ratio.

Keep these two links in mind and their links as well.  Ton of good
info and practical application.


Re: End Fed antennas w/ uBITX #ubitx

Chuck, N1KGY

On Wed, Jul 25, 2018 at 11:52 AM, Dennis Yancey wrote:
Take a look at the ARRL Antenna Book,  endfed zep antennas. You will find the length of the wire is not critical nor is the length of the feed line, as long as you are feeding it with 300 ohm twin lead or 450 ohm adder line. Your tuner must gave inputs for Balanced Line and you are all set. The Balun in the tuner will take care of you.


Apologies but I must disagree.  Particularly, the use of the phrase "not critical" with regard to the length of the antenna wire and/or parallel "matching section" of a Zepp antenna is grossly misleading.  Where the actual length of the [1/2 wave radiating element] deviates substantially from the nominal 1/2 wavelength figure, and/or the [1/4 wavelength * VF] of the parallel-line matching section deviates from the formula, the antenna is no longer a "Zepp" antenna by definition... even if it operates with reasonable efficiency.  This lack of rigor in your primary hypothesis reduces any subsequent conclusions to a plainly anecdotal nature, which doesn't serve the topic under discussion.

Further, there are definitive differences between a true EFHW and a "traditional Zepp Antenna" per your own citation, above.

In the simplest language, the unterminated side of the parallel transmission line (matching section) used with a Zepp antenna, being very near (1/4 wavelength * VF long) is in fact acting as a 1/4 wave counterpoise, and thus forces the feedpoint impedance to present a nominally low impedance due to this effect (though the reactive component of the feedpoint Z0 can still be quite high) . 

Continuing this very simple analysis of a "Zepp antenna" we find that the combination of the "hot side" of the parallel transmission line (being 1/4 wavelength * VF) PLUS the 1/2 wavelength radiating element ~= 3/4 wavelength... another length of wire which forces a nominally low impedance at the terminated end, although again, a substantial reactive component may still be present.  So we have two terminals which 'agree' to present a low impedance to the matching unit at the end of the "matching section" of the Zepp antenna.  Without the "1/4 wave matching section", the input impedance is totally different, and thus the parallel-line "matching section" is a mandatory component of a "Zepp Antenna"...

An EFHW has two terminals (Antenna plus GND/Counterpoise) which 'agree to present a very high impedance' to the matching unit.  The Zepp and EFHW are very different antennas, and require very different matching units.  But considering the antenna plus matching unit together in each case, there are similarities which help us to understand the differences between a Zepp and an EFHW, and to choose our best option, given the circumstances -

1. Both the Zepp and EFHW are the very opposite of being "balanced" antennas - very nearly ALL of the antenna impedance is present at ONE terminal of the parallel line (or EFHW matching unit), while the other is effectively an "open terminal" before we examine Displacement Currents" and parasitic coupling effects.

2. Using a Zepp antenna as described in the cited text is substantially different from what is being described by the OP, because, the "balanced line" of a Zepp antenna -- more correctly "parallel line matching section" in this instance -- doesn't even exist in the EFHW design.  Any matching effects attributed to the parallel-line matching section of a Zepp antenna, plus its function as a counterpoise, must be performed by the "matching unit" of lumped-value inductors and/or capacitors, <b>plus a functional counterpoise/ground</b> to absorb the displacement currents present - else no antenna current can flow.

The Zepp is as far from being a "balanced antenna" as is possible; the "balanced feed line" of the Zepp isn't a feedline at all.  It functions as -

A. a transmission-line transformer;
B. a counterpoise;
C. a loss element (skin effect plus mismatch losses)

Each of these effects serve to reduce the apparent VSWR at the feed point. 
The very same functions are operative in an EFHW, however the (1/4 wavelength * VF) "matching section" of the Zepp is replaced by the following -

<b>1. an "untuned"  ferite-core transformer</b> of very high surge impedance (across all bands of intended operation), used with a "resonant half-wavelength of wire" for the antenna element, plus a counterpoise. In this instance, whatever antenna reactance is presented at the end of the antenna wire is -NOT- compensated for by the transformer, but will pass the uncompensated reactance on to the coaxial feeder -- which will then act as a separate transmission-line transformer, bringing the rig-end of the coaxial line either closer to, or further away from, the intended 50+j0 ohm match. This is why you'll very often see folks recommend particular lengths of a particular coax for use with such EFHW antenna kits/designs.

