Tuesday, January 26, 2016

Simpleceiver ~ Part 27

Some light at the end of the Tunnel!

 
My XYL has been steadily improving and hopefully we expect her release from the nursing facility in about a week or two. That means I can get back to heating up the iron and continuing with this project. I have not had any comments regarding those who forged ahead and started work on the transmitter stage and so that either means there is a complete loss of interest or the project is dead. I sure hope not as I think given what I have observed from the receiver performance this project could rival some of the currently popular homebrew radios.
 
The basic transmitter board had 4 elements including the microphone amplifier, the carrier oscillator, the SBL-1 and the IF Amp/Filter block. The output is of course at 12.096 MHz which is the filter frequency and just so we all are on the same page the Zout is 50 Ohms. The block which will follow the basic transmitter board is the frequency translation board which performs some very specific functions. The 12.096 MHz SSB signal must be mixed with the local oscillator so that one of the resultant frequencies is on the band of choice. From our earlier discussions the mixing process results in sum and difference frequencies and a second effect is depending whether the Local Oscillator is above or below the filter frequency there may be a sideband inversion. We certainly do not want the case where we have the receiver on LSB and the transmitter is on USB.
 
The Simpleceiver receiver used an LO at 5.0 MHz so that the one of the resultant frequencies is in the 7 MHz range. Here it is the IF- the LO result we want (12.096 - 5.096 = 7.0). But there also is a component where you have 12.096 + 5.096 = 17.192 MHz which must be filtered from the output. So following our frequency translation we must have another Band Pass Filter centered on 7.150 MHz so that only the subtractive mix is in the output.  Given that we will also be using an SBL-1 for the frequency translation there will be a need for several amplifier stages to boost that signal to something directly usable on 40 Meters. A J310 based Dual Gate MOSFET would serve this purpose nicely for the 1st stage of amplification. Thus we can actually use the Simpleceiver RF amplifier stage as a circuit block following the Band Pass Filter. Again we stated that many of the circuit blocks built for the receiver could be simply reused for the transmitter and so this is what we will do.
 
I hope in the next week to post an amplified block diagram of the transmitter circuits that will include the follow on boards to the basic transmitter.
 
 
73's
Pete N6QW

Thursday, January 14, 2016

Simpleceiver ~ Part 26

 N6QW Back on the Blog (at least part time)!

A family medical emergency had caused me to stop blogging but that situation is somewhat improving and so I am now back on the keyboard. After three weeks of being turned off, I can see where my soldering iron actually has rust on the tip --that is not good and I soon hope to be soldering some parts on island squares. I might also have to recalibrate the CNC Mill.
 
Our plan is to now focus on the building the transmitter and then take the next step of configuring the blocks (receiver, transmitter, control and LPF + TR) into a working transceiver. During the last three weeks KK6FUT has built the DCR version and has a you tube video -- it sounds really good as does the superhet Simpleceiver  built by Mikele, 9A3XZ. I do know that others are building portions of the project so if anyone has built either the DCR or superhet variant please let us know.
 
In moving on the transmitter, it would be well to review the final configuration. Essentially the design intent was to build completely separate receiver and transmitter boards and to share a common VFO and BFO and thus one would have a transreceiver. There are benefits as well as potential problems in this approach. The chief benefit is that you can optimize the receiver for best performance while not impacting the transmitted signal.
 
[Many older technology transceivers had this problem and frequently the alignment instructions would strongly suggest that the final alignment was a compromise between that which could be achieved with separate circuits. Today that is not so much a problem and it is probably for the second reason that separates were chosen and that is experimentation. With the separate approach it would be possible to test variants individually without compromising overall performance. For instance one change might be to put a 6 pole filter in the receiver and only use a 4 pole for the transmitter. Or to even have switchable filters, SSB/CW, in the receiver. You get the idea.]
 
One other problem rears its head especially in the older technology transceivers and that was circuit loading that caused frequency shifts when going from transmit to receive. Imagine my surprise when I was working with the Ten Tec Model 150A SSB/CW transceiver to find there was a compensating circuit in the BFO circuit so that the BFO was calibrated in receive and then using circuitry that was only switched into play on transmit was calibrated on transmit.
 
