开云体育

ATU-10 Random Wire Testing

Test completed with a Tufteln 9:1 QRP Antenna

Antenna configured with an elevated feed point slopper and counterpoise hanging straight down

Coax RG316 17'

ATU-10 Sourced from newdiytech.com, price was $120.24 shipped to me in GA USA, Ordered June 25, delivered July 8th.

ATU-10 Features

0.91" OLED Display that shows Power, SWR and internal battery remaining.

USB-C Rechargeable LiPo 1.7Ah

Grounding Lug

Bypass Mode

Latching Relays

Input port for communication with IC705 (and potentially others)

7 Inductors, 8 capacitors (let me know if this is not right, clearly 7 inductors, 15 relays...)

Inductor and Capacitance Range -?

ATU-10 with Protection Cover

SWR measured with a RigExpert RigStick 320, Lab599 TX500 and the ATU-10

The test process was to first check the SWR on the antenna with no tuner. SWR values recorded from the TX500 and RigExpert320. Values recorded in the 2 columns under the "No Tuner" section. This was completed for each of the Bands listed in the table rows. SWR values were the lowest in the band range for all recorded numbers.

The second test was with the ATU-10 tuned. I tuned the antenna with the ATU-10 from the tune function on the TX500 and recorded the value on the ATU-10 and the TX500. I then disconnected the radio and checked the SWR with the RigExpert 320 through the ATU-10. These three values are recorded in the "ATU-10 Tuned" columns.

The third test was the same as the previous but after tuning with the Elecraft T1.

These tests were then completed with the same setup with only a different radiator length. See the table name in each for the length.

Test Conclusions

I will keep this brief as the data is self explanatory. The ATU-10 is not a great tuner. Maybe it is just the one I got? The Elecraft T1 is excellent (not the intention of this testing but of course no surprise on the results). The ATU-10 i received from newdiytech.com was not DOA. It reads power level, shows an SWR level, Charges, etc. but it just doesn’t do a good job tuning the antenna. I will have to do some further testing and research on the device and see where this goes.

Other thoughts

Yes, a brief conclusion for sure. I am not very familiar with this ATU and to be honest I didn't have very high expectations. The display with power and SWR reading was very attractive for me as well as the opportunity to offer another protection cover like I make for the T1. One reason I chose newdiytech.com was that they are one of the very few approved vendors for the (tr)uSDX and I have seen several times there they appear to have great customer support. Did I mention the T1 is great? Currently they have shipping time from Elecraft listed as TBD due to ship shortages I believe. The ATU-10 I received from newdiytech.com was not DOA. It reads power level, shows an SWR level, Charges, etc. but it just doesn’t do a good job tuning the antenna. I will have to do some further testing and research on the device and see where this goes.

Joshua

KO4AWH

For those still Interested a case:

Hi, Joshua.?
Thank you, nice job.
ATU-10 requires new his own tuning algorithm.
The old one using in ATU-100 is not working properly there. I will work on it soon.?

Dear Joshua,
Thanks for the report!

Dear David,
I think most of the problems in the ATU-10 are not for the tuning algorithm deficiences (they are not proven to exist), in my opinion the root of the problem is the set of caps' and inductances' values chosen.
They heavily deviate from the powers of two line. And what's more - the 22pF is far too much for the smallest cap there. Should be 10pF, or better yet 5pF.
Here is . Notice the gaps and the irregular "holes" in the shaded areas. The lower half of the page has the caps' ans inds' values equal to the powers of two. Much more sane behaviour, - but the 22pF as the least capacitance still creates large gaps on the edges of the graph.
These (let them call it) bad values would create problems in optimum finding for any algorithm, even for the true brute-force one.

Warm regards,
Tony

Hi, Tony

Dear David,

Your point that ATUs are most valuable at lower frequencies is perfectly valid. I do not object to that.
I'll try (ASAP) to run the analyser script with slightly different set of the capacitances - to check if it is possible to optimise the performance at the lower end not compromising the performance at the higher end.

Regarding the algorithm, - my opinion is that those gaps and holes "confuse" the otherwise workable algo. The fact that a manual tuning allows to find a better combinations is solely to the obvious superiority of human brain to that tiny PIC's.

BRB with the analyser's plots.

Warm regards,
Tony

Hello!

I tried to add a static capacitance that would be there in parallels with the seven tunable capacitors.
Alas, it works only at one chosen band. The other bands suffer from significant degradation...

The only option is to make the switching matrix larger, - 8*8 or more.

Looks like a user should choose where he wants ATU to operate, - and adapt the reactances accordingly.