A counterpoise is required for correct operation, altough it will typically function with reasonable efficiency using a counterpoise much shorter than 1/4 wavelength, when the antenna presents little or no reactance (i.e. AT or VERY NEAR the natural resonant point of the antenna wire).

<b>In this scenario, losses are almost directly proportional to the reactance presented to the transformer by the antenna+counterpoise;</b> thus the only way to minimize losses and optimize efficiency with such designs is to adjust the wire to a "resonant length", i.e. a length which presents very little or no reactive component to the transformer, which is exactly how you "tune" a Zepp antenna to resonance... by "pruning" the wire.  Like the Zepp antenna, an EFHW is typically a monoband antenna, although reasonably efficient operation on the second harmonic can be achieved if a good quality ATU is used at the rig-end of the coax.

While this is a "no tune" setup, it typically has far higher losses than #2 or #3, below.

<b>2. a "parallel resonant" LC circuit where either the L or C is adjustable to resonate the entire antenna system</b>.  Stated differently, the antenna + matching unit + ground/counterpoise presents little or no reactance to the coaxial feedline (once tuned for the desired frequency of operation); there is some form of impedance transformer to convert the very high Z0 at the end of the wire to a (nearly) 50 ohm unbalanced input. 

The "matching range" of such an antenna system is very tightly associated with both the length of the antenna wire, as well as the range of the variable component(s) in the matching unit (most typically the capacitor).

The most common approach for the impedance conversion in an EFHW matching unit is a resonant transformer using a tapped coil (a.k.a. an "auto-transformer") or an isolated input winding on the coil of the resonant circuit.  Alternatively, a capcitive divider can be used to obtain a similar impedance transformation to the desired input impedance.  Where one (or both) of the L and C components are variable, this type of tuner is capable of matching wires which are not "nearly resonant" (again, integral of N * (Lambda/2), but still present a suitably high resistive component of Z0 to the antenna terminals of the matching unit.

With a parallel-tuned LC matching circuit, suitably designed, it is very reasonable to expect an efficient match to be possible across two bands, where the higher band is (roughly) 1.5X or 2X the frequency of the lower band, without using tapped or switched coils.  I.E. 40 and 20 meters, 20 and 15 meters, or 40 and 30 Meters (with some compromise).  Such a setup must be re-tuned when changing bands, and may require "touch up tuning" even when QSYing within the primary (lower) band for which the unit is designed.

<b>3. an L-Network (or Pi or Tee network)</b> may also be employed to provide the necessary impedance transformation between the antenna and the feedline, as well as compensating for ractance present at the end of the antenna wire with good efficiency. 
In all of these cases (L-net, Pi-net, Tee), as with the "parallel resonant LC" match described above), the variable capacitance on the antenna side of the matching coil can be used to "tune out" a certain amount of reactance at the antenna terminals to optimize the match presented to the coaxial feedline (a variable/switched inductance can in the matching unit may also serve this function).  Having a variable capacitance on the antenna side of the matching network, plus a switch-selected range of inductance, expands the usable matching range of this (these) matching units even further than what is easily achievable with the "resonant LC" matching units described in #2, above.

I have spent a substantial amount of time and effort over the last 8 years or so building and testing all of the above described antennas, specifically for portable QRP use.  When properly designed and constructed, the traditional "Zepp" antenna (including it's 1/4 wave matching section, and the EFHW + L-Net tuner yield the best efficiency - nearly always above 90%, even when very lightweight and compact materials are used.  The Parallel-tuned LC EFHW is almost as efficient on the primary band even when thin (#24~#28 gauge) antenna wire and small toroidal coils {T50 and even T37 sized cores] are employed - however, efficiency will be substantially lower [on the order of 2 to 6 dB) on the second band you choose to design for.  To obtain highest efficiency across two bands (or more), toroidal cores should be avoided, and every effort should be made to maximize the Q of the air-wound coil across the bands of desired operation - where operation on more than two bands (as described above) is desired, entirely separate units may be constructed: each covering two bands with reasonably high efficiency - while this may not be the compromise you're looking for, it is one which should be considered before being discarded. 