Being on the same frequency in both transmit and receive is critical! Typically where there are crystal oscillators involved, variable circuit loading can cause a shift in the fundamental frequency. A shift of 100 Hz can be picked off with some of the new crop SDR radios --and if you are on 40 Meters many times the newly minted Extra's running those $10 K SDR radio using a 37 inch flat screen for a waterfall display will tell you that you are 100 Hz low (or high) followed quickly by a command to get on frequency --you are on frequency it is just the BFO shifted.
 
My original thoughts were to share a common BFO and VFO for the receiver and transmitter. Because of what was just described I am now leaning to separate BFO's and a common VFO. The VFO if you use the AD9850 or the Si5351 is fairly immune to frequency shifts caused by circuit loading since the frequency being generated is a result of mathematics and not an inductor / capacitor. Thus the plan is to have separate BFO's that can be individually "netted" (an old term for being on the same output frequency). If the Si5351 is used it may be even a lesser problem as the BFO like wise could be generated mathematically and not subject to loading. But with the AD9850 (or LC VFO or VXO) you will have to supply a BFO signal. In my current build I am using the AD9850.
 
So Ok how do you get two BFO's on the same frequency and how do you switch them on and off. Having two BFO's running continuously is more of a problem on the receive side than the transmit side. If the transmit BFO was only 20 Hz off (not enough for the SDR guy to scream at you) but is sufficient to be heard in the receiver. So the plan is to leave the receive BFO run all of the time and to only run the transmit BFO when in transmit. In our earlier post on control circuits which uses a NE555 that will work nicely to power on the transmit BFO. So that part of the problem may be minimized.
 
The real nut to crack is the netting. Assuming you have the receiver working properly and on the correct frequency then it is a matter of a little cut and try. In my receiver build I used a socket for the BFO crystal (three pins of a SIP socket) and this functionality will provide a means of finding two crystals that will be very close in frequency. Before I started this project I had a batch of 12.096 MHz crystals and then purchased another batch of Series type crystals. My process for finding two close crystals was to tune in say a net on 40M with the normally used BFO crystal and then to simply cycle through a batch of crystals that I had. It is a pretty simple mater to find ones that are higher or lower in frequency as the pitch of the voices will change. There will be a grouping that then seems to produce the same pitched sound. Run this group through a second time and "listen" for the closest ones. The ear is quite good at this process. The ones that pass this test are candidates for the transmitter carrier oscillator.
 
What is significant here is that the crystals are tested in a circuit that is used in the receiver and that same circuit will be used in the transmitter. The feature of this circuit is a small trimmer cap to put the crystal on the exact frequency needed. The same value trimmer is used in the carrier oscillator. Thus it will be possible later (just like the Ten Tec Model 150A) to calibrate the transmitter carrier oscillator on the same frequency thereby bypassing the wrath of the 40M SDR Frequency Police!
 
I will stop here and pick up next time with more discussion of the transmitter circuits.
 
73's
Pete N6QW

Monday, January 4, 2016

Simpleceiver~ Part 25

Temporarily QRT @ N6QW

 
 
Due a family medical emergency there will not be much coming out of the pipeline from N6QW for a short period of time. However there has been enough information presented to proceed with the transmitter portion. You would only need to add a pre-driver and driver stage plus the final amp and the low pass filter and you are there.
 
Here are some circuits to consider for the final pieces of the transmitter portion. These are proven circuits having been used in the 2009 Tri-bander, the JABOM and the KWM-4. I strongly recommend NOT using the IRF510 as is the favorite of the BITX crowd but instead put a real RF device in the final amp. For those who want to pick apart the pre-driver/driver check EMRFD for an explanation and specifications of the circuit. The RF Final is my own design and the LPF you can simulate in LT Spice. You will need to fit in a TR Relay and the final integration wiring. If  at this point you are not capable of doing that then I suggest you do some Internet research on how this is done. I just do not have the time right now to provide that information.
 


 
 
Good Luck and I hope this temporary cessation is in fact temporary. That said  the objective has been to share my experience and knowledge of how to scratch build a radio. The over documentation in this project was so that the project didn't end up like so many kits --solder all the resistors first and then .. Hopefully I have provided information on the functions of the circuits and how with tools like LT Spice it is possible for you the homebrewer to make changes. I know how to do it -- the important point is that you know how.
 
 
73's
Pete N6QW