开云体育

ATU-10: an alternative PCB layout?

ZAO All Messages By This Member #1027 Hello! First of all - thanks, David, for this awesome project! My attempt is about bettering the ground at the RF-power part of the board. Originally it is too much cut-through, forcing more RF-currents to round the board - dangerously approaching the small-signal part of the circuitry. Please have a look at what I've done so far: input and output BNCs are moved over to the rear panel PCB. As I have the through-hole type of vertical BNCs - I used those, but they could be more usual - hole and nut type which are easier to get. the ground on the main board now is not encircling the circuitry - preventing the stray RF-currents from entering the small-signal area. the input and output signals are to be fed via short bits of coaxial cable (RG-174 may be). the input coax goes directly into the TransMatch binocular, coax's braid is grounded just before the ferrite core. theoretically - the input coax WITH its' braid could pass all the way to the other side of the binocular creating an electric shield between the windings the similar is to be done with the other current TM's winding: it is made of RG-174 too, also braid-grounded at one side. to prevent current loops near the BNCs - the ground on the real panel is also separated. Please comment on the above. Also I'm trying to add the TQFP-44 pads around the original UQFN-40 thus making it easier to get the microcontroller, this is yet to be done. And again - thank, David!
ZAO All Messages By This Member #1054 Hello! Happy New Year! is my attempt to add the TQFP-44 sitting option around the original UQFN-40 pads. Also - I cleared as much ground near the Tandem Match as possible. If you want the on-board BNCs - there it is with its' rear panel. I prefer the off-board BNCs sitting perpendicularly on the rear board itself - there is such a PCB variant as well.
ZAO All Messages By This Member #1172 Hello! I've uploaded to JLCPCB for production (ATU-10_v1-4c2_jlcpcb.lay6 in Sprint Layout format). AD8361 as RF detectors more lean ground everywhere TQFP-44/UQFN-40 pads for the PIC16F18877 chip As I now have the microC for PIC v7.6 and David's sources with the AD8361 support (BIG THANKS, N7DDC!) - I expect build the ATU capable of QRPP. The firmware sources will be posted on the GitHub.
Michael Frachel All Messages By This Member #1173 开云体育 How much have these changes impacted the initial BOM for the? Glenn's V1.3 ATU10 board?? I'm still scraping together all the pieces needed to make that one. I'm still learning how to read and build these things, so everything here is a challenge.? What are the advantages or disadvantages? mike KO4BCN On 2/22/2022 9:42 AM, ZAO via groups.io wrote: toggle quoted message Show quoted text Hello! I've uploaded to JLCPCB for production (ATU-10_v1-4c2_jlcpcb.lay6 in Sprint Layout format). AD8361 as RF detectors more lean ground everywhere TQFP-44/UQFN-40 pads for the PIC16F18877 chip As I now have the microC for PIC v7.6 and David's sources with the AD8361 support (BIG THANKS, N7DDC!) - I expect build the ATU capable of QRPP. The firmware sources will be posted on the GitHub.
Michael Frachel All Messages By This Member #1174 开云体育 To be more specific:? I picked up the board,? PIC programmer and some chips, starting to get all the other components. On 2/22/2022 1:15 PM, Michael Frachel via groups.io wrote: toggle quoted message Show quoted text How much have these changes impacted the initial BOM for the? Glenn's V1.3 ATU10 board?? I'm still scraping together all the pieces needed to make that one. I'm still learning how to read and build these things, so everything here is a challenge.? What are the advantages or disadvantages? mike KO4BCN On 2/22/2022 9:42 AM, ZAO via groups.io wrote: Hello! I've uploaded to JLCPCB for production (ATU-10_v1-4c2_jlcpcb.lay6 in Sprint Layout format). AD8361 as RF detectors more lean ground everywhere TQFP-44/UQFN-40 pads for the PIC16F18877 chip As I now have the microC for PIC v7.6 and David's sources with the AD8361 support (BIG THANKS, N7DDC!) - I expect build the ATU capable of QRPP. The firmware sources will be posted on the GitHub.