As for me, - I'm planning to use ATU-10 on outdoors hikes where I will be able to use simple antennas on 14MHz band. All lower bands are closed for me - because of the antenna size. And while at home - I blocked off the air by huge level of industrial noise.
:(

Warm regards,
Tony

I believe my use for this ATU matches quite well with you Tony.

Is there any reason to not simply swap out the capacitors and inductors to be closer to say the Elecraft T1 C and L values if the matching ability then works better for the user's specific purpose?

Very nice job on this David, what a great concept this ATU is with its feature set.? I may have to give the manual mode a try with a few of the test that failed to provide a lower SWR. Do you recommend adjusting L or C first? I can imagine how to do this but will have to play with the manual mode Firmware a bit more to understand how it work. Assuming it starts each L and C network in bypass then steps through but not sure if there is an easy way to "reset" each to start the manual matching process over again. Although it sounds like an updated algorithm is maybe all that is needed.

Joshua
KO4AWH

Dear Joshua,

As of today the only thing left to do here (for me) - is to get the proper values of HV-capacitors. I finally got the aliexpress contact that is ready to sell the 250Volts 0805 caps with the values I prefer. Hope to get a hold of them in ~two~three weeks. Then I will be able to tell - doing the field test - what's wrong with the original ATU-10 v1.4, - the algo or its' non-following the powers of two in the reactances. Also I'm eager to try the AD8361 TruPWR detectors instead of diode RF-rectifiers.

Regards,
Tony

Tony

I actually designed a PCB for the 8x8 format on the ATU-10 board. Based on the 7x7 of course.

I never had any actually made though, as David said that he doesn't support that format.?? Perhaps somebody who is developing code for the ATU-10 might add in an option for 8 x 8 at some point.

Glenn
vk3pe

Dear Glenn,

Yes, as soon as I noticed that there are two spare pins left on the PIC16F18877 located exactly where they could be used to control the relays, - I got an idea to try an 8*8.
Most probably I'll do this - on the original PCB with AD8361 capability, looks not so difficult.

Another promising option to enhance ATU-10 is to make it a T-type high-pass circuit - with a "differential" step capacitor. All the unused capacitors on one side of the inductance to be switched to the other side. Really easy PCB alteration required. There will be no need in LC/CL switching, that relay could be used as the eighth stage of the C-matrix. So, with today's fifteen relays it would be possible to obtain 8C*7L matrix. gives really . With one more relay crammed in - in the L row - the T-type would be 8*8.
I saw an ATU with just two knobs - one for the vary-cap, the other for the vary-ind, forgot its' brand name.

I'm really set and dedicated to extract the very best from this ATU-10 project.

Warm regards,
Tony

Tony
These are the pins I used.

Glenn
vk3pe

Hello!

What I mean by the "T-type differential capacitor ATU" is the .
Yes, even with 256 steps of variable capacitance and inductance - provided we make the 8*8 relay matrix - it is still not the same as truely smooth adjustments of the AT2KD's, - but I think it is worth a try. In such a T-type the capacitance is reasonably small (13pF~430pF), and the inductor is not too huge, - within 25?H.

As I can imaging, - the optimum finding with the T-type-differential configuration should be much more straightforward, - first you seek the SWR minimum with the coil, then - with the capacitor.
The required PCB alteration is almost negligible and will simplify the layout.

What do you think?

Warm regards,
Tony

Hi, Tony

Dear David,

Did you simulate the T-types as well?
┌─窜颁1─┬─窜颁2─
Zapp????? ZL??????? <─┐
└─────┴────? Zload
where C1+C2==const, i.e. the "differential" capacitor

As per my simulations - the T-type with the differential capacitor looks almost perfect.

Warm regards,
Tony

Hi, Tony

I reached out to the vendor on the performance of this ATU. They asked for a video. Here is that video of the operation of the ATU.??

I will echo your experience, my main antenna tunes to almost 1:1 with my FT-950 internal tuner and that has a limited range. I expected this to be able to tune that antenna with no issues, but like you have shown, 20m best I can get is in the mid 2’s sometimes 3. Other bands are OK but not great either.

@Tony:

I tried to use your script on my laptop running Python 3.10 and I I get the following error:

"TypeError: 'numpy.float64' object cannot be interpreted as an integer"

Not really well versed in Python yet, any thoughts on getting this working?

Dear Lou,

Yes, this is the well-known and long-outstanding problem, - there is the patch that solves this. But - unfortunately - the Smithplot author - for some unknown reason - does not apply it officially.

Here is the patch (attached).

Warm regards,
Tony