Antennas which employ a "9:1 untuned transformer plus a "non-resonant" length of antenna wire aren't an EFHW to begin with, and shouldn't be lumped in with EFHW (or Zepp) antennas as a category, to begin with. 

These non-resonant wire + non-resonant transformer arrangements are going to have higher losses than any of the above EFHW configurations, regardless of any claims to the contrary, because uncompensated reactances always increase losses in an antenna system - the Laws of Physics simply aren't negotiable - power not delivered to the load must go somewhere else, and that "somewhere" is always in the  "losses" category.  These EFRW (End-Fed Random Wire) antennas often serve "well enough" to make contacts, but unless we know what compromises (a.k.a. losses) are designed into the matching unit to obtain a low VSWR, and what the actual efficiency of the system is at the desired operating frequencies, a quantitative analysis isn't possible.

On Wed, Jul 25, 2018 at 11:52 AM, Dennis Yancey wrote:
Your tuner must gave inputs for Balanced Line and you are all set. The Balun in the tuner will take care of you.
I won't even touch this tar baby of  a statement.

In conclusion, the "All-Band, No Tune Antenna" is just another fools dream, upon which countless hours and dollars have been spent by countless hams, for no other reason than that people don't take the time to understand the compromises which such a product must incorporate - i.e. severely attenuated efficiency on some or (typically) all of the bands which they claim to operate on.  Choosing compromises doesn't automatically make a person a fool - I don't mean to suggest such a premise - but choosing compromises without first understanding them has certainly opened the door to all sorts of foolishness in all sorts of human pursuits, not just ham radio.

uBITx PCB power connector

Chris Clarke G3SQU


Been searching around and can't seem to identify the type of 3-pin connector used to supply the 12v and 15v power to the uBITx main board.

I want to make up a new power lead (so I need the female socket wired for preference). Can anyone tell me what type of connector it is?

Chris G3SQU

Re: End Fed antennas w/ uBITX #ubitx

John Smith

Excellent, thank you Jerry, and Allison. That nugget got lost in the rumble. I guess I got the impedance flipped in my mind, and it's probably why I've been hearing that a 49:1 is usually used for the half wave antenna. I think I may have picked a complicated antenna to try out for easy deployment and use. Observing that "discussion" earlier was a bit discouraging, but it's a good thing I am stubborn enough to keep looking for a good starting point and clarification. I'll write my data down so I can flash my cred on this topic next year.

Re: IRF510 ubitx failure


There will be signs if you look.

THe IRF510 is rated for 43 watts dissipation, specsheet and all.

The issue is that you have to be able to get the heat generated by that power out of the part.
Therein lies the problem as Any (and all) TO220 parts have lousy thermal resistance so it
can get very hot before the heat sink is really cooking.   Max temp is 150C at the die inside
the plastic that is 300F!  However keeping under 100C is better for its lifetime.  It takes
a lot of heatsink to do that.

The other issue, not paid attention too.  We know they can get hot... in a closed box where
does the heat go?  Unfortunately the answer is nowhere it just gets hotter.  I noted this with
mine and other IRF510 amps.  I could key down for 10 minutes easily on the bench in
the open and the heatsinks might hit 125F.  Put a card board box over it and it would get
to 175F after 10 minutes.   Airflow is important with those heatsinks 10W is fairly easy.

Hint: if the transistors are mounted to the back of the box with the right insulators
and a large flat heatsink it helps.  The other is vent hole in the box and a small 2"
fan to move the air though to the outside can also be effective.

To further make the point I have a IRF510 push pull amp (WA5EBY design) with a
8x4x1.25 inch heatsink.  With the 28V power supply it runs on 40m and 20m about
50W and 37W at 10M out and key down for 5 minutes the heatsink (on the back
of the box out in the air) gets to 115-120F which is not very warm.  Those IRF510s 
have survived shorted coax, shorted antennas, wrong antenna for the band, 
and no coax at all at full power and continues to do so since I built it back
in 2006 on the same parts.   With 50W out the input power to the pair of
transistors is about 100W so theres about 50W of heat (or 25W per IRF510)
the heatsink has to remove.

So they are not fragile if not allowed to cook.