ZAO All Messages By This Member #1193 Dear Michael, I did not change much - there are the same 0603 capacitors and resistors, the board is mostly the same as the original. Only the diodes are replaced with the AD8361 detectors and an additional PCB trace is provided to control them. This change requires to amend the firmware, that's what I'm trying to do right now as I wait for the PCBs to arrive. Regards, Tony
Michael Frachel All Messages By This Member #1194 开云体育 Thanks Tony - Forgive my ignorance, how does this improve the system?? I'm learning as I go here. On 2/24/2022 11:40 AM, ZAO via groups.io wrote: toggle quoted message Show quoted text Dear Michael, I did not change much - there are the same 0603 capacitors and resistors, the board is mostly the same as the original. Only the diodes are replaced with the AD8361 detectors and an additional PCB trace is provided to control them. This change requires to amend the firmware, that's what I'm trying to do right now as I wait for the PCBs to arrive. Regards, Tony _._
ZAO All Messages By This Member #1195 Dear Michael, My goal is an attempt to make it possible to tune with as little RF-power as possible. The original project (just as most other SWR-meters and ATUs) employs diodes as RF-detectors. This is OK for higher-power devices, and it requires some sort of diodes' Volt-Amp curve linearisation - because diodes do not rectify well if small signal is applied. The quality of any ATU is (among other things) rated by its' ability to determinate the between the forward and the reflected RF-power; and by the insertion loss it causes. These parameters are called the directivity and the isolation factors. Both of them depend on design of the RF-transformer called the Trandem Match or the binocular". The Tandem Match or equivalent transformers' in the high-power ATUs have windings ratio around 1:20 which provides for better isolation factor (less losses in the direct path). The original ATU-10 has 1:5 ratio - because of QRP nature of that project. The isolation factor is not sufficient (by my liking!). The 1:5 ratio is - as I think - a compromise between the isolation and sensitivity. Make it 1:10 - the isolation improves but the RF-power to be rectified by diodes gets too low (in this QRP project). By using the AD8361 RF-detector instead of more common diodes allows (as I hope) to achieve good isolation and have good sensitivity as well. David has built an ATU with AD8361 before, he recommends to wind 1:10 ratio at the Tandem Match. By guarding the AD8361 inputs with 15dB attenuators it is possible to achieve very good balance of the windings - this assures high directivity factor in the ATU. The directivity is THE factor most important for an ATU because in the process of tuning the less and less power reflects from the load. If it were diodes - they would stop giving the voltage sooner. The AD8361 RF-detectors will provide signals at much lower reflected power levels - thus THEORETICALLY the ATU could tune to better SWRs. Hope to get the PCBs from the factory soonest to give that modification a try. Regards, Tony
Rolf Schroeder All Messages By This Member #1196 开云体育 Hi, I think that it is a very interesting concept and as soon as I can get hold of a PCB I will definitely try that.? Vy 73 de? Rolf, DL8BAG? toggle quoted message Show quoted text Am 28.02.2022 um 18:07 schrieb ZAO via groups.io <zao@...>: ?Dear Michael, My goal is an attempt to make it possible to tune with as little RF-power as possible. The original project (just as most other SWR-meters and ATUs) employs diodes as RF-detectors. This is OK for higher-power devices, and it requires some sort of diodes' Volt-Amp curve linearisation - because diodes do not rectify well if small signal is applied. The quality of any ATU is (among other things) rated by its' ability to determinate the between the forward and the reflected RF-power; and by the insertion loss it causes. These parameters are called the directivity and the isolation factors. Both of them depend on design of the RF-transformer called the Trandem Match or the binocular". The Tandem Match or equivalent transformers' in the high-power ATUs have windings ratio around 1:20 which provides for better isolation factor (less losses in the direct path). The original ATU-10 has 1:5 ratio - because of QRP nature of that project. The isolation factor is not sufficient (by my liking!). The 1:5 ratio is - as I think - a compromise between the isolation and sensitivity. Make it 1:10 - the isolation improves but the RF-power to be rectified by diodes gets too low (in this QRP project). By using the AD8361 RF-detector instead of more common diodes allows (as I hope) to achieve good isolation and have good sensitivity as well. David has built an ATU with AD8361 before, he recommends to wind 1:10 ratio at the Tandem Match. By guarding the AD8361 inputs with 15dB attenuators it is possible to achieve very good balance of the windings - this assures high directivity factor in the ATU. The directivity is THE factor most important for an ATU because in the process of tuning the less and less power reflects from the load. If it were diodes - they would stop giving the voltage sooner. The AD8361 RF-detectors will provide signals at much lower reflected power levels - thus THEORETICALLY the ATU could tune to better SWRs. Hope to get the PCBs from the factory soonest to give that modification a try. Regards, Tony
