Both messages from the BITX20.io email group.
26 July 2019
Peeps:
The Antuino has been in use for over a few months now. We had produced 100 of them that were sold at the FDIM. These work quite well for an SWR meter and Antenna Analyzer, but they were sub-optimal for serious RF work. For those who bought it for SWR measurements, you can continue to use it. For those who want to improve it, read on.
It was noticed that the db scale was not accurate. Antuino was designed to be an accurate and precise instrument. The db readings should be within +/- 1 db accuracy. However, they tended to vary by as much as 2 db on the upper range of power measurments. This was finally tracked down to having too much gain. I had prototyped the original with 2N3904 transistors but in production we used BFR93W as I guessed it would have 'better' performance. It turned to be a bad choice. The higher gain resulted in compression of signals above -30 dbm. This restricted the useful range of the Antuino to about 50 db.
I am attaching the new (Version 2.1, it is a minor edit of the V2 that was sold at the FDIM). In summary, these are the changes:
1. The new software with a cleaner way to calibrate is on https://github.com/afarhan/antuinov2.1. Please upgrade to it even if you don't want to carry out the other hardware modifications.
2. Replace the Q1 and Q3 transistors to 2N3904. Although the PCB pads are SMT, you can solder the leaded type by twisting the legs around.
3. In the first IF amplifier (that immediately follows the mixer), we decrease the emitter resistor (R18) from 100 ohms to 51 ohms (you can also parallel a regular 1/4 watt 100 ohms with the original 100 ohms to get to 51 ohms). We also parallel the (R19) 220 ohms collector load with a 10 uh inductor (you can use 10 turns on FT 37-43), the exact value is not important.
4. In the second and third IF amplifiers, we replace the 100 ohms emitter resistors (R7, R34) with 10 ohms and remove the 4.7 ohm resistors (R32, R4).
5. In the last IF amplifier, we change the load resistor(R33) from 51 ohms to 220 ohms.
You need a few 10 ohms resistors and a 220 ohm resistor. You can resuse the R31 at R18.
I have detailed the modifications at http://www.hfsignals.com/index.php/2019/07/26/modifications-to-the-daytonfdim-2019-antuino/
Attached are the images and the circuit.
73, f
_._,_._,_
August 3, 2019
Every instrument has limits on its accuracy. While making the Antuino, I was well aware of its deficiencies.
I made a decision to keep it simple to a point where a radio ham could throw this thing together in an evening or two.
In order to overcome the limitations inherent in the Antuino design, the complexity could have been prohibitively complicated and expensive.
On the other hand, it is an extremely useful instrument that grows on you. I no longer use a frequency counter or the specan. I rarely use the oscilloscope. Antuino does most of my measurements.
There are two very important things you must be aware of while using the Antuino :
1. In the power measurement mode (the problem is non existent on swr or sweep mode), any reading above 25 mhz could be an image. Thus, if you see something at 35 mhz, you will have to do some mental math to figure out if it is not an image. An easy way to know is to add an external low pass filter with 25 mhz cutoff.
You have to use it like a radio in this mode. If you want to measure the harmonics from your pixie radio with 7040 crystal, tune to 14080, 21120, 28160 and measure. It is as accurate as any spectrum analyzer with more than 80 db of usable bandwidth. Raj and I struggled to get this for a month.
2. Unlike a full fledged spectrum analyzer, Antuino has just one bandwidth of about 7 KHz. This is enough to made IMD measurements at 20 KHz tone separation. The sweep plot does only 128 readings. Thus, if you sweep a low pass filter from 0 to 20 mhz, it will measure the filter response every 120 khz. If there is something lurking between the steps, it will miss it. This is a common challenge with spectrum analyzers. So, a crystal filter should be swept at less than 100 khz.
There is a software hack to mitigate this. First : introduce another control for step size. This can slow down the plot. A 30 Mhz sweep at 5 khz steps will involve 30 x 200 =60000 readings.It could take minutes. The Second : write a more optimal Si5351 routine that changes frequencies faster. I know that smaller jumps can be instantaneous on Si5351. I don't know that hack. If someone wants to take a stab at it, I am willing to work with them.
In a nutshell, Antuino is a very useful instrument. You don't have to buy it. You can build it. It is just as challenging as a direct conversion receiver. It does a fabulous job though. It can measure oscillator frequencies, it can measure amplifier gain, distortion, frequency response, it can measure filter response, it can show mixer behaviour, it can tune your antenna, it can measure power from a few uV to 100 mv and more with attenuators. It does all this slowly but surely. Like any precision tool, you must know its limitations and use them as an aware user. I would wager that if you have to choose just instrument for your lab, it would be this; Apart from a DVM.
And (I repeat) don't buy it, build it (grin)
- f
Posts