Re: VK3YE AGC pcb.


That is hilarious, I never know they ever made anything like that, unless....... it was a April fools day joke.


Re: Receive pitch shifts when confirming.... #arduino #ubitx #ubitx-help

Ion Petroianu, VA3NOI

I do not know what number you shall write in uBitx manager calibration field for your particular clock crystal (25 MHz crystal for SI clock generator). Why don't you try calibration using the setup menu of the radio itself? If the UT beacons on 5 MHz, 10 MHz and 15MHz cannot be received at your QTH try a SW broadcast station on a known frequency.
At my QTH I can hear loud and clear a Spanish broadcast on 7.210 MHz almost every day around 7:00 p.m. EDT. 
To be able to use the S-meter you should have a voltage detector of the audio signal installed ay the hot end of the volume potentiometer.
You already have that if you install Don/Kees AGC kit.
You will connect your unused purple wire on the digital connector to the output of the voltage detector and enable S-meter function through uBitx manager.
To read more about that check Ian's website


Re: IRF510 ubitx failure


I think IRF510's were originally designed for the automotive industry for switching applications like turn signals and relay drivers. I have them in homebrew radios delivering around 10 watts and they are still working on the original part years later. There are a lot of poor quality parts around anymore me thinks.


Re: uBitX SSB transmit oscillations on 40m only #ubitx


First a correction. The distorted output is on 20m not 40m. Now that's embarrassing! A form of ham dyslexia perhaps?

Audio feedback? Although I had gone through all the bands with no distorted output, I have been mostly comparing 20m with 40m. Distortion-feedback on 20m, clean on 40m. I just went back through all the bands and noticed that I hear a little audio from my speaker as I go up the bands. 80m, 60m, and 40m are good with a clean signal on the 'scope, nothing from the speaker. 20m is badly distorted with feedback and garbled audio from the speaker. 17m and up (in frequency) clean signal but faint audio from the speaker.

Re: IRF510 ubitx failure

Sean W7SKD

Bill - just read through your thread...mine was a bit different symptom wise.

I was able to power up just fine, display was working etc....the only thing I could notice immediately was that there was no power on transmit.  The OTHER thing I was noticing earlier is that it seemed like I was either unstable on transmit frequency or OFF frequency - on 40m I couldnt get a good match through my tuner, which isnt at all normal.

so...I'm not sure my irf510s were quite as cooked as yours, but they needed to be replaced for sure (proof is in the pudding).



Re: VK3YE AGC pcb.

Jerry Gaffke

If the clam shell gets that much attention, 
imagine having a cell phone from 1955:

On Fri, Jul 27, 2018 at 12:35 PM, Richard Spohn wrote:
"Was that the 1980's asking for their phone back?" - Jack.
Jack, a classic story! I did try texting on my rotary phone. No love. - Rich



As I thought there was a broken wire on one of the bias pots. Replaced and it setup as advised. Very quick test (blowing into mic)  showed ~12W out on 20m, ~15w on 80m. Yet to do the rest of the setup but the RX seemed lively and heard SSB on 20m with a short length of wire. The AF output is just about adequate and drives a small speaker. I found another poor joint, mine this time, which accounted for the low audio on the initial test. Rest of setup and wiring the audio CW filter is next, on another day!


Colin – M5FRA


Re: uBitX SSB transmit oscillations on 40m only #ubitx


The latest:

If the microphone jack is mounted in the enclosure, I get the distorted feedback on 40m, but clean on other bands.
If the microphone jack is removed from the enclosure but held near its mounting point, I get a clean 40m output.
If the microphone jack is removed from the enclosure and grounded to the enclosure with a jumper, I get a clean 40m output.
With the microphone jack is anyplace on, over, beside, near the RF circuits, etc. I get a clean 40m output
If I cover the microphone jack in insulation and put it in the enclosure, I get the distorted feedback on 40m.
Just for grins I removed a jack from my BitX40 to replace the one delivered with the uBitX and I get the distorted feedback on 40m if mounted in the enclosure.

Any ideas? Suggestions to diagnose?

And thanks Allison. It is in a metal enclosure. I'll re-check grounds. Odd, though that the CW key and audio jacks also grounded to the enclosure work fine.