Kees T All Messages By This Member #1197 Edited The AD8361 is a great part and ideal for this application or as a mWattmeter. Are you testing the SOIC or SOT23 package ? You might have to add a better (more bits) ADC with a I2C interface like the MCP3422A0. 73 Kees K5BCQ
ZAO All Messages By This Member #1198 Hello! The PCB is designed to accept SOT-23-6. My point of view - because the capacitors and inductors have only 128 steps (just seven relays, so 2[sup]7[/sup] available values), there is no sense in higher resolution ADCs. The built-in are 12 bits and David adjusts the reference voltage to accomodate the input signal level: for coarse measurements there is 2-volts VREF, for finer measurements of lower signals he switches to 1-volt VREF. This should be more than enough (I think). I'm winding the coils trying to have them doubling the inductance, so that the "power of two" set of values would be obtained. This is VERY difficult to achieve. The first three coils are easy, from there on - even a slightest movement of individual wires changes the resulting inductance severely - way much more that the by the value of the first (Least Significant Bit) coil. To obtain a smooth and monotonous "inductance vs control code" characteristic I need to somehow glue the wires to the cores. With capacitors it is possible to choose the values for the "power of two" set, and hope that the C0G material will be stable enough to keep it. Initially I was looking for 1% capacitors but higher-voltage ones are very expensive (not all values are available even at Mouser!). Now I'm peeking among those I've got so far. Regards, Tony
Robbie M1ROK All Messages By This Member #1199 I'm probably having a blonde moment here but as the?AD8361 is an RF detector for the 100MHz-2GHz range how can it be accurate at the lower HF frequencies? Just curious... Robbie M1ROK
G8DQX list All Messages By This Member #1200 Robbie, blonde is the new forgetful! Try , where AN-691 () discusses LF operation of the AD8361, and various other chips. HTH, 73, Stay Safe, Robin, G8DQX toggle quoted message Show quoted text On 01/03/2022 11:30, Robbie M1ROK wrote: I'm probably having a blonde moment here but as the?AD8361 is an RF detector for the 100MHz-2GHz range how can it be accurate at the lower HF frequencies? Just curious... Robbie M1ROK
Bob N3MNT All Messages By This Member #1202 I am working on similar project.? Does anyone have the EEPROM code? .? The tuner I am trying to modify has the locked the read function.
Kees T All Messages By This Member #1205 Edited Hello ZAO, If you assume a 50ohm system and max input of 10W (40dBm) and the max input of the AD8361 is 0.4Vrms (5dBm) that implies a coupling ratio and voltage divider attenuation of -35dBm ? What values are you using for the turns ratio and voltage divider ?? The coupling ratio effect of 10 turns would be 20 log 0,1 or -20dB. 73 Kees K5BCQ
ZAO All Messages By This Member #1209 Dear Kees, Yes, I expect the 40 dBm maximum input level, and to fit it into AD8361's 0,4 Vrms allowed signal I plan to wind 10 turns, then I've arranged a Pi matching attenuator (56,2 Ohm shunt in, 287 Ohm series, 787 Ohm shunt out) of ~15 dB. That attenuator also helps to alleviate the disbalance caused by +/-20% allowance of chip's input impedance (nominally equal to 225 Ohm). Warm regards, Tony
Kees T All Messages By This Member #1210 Hello Tony, Sounds great and awaiting your results. 73 Kees K5BCQ?
ZAO All Messages By This Member #1212 Hello! As I said earlier (or elswhere) - I work with microC for PIC, so this is my GitHub Just made initial changes to David's v1.4 code (using kindly provided samples of his so called "symmetric" ATU sources).
ZAO All Messages By This Member #1417 Hello! This is a cleanup version of the ATU-10 (w/ AD8361) PCB - with USB-C charging error fixed and the option to use PIC16F1454 in either the TSSOP or QFN packages. Sorry, I could not do a provision for the SOIC. Hope, this makes it easyer to assemble the device. Warm regards, Tony ATU-10_v1-4c5_jlcpcb.lay6 ATU-10_v1-4c5_jlcpcb.lay6
Glenn All Messages By This Member #1418 开云体育 Tony Can you post a pdf or jpg of the pcb.?? I am unable to open that type u posted. Glenn

Previous Topic Next Topic